19 files changed, 7326 insertions, 7437 deletions
diff --git a/mesalib/docs/MESA_agp_offset.spec b/mesalib/docs/MESA_agp_offset.spec
index bf3ea1748..06e1d902e 100644
--- a/mesalib/docs/MESA_agp_offset.spec
+++ b/mesalib/docs/MESA_agp_offset.spec
@@ -1,95 +1,95 @@
-Name
-
-    MESA_agp_offset
-
-Name Strings
-
-    GLX_MESA_agp_offset
-
-Contact
-
-    Brian Paul, Tungsten Graphics, Inc. (brian.paul 'at' tungstengraphics.com)
-    Keith Whitwell, Tungsten Graphics, Inc.  (keith 'at' tungstengraphics.com)
-
-Status
-
-    Shipping (Mesa 4.0.4 and later.  Only implemented in particular
-    XFree86/DRI drivers.)
-
-Version
-
-    1.0
-
-Number
-
-    TBD
-
-Dependencies
-
-    OpenGL 1.0 or later is required
-    GLX_NV_vertex_array_range is required.
-    This extensions is written against the OpenGL 1.4 Specification.
-
-Overview
-
-    This extensions provides a way to convert pointers in an AGP memory
-    region into byte offsets into the AGP aperture.
-    Note, this extension depends on GLX_NV_vertex_array_range, for which
-    no real specification exists.  See GL_NV_vertex_array_range for more
-    information.
-
-IP Status
-
-    None
-
-Issues
-
-    None
-
-New Procedures and Functions
-
-    unsigned int glXGetAGPOffsetMESA( const void *pointer )
-
-New Tokens
-
-    None
-
-Additions to the OpenGL 1.4 Specification
-
-    None
-
-Additions to Chapter 3 the GLX 1.4 Specification (Functions and Errors)
-
-    Add a new section, 3.6 as follows:
-
-    3.6 AGP Memory Access
-
-    On "PC" computers, AGP memory can be allocated with glXAllocateMemoryNV
-    and freed with glXFreeMemoryNV.  Sometimes it's useful to know where a
-    block of AGP memory is located with respect to the start of the AGP
-    aperture.  The function
-
-        GLuint glXGetAGPOffsetMESA( const GLvoid *pointer )
-
-    Returns the offset of the given memory block from the start of AGP
-    memory in basic machine units (i.e. bytes).  If pointer is invalid
-    the value ~0 will be returned.
-
-GLX Protocol
-
-    None.  This is a client side-only extension.
-
-Errors
-
-    glXGetAGPOffsetMESA will return ~0 if the pointer does not point to
-    an AGP memory region.
-
-New State
-
-    None
-
-Revision History
-
-    20 September 2002 - Initial draft
-    2 October 2002 - finished GLX chapter 3 additions
-    27 July 2004 - use unsigned int instead of GLuint, void instead of GLvoid
+Name
+
+    MESA_agp_offset
+
+Name Strings
+
+    GLX_MESA_agp_offset
+
+Contact
+
+    Brian Paul, Tungsten Graphics, Inc. (brian.paul 'at' tungstengraphics.com)
+    Keith Whitwell, Tungsten Graphics, Inc.  (keith 'at' tungstengraphics.com)
+
+Status
+
+    Shipping (Mesa 4.0.4 and later.  Only implemented in particular
+    XFree86/DRI drivers.)
+
+Version
+
+    1.0
+
+Number
+
+    TBD
+
+Dependencies
+
+    OpenGL 1.0 or later is required
+    GLX_NV_vertex_array_range is required.
+    This extensions is written against the OpenGL 1.4 Specification.
+
+Overview
+
+    This extensions provides a way to convert pointers in an AGP memory
+    region into byte offsets into the AGP aperture.
+    Note, this extension depends on GLX_NV_vertex_array_range, for which
+    no real specification exists.  See GL_NV_vertex_array_range for more
+    information.
+
+IP Status
+
+    None
+
+Issues
+
+    None
+
+New Procedures and Functions
+
+    unsigned int glXGetAGPOffsetMESA( const void *pointer )
+
+New Tokens
+
+    None
+
+Additions to the OpenGL 1.4 Specification
+
+    None
+
+Additions to Chapter 3 the GLX 1.4 Specification (Functions and Errors)
+
+    Add a new section, 3.6 as follows:
+
+    3.6 AGP Memory Access
+
+    On "PC" computers, AGP memory can be allocated with glXAllocateMemoryNV
+    and freed with glXFreeMemoryNV.  Sometimes it's useful to know where a
+    block of AGP memory is located with respect to the start of the AGP
+    aperture.  The function
+
+        GLuint glXGetAGPOffsetMESA( const GLvoid *pointer )
+
+    Returns the offset of the given memory block from the start of AGP
+    memory in basic machine units (i.e. bytes).  If pointer is invalid
+    the value ~0 will be returned.
+
+GLX Protocol
+
+    None.  This is a client side-only extension.
+
+Errors
+
+    glXGetAGPOffsetMESA will return ~0 if the pointer does not point to
+    an AGP memory region.
+
+New State
+
+    None
+
+Revision History
+
+    20 September 2002 - Initial draft
+    2 October 2002 - finished GLX chapter 3 additions
+    27 July 2004 - use unsigned int instead of GLuint, void instead of GLvoid
diff --git a/mesalib/docs/MESA_pack_invert.spec b/mesalib/docs/MESA_pack_invert.spec
index 0eeb58c13..33fb3c7bf 100644
--- a/mesalib/docs/MESA_pack_invert.spec
+++ b/mesalib/docs/MESA_pack_invert.spec
@@ -1,138 +1,138 @@
-Name
-
-    MESA_pack_invert
-
-Name Strings
-
-    GL_MESA_pack_invert
-
-Contact
-
-    Brian Paul, Tungsten Graphics, Inc. (brian.paul 'at' tungstengraphics.com)
-    Keith Whitwell, Tungsten Graphics, Inc.  (keith 'at' tungstengraphics.com)
-
-Status
-
-    Shipping (Mesa 4.0.4 and later)
-
-Version
-
-    1.0
-
-Number
-
-    TBD
-
-Dependencies
-
-    OpenGL 1.0 or later is required
-    This extensions is written against the OpenGL 1.4 Specification.
-
-Overview
-
-    This extension adds a new pixel storage parameter to indicate that
-    images are to be packed in top-to-bottom order instead of OpenGL's
-    conventional bottom-to-top order.  Only pixel packing can be
-    inverted (i.e. for glReadPixels, glGetTexImage, glGetConvolutionFilter,
-    etc).
-
-    Almost all known image file formats store images in top-to-bottom
-    order.  As it is, OpenGL reads images from the frame buffer in
-    bottom-to-top order.  Thus, images usually have to be inverted before
-    writing them to a file with image I/O libraries.  This extension
-    allows images to be read such that inverting isn't needed.
-
-IP Status
-
-    None
-
-Issues
-
-    1. Should we also define UNPACK_INVERT_MESA for glDrawPixels, etc?
-
-    Resolved:  No, we're only concerned with pixel packing.  There are other
-    solutions for inverting images when using glDrawPixels (negative Y pixel
-    zoom) or glTexImage (invert the vertex T coordinates).  It would be easy
-    enough to define a complementary extension for pixel packing in the
-    future if needed.
-
-New Procedures and Functions
-
-    None
-
-New Tokens
-
-    Accepted by the <pname> parameter of PixelStorei and PixelStoref
-    and the <pname> parameter of GetIntegerv, GetFloatv, GetDoublev
-    and GetBooleanv:
-
-        PACK_INVERT_MESA                   0x8758
-
-Additions to Chapter 2 of the OpenGL 1.4 Specification (OpenGL Operation)
-
-    None
-
-Additions to Chapter 3 of the OpenGL 1.4 Specification (Rasterization)
-
-    None
-
-Additions to Chapter 4 of the OpenGL 1.4 Specification (Per-Fragment
-Operations and the Frame Buffer)
-
-    Add the following entry to table 4.4 (PixelStore parameters) on page 182:
-
-    Parameter Name       Type    Initial Value    Valid Range
-    ---------------------------------------------------------
-    PACK_INVERT_MESA     boolean     FALSE        TRUE/FALSE
-
-    In the section labeled "Placement in Client Memory" on page 184
-    insert the following text into the paragraph before the sentence
-    that starts with "If the format is RED, GREEN, BLUE...":
-
-    "The parameter PACK_INVERT_MESA controls whether the image is packed
-     in bottom-to-top order (the default) or top-to-bottom order.  Equation
-     3.8 is modified as follows:
-
-     ... the first element of the Nth row is indicated by
-
-         p + Nk,                if PACK_INVERT_MESA is false
-         p + k * (H - 1) - Nk,  if PACK_INVERT_MESA is true, where H is the
-                                image height
-    "
-
-Additions to Chapter 5 of the OpenGL 1.4 Specification (Special Functions)
-
-    None
-
-Additions to Chapter 6 of the OpenGL 1.4 Specification (State and
-State Requests)
-
-    None
-
-Additions to Appendix A of the OpenGL 1.4 Specification (Invariance)
-
-    None
-
-Additions to the AGL/GLX/WGL Specifications
-
-    None
-
-GLX Protocol
-
-    None
-
-Errors
-
-    None
-
-New State
-
-    Add the following entry to table 6.20 (Pixels) on page 235:
-
-    Get Value         Type     Get Cmd    Initial Value  Description                Sec    Attribute
-    --------------------------------------------------------------------------------------------------
-    PACK_INVERT_MESA  boolean  GetBoolean  FALSE         Value of PACK_INVERT_MESA  4.3.2  pixel-store
-
-Revision History
-
-    21 September 2002 - Initial draft
+Name
+
+    MESA_pack_invert
+
+Name Strings
+
+    GL_MESA_pack_invert
+
+Contact
+
+    Brian Paul, Tungsten Graphics, Inc. (brian.paul 'at' tungstengraphics.com)
+    Keith Whitwell, Tungsten Graphics, Inc.  (keith 'at' tungstengraphics.com)
+
+Status
+
+    Shipping (Mesa 4.0.4 and later)
+
+Version
+
+    1.0
+
+Number
+
+    TBD
+
+Dependencies
+
+    OpenGL 1.0 or later is required
+    This extensions is written against the OpenGL 1.4 Specification.
+
+Overview
+
+    This extension adds a new pixel storage parameter to indicate that
+    images are to be packed in top-to-bottom order instead of OpenGL's
+    conventional bottom-to-top order.  Only pixel packing can be
+    inverted (i.e. for glReadPixels, glGetTexImage, glGetConvolutionFilter,
+    etc).
+
+    Almost all known image file formats store images in top-to-bottom
+    order.  As it is, OpenGL reads images from the frame buffer in
+    bottom-to-top order.  Thus, images usually have to be inverted before
+    writing them to a file with image I/O libraries.  This extension
+    allows images to be read such that inverting isn't needed.
+
+IP Status
+
+    None
+
+Issues
+
+    1. Should we also define UNPACK_INVERT_MESA for glDrawPixels, etc?
+
+    Resolved:  No, we're only concerned with pixel packing.  There are other
+    solutions for inverting images when using glDrawPixels (negative Y pixel
+    zoom) or glTexImage (invert the vertex T coordinates).  It would be easy
+    enough to define a complementary extension for pixel packing in the
+    future if needed.
+
+New Procedures and Functions
+
+    None
+
+New Tokens
+
+    Accepted by the <pname> parameter of PixelStorei and PixelStoref
+    and the <pname> parameter of GetIntegerv, GetFloatv, GetDoublev
+    and GetBooleanv:
+
+        PACK_INVERT_MESA                   0x8758
+
+Additions to Chapter 2 of the OpenGL 1.4 Specification (OpenGL Operation)
+
+    None
+
+Additions to Chapter 3 of the OpenGL 1.4 Specification (Rasterization)
+
+    None
+
+Additions to Chapter 4 of the OpenGL 1.4 Specification (Per-Fragment
+Operations and the Frame Buffer)
+
+    Add the following entry to table 4.4 (PixelStore parameters) on page 182:
+
+    Parameter Name       Type    Initial Value    Valid Range
+    ---------------------------------------------------------
+    PACK_INVERT_MESA     boolean     FALSE        TRUE/FALSE
+
+    In the section labeled "Placement in Client Memory" on page 184
+    insert the following text into the paragraph before the sentence
+    that starts with "If the format is RED, GREEN, BLUE...":
+
+    "The parameter PACK_INVERT_MESA controls whether the image is packed
+     in bottom-to-top order (the default) or top-to-bottom order.  Equation
+     3.8 is modified as follows:
+
+     ... the first element of the Nth row is indicated by
+
+         p + Nk,                if PACK_INVERT_MESA is false
+         p + k * (H - 1) - Nk,  if PACK_INVERT_MESA is true, where H is the
+                                image height
+    "
+
+Additions to Chapter 5 of the OpenGL 1.4 Specification (Special Functions)
+
+    None
+
+Additions to Chapter 6 of the OpenGL 1.4 Specification (State and
+State Requests)
+
+    None
+
+Additions to Appendix A of the OpenGL 1.4 Specification (Invariance)
+
+    None
+
+Additions to the AGL/GLX/WGL Specifications
+
+    None
+
+GLX Protocol
+
+    None
+
+Errors
+
+    None
+
+New State
+
+    Add the following entry to table 6.20 (Pixels) on page 235:
+
+    Get Value         Type     Get Cmd    Initial Value  Description                Sec    Attribute
+    --------------------------------------------------------------------------------------------------
+    PACK_INVERT_MESA  boolean  GetBoolean  FALSE         Value of PACK_INVERT_MESA  4.3.2  pixel-store
+
+Revision History
+
+    21 September 2002 - Initial draft
diff --git a/mesalib/docs/MESA_release_buffers.spec b/mesalib/docs/MESA_release_buffers.spec
index e906e1459..52d1e5a9c 100644
--- a/mesalib/docs/MESA_release_buffers.spec
+++ b/mesalib/docs/MESA_release_buffers.spec
@@ -1,85 +1,85 @@
-Name
-
-    MESA_release_buffers
-
-Name Strings
-
-    GLX_MESA_release_buffers
-
-Contact
-
-    Brian Paul (brian.paul 'at' tungstengraphics.com)
-
-Status
-
-    Shipping since Mesa 2.0 in October, 1996.
-
-Version
-
-    Last Modified Date:  8 June 2000
-
-Number
-
-    217
-
-Dependencies
-
-    OpenGL 1.0 or later is required.
-    GLX 1.0 or later is required.
-
-Overview
-
-    Mesa's implementation of GLX is entirely implemented on the client side.
-    Therefore, Mesa cannot immediately detect when an X window or pixmap is
-    destroyed in order to free any ancillary data associated with the window
-    or pixmap.
-
-    The glxMesaReleaseBuffers() function can be used to explicitly indicate
-    when the back color buffer, depth buffer, stencil buffer, and/or accumu-
-    lation buffer associated with a drawable can be freed.
-
-IP Status
-
-    Open-source; freely implementable.
-
-Issues
-
-    None.
-
-New Procedures and Functions
-
-    Bool glXReleaseBuffersMESA( Display *dpy, GLXDrawable d );
-
-New Tokens
-
-    None.
-
-Additions to Chapter 3 of the GLX 1.3 Specification (Functions and Errors)
-
-    The function
-
-	Bool glXReleaseBuffersMESA( Display *dpy, GLXDrawable d );
-
-    causes all software ancillary buffers (back buffer, depth, stencil,
-    accum, etc) associated with the named drawable to be immediately
-    deallocated.  True is returned if <d> is a valid Mesa GLX drawable,
-    else False is returned.  After calling glXReleaseBuffersMESA, the
-    drawable should no longer be used for GL rendering.  Results of
-    attempting to do so are undefined.
-
-
-GLX Protocol
-
-    None, since this is a client-side operation.
-
-Errors
-
-    None.
-
-New State
-
-    None.
-
-Revision History
-
-    8 June 2000 - initial specification
+Name
+
+    MESA_release_buffers
+
+Name Strings
+
+    GLX_MESA_release_buffers
+
+Contact
+
+    Brian Paul (brian.paul 'at' tungstengraphics.com)
+
+Status
+
+    Shipping since Mesa 2.0 in October, 1996.
+
+Version
+
+    Last Modified Date:  8 June 2000
+
+Number
+
+    217
+
+Dependencies
+
+    OpenGL 1.0 or later is required.
+    GLX 1.0 or later is required.
+
+Overview
+
+    Mesa's implementation of GLX is entirely implemented on the client side.
+    Therefore, Mesa cannot immediately detect when an X window or pixmap is
+    destroyed in order to free any ancillary data associated with the window
+    or pixmap.
+
+    The glxMesaReleaseBuffers() function can be used to explicitly indicate
+    when the back color buffer, depth buffer, stencil buffer, and/or accumu-
+    lation buffer associated with a drawable can be freed.
+
+IP Status
+
+    Open-source; freely implementable.
+
+Issues
+
+    None.
+
+New Procedures and Functions
+
+    Bool glXReleaseBuffersMESA( Display *dpy, GLXDrawable d );
+
+New Tokens
+
+    None.
+
+Additions to Chapter 3 of the GLX 1.3 Specification (Functions and Errors)
+
+    The function
+
+	Bool glXReleaseBuffersMESA( Display *dpy, GLXDrawable d );
+
+    causes all software ancillary buffers (back buffer, depth, stencil,
+    accum, etc) associated with the named drawable to be immediately
+    deallocated.  True is returned if <d> is a valid Mesa GLX drawable,
+    else False is returned.  After calling glXReleaseBuffersMESA, the
+    drawable should no longer be used for GL rendering.  Results of
+    attempting to do so are undefined.
+
+
+GLX Protocol
+
+    None, since this is a client-side operation.
+
+Errors
+
+    None.
+
+New State
+
+    None.
+
+Revision History
+
+    8 June 2000 - initial specification
diff --git a/mesalib/docs/MESA_resize_buffers.spec b/mesalib/docs/MESA_resize_buffers.spec
index 4d6e9f720..dabc7c421 100644
--- a/mesalib/docs/MESA_resize_buffers.spec
+++ b/mesalib/docs/MESA_resize_buffers.spec
@@ -1,81 +1,81 @@
-Name
-
-    MESA_resize_buffers
-
-Name Strings
-
-    GL_MESA_resize_buffers
-
-Contact
-
-    Brian Paul (brian.paul 'at' tungstengraphics.com)
-
-Status
-
-    Shipping (since Mesa version 2.2)
-
-Version
-
-
-Number
-
-    196
-
-Dependencies
-
-    Mesa 2.2 or later is required.
-
-Overview
-
-    Mesa is often used as a client library with no integration with
-    the computer's window system (an X server, for example).  And since
-    Mesa does not have an event loop nor window system callbacks, it
-    cannot properly respond to window system events.  In particular,
-    Mesa cannot automatically detect when a window has been resized.
-
-    Mesa's glViewport command queries the current window size and updates
-    its internal data structors accordingly.  This normally works fine
-    since most applications call glViewport in response to window size
-    changes.
-
-    In some situations, however, the application may not call glViewport
-    when a window size changes but would still like Mesa to adjust to
-    the new window size.  This extension exports a new function to solve
-    this problem.
-
-New Procedures and Functions
-
-    void glResizeBuffersMESA( void )
-
-New Tokens
-
-    none
-
-Additions to the OpenGL Specification (no particular section)
-
-    The glResizeBuffersMESA command may be called when the client
-    determines that a window has been resized.  Calling
-    glResizeBuffersMESA causes Mesa to query the current window size
-    and adjust its internal data structures.  This may include
-    reallocating depth, stencil, alpha and accumulation buffers.
-
-Additions to the AGL/GLX/WGL Specifications
-
-    None
-
-Errors
-
-    INVALID_OPERATION is generated if glResizeBuffersMESA is called between
-    Begin and End.
-
-New State
-
-    None.
-
-New Implementation Dependent State
-
-    None.
-
-Revision History
-
-  * Revision 1.0 - Initial specification
+Name
+
+    MESA_resize_buffers
+
+Name Strings
+
+    GL_MESA_resize_buffers
+
+Contact
+
+    Brian Paul (brian.paul 'at' tungstengraphics.com)
+
+Status
+
+    Shipping (since Mesa version 2.2)
+
+Version
+
+
+Number
+
+    196
+
+Dependencies
+
+    Mesa 2.2 or later is required.
+
+Overview
+
+    Mesa is often used as a client library with no integration with
+    the computer's window system (an X server, for example).  And since
+    Mesa does not have an event loop nor window system callbacks, it
+    cannot properly respond to window system events.  In particular,
+    Mesa cannot automatically detect when a window has been resized.
+
+    Mesa's glViewport command queries the current window size and updates
+    its internal data structors accordingly.  This normally works fine
+    since most applications call glViewport in response to window size
+    changes.
+
+    In some situations, however, the application may not call glViewport
+    when a window size changes but would still like Mesa to adjust to
+    the new window size.  This extension exports a new function to solve
+    this problem.
+
+New Procedures and Functions
+
+    void glResizeBuffersMESA( void )
+
+New Tokens
+
+    none
+
+Additions to the OpenGL Specification (no particular section)
+
+    The glResizeBuffersMESA command may be called when the client
+    determines that a window has been resized.  Calling
+    glResizeBuffersMESA causes Mesa to query the current window size
+    and adjust its internal data structures.  This may include
+    reallocating depth, stencil, alpha and accumulation buffers.
+
+Additions to the AGL/GLX/WGL Specifications
+
+    None
+
+Errors
+
+    INVALID_OPERATION is generated if glResizeBuffersMESA is called between
+    Begin and End.
+
+New State
+
+    None.
+
+New Implementation Dependent State
+
+    None.
+
+Revision History
+
+  * Revision 1.0 - Initial specification
diff --git a/mesalib/docs/MESA_swap_control.spec b/mesalib/docs/MESA_swap_control.spec
index 31340772f..a002563c9 100644
--- a/mesalib/docs/MESA_swap_control.spec
+++ b/mesalib/docs/MESA_swap_control.spec
@@ -1,129 +1,129 @@
-Name
-
-    MESA_swap_control
-
-Name Strings
-
-    GLX_MESA_swap_control
-
-Contact
-
-    Ian Romanick, IBM, idr at us.ibm.com
-
-Status
-
-    Deployed in DRI drivers post-XFree86 4.3.
-
-Version
-
-    Date: 5/1/2003   Revision: 1.1
-
-Number
-
-    ???
-
-Dependencies
-
-    None
-
-    Based on GLX_SGI_swap_control version 1.9 and WGL_EXT_swap_control
-    version 1.5.
-
-Overview
-
-    This extension allows an application to specify a minimum periodicity
-    of color buffer swaps, measured in video frame periods.
-
-Issues
-
-    * Should implementations that export GLX_MESA_swap_control also export
-      GL_EXT_swap_control for compatibility with WGL_EXT_swap_control?
-
-    UNRESOLVED.
-
-New Procedures and Functions
-
-    int glXSwapIntervalMESA(unsigned int interval)
-    int glXGetSwapIntervalMESA(void)
-
-New Tokens
-
-    None
-
-Additions to Chapter 2 of the 1.4 GL Specification (OpenGL Operation)
-
-    None
-
-Additions to Chapter 3 of the 1.4 GL Specification (Rasterization)
-
-    None
-
-Additions to Chapter 4 of the 1.4 GL Specification (Per-Fragment Operations
-and the Framebuffer)
-
-    None
-
-Additions to Chapter 5 of the 1.4 GL Specification (Special Functions)
-
-    None
-
-Additions to Chapter 6 of the 1.4 GL Specification (State and State Requests)
-
-    None
-
-Additions to the GLX 1.3 Specification
-
-    [Add the following to Section 3.3.10 of the GLX Specification (Double
-     Buffering)]
-
-    glXSwapIntervalMESA specifies the minimum number of video frame periods
-    per buffer swap.  (e.g. a value of two means that the color buffers
-    will be swapped at most every other video frame.)  A return value
-    of zero indicates success; otherwise an error occurred.  The interval
-    takes effect when glXSwapBuffers is first called subsequent to the
-    glXSwapIntervalMESA call.
-
-    A video frame period is the time required by the monitor to display a 
-    full frame of video data.  In the case of an interlaced monitor,
-    this is typically the time required to display both the even and odd 
-    fields of a frame of video data.
-
-    If <interval> is set to a value of 0, buffer swaps are not synchro-
-    nized to a video frame.  The <interval> value is silently clamped to
-    the maximum implementation-dependent value supported before being
-    stored.
-
-    The swap interval is not part of the render context state.  It cannot
-    be pushed or popped.  The current swap interval for the window
-    associated with the current context can be obtained by calling
-    glXGetSwapIntervalMESA.  The default swap interval is 0.
-
-    On XFree86, setting the environment variable LIBGL_THROTTLE_REFRESH sets
-    the swap interval to 1.
-
-Errors
-
-    glXSwapIntervalMESA returns GLX_BAD_CONTEXT if there is no current
-    GLXContext or if the current context is not a direct rendering context.
-
-GLX Protocol
-
-    None.  This extension only extends to direct rendering contexts.
-
-New State
-
-    Get Value		Get Command	Type	    Initial Value
-    ---------		-----------	----	    -------------
-    [swap interval]	GetSwapInterval	Z+	    0
-
-New Implementation Dependent State
-
-    None
-
-
-Revision History
-
-    1.1,  5/1/03   Added the issues section and contact information.
-    	  	   Changed the default swap interval to 0.
-    1.0,  3/17/03  Initial version based on GLX_SGI_swap_control and
-                   WGL_EXT_swap_control.
+Name
+
+    MESA_swap_control
+
+Name Strings
+
+    GLX_MESA_swap_control
+
+Contact
+
+    Ian Romanick, IBM, idr at us.ibm.com
+
+Status
+
+    Deployed in DRI drivers post-XFree86 4.3.
+
+Version
+
+    Date: 5/1/2003   Revision: 1.1
+
+Number
+
+    ???
+
+Dependencies
+
+    None
+
+    Based on GLX_SGI_swap_control version 1.9 and WGL_EXT_swap_control
+    version 1.5.
+
+Overview
+
+    This extension allows an application to specify a minimum periodicity
+    of color buffer swaps, measured in video frame periods.
+
+Issues
+
+    * Should implementations that export GLX_MESA_swap_control also export
+      GL_EXT_swap_control for compatibility with WGL_EXT_swap_control?
+
+    UNRESOLVED.
+
+New Procedures and Functions
+
+    int glXSwapIntervalMESA(unsigned int interval)
+    int glXGetSwapIntervalMESA(void)
+
+New Tokens
+
+    None
+
+Additions to Chapter 2 of the 1.4 GL Specification (OpenGL Operation)
+
+    None
+
+Additions to Chapter 3 of the 1.4 GL Specification (Rasterization)
+
+    None
+
+Additions to Chapter 4 of the 1.4 GL Specification (Per-Fragment Operations
+and the Framebuffer)
+
+    None
+
+Additions to Chapter 5 of the 1.4 GL Specification (Special Functions)
+
+    None
+
+Additions to Chapter 6 of the 1.4 GL Specification (State and State Requests)
+
+    None
+
+Additions to the GLX 1.3 Specification
+
+    [Add the following to Section 3.3.10 of the GLX Specification (Double
+     Buffering)]
+
+    glXSwapIntervalMESA specifies the minimum number of video frame periods
+    per buffer swap.  (e.g. a value of two means that the color buffers
+    will be swapped at most every other video frame.)  A return value
+    of zero indicates success; otherwise an error occurred.  The interval
+    takes effect when glXSwapBuffers is first called subsequent to the
+    glXSwapIntervalMESA call.
+
+    A video frame period is the time required by the monitor to display a 
+    full frame of video data.  In the case of an interlaced monitor,
+    this is typically the time required to display both the even and odd 
+    fields of a frame of video data.
+
+    If <interval> is set to a value of 0, buffer swaps are not synchro-
+    nized to a video frame.  The <interval> value is silently clamped to
+    the maximum implementation-dependent value supported before being
+    stored.
+
+    The swap interval is not part of the render context state.  It cannot
+    be pushed or popped.  The current swap interval for the window
+    associated with the current context can be obtained by calling
+    glXGetSwapIntervalMESA.  The default swap interval is 0.
+
+    On XFree86, setting the environment variable LIBGL_THROTTLE_REFRESH sets
+    the swap interval to 1.
+
+Errors
+
+    glXSwapIntervalMESA returns GLX_BAD_CONTEXT if there is no current
+    GLXContext or if the current context is not a direct rendering context.
+
+GLX Protocol
+
+    None.  This extension only extends to direct rendering contexts.
+
+New State
+
+    Get Value		Get Command	Type	    Initial Value
+    ---------		-----------	----	    -------------
+    [swap interval]	GetSwapInterval	Z+	    0
+
+New Implementation Dependent State
+
+    None
+
+
+Revision History
+
+    1.1,  5/1/03   Added the issues section and contact information.
+    	  	   Changed the default swap interval to 0.
+    1.0,  3/17/03  Initial version based on GLX_SGI_swap_control and
+                   WGL_EXT_swap_control.
diff --git a/mesalib/docs/MESA_texture_array.spec b/mesalib/docs/MESA_texture_array.spec
index 1cc2217f3..b146821f7 100644
--- a/mesalib/docs/MESA_texture_array.spec
+++ b/mesalib/docs/MESA_texture_array.spec
@@ -1,804 +1,804 @@
-Name
-
-    MESA_texture_array
-
-Name Strings
-
-    GL_MESA_texture_array
-
-Contact
-
-    Ian Romanick, IBM (idr 'at' us.ibm.com)
-
-IP Status
-
-    No known IP issues.
-
-Status
-
-    Shipping in Mesa 7.1
-
-Version
-
-
-Number
-
-    TBD
-
-Dependencies
-
-    OpenGL 1.2 or GL_EXT_texture3D is required.
-
-    Support for ARB_fragment_program is assumed, but not required.
-
-    Support for ARB_fragment_program_shadow is assumed, but not required.
-
-    Support for EXT_framebuffer_object is assumed, but not required.
-
-    Written based on the wording of the OpenGL 2.0 specification and
-    ARB_fragment_program_shadow but not dependent on them.
-
-Overview
-
-    There are a number of circumstances where an application may wish to
-    blend two textures out of a larger set of textures.  Moreover, in some
-    cases the selected textures may vary on a per-fragment basis within
-    a polygon.  Several examples include:
-
-       1. High dynamic range textures.  The application stores several
-       different "exposures" of an image as different textures.  On a
-       per-fragment basis, the application selects which exposures are
-       used.
-
-       2. A terrain engine where the altitude of a point determines the
-       texture applied to it.  If the transition is from beach sand to
-       grass to rocks to snow, the application will store each texture
-       in a different texture map, and dynamically select which two
-       textures to blend at run-time.
-
-       3. Storing short video clips in textures.  Each depth slice is a
-       single frame of video.
-
-    Several solutions to this problem have been proposed, but they either
-    involve using a separate texture unit for each texture map or using 3D
-    textures without mipmaps.  Both of these options have major drawbacks.
-
-    This extension provides a third alternative that eliminates the major
-    drawbacks of both previous methods.  A new texture target,
-    TEXTURE_2D_ARRAY, is added that functions identically to TEXTURE_3D in
-    all aspects except the sizes of the non-base level images.  In
-    traditional 3D texturing, the size of the N+1 LOD is half the size
-    of the N LOD in all three dimensions.  For the TEXTURE_2D_ARRAY target,
-    the height and width of the N+1 LOD is halved, but the depth is the
-    same for all levels of detail. The texture then becomes an array of
-    2D textures.  The per-fragment texel is selected by the R texture
-    coordinate.
-
-    References:
-
-        http://www.opengl.org/discussion_boards/cgi_directory/ultimatebb.cgi?ubb=get_topic;f=3;t=011557
-        http://www.opengl.org/discussion_boards/cgi_directory/ultimatebb.cgi?ubb=get_topic;f=3;t=000516
-        http://www.opengl.org/discussion_boards/cgi_directory/ultimatebb.cgi?ubb=get_topic;f=3;t=011903
-        http://www.delphi3d.net/articles/viewarticle.php?article=terraintex.htm
-
-New Procedures and Functions
-
-    All functions come directly from EXT_texture_array.
-
-    void FramebufferTextureLayerEXT(enum target, enum attachment,
-                                    uint texture, int level, int layer);
-
-New Tokens
-
-    All token names and values come directly from EXT_texture_array.
-
-    Accepted by the <cap> parameter of Enable, Disable, and IsEnabled, by
-    the <pname> parameter of GetBooleanv, GetIntegerv, GetFloatv, and
-    GetDoublev, and by the <target> parameter of TexImage3D, GetTexImage,
-    GetTexLevelParameteriv, GetTexLevelParameterfv, GetTexParameteriv, and
-    GetTexParameterfv:
-
-        TEXTURE_1D_ARRAY_EXT                            0x8C18
-        TEXTURE_2D_ARRAY_EXT                            0x8C1A
-
-    Accepted by the <target> parameter of TexImage2D, TexSubImage2D,
-    CopyTexImage2D, CopyTexSubImage2D, CompressedTexImage2D,
-    CompressedTexSubImage2D, GetTexLevelParameteriv, and 
-    GetTexLevelParameterfv:
-
-        TEXTURE_1D_ARRAY_EXT
-        PROXY_TEXTURE_1D_ARRAY_EXT                      0x8C19
-
-    Accepted by the <target> parameter of TexImage3D, TexSubImage3D,
-    CopyTexSubImage3D, CompressedTexImage3D, CompressedTexSubImage3D,
-    GetTexLevelParameteriv, and GetTexLevelParameterfv:
-
-        TEXTURE_2D_ARRAY_EXT
-        PROXY_TEXTURE_2D_ARRAY_EXT                      0x8C1B
-
-    Accepted by the <pname> parameter of GetBooleanv, GetIntegerv,
-    GetFloatv, and GetDoublev
-
-        TEXTURE_BINDING_1D_ARRAY_EXT                    0x8C1C
-        TEXTURE_BINDING_2D_ARRAY_EXT                    0x8C1D
-        MAX_ARRAY_TEXTURE_LAYERS_EXT                    0x88FF
-
-    Accepted by the <param> parameter of TexParameterf, TexParameteri,
-    TexParameterfv, and TexParameteriv when the <pname> parameter is
-    TEXTURE_COMPARE_MODE_ARB:
-
-        COMPARE_REF_DEPTH_TO_TEXTURE_EXT                0x884E
-
-    (Note:  COMPARE_REF_DEPTH_TO_TEXTURE_EXT is simply an alias for the
-    existing COMPARE_R_TO_TEXTURE token in OpenGL 2.0; the alternate name
-    reflects the fact that the R coordinate is not always used.)
-
-    Accepted by the <internalformat> parameter of TexImage3D and
-    CompressedTexImage3D, and by the <format> parameter of
-    CompressedTexSubImage3D:
-
-        COMPRESSED_RGB_S3TC_DXT1_EXT
-        COMPRESSED_RGBA_S3TC_DXT1_EXT
-        COMPRESSED_RGBA_S3TC_DXT3_EXT
-        COMPRESSED_RGBA_S3TC_DXT5_EXT
-
-    Accepted by the <pname> parameter of
-    GetFramebufferAttachmentParameterivEXT:
-
-        FRAMEBUFFER_ATTACHMENT_TEXTURE_LAYER_EXT          0x8CD4
-
-    (Note:  FRAMEBUFFER_ATTACHMENT_TEXTURE_LAYER is simply an alias for the
-    FRAMEBUFFER_ATTACHMENT_TEXTURE_3D_ZOFFSET_EXT token provided in
-    EXT_framebuffer_object.  This extension generalizes the notion of
-    "<zoffset>" to include layers of an array texture.)
-
-Additions to Chapter 2 of the OpenGL 2.0 Specification (OpenGL Operation)
-
-    None
-
-Additions to Chapter 3 of the OpenGL 2.0 Specification (Rasterization)
-
-    -- Section 3.8.1 "Texture Image Specification"
-
-       Change the first paragraph (page 150) to say (spec changes identical to
-       EXT_texture_array):
-
-       "The command
-
-         void TexImage3D(enum target, int level, int internalformat,
-                         sizei width, sizei height, sizei depth, int border,
-                         enum format, enum type, void *data);
-
-       is used to specify a three-dimensional texture image. target must be one
-       one of TEXTURE_3D for a three-dimensional texture or
-       TEXTURE_2D_ARRAY_EXT for an two-dimensional array texture.
-       Additionally, target may be either PROXY_TEXTURE_3D for a
-       three-dimensional proxy texture, or PROXY_TEXTURE_2D_ARRAY_EXT for a
-       two-dimensional proxy array texture."
-
-       Change the fourth paragraph on page 151 to say (spec changes identical
-       to EXT_texture_array):
-
-       "Textures with a base internal format of DEPTH_COMPONENT are supported
-       by texture image specification commands only if target is TEXTURE_1D,
-       TEXTURE_2D, TEXTURE_1D_ARRAY_EXT, TEXTURE_2D_ARRAY_EXT,
-       PROXY_TEXTURE_1D, PROXY_TEXTURE_2D, PROXY_TEXTURE_1D_ARRAY_EXT, or
-       PROXY_TEXTURE_2D_ARRAY_EXT. Using this format in conjunction with any
-       other target will result in an INVALID_OPERATION error."
-
-
-       Change the fourth paragraph on page 156 to say (spec changes identical
-       to EXT_texture_array):
-       
-       "The command
-
-         void TexImage2D(enum target, int level,
-                         int internalformat, sizei width, sizei height,
-                         int border, enum format, enum type, void *data);
-
-       is used to specify a two-dimensional texture image. target must be one
-       of TEXTURE_2D for a two-dimensional texture, TEXTURE_1D_ARRAY_EXT for a
-       one-dimensional array texture, or one of TEXTURE_CUBE_MAP_POSITIVE_X,
-       TEXTURE_CUBE_MAP_NEGATIVE_X, TEXTURE_CUBE_MAP_POSITIVE_Y,
-       TEXTURE_CUBE_MAP_NEGATIVE_Y, TEXTURE_CUBE_MAP_POSITIVE_Z, or
-       TEXTURE_CUBE_MAP_NEGATIVE_Z for a cube map texture. Additionally,
-       target may be either PROXY_TEXTURE_2D for a two-dimensional proxy
-       texture, PROXY_TEXTURE_1D_ARRAY_EXT for a one-dimensional proxy array
-       texture, or PROXY TEXTURE_CUBE_MAP for a cube map proxy texture in the
-       special case discussed in section 3.8.11.  The other parameters match
-       the corresponding parameters of TexImage3D.
-
-       For the purposes of decoding the texture image, TexImage2D is
-       equivalent to calling TexImage3D with corresponding arguments and depth
-       of 1, except that
-
-         * The border depth, d_b, is zero, and the depth of the image is
-           always 1 regardless of the value of border. 
-
-         * The border height, h_b, is zero if <target> is
-           TEXTURE_1D_ARRAY_EXT, and <border> otherwise.
-
-         * Convolution will be performed on the image (possibly changing its
-           width and height) if SEPARABLE 2D or CONVOLUTION 2D is enabled.
-
-         * UNPACK SKIP IMAGES is ignored."
-
-    -- Section 3.8.2 "Alternate Texture Image Specification Commands"
-
-       Change the second paragraph (page 159) (spec changes identical
-       to EXT_texture_array):
-
-       "The command
-
-         void CopyTexImage2D(enum target, int level,
-                             enum internalformat, int x, int y, sizei width,
-                             sizei height, int border);
-
-       defines a two-dimensional texture image in exactly the manner of
-       TexImage2D, except that the image data are taken from the framebuffer
-       rather than from client memory. Currently, target must be one of
-       TEXTURE_2D, TEXTURE_1D_ARRAY_EXT, TEXTURE_CUBE_MAP_POSITIVE_X,
-       TEXTURE_CUBE_MAP_NEGATIVE_X, TEXTURE_CUBE MAP_POSITIVE_Y,
-       TEXTURE_CUBE_MAP_NEGATIVE_Y, TEXTURE_CUBE_MAP_POSITIVE_Z, or
-       TEXTURE_CUBE_MAP_NEGATIVE_Z.
-
-
-       Change the last paragraph on page 160 to say (spec changes identical
-       to EXT_texture_array):
-
-       "Currently the target arguments of TexSubImage1D and CopyTexSubImage1D
-       must be TEXTURE_1D, the target arguments of TexSubImage2D and
-       CopyTexSubImage2D must be one of TEXTURE_2D, TEXTURE_1D_ARRAY_EXT,
-       TEXTURE_CUBE_MAP_POSITIVE_X, TEXTURE_CUBE_MAP_NEGATIVE_X,
-       TEXTURE_CUBE_MAP_POSITIVE_Y, TEXTURE_CUBE_MAP_NEGATIVE_Y,
-       TEXTURE_CUBE_MAP_POSITIVE_Z, or TEXTURE_CUBE_MAP_NEGATIVE_Z, and the
-       target arguments of TexSubImage3D and CopyTexSubImage3D must be
-       TEXTURE_3D or TEXTURE_2D_ARRAY_EXT. ..."
-
-
-    -- Section 3.8.4 "Texture Parameters"
-
-       Change the first paragraph (page 166) to say:
-
-       "Various parameters control how the texel array is treated when
-       specified or changed, and when applied to a fragment. Each parameter is
-       set by calling
-
-         void TexParameter{if}(enum target, enum pname, T param); 
-         void TexParameter{if}v(enum target, enum pname, T params);
-
-       target is the target, either TEXTURE_1D, TEXTURE_2D, TEXTURE_3D,
-       TEXTURE_CUBE_MAP, TEXTURE_1D_ARRAY_EXT, or TEXTURE_2D_ARRAY_EXT."
-
-
-    -- Section 3.8.8 "Texture Minification" in the section "Scale Factor and Level of Detail"
-
-       Change the first paragraph (page 172) to say:
-
-       "Let s(x,y) be the function that associates an s texture coordinate
-       with each set of window coordinates (x,y) that lie within a primitive;
-       define t(x,y) and r(x,y) analogously.  Let u(x,y) = w_t * s(x,y),
-       v(x,y) = h_t * t(x,y), and w(x,y) = d_t * r(x,y), where w_t, h_t,
-       and d_t are as defined by equations 3.15, 3.16, and 3.17 with
-       w_s, h_s, and d_s equal to the width, height, and depth of the
-       image array whose level is level_base.  For a one-dimensional
-       texture or a one-dimensional array texture, define v(x,y) = 0 and
-       w(x,y) = 0; for a two-dimensional texture or a two-dimensional array
-       texture, define w(x,y) = 0..."
-
-    -- Section 3.8.8 "Texture Minification" in the section "Mipmapping"
-
-       Change the third paragraph (page 174) to say:
-       
-       "For a two-dimensional texture, two-dimensional array texture, or
-       cube map texture,"
-
-       Change the fourth paragraph (page 174) to say:
-
-       "And for a one-dimensional texture or a one-dimensional array texture,"
-
-       After the first paragraph (page 175) add:
-
-       "For one-dimensional array textures, h_b and d_b are treated as 1,
-       regardless of the actual values, when performing mipmap calculations.
-       For two-dimensional array textures, d_b is always treated as one,
-       regardless of the actual value, when performing mipmap calculations."
-
-    -- Section 3.8.8 "Automatic Mipmap Generation" in the section "Mipmapping"
-
-       Change the third paragraph (page 176) to say (spec changes identical
-       to EXT_texture_array):
-
-       "The contents of the derived arrays are computed by repeated, filtered
-       reduction of the level_base array.  For one- and two-dimensional array
-       textures, each layer is filtered independently.  ..."
-
-    -- Section 3.8.8 "Manual Mipmap Generation" in the section "Mipmapping"
-
-       Change first paragraph to say (spec changes identical to
-       EXT_texture_array):
-
-       "Mipmaps can be generated manually with the command
-
-         void GenerateMipmapEXT(enum target);
-
-       where <target> is one of TEXTURE_1D, TEXTURE_2D, TEXTURE_CUBE_MAP,
-       TEXTURE_3D, TEXTURE_1D_ARRAY, or TEXTURE_2D_ARRAY.  Mipmap generation
-       affects the texture image attached to <target>.  ..."
-
-    -- Section 3.8.10 "Texture Completeness"
-
-       Change the second paragraph (page 177) to say (spec changes identical
-       to EXT_texture_array):
-
-       "For one-, two-, or three-dimensional textures and one- or
-       two-dimensional array textures, a texture is complete if the following
-       conditions all hold true:"
-
-    -- Section 3.8.11 "Texture State and Proxy State"
-
-       Change the second and third paragraphs (page 179) to say (spec changes
-       identical to EXT_texture_array):
-
-       "In addition to image arrays for one-, two-, and three-dimensional
-       textures, one- and two-dimensional array textures, and the six image
-       arrays for the cube map texture, partially instantiated image arrays
-       are maintained for one-, two-, and three-dimensional textures and one-
-       and two-dimensional array textures.  Additionally, a single proxy image
-       array is maintained for the cube map texture.  Each proxy image array
-       includes width, height, depth, border width, and internal format state
-       values, as well as state for the red, green, blue, alpha, luminance,
-       and intensity component resolutions. Proxy image arrays do not include
-       image data, nor do they include texture properties. When TexImage3D is
-       executed with target specified as PROXY_TEXTURE_3D, the
-       three-dimensional proxy state values of the specified level-of-detail
-       are recomputed and updated. If the image array would not be supported
-       by TexImage3D called with target set to TEXTURE 3D, no error is
-       generated, but the proxy width, height, depth, border width, and
-       component resolutions are set to zero. If the image array would be
-       supported by such a call to TexImage3D, the proxy state values are set
-       exactly as though the actual image array were being specified. No pixel
-       data are transferred or processed in either case.
-
-       Proxy arrays for one- and two-dimensional textures and one- and
-       two-dimensional array textures are operated on in the same way when
-       TexImage1D is executed with target specified as PROXY_TEXTURE_1D,
-       TexImage2D is executed with target specified as PROXY_TEXTURE_2D or
-       PROXY_TEXTURE_1D_ARRAY_EXT, or TexImage3D is executed with target
-       specified as PROXY_TETXURE_2D_ARRAY_EXT."
-
-    -- Section 3.8.12 "Texture Objects"
-
-       Change section (page 180) to say (spec changes identical to 
-       EXT_texture_array):
-
-       "In addition to the default textures TEXTURE_1D, TEXTURE_2D,
-       TEXTURE_3D, TEXTURE_CUBE_MAP, TEXTURE_1D_ARRAY_EXT, and TEXTURE_2D_EXT,
-       named one-, two-, and three-dimensional, cube map, and one- and
-       two-dimensional array texture objects can be created and operated upon.
-       The name space for texture objects is the unsigned integers, with zero
-       reserved by the GL.
-
-       A texture object is created by binding an unused name to TEXTURE_1D,
-       TEXTURE_2D, TEXTURE_3D, TEXTURE_CUBE_MAP, TEXTURE_1D_ARRAY_EXT, or
-       TEXTURE_2D_ARRAY_EXT. The binding is effected by calling
-
-         void BindTexture(enum target, uint texture);
-
-       with <target> set to the desired texture target and <texture> set to
-       the unused name.  The resulting texture object is a new state vector,
-       comprising all the state values listed in section 3.8.11, set to the
-       same initial values. If the new texture object is bound to TEXTURE_1D,
-       TEXTURE_2D, TEXTURE_3D, TEXTURE_CUBE_MAP, TEXTURE_1D_ARRAY_EXT, or
-       TEXTURE_2D_ARRAY_EXT, it is and remains a one-, two-,
-       three-dimensional, cube map, one- or two-dimensional array texture
-       respectively until it is deleted.
-
-       BindTexture may also be used to bind an existing texture object to
-       either TEXTURE_1D, TEXTURE_2D, TEXTURE_3D, TEXTURE_CUBE_MAP,
-       TEXTURE_1D_ARRAY_EXT, or TEXTURE_2D_ARRAY_EXT. The error
-       INVALID_OPERATION is generated if an attempt is made to bind a texture
-       object of different dimensionality than the specified target. If the
-       bind is successful no change is made to the state of the bound texture
-       object, and any previous binding to target is broken.
-
-       While a texture object is bound, GL operations on the target to which
-       it is bound affect the bound object, and queries of the target to which
-       it is bound return state from the bound object. If texture mapping of
-       the dimensionality of the target to which a texture object is bound is
-       enabled, the state of the bound texture object directs the texturing
-       operation.
-
-       In the initial state, TEXTURE_1D, TEXTURE_2D, TEXTURE_3D,
-       TEXTURE_CUBE_MAP, TEXTURE_1D_ARRAY_EXT, and TEXTURE_2D_ARRAY_EXT have
-       one-, two-, three-dimensional, cube map, and one- and two-dimensional
-       array texture state vectors respectively associated with them. In order
-       that access to these initial textures not be lost, they are treated as
-       texture objects all of whose names are 0. The initial one-, two-,
-       three-dimensional, cube map, one- and two-dimensional array textures
-       are therefore operated upon, queried, and applied as TEXTURE_1D,
-       TEXTURE_2D, TEXTURE_3D, TEXTURE_CUBE_MAP, TEXTURE_1D_ARRAY_EXT, and
-       TEXTURE_2D_ARRAY_EXT respectively while 0 is bound to the corresponding
-       targets.
-
-       Change second paragraph on page 181 to say (spec changes identical to 
-       EXT_texture_array):
-       
-       "...  If a texture that is currently bound to one of the targets
-       TEXTURE_1D, TEXTURE_2D, TEXTURE_3D, TEXTURE_CUBE_MAP,
-       TEXTURE_1D_ARRAY_EXT, or TEXTURE_2D_ARRAY_EXT is deleted, it is as
-       though BindTexture had been executed with the same target and texture
-       zero. ..."
-
-       Change second paragraph on page 182 to say (spec changes identical to 
-       EXT_texture_array):
-       
-       "The texture object name space, including the initial one-, two-, and
-       three dimensional, cube map, and one- and two-dimensional array texture
-       objects, is shared among all texture units. ..."
-
-
-    -- Section 3.8.14 "Depth Texture Comparison Modes" in "Texture Comparison Modes"
-
-       Change second through fourth paragraphs (page 188) to say:
-
-       "Let D_t be the depth texture value, in the range [0, 1].  For
-       texture lookups from one- and two-dimensional, rectangle, and
-       one-dimensional array targets, let R be the interpolated <r>
-       texture coordinate, clamped to the range [0, 1].  For texture lookups
-       from two-dimensional array texture targets, let R be the interpolated
-       <q> texture coordinate, clamped to the range [0, 1].  Then the
-       effective texture value L_t, I_t, or A_t is computed as follows:
-
-       If the value of TEXTURE_COMPARE_MODE is NONE, then
-
-          r = Dt
-
-       If the value of TEXTURE_COMPARE_MODE is
-       COMPARE_REF_DEPTH_TO_TEXTURE_EXT), then r depends on the texture
-       comparison function as shown in table 3.27."
-
-    -- Section 3.8.15 "Texture Application"
-
-       Change the first paragraph (page 189) to say:
-
-       "Texturing is enabled or disabled using the generic Enable and Disable
-       commands, respectively, with the symbolic constants TEXTURE_1D,
-       TEXTURE_2D, TEXTURE_3D, TEXTURE_CUBE_MAP, TEXTURE_1D_ARRAY_EXT, or
-       TEXTURE_2D_ARRAY_EXT to enable one-, two-, three-dimensional, cube
-       map, one-dimensional array, or two-dimensional array texture,
-       respectively.  If both two- and one-dimensional textures are enabled,
-       the two-dimensional texture is used.  If the three-dimensional and
-       either of the two- or one-dimensional textures is enabled, the
-       three-dimensional texture is used.  If the cube map texture and any of
-       the three-, two-, or one-dimensional textures is enabled, then cube map
-       texturing is used.  If one-dimensional array texture is enabled and any
-       of cube map, three-, two-, or one-dimensional textures is enabled, 
-       one-dimensional array texturing is used.  If two-dimensional array
-       texture is enabled and any of cube map, three-, two-, one-dimensional
-       textures or one-dimensional array texture is enabled, two-dimensional
-       array texturing is used..."
-
-    -- Section 3.11.2 of ARB_fragment_program (Fragment Program Grammar and Restrictions):
-
-       (mostly add to existing grammar rules)
-
-       <optionName>           ::= "MESA_texture_array"
-
-       <texTarget>            ::= "1D"
-                               | "2D"
-                               | "3D"
-                               | "CUBE"
-                               | "RECT"
-                               | <arrayTarget> (if program option is present)
-                               | <shadowTarget> (if program option is present)
-
-       <arrayTarget>          ::= "ARRAY1D"
-                               | "ARRAY2D"
-
-       <shadowTarget>         ::= "SHADOW1D"
-                               | "SHADOW2D"
-                               | "SHADOWRECT"
-                               | <shadowArrayTarget> (if program option is present)
-
-       <shadowArrayTarget>    ::= "SHADOWARRAY1D"
-                               | "SHADOWARRAY2D"
-
-
-    -- Add Section 3.11.4.5.4 "Texture Stack Option"
-
-       "If a fragment program specifies the "MESA_texture_array" program
-       option, the <texTarget> rule is modified to add the texture targets
-       ARRAY1D and ARRAY2D (See Section 3.11.2)."
-
-    -- Section 3.11.6 "Fragment Program Texture Instruction Set"
-
-       (replace 1st and 2nd paragraphs with the following paragraphs)
-
-       "The first three texture instructions described below specify the
-       mapping of 4-tuple input vectors to 4-tuple output vectors.
-       The sampling of the texture works as described in section 3.8,
-       except that texture environments and texture functions are not
-       applicable, and the texture enables hierarchy is replaced by explicit
-       references to the desired texture target (i.e., 1D, 2D, 3D, cube map,
-       rectangle, ARRAY1D, ARRAY2D).  These texture instructions specify
-       how the 4-tuple is mapped into the coordinates used for sampling.  The
-       following function is used to describe the texture sampling in the
-       descriptions below: 
-
-         vec4 TextureSample(vec4 coord, float lodBias, int texImageUnit,
-                            enum texTarget);
-
-       Note that not all four components of the texture coordinates <coord>
-       are used by all texture targets.  Component usage for each <texTarget>
-       is defined in table X.
-
-                                                        coordinates used
-         texTarget          Texture Type               s t r  layer  shadow
-         ----------------   ---------------------      -----  -----  ------
-         1D                 TEXTURE_1D                 x - -    -      -
-         2D                 TEXTURE_2D                 x y -    -      -
-         3D                 TEXTURE_3D                 x y z    -      -
-         CUBE               TEXTURE_CUBE_MAP           x y z    -      -
-         RECT               TEXTURE_RECTANGLE_ARB      x y -    -      -
-         ARRAY1D            TEXTURE_1D_ARRAY_EXT       x - -    y      -
-         ARRAY2D            TEXTURE_2D_ARRAY_EXT       x y -    z      -
-         SHADOW1D           TEXTURE_1D                 x - -    -      z
-         SHADOW2D           TEXTURE_2D                 x y -    -      z
-         SHADOWRECT         TEXTURE_RECTANGLE_ARB      x y -    -      z
-         SHADOWARRAY1D      TEXTURE_1D_ARRAY_EXT       x - -    y      z
-         SHADOWARRAY2D      TEXTURE_2D_ARRAY_EXT       x y -    z      w
-
-         Table X:  Texture types accessed for each of the <texTarget>, and
-         coordinate mappings.  The "coordinates used" column indicate the
-         input values used for each coordinate of the texture lookup, the
-         layer selector for array textures, and the reference value for
-         texture comparisons."
-
-    -- Section 3.11.6.2 "TXP: Project coordinate and map to color"
-    
-       Add to the end of the section:
-       
-       "A program will fail to load if the TXP instruction is used in
-       conjunction with the SHADOWARRAY2D target."
-
-Additions to Chapter 4 of the OpenGL 2.0 Specification (Per-Fragment Operations)
-
-    -- Section 4.4.2.3 "Attaching Texture Images to a Framebuffer"
-
-       Add to the end of the section (spec changes identical to
-       EXT_texture_array):
-
-       "The command
-
-         void FramebufferTextureLayerEXT(enum target, enum attachment,
-                                         uint texture, int level, int layer);
-
-       operates identically to FramebufferTexture3DEXT, except that it
-       attaches a single layer of a three-dimensional texture or a one- or
-       two-dimensional array texture.  <layer> is an integer indicating the
-       layer number, and is treated identically to the <zoffset> parameter in
-       FramebufferTexture3DEXT.  The error INVALID_VALUE is generated if
-       <layer> is negative.  The error INVALID_OPERATION is generated if
-       <texture> is non-zero and is not the name of a three dimensional
-       texture or one- or two-dimensional array texture.  Unlike
-       FramebufferTexture3D, no <textarget> parameter is accepted.
-
-       If <texture> is non-zero and the command does not result in an error,
-       the framebuffer attachment state corresponding to <attachment> is
-       updated as in the other FramebufferTexture commands, except that
-       FRAMEBUFFER_ATTACHMENT_TEXTURE_LAYER_EXT is set to <layer>."
-
-    -- Section 4.4.4.1 "Framebuffer Attachment Completeness"
-
-      Add to the end of the list of completeness rules (spec changes
-      identical to EXT_texture_array):
-
-        "* If FRAMEBUFFER_ATTACHMENT_OBJECT_TYPE_EXT is TEXTURE and
-           FRAMEBUFFER_ATTACHMENT_OBJECT_NAME_EXT names a one- or 
-           two-dimensional array texture, then
-           FRAMEBUFFER_ATTACHMENT_TEXTURE_LAYER_EXT must be smaller than the
-           number of layers in the texture."
-
-Additions to Chapter 5 of the OpenGL 2.0 Specification (Special Functions)
-
-    -- Section 5.4 "Display Lists"
-
-       Change the first paragraph on page 242 to say (spec changes
-       identical to EXT_texture_array):
-
-       "TexImage3D, TexImage2D, TexImage1D, Histogram, and ColorTable are
-       executed immediately when called with the corresponding proxy arguments
-       PROXY_TEXTURE_3D or PROXY_TEXTURE_2D_ARRAY_EXT; PROXY_TEXTURE_2D,
-       PROXY_TEXTURE_CUBE_MAP, or PROXY_TEXTURE_1D_ARRAY_EXT;
-       PROXY_TEXTURE_1D; PROXY_HISTOGRAM; and PROXY_COLOR_TABLE,
-       PROXY_POST_CONVOLUTION_COLOR_TABLE, or
-       PROXY_POST_COLOR_MATRIX_COLOR_TABLE."
-
-Additions to Chapter 6 of the OpenGL 2.0 Specification (State and State Requests)
-
-    -- Section 6.1.3 "Enumerated Queries"
-
-       Add after the line beginning "If the value of
-       FRAMEBUFFER_ATTACHMENT_OBJECT_TYPE_EXT is TEXTURE" (spec changes
-       identical to EXT_texture_array):
-
-       "If <pname> is FRAMEBUFFER_ATTACHMENT_TEXTURE_LAYER_EXT and the
-       texture object named FRAMEBUFFER_ATTACHMENT_OBJECT_NAME_EXT is a
-       three-dimensional texture or a one- or two-dimensional array texture,
-       then <params> will contain the number of texture layer attached to the
-       attachment point.  Otherwise, <params> will contain the value zero."
-
-    -- Section 6.1.4 "Texture Queries"
-    
-       Change the first three paragraphs (page 248) to say (spec changes
-       identical to EXT_texture_array):
-
-       "The command
-
-         void GetTexImage(enum tex, int lod, enum format,
-                          enum type, void *img);
-
-       is used to obtain texture images. It is somewhat different from the
-       other get commands; tex is a symbolic value indicating which texture
-       (or texture face in the case of a cube map texture target name) is to
-       be obtained.  TEXTURE_1D, TEXTURE_2D, TEXTURE_3D, TEXTURE_1D_ARRAY_EXT,
-       and TEXTURE_2D_ARRAY_EXT indicate a one-, two-, or three-dimensional
-       texture, or one- or two-dimensional array texture, respectively.
-       TEXTURE_CUBE_MAP_POSITIVE_X, ...
-
-       GetTexImage obtains... from the first image to the last for
-       three-dimensional textures.  One- and two-dimensional array textures
-       are treated as two- and three-dimensional images, respectively, where
-       the layers are treated as rows or images.  These groups are then...
-
-       For three-dimensional and two-dimensional array textures, pixel storage
-       operations are applied as if the image were two-dimensional, except
-       that the additional pixel storage state values PACK_IMAGE_HEIGHT and
-       PACK_SKIP_IMAGES are applied. ..."
-
-Additions to Appendix A of the OpenGL 2.0 Specification (Invariance)
-
-    None
-
-Additions to the AGL/GLX/WGL Specifications
-
-    None
-
-GLX Protocol
-
-    None
-
-Dependencies on ARB_fragment_program
-
-    If ARB_fragment_program is not supported, the changes to section 3.11
-    should be ignored.
-
-Dependencies on EXT_framebuffer_object
-
-    If EXT_framebuffer_object is not supported, the changes to section
-    3.8.8 ("Manual Mipmap Generation"), 4.4.2.3, and 6.1.3 should be ignored.
-
-Dependencies on EXT_texture_compression_s3tc and NV_texture_compression_vtc
-
-    (Identical dependency as EXT_texture_array.)
-
-    S3TC texture compression is supported for two-dimensional array textures.
-    When <target> is TEXTURE_2D_ARRAY_EXT, each layer is stored independently
-    as a compressed two-dimensional textures.  When specifying or querying
-    compressed images using one of the S3TC formats, the images are provided
-    and/or returned as a series of two-dimensional textures stored
-    consecutively in memory, with the layer closest to zero specified first.
-    For array textures, images are not arranged in 4x4x4 or 4x4x2 blocks as in
-    the three-dimensional compression format provided in the
-    EXT_texture_compression_vtc extension.  Pixel store parameters, including
-    those specific to three-dimensional images, are ignored when compressed
-    image data are provided or returned, as in the
-    EXT_texture_compression_s3tc extension.
-
-    S3TC compression is not supported for one-dimensional texture targets in
-    EXT_texture_compression_s3tc, and is not supported for one-dimensional
-    array textures in this extension.  If compressed one-dimensional arrays
-    are needed, use a two-dimensional texture with a height of one.
-
-    This extension allows the use of the four S3TC internal format types in
-    TexImage3D, CompressedTexImage3D, and CompressedTexSubImage3D calls.
-
-Errors
-
-    None
-
-New State
-
-    (add to table 6.15, p. 276)
-
-                                                     Initial
-    Get Value                     Type   Get Command  Value Description           Sec.    Attribute
-    ----------------------------  -----  -----------  ----- --------------------  ------  ---------
-    TEXTURE_BINDING_1D_ARRAY_EXT  2*xZ+  GetIntegerv    0   texture object bound  3.8.12  texture
-                                                            to TEXTURE_1D_ARRAY
-    TEXTURE_BINDING_2D_ARRAY_EXT  2*xZ+  GetIntegerv    0   texture object bound  3.8.12  texture
-                                                            to TEXTURE_2D_ARRAY
-
-
-New Implementation Dependent State
-
-    (add to Table 6.32, p. 293)
-
-                                                    Minimum
-    Get Value                     Type  Get Command  Value  Description         Sec.  Attribute
-    ----------------------------  ----  ----------- ------- ------------------  ----- ---------
-    MAX_TEXTURE_ARRAY_LAYERS_EXT   Z+   GetIntegerv   64    maximum number of   3.8.1     -
-                                                            layers for texture
-                                                            arrays
-
-Issues
-
-    (1) Is "texture stack" a good name for this functionality?
-
-        NO.  The name is changed to "array texture" to match the
-        nomenclature used by GL_EXT_texture_array.
-
-    (2) Should the R texture coordinate be treated as normalized or
-    un-normalized?  If it were un-normalized, floor(R) could be thought
-    of as a direct index into the array texture.  This may be more
-    convenient for applications.
-
-        RESOLVED.  All texture coordinates are normalized.  The issue of
-        un-normalized texture coordinates has been discussed in the ARB
-        before and should be left for a layered extension.
-
-        RE-RESOLVED.  The R coordinate is un-normalized.  Accessing an array
-        using [0, layers-1] coordinates is much more natural.
-
-    (3) How does LOD selection work for stacked textures?
-
-        RESOLVED.  For 2D array textures the R coordinate is ignored, and
-        the LOD selection equations for 2D textures are used.  For 1D
-        array textures the T coordinate is ignored, and the LOD selection
-        equations for 1D textures are used.  The expected usage is in a
-        fragment program with an explicit LOD selection.
-
-    (4) What is the maximum size of a 2D array texture?  Is it the same
-    as for a 3D texture, or should a new query be added?  How about for 1D
-    array textures?
-
-        RESOLVED.  A new query is added.
-
-    (5) How are array textures exposed in GLSL?
-    
-        RESOLVED.  Use GL_EXT_texture_array.
-        
-    (6) Should a 1D array texture also be exposed?
-
-        RESOLVED.  For orthogonality, yes.
-
-    (7) How are stacked textures attached to framebuffer objects?
-
-        RESOLVED.  Layers of both one- and two-dimensional array textures
-        are attached using FreambufferTextureLayerEXT.  Once attached, the
-        array texture layer behaves exactly as either a one- or
-        two-dimensional texture.
-
-    (8) How is this extension related to GL_EXT_texture_array?
-    
-        This extension adapats GL_MESAX_texture_stack to the notation,
-        indexing, and FBO access of GL_EXT_texture_array.  This extension
-        replaces the GLSL support of GL_EXT_texture_array with
-        GL_ARB_fragment_program support.
-
-        Assembly program support is also provided by GL_NV_gpu_program4.
-        GL_NV_gpu_program4 also adds support for other features that are
-        specific to Nvidia hardware, while this extension adds only support
-        for array textures.
-
-        Much of text of this extension that has changed since
-        GL_MESAX_texture_stack comes directly from either
-        GL_EXT_texture_array or GL_NV_gpu_program4.
-
-Revision History
-
-    ||2005/11/15||0.1||idr||Initial draft MESAX version.||
-    ||2005/12/07||0.2||idr||Added framebuffer object interactions.||
-    ||2005/12/12||0.3||idr||Updated fragment program interactions.||
-    ||2007/05/16||0.4||idr||Converted to MESA_texture_array.  Brought in line with EXT_texture_array and NV_gpu_program4.||
+Name
+
+    MESA_texture_array
+
+Name Strings
+
+    GL_MESA_texture_array
+
+Contact
+
+    Ian Romanick, IBM (idr 'at' us.ibm.com)
+
+IP Status
+
+    No known IP issues.
+
+Status
+
+    Shipping in Mesa 7.1
+
+Version
+
+
+Number
+
+    TBD
+
+Dependencies
+
+    OpenGL 1.2 or GL_EXT_texture3D is required.
+
+    Support for ARB_fragment_program is assumed, but not required.
+
+    Support for ARB_fragment_program_shadow is assumed, but not required.
+
+    Support for EXT_framebuffer_object is assumed, but not required.
+
+    Written based on the wording of the OpenGL 2.0 specification and
+    ARB_fragment_program_shadow but not dependent on them.
+
+Overview
+
+    There are a number of circumstances where an application may wish to
+    blend two textures out of a larger set of textures.  Moreover, in some
+    cases the selected textures may vary on a per-fragment basis within
+    a polygon.  Several examples include:
+
+       1. High dynamic range textures.  The application stores several
+       different "exposures" of an image as different textures.  On a
+       per-fragment basis, the application selects which exposures are
+       used.
+
+       2. A terrain engine where the altitude of a point determines the
+       texture applied to it.  If the transition is from beach sand to
+       grass to rocks to snow, the application will store each texture
+       in a different texture map, and dynamically select which two
+       textures to blend at run-time.
+
+       3. Storing short video clips in textures.  Each depth slice is a
+       single frame of video.
+
+    Several solutions to this problem have been proposed, but they either
+    involve using a separate texture unit for each texture map or using 3D
+    textures without mipmaps.  Both of these options have major drawbacks.
+
+    This extension provides a third alternative that eliminates the major
+    drawbacks of both previous methods.  A new texture target,
+    TEXTURE_2D_ARRAY, is added that functions identically to TEXTURE_3D in
+    all aspects except the sizes of the non-base level images.  In
+    traditional 3D texturing, the size of the N+1 LOD is half the size
+    of the N LOD in all three dimensions.  For the TEXTURE_2D_ARRAY target,
+    the height and width of the N+1 LOD is halved, but the depth is the
+    same for all levels of detail. The texture then becomes an array of
+    2D textures.  The per-fragment texel is selected by the R texture
+    coordinate.
+
+    References:
+
+        http://www.opengl.org/discussion_boards/cgi_directory/ultimatebb.cgi?ubb=get_topic;f=3;t=011557
+        http://www.opengl.org/discussion_boards/cgi_directory/ultimatebb.cgi?ubb=get_topic;f=3;t=000516
+        http://www.opengl.org/discussion_boards/cgi_directory/ultimatebb.cgi?ubb=get_topic;f=3;t=011903
+        http://www.delphi3d.net/articles/viewarticle.php?article=terraintex.htm
+
+New Procedures and Functions
+
+    All functions come directly from EXT_texture_array.
+
+    void FramebufferTextureLayerEXT(enum target, enum attachment,
+                                    uint texture, int level, int layer);
+
+New Tokens
+
+    All token names and values come directly from EXT_texture_array.
+
+    Accepted by the <cap> parameter of Enable, Disable, and IsEnabled, by
+    the <pname> parameter of GetBooleanv, GetIntegerv, GetFloatv, and
+    GetDoublev, and by the <target> parameter of TexImage3D, GetTexImage,
+    GetTexLevelParameteriv, GetTexLevelParameterfv, GetTexParameteriv, and
+    GetTexParameterfv:
+
+        TEXTURE_1D_ARRAY_EXT                            0x8C18
+        TEXTURE_2D_ARRAY_EXT                            0x8C1A
+
+    Accepted by the <target> parameter of TexImage2D, TexSubImage2D,
+    CopyTexImage2D, CopyTexSubImage2D, CompressedTexImage2D,
+    CompressedTexSubImage2D, GetTexLevelParameteriv, and 
+    GetTexLevelParameterfv:
+
+        TEXTURE_1D_ARRAY_EXT
+        PROXY_TEXTURE_1D_ARRAY_EXT                      0x8C19
+
+    Accepted by the <target> parameter of TexImage3D, TexSubImage3D,
+    CopyTexSubImage3D, CompressedTexImage3D, CompressedTexSubImage3D,
+    GetTexLevelParameteriv, and GetTexLevelParameterfv:
+
+        TEXTURE_2D_ARRAY_EXT
+        PROXY_TEXTURE_2D_ARRAY_EXT                      0x8C1B
+
+    Accepted by the <pname> parameter of GetBooleanv, GetIntegerv,
+    GetFloatv, and GetDoublev
+
+        TEXTURE_BINDING_1D_ARRAY_EXT                    0x8C1C
+        TEXTURE_BINDING_2D_ARRAY_EXT                    0x8C1D
+        MAX_ARRAY_TEXTURE_LAYERS_EXT                    0x88FF
+
+    Accepted by the <param> parameter of TexParameterf, TexParameteri,
+    TexParameterfv, and TexParameteriv when the <pname> parameter is
+    TEXTURE_COMPARE_MODE_ARB:
+
+        COMPARE_REF_DEPTH_TO_TEXTURE_EXT                0x884E
+
+    (Note:  COMPARE_REF_DEPTH_TO_TEXTURE_EXT is simply an alias for the
+    existing COMPARE_R_TO_TEXTURE token in OpenGL 2.0; the alternate name
+    reflects the fact that the R coordinate is not always used.)
+
+    Accepted by the <internalformat> parameter of TexImage3D and
+    CompressedTexImage3D, and by the <format> parameter of
+    CompressedTexSubImage3D:
+
+        COMPRESSED_RGB_S3TC_DXT1_EXT
+        COMPRESSED_RGBA_S3TC_DXT1_EXT
+        COMPRESSED_RGBA_S3TC_DXT3_EXT
+        COMPRESSED_RGBA_S3TC_DXT5_EXT
+
+    Accepted by the <pname> parameter of
+    GetFramebufferAttachmentParameterivEXT:
+
+        FRAMEBUFFER_ATTACHMENT_TEXTURE_LAYER_EXT          0x8CD4
+
+    (Note:  FRAMEBUFFER_ATTACHMENT_TEXTURE_LAYER is simply an alias for the
+    FRAMEBUFFER_ATTACHMENT_TEXTURE_3D_ZOFFSET_EXT token provided in
+    EXT_framebuffer_object.  This extension generalizes the notion of
+    "<zoffset>" to include layers of an array texture.)
+
+Additions to Chapter 2 of the OpenGL 2.0 Specification (OpenGL Operation)
+
+    None
+
+Additions to Chapter 3 of the OpenGL 2.0 Specification (Rasterization)
+
+    -- Section 3.8.1 "Texture Image Specification"
+
+       Change the first paragraph (page 150) to say (spec changes identical to
+       EXT_texture_array):
+
+       "The command
+
+         void TexImage3D(enum target, int level, int internalformat,
+                         sizei width, sizei height, sizei depth, int border,
+                         enum format, enum type, void *data);
+
+       is used to specify a three-dimensional texture image. target must be one
+       one of TEXTURE_3D for a three-dimensional texture or
+       TEXTURE_2D_ARRAY_EXT for an two-dimensional array texture.
+       Additionally, target may be either PROXY_TEXTURE_3D for a
+       three-dimensional proxy texture, or PROXY_TEXTURE_2D_ARRAY_EXT for a
+       two-dimensional proxy array texture."
+
+       Change the fourth paragraph on page 151 to say (spec changes identical
+       to EXT_texture_array):
+
+       "Textures with a base internal format of DEPTH_COMPONENT are supported
+       by texture image specification commands only if target is TEXTURE_1D,
+       TEXTURE_2D, TEXTURE_1D_ARRAY_EXT, TEXTURE_2D_ARRAY_EXT,
+       PROXY_TEXTURE_1D, PROXY_TEXTURE_2D, PROXY_TEXTURE_1D_ARRAY_EXT, or
+       PROXY_TEXTURE_2D_ARRAY_EXT. Using this format in conjunction with any
+       other target will result in an INVALID_OPERATION error."
+
+
+       Change the fourth paragraph on page 156 to say (spec changes identical
+       to EXT_texture_array):
+       
+       "The command
+
+         void TexImage2D(enum target, int level,
+                         int internalformat, sizei width, sizei height,
+                         int border, enum format, enum type, void *data);
+
+       is used to specify a two-dimensional texture image. target must be one
+       of TEXTURE_2D for a two-dimensional texture, TEXTURE_1D_ARRAY_EXT for a
+       one-dimensional array texture, or one of TEXTURE_CUBE_MAP_POSITIVE_X,
+       TEXTURE_CUBE_MAP_NEGATIVE_X, TEXTURE_CUBE_MAP_POSITIVE_Y,
+       TEXTURE_CUBE_MAP_NEGATIVE_Y, TEXTURE_CUBE_MAP_POSITIVE_Z, or
+       TEXTURE_CUBE_MAP_NEGATIVE_Z for a cube map texture. Additionally,
+       target may be either PROXY_TEXTURE_2D for a two-dimensional proxy
+       texture, PROXY_TEXTURE_1D_ARRAY_EXT for a one-dimensional proxy array
+       texture, or PROXY TEXTURE_CUBE_MAP for a cube map proxy texture in the
+       special case discussed in section 3.8.11.  The other parameters match
+       the corresponding parameters of TexImage3D.
+
+       For the purposes of decoding the texture image, TexImage2D is
+       equivalent to calling TexImage3D with corresponding arguments and depth
+       of 1, except that
+
+         * The border depth, d_b, is zero, and the depth of the image is
+           always 1 regardless of the value of border. 
+
+         * The border height, h_b, is zero if <target> is
+           TEXTURE_1D_ARRAY_EXT, and <border> otherwise.
+
+         * Convolution will be performed on the image (possibly changing its
+           width and height) if SEPARABLE 2D or CONVOLUTION 2D is enabled.
+
+         * UNPACK SKIP IMAGES is ignored."
+
+    -- Section 3.8.2 "Alternate Texture Image Specification Commands"
+
+       Change the second paragraph (page 159) (spec changes identical
+       to EXT_texture_array):
+
+       "The command
+
+         void CopyTexImage2D(enum target, int level,
+                             enum internalformat, int x, int y, sizei width,
+                             sizei height, int border);
+
+       defines a two-dimensional texture image in exactly the manner of
+       TexImage2D, except that the image data are taken from the framebuffer
+       rather than from client memory. Currently, target must be one of
+       TEXTURE_2D, TEXTURE_1D_ARRAY_EXT, TEXTURE_CUBE_MAP_POSITIVE_X,
+       TEXTURE_CUBE_MAP_NEGATIVE_X, TEXTURE_CUBE MAP_POSITIVE_Y,
+       TEXTURE_CUBE_MAP_NEGATIVE_Y, TEXTURE_CUBE_MAP_POSITIVE_Z, or
+       TEXTURE_CUBE_MAP_NEGATIVE_Z.
+
+
+       Change the last paragraph on page 160 to say (spec changes identical
+       to EXT_texture_array):
+
+       "Currently the target arguments of TexSubImage1D and CopyTexSubImage1D
+       must be TEXTURE_1D, the target arguments of TexSubImage2D and
+       CopyTexSubImage2D must be one of TEXTURE_2D, TEXTURE_1D_ARRAY_EXT,
+       TEXTURE_CUBE_MAP_POSITIVE_X, TEXTURE_CUBE_MAP_NEGATIVE_X,
+       TEXTURE_CUBE_MAP_POSITIVE_Y, TEXTURE_CUBE_MAP_NEGATIVE_Y,
+       TEXTURE_CUBE_MAP_POSITIVE_Z, or TEXTURE_CUBE_MAP_NEGATIVE_Z, and the
+       target arguments of TexSubImage3D and CopyTexSubImage3D must be
+       TEXTURE_3D or TEXTURE_2D_ARRAY_EXT. ..."
+
+
+    -- Section 3.8.4 "Texture Parameters"
+
+       Change the first paragraph (page 166) to say:
+
+       "Various parameters control how the texel array is treated when
+       specified or changed, and when applied to a fragment. Each parameter is
+       set by calling
+
+         void TexParameter{if}(enum target, enum pname, T param); 
+         void TexParameter{if}v(enum target, enum pname, T params);
+
+       target is the target, either TEXTURE_1D, TEXTURE_2D, TEXTURE_3D,
+       TEXTURE_CUBE_MAP, TEXTURE_1D_ARRAY_EXT, or TEXTURE_2D_ARRAY_EXT."
+
+
+    -- Section 3.8.8 "Texture Minification" in the section "Scale Factor and Level of Detail"
+
+       Change the first paragraph (page 172) to say:
+
+       "Let s(x,y) be the function that associates an s texture coordinate
+       with each set of window coordinates (x,y) that lie within a primitive;
+       define t(x,y) and r(x,y) analogously.  Let u(x,y) = w_t * s(x,y),
+       v(x,y) = h_t * t(x,y), and w(x,y) = d_t * r(x,y), where w_t, h_t,
+       and d_t are as defined by equations 3.15, 3.16, and 3.17 with
+       w_s, h_s, and d_s equal to the width, height, and depth of the
+       image array whose level is level_base.  For a one-dimensional
+       texture or a one-dimensional array texture, define v(x,y) = 0 and
+       w(x,y) = 0; for a two-dimensional texture or a two-dimensional array
+       texture, define w(x,y) = 0..."
+
+    -- Section 3.8.8 "Texture Minification" in the section "Mipmapping"
+
+       Change the third paragraph (page 174) to say:
+       
+       "For a two-dimensional texture, two-dimensional array texture, or
+       cube map texture,"
+
+       Change the fourth paragraph (page 174) to say:
+
+       "And for a one-dimensional texture or a one-dimensional array texture,"
+
+       After the first paragraph (page 175) add:
+
+       "For one-dimensional array textures, h_b and d_b are treated as 1,
+       regardless of the actual values, when performing mipmap calculations.
+       For two-dimensional array textures, d_b is always treated as one,
+       regardless of the actual value, when performing mipmap calculations."
+
+    -- Section 3.8.8 "Automatic Mipmap Generation" in the section "Mipmapping"
+
+       Change the third paragraph (page 176) to say (spec changes identical
+       to EXT_texture_array):
+
+       "The contents of the derived arrays are computed by repeated, filtered
+       reduction of the level_base array.  For one- and two-dimensional array
+       textures, each layer is filtered independently.  ..."
+
+    -- Section 3.8.8 "Manual Mipmap Generation" in the section "Mipmapping"
+
+       Change first paragraph to say (spec changes identical to
+       EXT_texture_array):
+
+       "Mipmaps can be generated manually with the command
+
+         void GenerateMipmapEXT(enum target);
+
+       where <target> is one of TEXTURE_1D, TEXTURE_2D, TEXTURE_CUBE_MAP,
+       TEXTURE_3D, TEXTURE_1D_ARRAY, or TEXTURE_2D_ARRAY.  Mipmap generation
+       affects the texture image attached to <target>.  ..."
+
+    -- Section 3.8.10 "Texture Completeness"
+
+       Change the second paragraph (page 177) to say (spec changes identical
+       to EXT_texture_array):
+
+       "For one-, two-, or three-dimensional textures and one- or
+       two-dimensional array textures, a texture is complete if the following
+       conditions all hold true:"
+
+    -- Section 3.8.11 "Texture State and Proxy State"
+
+       Change the second and third paragraphs (page 179) to say (spec changes
+       identical to EXT_texture_array):
+
+       "In addition to image arrays for one-, two-, and three-dimensional
+       textures, one- and two-dimensional array textures, and the six image
+       arrays for the cube map texture, partially instantiated image arrays
+       are maintained for one-, two-, and three-dimensional textures and one-
+       and two-dimensional array textures.  Additionally, a single proxy image
+       array is maintained for the cube map texture.  Each proxy image array
+       includes width, height, depth, border width, and internal format state
+       values, as well as state for the red, green, blue, alpha, luminance,
+       and intensity component resolutions. Proxy image arrays do not include
+       image data, nor do they include texture properties. When TexImage3D is
+       executed with target specified as PROXY_TEXTURE_3D, the
+       three-dimensional proxy state values of the specified level-of-detail
+       are recomputed and updated. If the image array would not be supported
+       by TexImage3D called with target set to TEXTURE 3D, no error is
+       generated, but the proxy width, height, depth, border width, and
+       component resolutions are set to zero. If the image array would be
+       supported by such a call to TexImage3D, the proxy state values are set
+       exactly as though the actual image array were being specified. No pixel
+       data are transferred or processed in either case.
+
+       Proxy arrays for one- and two-dimensional textures and one- and
+       two-dimensional array textures are operated on in the same way when
+       TexImage1D is executed with target specified as PROXY_TEXTURE_1D,
+       TexImage2D is executed with target specified as PROXY_TEXTURE_2D or
+       PROXY_TEXTURE_1D_ARRAY_EXT, or TexImage3D is executed with target
+       specified as PROXY_TETXURE_2D_ARRAY_EXT."
+
+    -- Section 3.8.12 "Texture Objects"
+
+       Change section (page 180) to say (spec changes identical to 
+       EXT_texture_array):
+
+       "In addition to the default textures TEXTURE_1D, TEXTURE_2D,
+       TEXTURE_3D, TEXTURE_CUBE_MAP, TEXTURE_1D_ARRAY_EXT, and TEXTURE_2D_EXT,
+       named one-, two-, and three-dimensional, cube map, and one- and
+       two-dimensional array texture objects can be created and operated upon.
+       The name space for texture objects is the unsigned integers, with zero
+       reserved by the GL.
+
+       A texture object is created by binding an unused name to TEXTURE_1D,
+       TEXTURE_2D, TEXTURE_3D, TEXTURE_CUBE_MAP, TEXTURE_1D_ARRAY_EXT, or
+       TEXTURE_2D_ARRAY_EXT. The binding is effected by calling
+
+         void BindTexture(enum target, uint texture);
+
+       with <target> set to the desired texture target and <texture> set to
+       the unused name.  The resulting texture object is a new state vector,
+       comprising all the state values listed in section 3.8.11, set to the
+       same initial values. If the new texture object is bound to TEXTURE_1D,
+       TEXTURE_2D, TEXTURE_3D, TEXTURE_CUBE_MAP, TEXTURE_1D_ARRAY_EXT, or
+       TEXTURE_2D_ARRAY_EXT, it is and remains a one-, two-,
+       three-dimensional, cube map, one- or two-dimensional array texture
+       respectively until it is deleted.
+
+       BindTexture may also be used to bind an existing texture object to
+       either TEXTURE_1D, TEXTURE_2D, TEXTURE_3D, TEXTURE_CUBE_MAP,
+       TEXTURE_1D_ARRAY_EXT, or TEXTURE_2D_ARRAY_EXT. The error
+       INVALID_OPERATION is generated if an attempt is made to bind a texture
+       object of different dimensionality than the specified target. If the
+       bind is successful no change is made to the state of the bound texture
+       object, and any previous binding to target is broken.
+
+       While a texture object is bound, GL operations on the target to which
+       it is bound affect the bound object, and queries of the target to which
+       it is bound return state from the bound object. If texture mapping of
+       the dimensionality of the target to which a texture object is bound is
+       enabled, the state of the bound texture object directs the texturing
+       operation.
+
+       In the initial state, TEXTURE_1D, TEXTURE_2D, TEXTURE_3D,
+       TEXTURE_CUBE_MAP, TEXTURE_1D_ARRAY_EXT, and TEXTURE_2D_ARRAY_EXT have
+       one-, two-, three-dimensional, cube map, and one- and two-dimensional
+       array texture state vectors respectively associated with them. In order
+       that access to these initial textures not be lost, they are treated as
+       texture objects all of whose names are 0. The initial one-, two-,
+       three-dimensional, cube map, one- and two-dimensional array textures
+       are therefore operated upon, queried, and applied as TEXTURE_1D,
+       TEXTURE_2D, TEXTURE_3D, TEXTURE_CUBE_MAP, TEXTURE_1D_ARRAY_EXT, and
+       TEXTURE_2D_ARRAY_EXT respectively while 0 is bound to the corresponding
+       targets.
+
+       Change second paragraph on page 181 to say (spec changes identical to 
+       EXT_texture_array):
+       
+       "...  If a texture that is currently bound to one of the targets
+       TEXTURE_1D, TEXTURE_2D, TEXTURE_3D, TEXTURE_CUBE_MAP,
+       TEXTURE_1D_ARRAY_EXT, or TEXTURE_2D_ARRAY_EXT is deleted, it is as
+       though BindTexture had been executed with the same target and texture
+       zero. ..."
+
+       Change second paragraph on page 182 to say (spec changes identical to 
+       EXT_texture_array):
+       
+       "The texture object name space, including the initial one-, two-, and
+       three dimensional, cube map, and one- and two-dimensional array texture
+       objects, is shared among all texture units. ..."
+
+
+    -- Section 3.8.14 "Depth Texture Comparison Modes" in "Texture Comparison Modes"
+
+       Change second through fourth paragraphs (page 188) to say:
+
+       "Let D_t be the depth texture value, in the range [0, 1].  For
+       texture lookups from one- and two-dimensional, rectangle, and
+       one-dimensional array targets, let R be the interpolated <r>
+       texture coordinate, clamped to the range [0, 1].  For texture lookups
+       from two-dimensional array texture targets, let R be the interpolated
+       <q> texture coordinate, clamped to the range [0, 1].  Then the
+       effective texture value L_t, I_t, or A_t is computed as follows:
+
+       If the value of TEXTURE_COMPARE_MODE is NONE, then
+
+          r = Dt
+
+       If the value of TEXTURE_COMPARE_MODE is
+       COMPARE_REF_DEPTH_TO_TEXTURE_EXT), then r depends on the texture
+       comparison function as shown in table 3.27."
+
+    -- Section 3.8.15 "Texture Application"
+
+       Change the first paragraph (page 189) to say:
+
+       "Texturing is enabled or disabled using the generic Enable and Disable
+       commands, respectively, with the symbolic constants TEXTURE_1D,
+       TEXTURE_2D, TEXTURE_3D, TEXTURE_CUBE_MAP, TEXTURE_1D_ARRAY_EXT, or
+       TEXTURE_2D_ARRAY_EXT to enable one-, two-, three-dimensional, cube
+       map, one-dimensional array, or two-dimensional array texture,
+       respectively.  If both two- and one-dimensional textures are enabled,
+       the two-dimensional texture is used.  If the three-dimensional and
+       either of the two- or one-dimensional textures is enabled, the
+       three-dimensional texture is used.  If the cube map texture and any of
+       the three-, two-, or one-dimensional textures is enabled, then cube map
+       texturing is used.  If one-dimensional array texture is enabled and any
+       of cube map, three-, two-, or one-dimensional textures is enabled, 
+       one-dimensional array texturing is used.  If two-dimensional array
+       texture is enabled and any of cube map, three-, two-, one-dimensional
+       textures or one-dimensional array texture is enabled, two-dimensional
+       array texturing is used..."
+
+    -- Section 3.11.2 of ARB_fragment_program (Fragment Program Grammar and Restrictions):
+
+       (mostly add to existing grammar rules)
+
+       <optionName>           ::= "MESA_texture_array"
+
+       <texTarget>            ::= "1D"
+                               | "2D"
+                               | "3D"
+                               | "CUBE"
+                               | "RECT"
+                               | <arrayTarget> (if program option is present)
+                               | <shadowTarget> (if program option is present)
+
+       <arrayTarget>          ::= "ARRAY1D"
+                               | "ARRAY2D"
+
+       <shadowTarget>         ::= "SHADOW1D"
+                               | "SHADOW2D"
+                               | "SHADOWRECT"
+                               | <shadowArrayTarget> (if program option is present)
+
+       <shadowArrayTarget>    ::= "SHADOWARRAY1D"
+                               | "SHADOWARRAY2D"
+
+
+    -- Add Section 3.11.4.5.4 "Texture Stack Option"
+
+       "If a fragment program specifies the "MESA_texture_array" program
+       option, the <texTarget> rule is modified to add the texture targets
+       ARRAY1D and ARRAY2D (See Section 3.11.2)."
+
+    -- Section 3.11.6 "Fragment Program Texture Instruction Set"
+
+       (replace 1st and 2nd paragraphs with the following paragraphs)
+
+       "The first three texture instructions described below specify the
+       mapping of 4-tuple input vectors to 4-tuple output vectors.
+       The sampling of the texture works as described in section 3.8,
+       except that texture environments and texture functions are not
+       applicable, and the texture enables hierarchy is replaced by explicit
+       references to the desired texture target (i.e., 1D, 2D, 3D, cube map,
+       rectangle, ARRAY1D, ARRAY2D).  These texture instructions specify
+       how the 4-tuple is mapped into the coordinates used for sampling.  The
+       following function is used to describe the texture sampling in the
+       descriptions below: 
+
+         vec4 TextureSample(vec4 coord, float lodBias, int texImageUnit,
+                            enum texTarget);
+
+       Note that not all four components of the texture coordinates <coord>
+       are used by all texture targets.  Component usage for each <texTarget>
+       is defined in table X.
+
+                                                        coordinates used
+         texTarget          Texture Type               s t r  layer  shadow
+         ----------------   ---------------------      -----  -----  ------
+         1D                 TEXTURE_1D                 x - -    -      -
+         2D                 TEXTURE_2D                 x y -    -      -
+         3D                 TEXTURE_3D                 x y z    -      -
+         CUBE               TEXTURE_CUBE_MAP           x y z    -      -
+         RECT               TEXTURE_RECTANGLE_ARB      x y -    -      -
+         ARRAY1D            TEXTURE_1D_ARRAY_EXT       x - -    y      -
+         ARRAY2D            TEXTURE_2D_ARRAY_EXT       x y -    z      -
+         SHADOW1D           TEXTURE_1D                 x - -    -      z
+         SHADOW2D           TEXTURE_2D                 x y -    -      z
+         SHADOWRECT         TEXTURE_RECTANGLE_ARB      x y -    -      z
+         SHADOWARRAY1D      TEXTURE_1D_ARRAY_EXT       x - -    y      z
+         SHADOWARRAY2D      TEXTURE_2D_ARRAY_EXT       x y -    z      w
+
+         Table X:  Texture types accessed for each of the <texTarget>, and
+         coordinate mappings.  The "coordinates used" column indicate the
+         input values used for each coordinate of the texture lookup, the
+         layer selector for array textures, and the reference value for
+         texture comparisons."
+
+    -- Section 3.11.6.2 "TXP: Project coordinate and map to color"
+    
+       Add to the end of the section:
+       
+       "A program will fail to load if the TXP instruction is used in
+       conjunction with the SHADOWARRAY2D target."
+
+Additions to Chapter 4 of the OpenGL 2.0 Specification (Per-Fragment Operations)
+
+    -- Section 4.4.2.3 "Attaching Texture Images to a Framebuffer"
+
+       Add to the end of the section (spec changes identical to
+       EXT_texture_array):
+
+       "The command
+
+         void FramebufferTextureLayerEXT(enum target, enum attachment,
+                                         uint texture, int level, int layer);
+
+       operates identically to FramebufferTexture3DEXT, except that it
+       attaches a single layer of a three-dimensional texture or a one- or
+       two-dimensional array texture.  <layer> is an integer indicating the
+       layer number, and is treated identically to the <zoffset> parameter in
+       FramebufferTexture3DEXT.  The error INVALID_VALUE is generated if
+       <layer> is negative.  The error INVALID_OPERATION is generated if
+       <texture> is non-zero and is not the name of a three dimensional
+       texture or one- or two-dimensional array texture.  Unlike
+       FramebufferTexture3D, no <textarget> parameter is accepted.
+
+       If <texture> is non-zero and the command does not result in an error,
+       the framebuffer attachment state corresponding to <attachment> is
+       updated as in the other FramebufferTexture commands, except that
+       FRAMEBUFFER_ATTACHMENT_TEXTURE_LAYER_EXT is set to <layer>."
+
+    -- Section 4.4.4.1 "Framebuffer Attachment Completeness"
+
+      Add to the end of the list of completeness rules (spec changes
+      identical to EXT_texture_array):
+
+        "* If FRAMEBUFFER_ATTACHMENT_OBJECT_TYPE_EXT is TEXTURE and
+           FRAMEBUFFER_ATTACHMENT_OBJECT_NAME_EXT names a one- or 
+           two-dimensional array texture, then
+           FRAMEBUFFER_ATTACHMENT_TEXTURE_LAYER_EXT must be smaller than the
+           number of layers in the texture."
+
+Additions to Chapter 5 of the OpenGL 2.0 Specification (Special Functions)
+
+    -- Section 5.4 "Display Lists"
+
+       Change the first paragraph on page 242 to say (spec changes
+       identical to EXT_texture_array):
+
+       "TexImage3D, TexImage2D, TexImage1D, Histogram, and ColorTable are
+       executed immediately when called with the corresponding proxy arguments
+       PROXY_TEXTURE_3D or PROXY_TEXTURE_2D_ARRAY_EXT; PROXY_TEXTURE_2D,
+       PROXY_TEXTURE_CUBE_MAP, or PROXY_TEXTURE_1D_ARRAY_EXT;
+       PROXY_TEXTURE_1D; PROXY_HISTOGRAM; and PROXY_COLOR_TABLE,
+       PROXY_POST_CONVOLUTION_COLOR_TABLE, or
+       PROXY_POST_COLOR_MATRIX_COLOR_TABLE."
+
+Additions to Chapter 6 of the OpenGL 2.0 Specification (State and State Requests)
+
+    -- Section 6.1.3 "Enumerated Queries"
+
+       Add after the line beginning "If the value of
+       FRAMEBUFFER_ATTACHMENT_OBJECT_TYPE_EXT is TEXTURE" (spec changes
+       identical to EXT_texture_array):
+
+       "If <pname> is FRAMEBUFFER_ATTACHMENT_TEXTURE_LAYER_EXT and the
+       texture object named FRAMEBUFFER_ATTACHMENT_OBJECT_NAME_EXT is a
+       three-dimensional texture or a one- or two-dimensional array texture,
+       then <params> will contain the number of texture layer attached to the
+       attachment point.  Otherwise, <params> will contain the value zero."
+
+    -- Section 6.1.4 "Texture Queries"
+    
+       Change the first three paragraphs (page 248) to say (spec changes
+       identical to EXT_texture_array):
+
+       "The command
+
+         void GetTexImage(enum tex, int lod, enum format,
+                          enum type, void *img);
+
+       is used to obtain texture images. It is somewhat different from the
+       other get commands; tex is a symbolic value indicating which texture
+       (or texture face in the case of a cube map texture target name) is to
+       be obtained.  TEXTURE_1D, TEXTURE_2D, TEXTURE_3D, TEXTURE_1D_ARRAY_EXT,
+       and TEXTURE_2D_ARRAY_EXT indicate a one-, two-, or three-dimensional
+       texture, or one- or two-dimensional array texture, respectively.
+       TEXTURE_CUBE_MAP_POSITIVE_X, ...
+
+       GetTexImage obtains... from the first image to the last for
+       three-dimensional textures.  One- and two-dimensional array textures
+       are treated as two- and three-dimensional images, respectively, where
+       the layers are treated as rows or images.  These groups are then...
+
+       For three-dimensional and two-dimensional array textures, pixel storage
+       operations are applied as if the image were two-dimensional, except
+       that the additional pixel storage state values PACK_IMAGE_HEIGHT and
+       PACK_SKIP_IMAGES are applied. ..."
+
+Additions to Appendix A of the OpenGL 2.0 Specification (Invariance)
+
+    None
+
+Additions to the AGL/GLX/WGL Specifications
+
+    None
+
+GLX Protocol
+
+    None
+
+Dependencies on ARB_fragment_program
+
+    If ARB_fragment_program is not supported, the changes to section 3.11
+    should be ignored.
+
+Dependencies on EXT_framebuffer_object
+
+    If EXT_framebuffer_object is not supported, the changes to section
+    3.8.8 ("Manual Mipmap Generation"), 4.4.2.3, and 6.1.3 should be ignored.
+
+Dependencies on EXT_texture_compression_s3tc and NV_texture_compression_vtc
+
+    (Identical dependency as EXT_texture_array.)
+
+    S3TC texture compression is supported for two-dimensional array textures.
+    When <target> is TEXTURE_2D_ARRAY_EXT, each layer is stored independently
+    as a compressed two-dimensional textures.  When specifying or querying
+    compressed images using one of the S3TC formats, the images are provided
+    and/or returned as a series of two-dimensional textures stored
+    consecutively in memory, with the layer closest to zero specified first.
+    For array textures, images are not arranged in 4x4x4 or 4x4x2 blocks as in
+    the three-dimensional compression format provided in the
+    EXT_texture_compression_vtc extension.  Pixel store parameters, including
+    those specific to three-dimensional images, are ignored when compressed
+    image data are provided or returned, as in the
+    EXT_texture_compression_s3tc extension.
+
+    S3TC compression is not supported for one-dimensional texture targets in
+    EXT_texture_compression_s3tc, and is not supported for one-dimensional
+    array textures in this extension.  If compressed one-dimensional arrays
+    are needed, use a two-dimensional texture with a height of one.
+
+    This extension allows the use of the four S3TC internal format types in
+    TexImage3D, CompressedTexImage3D, and CompressedTexSubImage3D calls.
+
+Errors
+
+    None
+
+New State
+
+    (add to table 6.15, p. 276)
+
+                                                     Initial
+    Get Value                     Type   Get Command  Value Description           Sec.    Attribute
+    ----------------------------  -----  -----------  ----- --------------------  ------  ---------
+    TEXTURE_BINDING_1D_ARRAY_EXT  2*xZ+  GetIntegerv    0   texture object bound  3.8.12  texture
+                                                            to TEXTURE_1D_ARRAY
+    TEXTURE_BINDING_2D_ARRAY_EXT  2*xZ+  GetIntegerv    0   texture object bound  3.8.12  texture
+                                                            to TEXTURE_2D_ARRAY
+
+
+New Implementation Dependent State
+
+    (add to Table 6.32, p. 293)
+
+                                                    Minimum
+    Get Value                     Type  Get Command  Value  Description         Sec.  Attribute
+    ----------------------------  ----  ----------- ------- ------------------  ----- ---------
+    MAX_TEXTURE_ARRAY_LAYERS_EXT   Z+   GetIntegerv   64    maximum number of   3.8.1     -
+                                                            layers for texture
+                                                            arrays
+
+Issues
+
+    (1) Is "texture stack" a good name for this functionality?
+
+        NO.  The name is changed to "array texture" to match the
+        nomenclature used by GL_EXT_texture_array.
+
+    (2) Should the R texture coordinate be treated as normalized or
+    un-normalized?  If it were un-normalized, floor(R) could be thought
+    of as a direct index into the array texture.  This may be more
+    convenient for applications.
+
+        RESOLVED.  All texture coordinates are normalized.  The issue of
+        un-normalized texture coordinates has been discussed in the ARB
+        before and should be left for a layered extension.
+
+        RE-RESOLVED.  The R coordinate is un-normalized.  Accessing an array
+        using [0, layers-1] coordinates is much more natural.
+
+    (3) How does LOD selection work for stacked textures?
+
+        RESOLVED.  For 2D array textures the R coordinate is ignored, and
+        the LOD selection equations for 2D textures are used.  For 1D
+        array textures the T coordinate is ignored, and the LOD selection
+        equations for 1D textures are used.  The expected usage is in a
+        fragment program with an explicit LOD selection.
+
+    (4) What is the maximum size of a 2D array texture?  Is it the same
+    as for a 3D texture, or should a new query be added?  How about for 1D
+    array textures?
+
+        RESOLVED.  A new query is added.
+
+    (5) How are array textures exposed in GLSL?
+    
+        RESOLVED.  Use GL_EXT_texture_array.
+        
+    (6) Should a 1D array texture also be exposed?
+
+        RESOLVED.  For orthogonality, yes.
+
+    (7) How are stacked textures attached to framebuffer objects?
+
+        RESOLVED.  Layers of both one- and two-dimensional array textures
+        are attached using FreambufferTextureLayerEXT.  Once attached, the
+        array texture layer behaves exactly as either a one- or
+        two-dimensional texture.
+
+    (8) How is this extension related to GL_EXT_texture_array?
+    
+        This extension adapats GL_MESAX_texture_stack to the notation,
+        indexing, and FBO access of GL_EXT_texture_array.  This extension
+        replaces the GLSL support of GL_EXT_texture_array with
+        GL_ARB_fragment_program support.
+
+        Assembly program support is also provided by GL_NV_gpu_program4.
+        GL_NV_gpu_program4 also adds support for other features that are
+        specific to Nvidia hardware, while this extension adds only support
+        for array textures.
+
+        Much of text of this extension that has changed since
+        GL_MESAX_texture_stack comes directly from either
+        GL_EXT_texture_array or GL_NV_gpu_program4.
+
+Revision History
+
+    ||2005/11/15||0.1||idr||Initial draft MESAX version.||
+    ||2005/12/07||0.2||idr||Added framebuffer object interactions.||
+    ||2005/12/12||0.3||idr||Updated fragment program interactions.||
+    ||2007/05/16||0.4||idr||Converted to MESA_texture_array.  Brought in line with EXT_texture_array and NV_gpu_program4.||
diff --git a/mesalib/docs/MESA_texture_signed_rgba.spec b/mesalib/docs/MESA_texture_signed_rgba.spec
index bcc4d4203..e3a6b59af 100644
--- a/mesalib/docs/MESA_texture_signed_rgba.spec
+++ b/mesalib/docs/MESA_texture_signed_rgba.spec
@@ -1,214 +1,214 @@
-Name
-
-    MESA_texture_signed_rgba
-
-Name Strings
-
-    GL_MESA_texture_signed_rgba
-
-Contact
-
-
-
-Notice
-
-
-
-IP Status
-
-    No known IP issues
-
-Status
-
-
-
-Version
-
-    0.3, 2009-03-24
-
-Number
-
-    Not assigned ?
-
-Dependencies
-
-    Written based on the wording of the OpenGL 2.0 specification.
-
-    This extension trivially interacts with ARB_texture_float.
-    This extension shares some language with ARB_texture_compression_rgtc
-    but does not depend on it.
-
-Overview
-
-    OpenGL prior to 3.1 does not support any signed texture formats.
-    ARB_texture_compression_rgtc introduces some compressed red and
-    red_green signed formats but no uncompressed ones, which might
-    still be useful. NV_texture_shader adds signed texture formats,
-    but also a lot of functionality which has been superseded by fragment
-    shaders.
-    It is usually possible to get the same functionality
-    using a unsigned format by doing scale and bias in a shader, but this
-    is undesirable since modern hardware has direct support for this.
-    This extension adds a signed 4-channel texture format by backporting
-    the relevant features from OpenGL 3.1, as a means to support this in
-    OpenGL implementations only supporting older versions.
-
-Issues
-
-    1) What should this extension be called?
-
-       RESOLVED: MESA_texture_signed_rgba seems reasonable.
-       The rgba part is there because only 4 channel format is supported.
-
-
-    2) Should the full set of signed formats (alpha, luminance, rgb, etc.)
-       be supported?
-
-       RESOLVED: NO. To keep this extension simple, only add the most
-       universal format, rgba. alpha/luminance can't be trivially supported
-       since OpenGL 3.1 does not support them any longer, and there is some
-       implied dependency on ARB_texture_rg for red/red_green formats so
-       avoid all this. Likewise, only 8 bits per channel is supported.
-
-
-    3) Should this extension use new enums for the texture formats?
-
-       RESOLVED: NO. Same enums as those used in OpenGL 3.1.
-
-
-    4) How are signed integer values mapped to floating-point values?
-
-       RESOLVED: Same as described in issue 5) of
-       ARB_texture_compression_rgtc (quote):
-       A signed 8-bit two's complement value X is computed to
-       a floating-point value Xf with the formula:
-
-                { X / 127.0, X > -128
-           Xf = {
-                { -1.0,      X == -128
-
-       This conversion means -1, 0, and +1 are all exactly representable,
-       however -128 and -127 both map to -1.0.  Mapping -128 to -1.0
-       avoids the numerical awkwardness of have a representable value
-       slightly more negative than -1.0.
-
-       This conversion is intentionally NOT the "byte" conversion listed
-       in Table 2.9 for component conversions.  That conversion says:
-
-           Xf = (2*X + 1) / 255.0
-
-       The Table 2.9 conversion is incapable of exactly representing
-       zero.
-
-       (Difference to ARB_texture_compression_rgtc):
-       This is the same mapping as OpenGL 3.1 uses.
-       This is also different to what NV_texture_shader used.
-       The above mapping should be considered the reference, but there
-       is some leeway so other mappings are allowed for implementations which
-       cannot do this. Particularly the mapping given in NV_texture_shader or
-       the standard OpenGL byte/float mapping is considered acceptable too, as
-       might be a mapping which represents -1.0 by -128, 0.0 by 0 and 1.0 by
-       127 (that is, uses different scale factors for negative and positive
-       numbers).
-       Also, it is ok to store incoming GL_BYTE user data as-is, without
-       converting to GL_FLOAT (using the standard OpenGL float/byte mapping)
-       and converting back (using the mapping described here).
-       Other than those subtle issues there are no other non-standard
-       conversions used, so when using for instance CopyTexImage2D with
-       a framebuffer clamped to [0,1] all converted numbers will be in the range
-       [0, 127] (and not scaled and biased).
-
-
-    5) How will signed components resulting from RGBA8_SNORM texture
-       fetches interact with fragment coloring?
-
-       RESOLVED: Same as described in issue 6) of
-       ARB_texture_compression_rgtc (quote):
-       The specification language for this extension is silent
-       about clamping behavior leaving this to the core specification
-       and other extensions.  The clamping or lack of clamping is left
-       to the core specification and other extensions.
-
-       For assembly program extensions supporting texture fetches
-       (ARB_fragment_program, NV_fragment_program, NV_vertex_program3,
-       etc.) or the OpenGL Shading Language, these signed formats will
-       appear as expected with unclamped signed components as a result
-       of a texture fetch instruction.
-
-       If ARB_color_buffer_float is supported, its clamping controls
-       will apply.
-
-       NV_texture_shader extension, if supported, adds support for
-       fixed-point textures with signed components and relaxed the
-       fixed-function texture environment clamping appropriately.  If the
-       NV_texture_shader extension is supported, its specified behavior
-       for the texture environment applies where intermediate values
-       are clamped to [-1,1] unless stated otherwise as in the case
-       of explicitly clamped to [0,1] for GL_COMBINE.  or clamping the
-       linear interpolation weight to [0,1] for GL_DECAL and GL_BLEND.
-
-       Otherwise, the conventional core texture environment clamps
-       incoming, intermediate, and output color components to [0,1].
-
-       This implies that the conventional texture environment
-       functionality of unextended OpenGL 1.5 or OpenGL 2.0 without
-       using GLSL (and with none of the extensions referred to above)
-       is unable to make proper use of the signed texture formats added
-       by this extension because the conventional texture environment
-       requires texture source colors to be clamped to [0,1].  Texture
-       filtering of these signed formats would be still signed, but
-       negative values generated post-filtering would be clamped to
-       zero by the core texture environment functionality.  The
-       expectation is clearly that this extension would be co-implemented
-       with one of the previously referred to extensions or used with
-       GLSL for the new signed formats to be useful.
-
-
-    6) Should the RGBA_SNORM tokens also be accepted by CopyTexImage
-       functions?
-
-       RESOLVED: YES.
-
-
-    7) What to do with GetTexParameter if ARB_texture_float is supported,
-       in particular what datatype should this return for TEXTURE_RED_TYPE_ARB,
-       TEXTURE_GREEN_TYPE_ARB, TEXTURE_BLUE_TYPE_ARB, TEXTURE_ALPHA_TYPE_ARB?
-
-       RESOLVED: ARB_texture_float states type is either NONE,
-       UNSIGNED_NORMALIZED_ARB, or FLOAT. This extension adds a new enum,
-       SIGNED_NORMALIZED, which will be returned accordingly. This is the
-       same behaviour as in OpenGL 3.1.
-
-
-New Tokens
-
-
-    Accepted by the <internalformat> parameter of
-    TexImage1D, TexImage2D, TexImage3D, CopyTexImage1D, and CopyTexImage2D:
-
-        RGBA_SNORM                                0x8F93
-        RGBA8_SNORM                               0x8F97
-
-    Returned by the <params> parameter of GetTexLevelParameter:
-
-        SIGNED_NORMALIZED                         0x8F9C
-
-
-Additions to Chapter 3 of the OpenGL 2.0 Specification (Rasterization):
-
- -- Section 3.8.1, Texture Image Specification
-
-    Add to Table 3.16 (page 154): Sized internal formats
-
-    Sized             Base             R    G    B    A    L    I    D
-    Internal Format   Internal Format bits bits bits bits bits bits bits
-    ---------------   --------------- ---- ---- ---- ---- ---- ---- ----
-    RGBA8_SNORM       RGBA             8    8    8    8    0    0    0
-
-
-Dependencies on ARB_texture_float extension:
-
-    If ARB_texture_float is supported, GetTexParameter queries with <value>
-    of TEXTURE_RED_TYPE_ARB, TEXTURE_GREEN_TYPE_ARB, TEXTURE_BLUE_TYPE_ARB or
-    TEXTURE_ALPHA_TYPE_ARB return SIGNED_NORMALIZED if
-    the base internal format is RGBA_SNORM.
+Name
+
+    MESA_texture_signed_rgba
+
+Name Strings
+
+    GL_MESA_texture_signed_rgba
+
+Contact
+
+
+
+Notice
+
+
+
+IP Status
+
+    No known IP issues
+
+Status
+
+
+
+Version
+
+    0.3, 2009-03-24
+
+Number
+
+    Not assigned ?
+
+Dependencies
+
+    Written based on the wording of the OpenGL 2.0 specification.
+
+    This extension trivially interacts with ARB_texture_float.
+    This extension shares some language with ARB_texture_compression_rgtc
+    but does not depend on it.
+
+Overview
+
+    OpenGL prior to 3.1 does not support any signed texture formats.
+    ARB_texture_compression_rgtc introduces some compressed red and
+    red_green signed formats but no uncompressed ones, which might
+    still be useful. NV_texture_shader adds signed texture formats,
+    but also a lot of functionality which has been superseded by fragment
+    shaders.
+    It is usually possible to get the same functionality
+    using a unsigned format by doing scale and bias in a shader, but this
+    is undesirable since modern hardware has direct support for this.
+    This extension adds a signed 4-channel texture format by backporting
+    the relevant features from OpenGL 3.1, as a means to support this in
+    OpenGL implementations only supporting older versions.
+
+Issues
+
+    1) What should this extension be called?
+
+       RESOLVED: MESA_texture_signed_rgba seems reasonable.
+       The rgba part is there because only 4 channel format is supported.
+
+
+    2) Should the full set of signed formats (alpha, luminance, rgb, etc.)
+       be supported?
+
+       RESOLVED: NO. To keep this extension simple, only add the most
+       universal format, rgba. alpha/luminance can't be trivially supported
+       since OpenGL 3.1 does not support them any longer, and there is some
+       implied dependency on ARB_texture_rg for red/red_green formats so
+       avoid all this. Likewise, only 8 bits per channel is supported.
+
+
+    3) Should this extension use new enums for the texture formats?
+
+       RESOLVED: NO. Same enums as those used in OpenGL 3.1.
+
+
+    4) How are signed integer values mapped to floating-point values?
+
+       RESOLVED: Same as described in issue 5) of
+       ARB_texture_compression_rgtc (quote):
+       A signed 8-bit two's complement value X is computed to
+       a floating-point value Xf with the formula:
+
+                { X / 127.0, X > -128
+           Xf = {
+                { -1.0,      X == -128
+
+       This conversion means -1, 0, and +1 are all exactly representable,
+       however -128 and -127 both map to -1.0.  Mapping -128 to -1.0
+       avoids the numerical awkwardness of have a representable value
+       slightly more negative than -1.0.
+
+       This conversion is intentionally NOT the "byte" conversion listed
+       in Table 2.9 for component conversions.  That conversion says:
+
+           Xf = (2*X + 1) / 255.0
+
+       The Table 2.9 conversion is incapable of exactly representing
+       zero.
+
+       (Difference to ARB_texture_compression_rgtc):
+       This is the same mapping as OpenGL 3.1 uses.
+       This is also different to what NV_texture_shader used.
+       The above mapping should be considered the reference, but there
+       is some leeway so other mappings are allowed for implementations which
+       cannot do this. Particularly the mapping given in NV_texture_shader or
+       the standard OpenGL byte/float mapping is considered acceptable too, as
+       might be a mapping which represents -1.0 by -128, 0.0 by 0 and 1.0 by
+       127 (that is, uses different scale factors for negative and positive
+       numbers).
+       Also, it is ok to store incoming GL_BYTE user data as-is, without
+       converting to GL_FLOAT (using the standard OpenGL float/byte mapping)
+       and converting back (using the mapping described here).
+       Other than those subtle issues there are no other non-standard
+       conversions used, so when using for instance CopyTexImage2D with
+       a framebuffer clamped to [0,1] all converted numbers will be in the range
+       [0, 127] (and not scaled and biased).
+
+
+    5) How will signed components resulting from RGBA8_SNORM texture
+       fetches interact with fragment coloring?
+
+       RESOLVED: Same as described in issue 6) of
+       ARB_texture_compression_rgtc (quote):
+       The specification language for this extension is silent
+       about clamping behavior leaving this to the core specification
+       and other extensions.  The clamping or lack of clamping is left
+       to the core specification and other extensions.
+
+       For assembly program extensions supporting texture fetches
+       (ARB_fragment_program, NV_fragment_program, NV_vertex_program3,
+       etc.) or the OpenGL Shading Language, these signed formats will
+       appear as expected with unclamped signed components as a result
+       of a texture fetch instruction.
+
+       If ARB_color_buffer_float is supported, its clamping controls
+       will apply.
+
+       NV_texture_shader extension, if supported, adds support for
+       fixed-point textures with signed components and relaxed the
+       fixed-function texture environment clamping appropriately.  If the
+       NV_texture_shader extension is supported, its specified behavior
+       for the texture environment applies where intermediate values
+       are clamped to [-1,1] unless stated otherwise as in the case
+       of explicitly clamped to [0,1] for GL_COMBINE.  or clamping the
+       linear interpolation weight to [0,1] for GL_DECAL and GL_BLEND.
+
+       Otherwise, the conventional core texture environment clamps
+       incoming, intermediate, and output color components to [0,1].
+
+       This implies that the conventional texture environment
+       functionality of unextended OpenGL 1.5 or OpenGL 2.0 without
+       using GLSL (and with none of the extensions referred to above)
+       is unable to make proper use of the signed texture formats added
+       by this extension because the conventional texture environment
+       requires texture source colors to be clamped to [0,1].  Texture
+       filtering of these signed formats would be still signed, but
+       negative values generated post-filtering would be clamped to
+       zero by the core texture environment functionality.  The
+       expectation is clearly that this extension would be co-implemented
+       with one of the previously referred to extensions or used with
+       GLSL for the new signed formats to be useful.
+
+
+    6) Should the RGBA_SNORM tokens also be accepted by CopyTexImage
+       functions?
+
+       RESOLVED: YES.
+
+
+    7) What to do with GetTexParameter if ARB_texture_float is supported,
+       in particular what datatype should this return for TEXTURE_RED_TYPE_ARB,
+       TEXTURE_GREEN_TYPE_ARB, TEXTURE_BLUE_TYPE_ARB, TEXTURE_ALPHA_TYPE_ARB?
+
+       RESOLVED: ARB_texture_float states type is either NONE,
+       UNSIGNED_NORMALIZED_ARB, or FLOAT. This extension adds a new enum,
+       SIGNED_NORMALIZED, which will be returned accordingly. This is the
+       same behaviour as in OpenGL 3.1.
+
+
+New Tokens
+
+
+    Accepted by the <internalformat> parameter of
+    TexImage1D, TexImage2D, TexImage3D, CopyTexImage1D, and CopyTexImage2D:
+
+        RGBA_SNORM                                0x8F93
+        RGBA8_SNORM                               0x8F97
+
+    Returned by the <params> parameter of GetTexLevelParameter:
+
+        SIGNED_NORMALIZED                         0x8F9C
+
+
+Additions to Chapter 3 of the OpenGL 2.0 Specification (Rasterization):
+
+ -- Section 3.8.1, Texture Image Specification
+
+    Add to Table 3.16 (page 154): Sized internal formats
+
+    Sized             Base             R    G    B    A    L    I    D
+    Internal Format   Internal Format bits bits bits bits bits bits bits
+    ---------------   --------------- ---- ---- ---- ---- ---- ---- ----
+    RGBA8_SNORM       RGBA             8    8    8    8    0    0    0
+
+
+Dependencies on ARB_texture_float extension:
+
+    If ARB_texture_float is supported, GetTexParameter queries with <value>
+    of TEXTURE_RED_TYPE_ARB, TEXTURE_GREEN_TYPE_ARB, TEXTURE_BLUE_TYPE_ARB or
+    TEXTURE_ALPHA_TYPE_ARB return SIGNED_NORMALIZED if
+    the base internal format is RGBA_SNORM.
diff --git a/mesalib/docs/MESA_window_pos.spec b/mesalib/docs/MESA_window_pos.spec
index ac3e5b1b3..9e81e9c4d 100644
--- a/mesalib/docs/MESA_window_pos.spec
+++ b/mesalib/docs/MESA_window_pos.spec
@@ -1,126 +1,126 @@
-Name
-
-    MESA_window_pos
-
-Name Strings
-
-    GL_MESA_window_pos
-
-Contact
-
-    Brian Paul, brian.paul 'at' tungstengraphics.com
-
-Status
-
-    Shipping (since Mesa version 1.2.8)
-
-Version
-
-
-Number
-
-    197
-
-Dependencies
-
-    OpenGL 1.0 is required.
-    The extension is written against the OpenGL 1.2 Specification
-
-Overview
-
-    In order to set the current raster position to a specific window
-    coordinate with the RasterPos command, the modelview matrix, projection
-    matrix and viewport must be set very carefully.  Furthermore, if the
-    desired window coordinate is outside of the window's bounds one must
-    rely on a subtle side-effect of the Bitmap command in order to circumvent
-    frustum clipping.
-
-    This extension provides a set of functions to directly set the
-    current raster position, bypassing the modelview matrix, the
-    projection matrix and the viewport to window mapping.  Furthermore,
-    clip testing is not performed.
-
-    This greatly simplifies the process of setting the current raster
-    position to a specific window coordinate prior to calling DrawPixels,
-    CopyPixels or Bitmap.
-
-New Procedures and Functions
-
-    void WindowPos2dMESA(double x, double y)
-    void WindowPos2fMESA(float x, float y)
-    void WindowPos2iMESA(int x, int y)
-    void WindowPos2sMESA(short x, short y)
-    void WindowPos2ivMESA(const int *p)
-    void WindowPos2svMESA(const short *p)
-    void WindowPos2fvMESA(const float *p)
-    void WindowPos2dvMESA(const double *p)
-    void WindowPos3iMESA(int x, int y, int z)
-    void WindowPos3sMESA(short x, short y, short z)
-    void WindowPos3fMESA(float x, float y, float z)
-    void WindowPos3dMESA(double x, double y, double z)
-    void WindowPos3ivMESA(const int *p)
-    void WindowPos3svMESA(const short *p)
-    void WindowPos3fvMESA(const float *p)
-    void WindowPos3dvMESA(const double *p)
-    void WindowPos4iMESA(int x, int y, int z, int w)
-    void WindowPos4sMESA(short x, short y, short z, short w)
-    void WindowPos4fMESA(float x, float y, float z, float w)
-    void WindowPos4dMESA(double x, double y, double z, double )
-    void WindowPos4ivMESA(const int *p)
-    void WindowPos4svMESA(const short *p)
-    void WindowPos4fvMESA(const float *p)
-    void WindowPos4dvMESA(const double *p)
-
-New Tokens
-
-    none
-
-Additions to Chapter 2 of the OpenGL 1.2 Specification (OpenGL Operation)
-
-  - (2.12, p. 41) Insert after third paragraph:
-
-      Alternately, the current raster position may be set by one of the
-      WindowPosMESA commands:
-
-         void WindowPos{234}{sidf}MESA( T coords );
-         void WindowPos{234}{sidf}vMESA( T coords );
-
-      WindosPos4MESA takes four values indicating x, y, z, and w.
-      WindowPos3MESA (or WindowPos2MESA) is analaguos, but sets only
-      x, y, and z with w implicitly set to 1 (or only x and y with z
-      implicitly set to 0 and w implicitly set to 1).
-
-      WindowPosMESA operates like RasterPos except that the current modelview
-      matrix, projection matrix and viewport parameters are ignored and the
-      clip test operation always passes.  The current raster position values
-      are directly set to the parameters passed to WindowPosMESA.  The current
-      color, color index and texture coordinate update the current raster
-      position's associated data.
-
-Additions to the AGL/GLX/WGL Specifications
-
-    None
-
-GLX Protocol
-
-    Not specified at this time.  However, a protocol message very similar
-    to that of RasterPos is expected.
-
-Errors
-
-    INVALID_OPERATION is generated if WindowPosMESA is called between
-    Begin and End.
-
-New State
-
-    None.
-
-New Implementation Dependent State
-
-    None.
-
-Revision History
-
-  * Revision 1.0 - Initial specification
-  * Revision 1.1 - Minor clean-up  (7 Jan 2000, Brian Paul)
-
+Name
+
+    MESA_window_pos
+
+Name Strings
+
+    GL_MESA_window_pos
+
+Contact
+
+    Brian Paul, brian.paul 'at' tungstengraphics.com
+
+Status
+
+    Shipping (since Mesa version 1.2.8)
+
+Version
+
+
+Number
+
+    197
+
+Dependencies
+
+    OpenGL 1.0 is required.
+    The extension is written against the OpenGL 1.2 Specification
+
+Overview
+
+    In order to set the current raster position to a specific window
+    coordinate with the RasterPos command, the modelview matrix, projection
+    matrix and viewport must be set very carefully.  Furthermore, if the
+    desired window coordinate is outside of the window's bounds one must
+    rely on a subtle side-effect of the Bitmap command in order to circumvent
+    frustum clipping.
+
+    This extension provides a set of functions to directly set the
+    current raster position, bypassing the modelview matrix, the
+    projection matrix and the viewport to window mapping.  Furthermore,
+    clip testing is not performed.
+
+    This greatly simplifies the process of setting the current raster
+    position to a specific window coordinate prior to calling DrawPixels,
+    CopyPixels or Bitmap.
+
+New Procedures and Functions
+
+    void WindowPos2dMESA(double x, double y)
+    void WindowPos2fMESA(float x, float y)
+    void WindowPos2iMESA(int x, int y)
+    void WindowPos2sMESA(short x, short y)
+    void WindowPos2ivMESA(const int *p)
+    void WindowPos2svMESA(const short *p)
+    void WindowPos2fvMESA(const float *p)
+    void WindowPos2dvMESA(const double *p)
+    void WindowPos3iMESA(int x, int y, int z)
+    void WindowPos3sMESA(short x, short y, short z)
+    void WindowPos3fMESA(float x, float y, float z)
+    void WindowPos3dMESA(double x, double y, double z)
+    void WindowPos3ivMESA(const int *p)
+    void WindowPos3svMESA(const short *p)
+    void WindowPos3fvMESA(const float *p)
+    void WindowPos3dvMESA(const double *p)
+    void WindowPos4iMESA(int x, int y, int z, int w)
+    void WindowPos4sMESA(short x, short y, short z, short w)
+    void WindowPos4fMESA(float x, float y, float z, float w)
+    void WindowPos4dMESA(double x, double y, double z, double )
+    void WindowPos4ivMESA(const int *p)
+    void WindowPos4svMESA(const short *p)
+    void WindowPos4fvMESA(const float *p)
+    void WindowPos4dvMESA(const double *p)
+
+New Tokens
+
+    none
+
+Additions to Chapter 2 of the OpenGL 1.2 Specification (OpenGL Operation)
+
+  - (2.12, p. 41) Insert after third paragraph:
+
+      Alternately, the current raster position may be set by one of the
+      WindowPosMESA commands:
+
+         void WindowPos{234}{sidf}MESA( T coords );
+         void WindowPos{234}{sidf}vMESA( T coords );
+
+      WindosPos4MESA takes four values indicating x, y, z, and w.
+      WindowPos3MESA (or WindowPos2MESA) is analaguos, but sets only
+      x, y, and z with w implicitly set to 1 (or only x and y with z
+      implicitly set to 0 and w implicitly set to 1).
+
+      WindowPosMESA operates like RasterPos except that the current modelview
+      matrix, projection matrix and viewport parameters are ignored and the
+      clip test operation always passes.  The current raster position values
+      are directly set to the parameters passed to WindowPosMESA.  The current
+      color, color index and texture coordinate update the current raster
+      position's associated data.
+
+Additions to the AGL/GLX/WGL Specifications
+
+    None
+
+GLX Protocol
+
+    Not specified at this time.  However, a protocol message very similar
+    to that of RasterPos is expected.
+
+Errors
+
+    INVALID_OPERATION is generated if WindowPosMESA is called between
+    Begin and End.
+
+New State
+
+    None.
+
+New Implementation Dependent State
+
+    None.
+
+Revision History
+
+  * Revision 1.0 - Initial specification
+  * Revision 1.1 - Minor clean-up  (7 Jan 2000, Brian Paul)
+
diff --git a/mesalib/docs/MESA_ycbcr_texture.spec b/mesalib/docs/MESA_ycbcr_texture.spec
index e865039fb..6a730e81c 100644
--- a/mesalib/docs/MESA_ycbcr_texture.spec
+++ b/mesalib/docs/MESA_ycbcr_texture.spec
@@ -1,204 +1,204 @@
-Name
-
-    MESA_ycbcr_texture
-
-Name Strings
-
-    GL_MESA_ycbcr_texture
-
-Contact
-
-    Brian Paul, Tungsten Graphics, Inc. (brian.paul 'at' tungstengraphics.com)
-    Keith Whitwell, Tungsten Graphics, Inc.  (keith 'at' tungstengraphics.com)
-
-Status
-
-    Shipping (Mesa 4.0.4 and later)
-
-Version
-
-    1.0
-
-Number
-
-    TBD
-
-Dependencies
-
-    OpenGL 1.0 or later is required
-    This extension is written against the OpenGL 1.4 Specification.
-    NV_texture_rectangle effects the definition of this extension.
-
-Overview
-
-    This extension supports texture images stored in the YCbCr format.
-    There is no support for converting YCbCr images to RGB or vice versa
-    during pixel transfer.  The texture's YCbCr colors are converted to
-    RGB during texture sampling, after-which, all the usual per-fragment
-    operations take place.  Only 2D texture images are supported (not
-    glDrawPixels, glReadPixels, etc).
-
-    A YCbCr pixel (texel) is a 16-bit unsigned short with two components.
-    The first component is luminance (Y).  For pixels in even-numbered
-    image columns, the second component is Cb.  For pixels in odd-numbered
-    image columns, the second component is Cr.  If one were to convert the
-    data to RGB one would need to examine two pixels from columns N and N+1
-    (where N is even) to deduce the RGB color.
-
-IP Status
-
-    None
-
-Issues
-
-    None
-
-New Procedures and Functions
-
-    None
-
-New Tokens
-
-    Accepted by the <internalFormat> and <format> parameters of
-    TexImage2D and TexSubImage2D:
-
-        YCBCR_MESA                   0x8757
-
-    Accepted by the <type> parameter of TexImage2D and TexSubImage2D:
-
-        UNSIGNED_SHORT_8_8_MESA      0x85BA /* same as Apple's */
-        UNSIGNED_SHORT_8_8_REV_MESA  0x85BB /* same as Apple's */
-
-Additions to Chapter 2 of the OpenGL 1.4 Specification (OpenGL Operation)
-
-    None
-
-Additions to Chapter 3 of the OpenGL 1.4 Specification (Rasterization)
-
-    In section 3.6.4, Rasterization of Pixel Rectangles, on page 101,
-    add the following to Table 3.8 (Packed pixel formats):
-    
-    type Parameter                GL Data   Number of        Matching
-     Token Name                    Type     Components     Pixel Formats
-    --------------                -------   ----------     -------------
-    UNSIGNED_SHORT_8_8_MESA       ushort         2         YCBCR_MESA
-    UNSIGNED_SHORT_8_8_REV_MESA   ushort         2         YCBCR_MESA
-
-
-    In section 3.6.4, Rasterization of Pixel Rectangles, on page 102,
-    add the following to Table 3.10 (UNSIGNED_SHORT formats):
-
-    UNSIGNED_SHORT_8_8_MESA:
-
-      15  14  13  12  11  10  9   8   7   6   5   4   3   2   1   0
-    +-------------------------------+-------------------------------+
-    |              1st              |              2nd              |
-    +-------------------------------+-------------------------------+
-                        
-    UNSIGNED_SHORT_8_8_REV_MESA:
-
-      15  14  13  12  11  10  9   8   7   6   5   4   3   2   1   0
-    +-------------------------------+-------------------------------+
-    |              2nd              |              1st              |
-    +-------------------------------+-------------------------------+
-
-
-    In section 3.6.4, Rasterization of Pixel Rectangles, on page 104,
-    add the following to Table 3.12 (Packed pixel field assignments):
-
-                       First       Second     Third      Fourth
-    Format             Element     Element    Element    Element
-    ------             -------     -------    -------    -------
-    YCBCR_MESA         luminance   chroma
-
-
-    In section 3.8.1, Texture Image Specification, on page 125, add
-    another item to the list of TexImage2D and TexImage3D equivalence
-    exceptions:
-
-    * The value of internalformat and format may be YCBCR_MESA to
-      indicate that the image data is in YCbCr format.  type must
-      be either UNSIGNED_SHORT_8_8_MESA or UNSIGNED_SHORT_8_8_REV_MESA
-      as seen in tables 3.8 and 3.10.  Table 3.12 describes the mapping
-      between Y and Cb/Cr to the components.
-      If NV_texture_rectangle is supported target may also be
-      TEXTURE_RECTANGLE_NV or PROXY_TEXTURE_RECTANGLE_NV.
-      All pixel transfer operations are bypassed.  The texture is stored as
-      YCbCr, not RGB.  Queries of the texture's red, green and blue component
-      sizes will return zero.  The YCbCr colors are converted to RGB during
-      texture sampling using an implementation dependent conversion.
-
-
-    In section 3.8.1, Texture Image Specification, on page 126, add
-    another item to the list of TexImage1D and TexImage2D equivalence
-    exceptions:
-
-    * The value of internalformat and format can not be YCBCR_MESA.
-
-
-    In section 3.8.2, Alternate Texture Image Specification Commands, on
-    page 129, insert this paragraph after the first full paragraph on the
-    page:
-
-         "If the internal storage format of the image being updated by
-    TexSubImage2D is YCBCR_MESA then format must be YCBCR_MESA.
-    The error INVALID_OPERATION will be generated otherwise."
-
-
-Additions to Chapter 4 of the OpenGL 1.4 Specification (Per-Fragment
-Operations and the Frame Buffer)
-
-    None
-
-Additions to Chapter 5 of the OpenGL 1.4 Specification (Special Functions)
-
-    None
-
-Additions to Chapter 6 of the OpenGL 1.4 Specification (State and
-State Requests)
-
-    None
-
-Additions to Appendix A of the OpenGL 1.4 Specification (Invariance)
-
-    None
-
-Additions to the AGL/GLX/WGL Specifications
-
-    None
-
-GLX Protocol
-
-    None
-
-Errors
-
-    INVALID_ENUM is generated by TexImage2D if <internalFormat> is
-    MESA_YCBCR but <format> is not MESA_YCBCR.
-
-    INVALID_ENUM is generated by TexImage2D if <format> is MESA_YCBCR but
-    <internalFormat> is not MESA_YCBCR.
-
-    INVALID_VALUE is generated by TexImage2D if <format> is MESA_YCBCR and
-    <internalFormat> is MESA_YCBCR and <border> is not zero.
-
-    INVALID_OPERATION is generated by TexSubImage2D if the internal image
-    format is YCBCR_MESA and <format> is not YCBCR_MESA.
-
-    INVALID_OPERATION is generated by CopyTexSubImage2D if the internal
-    image is YCBCR_MESA.
-    
-New State
-
-    Edit table 6.16 on page 231: change the type of TEXTURE_INTERNAL_FORMAT
-    from n x Z42 to n x Z43 to indicate that internal format may also be
-    YCBCR_MESA.
-
-Revision History
-
-    20 September 2002 - Initial draft
-    29 April 2003 - minor updates
-     3 September 2003 - further clarify when YCbCr->RGB conversion takes place
-    19 September 2003 - a few more updates prior to submitting to extension
-                        registry.
-     3 April 2004 - fix assorted inaccuracies
+Name
+
+    MESA_ycbcr_texture
+
+Name Strings
+
+    GL_MESA_ycbcr_texture
+
+Contact
+
+    Brian Paul, Tungsten Graphics, Inc. (brian.paul 'at' tungstengraphics.com)
+    Keith Whitwell, Tungsten Graphics, Inc.  (keith 'at' tungstengraphics.com)
+
+Status
+
+    Shipping (Mesa 4.0.4 and later)
+
+Version
+
+    1.0
+
+Number
+
+    TBD
+
+Dependencies
+
+    OpenGL 1.0 or later is required
+    This extension is written against the OpenGL 1.4 Specification.
+    NV_texture_rectangle effects the definition of this extension.
+
+Overview
+
+    This extension supports texture images stored in the YCbCr format.
+    There is no support for converting YCbCr images to RGB or vice versa
+    during pixel transfer.  The texture's YCbCr colors are converted to
+    RGB during texture sampling, after-which, all the usual per-fragment
+    operations take place.  Only 2D texture images are supported (not
+    glDrawPixels, glReadPixels, etc).
+
+    A YCbCr pixel (texel) is a 16-bit unsigned short with two components.
+    The first component is luminance (Y).  For pixels in even-numbered
+    image columns, the second component is Cb.  For pixels in odd-numbered
+    image columns, the second component is Cr.  If one were to convert the
+    data to RGB one would need to examine two pixels from columns N and N+1
+    (where N is even) to deduce the RGB color.
+
+IP Status
+
+    None
+
+Issues
+
+    None
+
+New Procedures and Functions
+
+    None
+
+New Tokens
+
+    Accepted by the <internalFormat> and <format> parameters of
+    TexImage2D and TexSubImage2D:
+
+        YCBCR_MESA                   0x8757
+
+    Accepted by the <type> parameter of TexImage2D and TexSubImage2D:
+
+        UNSIGNED_SHORT_8_8_MESA      0x85BA /* same as Apple's */
+        UNSIGNED_SHORT_8_8_REV_MESA  0x85BB /* same as Apple's */
+
+Additions to Chapter 2 of the OpenGL 1.4 Specification (OpenGL Operation)
+
+    None
+
+Additions to Chapter 3 of the OpenGL 1.4 Specification (Rasterization)
+
+    In section 3.6.4, Rasterization of Pixel Rectangles, on page 101,
+    add the following to Table 3.8 (Packed pixel formats):
+    
+    type Parameter                GL Data   Number of        Matching
+     Token Name                    Type     Components     Pixel Formats
+    --------------                -------   ----------     -------------
+    UNSIGNED_SHORT_8_8_MESA       ushort         2         YCBCR_MESA
+    UNSIGNED_SHORT_8_8_REV_MESA   ushort         2         YCBCR_MESA
+
+
+    In section 3.6.4, Rasterization of Pixel Rectangles, on page 102,
+    add the following to Table 3.10 (UNSIGNED_SHORT formats):
+
+    UNSIGNED_SHORT_8_8_MESA:
+
+      15  14  13  12  11  10  9   8   7   6   5   4   3   2   1   0
+    +-------------------------------+-------------------------------+
+    |              1st              |              2nd              |
+    +-------------------------------+-------------------------------+
+                        
+    UNSIGNED_SHORT_8_8_REV_MESA:
+
+      15  14  13  12  11  10  9   8   7   6   5   4   3   2   1   0
+    +-------------------------------+-------------------------------+
+    |              2nd              |              1st              |
+    +-------------------------------+-------------------------------+
+
+
+    In section 3.6.4, Rasterization of Pixel Rectangles, on page 104,
+    add the following to Table 3.12 (Packed pixel field assignments):
+
+                       First       Second     Third      Fourth
+    Format             Element     Element    Element    Element
+    ------             -------     -------    -------    -------
+    YCBCR_MESA         luminance   chroma
+
+
+    In section 3.8.1, Texture Image Specification, on page 125, add
+    another item to the list of TexImage2D and TexImage3D equivalence
+    exceptions:
+
+    * The value of internalformat and format may be YCBCR_MESA to
+      indicate that the image data is in YCbCr format.  type must
+      be either UNSIGNED_SHORT_8_8_MESA or UNSIGNED_SHORT_8_8_REV_MESA
+      as seen in tables 3.8 and 3.10.  Table 3.12 describes the mapping
+      between Y and Cb/Cr to the components.
+      If NV_texture_rectangle is supported target may also be
+      TEXTURE_RECTANGLE_NV or PROXY_TEXTURE_RECTANGLE_NV.
+      All pixel transfer operations are bypassed.  The texture is stored as
+      YCbCr, not RGB.  Queries of the texture's red, green and blue component
+      sizes will return zero.  The YCbCr colors are converted to RGB during
+      texture sampling using an implementation dependent conversion.
+
+
+    In section 3.8.1, Texture Image Specification, on page 126, add
+    another item to the list of TexImage1D and TexImage2D equivalence
+    exceptions:
+
+    * The value of internalformat and format can not be YCBCR_MESA.
+
+
+    In section 3.8.2, Alternate Texture Image Specification Commands, on
+    page 129, insert this paragraph after the first full paragraph on the
+    page:
+
+         "If the internal storage format of the image being updated by
+    TexSubImage2D is YCBCR_MESA then format must be YCBCR_MESA.
+    The error INVALID_OPERATION will be generated otherwise."
+
+
+Additions to Chapter 4 of the OpenGL 1.4 Specification (Per-Fragment
+Operations and the Frame Buffer)
+
+    None
+
+Additions to Chapter 5 of the OpenGL 1.4 Specification (Special Functions)
+
+    None
+
+Additions to Chapter 6 of the OpenGL 1.4 Specification (State and
+State Requests)
+
+    None
+
+Additions to Appendix A of the OpenGL 1.4 Specification (Invariance)
+
+    None
+
+Additions to the AGL/GLX/WGL Specifications
+
+    None
+
+GLX Protocol
+
+    None
+
+Errors
+
+    INVALID_ENUM is generated by TexImage2D if <internalFormat> is
+    MESA_YCBCR but <format> is not MESA_YCBCR.
+
+    INVALID_ENUM is generated by TexImage2D if <format> is MESA_YCBCR but
+    <internalFormat> is not MESA_YCBCR.
+
+    INVALID_VALUE is generated by TexImage2D if <format> is MESA_YCBCR and
+    <internalFormat> is MESA_YCBCR and <border> is not zero.
+
+    INVALID_OPERATION is generated by TexSubImage2D if the internal image
+    format is YCBCR_MESA and <format> is not YCBCR_MESA.
+
+    INVALID_OPERATION is generated by CopyTexSubImage2D if the internal
+    image is YCBCR_MESA.
+    
+New State
+
+    Edit table 6.16 on page 231: change the type of TEXTURE_INTERNAL_FORMAT
+    from n x Z42 to n x Z43 to indicate that internal format may also be
+    YCBCR_MESA.
+
+Revision History
+
+    20 September 2002 - Initial draft
+    29 April 2003 - minor updates
+     3 September 2003 - further clarify when YCbCr->RGB conversion takes place
+    19 September 2003 - a few more updates prior to submitting to extension
+                        registry.
+     3 April 2004 - fix assorted inaccuracies
diff --git a/mesalib/docs/VERSIONS b/mesalib/docs/VERSIONS
index f72679fbe..0ebcd1469 100644
--- a/mesalib/docs/VERSIONS
+++ b/mesalib/docs/VERSIONS
@@ -1,1514 +1,1514 @@
-
-
-Mesa Version History
-====================
-
-1.0 beta   February 1995
-    Initial release
-
-1.1 beta   March 4, 1995
-    Changes:
-	faster point and line drawing (2x faster)
-	more systems supported, better Makefiles
-	Renamed lib*.a files to avoid collisions
-	many small bug fixes
-    New:
-	pseudo-GLX functions added
-	new implementation of evaluators (eval2.c)
-	GLUT support
-
-1.1.1 beta   March 7, 1995
-    Changes:
-	Reverted from eval2.c to eval.c due to FPE on Linux
-	more speed improvements
-	more Makefile changes
-
-1.1.2 beta   March 14, 1995
-    New:
-	implementation of SGI's blending extensions
-	glXUseXFont implemented
-	added MESA_DEBUG environment variable support
-    Changes:
-	Using eval2.c again
-	more FPE-prevention checks (0-length normals are OK)
-	a few small bug fixes
-	much faster pixel logic ops!
-	faster transformation arithmetic
-
-1.1.3 beta   March 31, 1995
-    New:
-	gluScaleImage() and gluBuild2DMipMaps() implemented
-	Mesa widgets for Xt/Motif
-	blendEXT demos
-	added environment variables for selecting visuals
-    Changes:
-	almost all GLUT demos work correctly now
-	faster X device driver functions
-	more bug fixes
-
-1.1.4 beta   April 20, 1995
-    Bug fixes:
-	- missing #define SEEK_SET in src-tk/image.c
-	- compile glShadeModel into display lists
-	- fixed pow() domain error in src/light.c
-	- fixed "flickering bitmaps" in double buffer mode
-	- fixed tk.h and aux.h for C++
-	- state of LIGHT_MODEL_LOCAL_VIEWER was inverted
-    New features:
-	- MUCH, MUCH nicer dithering in 8-bit RGB mode
-	- updated widgets and widget demos
-	- Implemented GLXPixmap functions
-	- Added GLU 1.1 and GLX 1.1 functions
-	- Changed the X/Mesa interface API, more versatile
-	- Implemented gluPartialDisk()
-
-1.2   May 22, 1995
-    Bug fixes:
-	- IRIX 4.x makefile problem
-	- modified tk to share root colormap as needed
-	- gluLookAt normalization problem
-	- suppress Expose, NoExpose events in swapbuffers
-	- glBitmap() and glDrawPixels() clipping
-    New features:
-	- GL_BLEND, GL_MODULATE, GL_DECAL, and GL_REPLACE_EXT texture
-	      modes implemented
-	- texture maps stored more efficiently
-	- texture maps can be compiled into display lists
-	- Bogdan Sikorski's GLU polygon tesselation code
-	- Linas Vepstas's sweep and extrusion library
-	- glXCreateContext()'s shareList parameter works as it's supposed to.
-	      XMesaCreateContext() updated to accept a shareList parameter too.
-	- Mesa can be compiled with real OpenGL .h files
-	- MESA_BACK_BUFFER environment variable
-	- better GLX error checking
-
-1.2.1   June 22, 1995
-    Bug fixes:
-	- X/Mesa double buffer window resize crash
-	- widgets now pass PointerMotion events
-	- X/Mesa incorrect default clear color and drawing color
-	- more robust X MIT-SHM support in X/Mesa
-	- glTexImage( format=GL_LUMINANCE ) didn't work
-	- GL_LINE mode polygons with line width > 1.0 could cause a crash
-	- numerous feedback bugs
-	- glReadPixels() from depth buffer was wrong
-	- error prone depth and stencil buffer allocation
-    New features:
-	- Preliminary Microsoft Windows driver
-	- Implemented a number of missing functions: glEvalCoord[12][df]v(),
-	    glGet...(), etc.
-	- Added a few missing symbols to gl.h and glu.h
-	- Faster rendering of smooth-shaded, RGBA, depth-buffered polygons.
-	- Faster rendering of lines when width=2.0
-	- Stencil-related functions now work in display lists
-    Changes:
-	- renamed aux.h as glaux.h (MS-DOS names can't start with aux)
-	- most filenames are in 8.3 format to accomodate MS-DOS
-	- use GLubytes to store arrays of colors instead of GLints
-
-1.2.2   August 2, 1995
-    New features:
-	- texture mapped points and lines
-	- NURBS! (but not 100% complete)
-	- viewports may safely extend beyond window boundaries
-	- MESA_PRIVATE_CMAP environment variable
-	- Grayscale X display support
-	- two new demos:  demos/gears.c and demos/shadow.c
-	- MachTen for Macintosh configuration
-    Bug fixes:
-	- glGet*(GL_DEPTH_BITS) returned bytes, not bits
-	- point, line, and bitmap rasterization suffered from roundoff errors
-	- fixed a division by zero error in line clippping
-	- occasional wrong default background color really fixed!
-	- glDepthFunc(GL_ALWAYS) with glDepthMask(GL_FALSE) didn't work
-	- gluBuild2DMipmaps malloc problem fixed
-	- view volume clipping of smooth shaded lines resulted in bad colors
-    Changes:
-	- new visual selection method in glXChooseVisual()
-	- improved GLU quadric functions
-	- call XSync for glFinish and XFlush for glFlush
-	- glVertex() calls now use a function pointer to avoid conditionals
-	- removed contrib directory from Mesa tar file (available on ftp site)
-	- AIX shared library support
-	- Removed GLUenum type as it's not in OpenGL
-
-1.2.3   September 26, 1995
-    New features:
-	- Mesa header files now equivalent to SGI OpenGL headers
-	- Support for HP's Color Recovery dithering displays
-	- Faster vertex transformation
-	- Faster raster operations into X windows under certain conditions
-	- New configurations:  HP w/ shared libs, Ultrix w/ GCC, Data General
-	- 4-bit visuals now supported
-    Bug fixes:
-	- glScissor bug fixed
-	- round-off errors in clipping lines against clip planes fixed
-	- byte swapping between hosts and display servers implemented
-	- glGetError() can be called without a current rendering context
-	- problem with accidentally culled polygons is fixed
-	- fixed some widget compilation problems
-
-1.2.4   November 17, 1995
-    New features:
-	- More speed improvements (lighting, fogging, polygon drawing)
-	- Window system and OS-independent off-screen rendering
-	- Preliminary Fortran bindings
-	- glPolygonOffsetEXT implemented
-	- glColorMask and glIndexMask now fully implemented
-	- glPixelZoom implemented
-	- display lists fully implemented
-	- gamma correction
-	- dithering in 8-bit TrueColor/DirectColor visuals
-    Changes:
-	- Improved device driver interface
-	- tk.h renamed to gltk.h to avoid conflicts with Tcl's Tk
-	- Dithering support moved from core into device driver
-    Bug fixes:
-	- glEnable/Disable( GL_LIGHTING ) didn't always take effect
-	- glReadPixels byte swapping was broken
-	- glMaterial with pname==GL_AMBIENT_AND_DIFFUSE was broken
-	- duplicate glColor4b() prototype in GL/gl.h removed
-	- stripes in wave -ci demo fixed
-	- GL_LINEAR_MIPMAP_NEAREST had wrong value
-	- bugs in HP Color Recovery support fixed
-	- fixed bug when blending lines, points, bitmaps outside of window
-
-1.2.5   November 30, 1995
-    New Features:
-	- updated MS Windows driver
-	- new implementation of StaticGray/GrayScale visual support
-    Bug fixes:
-	- pixelzooming with gamma correction or blending didn't work
-	- HP color recovery visual wasn't being picked by glXChooseVisual
-	- glClear didn't always observe glColorMask changes
-	- olympic and offset demos didn't compile on some Suns
-	- texcoord clamping wasn't correct
-	- a polygon optimization introduced an occasional sampling problem
-
-1.2.6   January 26, 1996
-    New Features:
-	- faster line and polygon rendering under certain conditions.  See
-	  Performance Tips 9 and 10 in README
-	- profiling
-	- lighting is a bit faster
-	- better perspective corrected texture mapping
-	- Amiga AmiWin (X11) support
-	- preliminary Linux SVGA driver
-    Changes:
-	- now using a 16-bit depth buffer, faster, smaller
-	- GL_NORMALIZE is disabled by default
-    Bug fixes:
-	- projective texture mapping
-	- fixed a memory leak in the context destroy function
-	- GL_POLYGON with less than 3 vertices caused a crash
-	- glGet*() returned wrong result for GL_INDEX_MODE
-	- reading pixels from an unmapped X window caused a BadMatch error
-
-1.2.7   March 5, 1996
-    New:
-	- faster lighting
-	- faster 16-bit TrueColor rendering on Linux
-	- faster 32-bit TrueColor rendering on Linux, HP, IBM
-	- non-depth-buffered XImage polygons are faster
-	- vertex array extension
-	- software alpha planes
-	- updated Macintosh driver
-	- new NeXT driver
-	- GLU quadric functions generate texture coordinates
-	- reflect.c demo - reflective, textured surface demo
-    Changes:
-	- gamma correction code moved into the X driver for better performance
-    Bug fixes:
-	- multiple glClipPlane()'s didn't work reliably
-	- glPolygonMode() didn't always work
-	- glCullFace( GL_FRONT_AND_BACK ) didn't work
-	- texture mapping with gamma correction was buggy
-	- floating point exceptions in texture coordinate interpolation
-	- XImage byte swapping didn't always work
-	- polygon edge flags weren't always used correctly
-
-1.2.8   May 22, 1996
-    New:
-	- overlay planes on X servers with the SERVER_OVERLAY_VISUALS property
-	- better monochrome output
-	- more IRIX 6.x configurations
-	- more robust RGB mode color allocation
-	- added MESA_XSYNC environment variable
-	- GLX_MESA_pixmap_colormap and GLX_EXT_visual_info extensions
-	- GL_MESA_window_pos extension
-	- faster glReadPixels/glDrawPixels for GL_DEPTH and GL_UNSIGNED_SHORT
-	  and GL_UNSIGNED_INT
-        - driver for prototype Cirrus Mondello 3-D board
-	- updated AmigaDOS driver
-        - a few small speed optimizations in polygon rendering
-    Changes:
-	- internal device driver interface modified to simplify device
-	  driver implementations and to support hardware Z buffers
-	- several changes to the X/Mesa interface (xmesa.h)
-    Bug fixes:
-	- fixed pow(0,0) domain error triggered on some systems
-	- glStencilClear() in a display list caused an infinite loop
-	- glRasterPos*() was sometimes off by +/-0.5 in X and Y
-	- color masking and blending were performed in wrong order
-	- auxSolidCylinder() sometimes drew a wire-frame cylinder
-	- fixed file writing bug in osdemo.c
-	- pixel mapping didn't always work
-	- the GL_GEQUAL stencil func didn't work
-	- the GL_INVERT stencil op didn't work
-	- the stencil write mask didn't work
-	- glPush/PopAttrib() didn't do enough error checking
-	- glIsList() didn't always work correctly
-
-2.0  October 10, 1996
-    New:
-	- Implements OpenGL 1.1 API functions
-	- all texture filtering modes supported (mipmapping)
-	- faster texture mapping, see Performance Tip 11 in README
-	- antialiased RGB points
-	- X support for line and polygon stippling
-	- glDrawBuffer( GL_FRONT_AND_BACK ) works
-	- util/ directory of useful stuff
-	- demos/texobj demo of texture objects
-    Changes:
-	- major internal changes for thread-safeness
-	- new device driver interface
-	- MESA_ALPHA env variable removed
-	- triangle rasterizer replaces polygon rasterizer
-    Bug fixes:
-	- glPopAttrib() bug
-	- glDrawBuffer(GL_NONE) works now
-
-2.1  December 14, 1996
-    New:
-	- VMS support
-	- MS-DOS driver
-	- OpenStep support
-	- updated, combined Windows 95/NT driver
-	- implemented glGetLighti() and glGetTexGen*()
-	- GLX does garbage collection of ancillary buffers
-    Bug fixes:
-	- removed unused _EXT constants from gl.h
-	- fixed polygon offset bugs
-	- Z coordinates of clipped lines were incorrect
-	- glEdgeFlag() in display lists didn't always work
-	- glLight*() in display lists didn't work
-	- fixed X line stipple bugs (Michael Pichler)
-	- glXUseXfonts XFreeFont/XFreeFontInfo bug fixed
-	- fixed a feedback bug
-	- glTexGen*() now transforms GL_EYE_PLANE by inverse modelview matrix
-	- polygons were sometimes culled instead of clipped
-	- triangle rasterizer suffered from float/int overflow exceptions
-	- fixed FP underflow exception in lighting (specular exponent)
-	- glEnable/glDisable of GL_EXT_vertex_array enums didn't work
-	- fixed free(NULL) in GLU tesselator code
-	- using 24-bit color on some X servers resulted in garbage rendering
-	- 32-bit per pixel mode for XFree86 now works
-	- glRotate(a,0,0,0) gave unpredictable results
-	- GL_LINE_STRIP with > 480 vertices had occasional clipping problems
-	- 8-bit TrueColor GLXPixmap rendering incorrectly required a colormap
-	- glMaterial() wasn't ignored when GL_COLOR_MATERIAL was enabled
-	- glEnable(GL_COLOR_MATERIAL) followed by glColor() didn't work right
-	- accumulation buffer was limited to positive values
-	- projective textures didn't work
-	- selection buffer overflows weren't handled correctly
-    Changes:
-	- restored the GL_EXT_polygon_offset extension
-	- slightly faster RGB dithering
-	- the SVGA driver works again
-	- Amiga driver now distributed separately
-	- NeXT driver updated for Mesa 2.x
-
-2.2  March 14, 1997
-    New:
-	- better color selection when dithering
-	- added GL_EXT_texture_object extension
-	- updated MS-DOS driver for DJGPP
-	- added openbsd make configuration
-	- faster dithered flat-shaded triangles
-	- various compilation problems with Motif widgets fixed
-	- gl.h, glx.h and glu.h name mangling option
-	- BeOS driver
-	- 3D texture mapping extension
-	- GL_MESA_resize_buffers extension
-	- morph3d, stex3d and spectex demos
-	- 3Dfx support
-    Bug fixes:
-	- glColorMaterial should finally work right in all respects
-	- linear interpolation of mipmap levels was incorrectly weighted
-	- readpix.c didn't compile on Macintosh
-	- GL_INVERT and related logic ops didn't work right
-	- glTexImage[12]D() didn't check its parameters consistantly
-	- fixed a memory leak in glTexImage[12]D()
-	- kludged around a SunOS 5.x/GCC compiler bug in the feedback code
-	- glReadPixels aborted instead of normally catching some errors
-	- a few 1.1 constants were missing or misnamed in gl.h
-	- glBegin(p); glBegin(q); didn't generate an error
-	- fixed a memory leak in GLX code
-	- clipping of concave polygons could cause a core dump
-	- 1-component alpha texture maps didn't work
-	- fixed a GLU polygon tesselator bug
-	- polygons with colinear vertices were sometimes culled
-	- feedback triangle colors were wrong when using smooth shading
-	- textures with borders didn't work correctly
-	- colors returned in feedback mode were wrong when using lighting
-	- spotlights didn't effect ambient lighting correctly
-	- gluPartialDisk() had a few bugs
-    Changes:
-	- device driver interface expanded to support texture mapping
-	- faster matrix inversion subroutine
-	- commented out #include "wmesa_extend.h" from src/wmesa.c
-	- fixed many compiler warnings in the demo programs
-
-2.3  June 30, 1997
-    New:
-	- Mesa distribution divided into two pieces:  library code and demos
-	- faster vertex transformation, clip testing, lighting
-	- faster line drawing
-	- TrueColor visuals how have dithering (for depths < 24 bits)
-	- added MESA_NO_DITHER environment variable
-	- new device driver function: NearFar(), RenderVB(), RasterSetup()
-	- added LynxOS configuration
-	- added cygnus Win32 configuration
-	- added texcyl.c GLUT demo
-	- added XMesaDitherColor() to X/Mesa interface
-	- new NURBS code from Bogdan Sikorski
-	- added demos/shape.c (non-rectangular X window!)
-    Bug fixes:
-	- glEnable/DisableClientState() were missing from GL/gl.h
-	- GL_SPHERE_MAP texcoord generation didn't work correctly
-	- glXGetConfig() returned wrong number of depth, stencil, accum bits
-	- glDrawPixels feedback/selection didn't examine RasterPos valid bit
-	- black and white were reversed on some monochrome displays
-	- fixed potential image memory leak (wasn't setting reference counter)
-	- glDrawPixels sometimes didn't recognize some GL state changes
-	- gluProject/UnProject() didn't check for divide by zero
-	- stex3d demo called random() and srandom(), not portable
-	- fixed memory leaks in context.c and drawpix.c
-	- fixed NULL dereferencing problem in gl_update_texture_state()
-	- glReadPixels between glBegin/glEnd didn't generate an error.
-	- fixed memory leak in polygon tesselator (Randy Frank)
-	- fixed seg fault bug drawing flat-shaded, depth-tested lines
-	- clipped GL_TRIANGLE_STRIPs sometimes had wrong color when flat-shaded
-	- glBindTexture sometimes didn't work
-	- fixed a bug deep in glXReleaseBuffersMESA()
-	- fog was mistakenly applied to alpha
-	- glPopMatrix didn't set "dirty matrix" flag
-	- glPolygonStipple pattern was sometimes wrong
-	- glClear wasn't disabled during feedback and selection
-	- fixed memory leak in glTexSubImage[123]D
-    Changes:
-	- many library source files reorganized
-	- faster X color allocation, colors also freed when finished with them
-	- new texture sampling function pointer in texture objects
-	- incorporated 3Dfx VooDoo driver v0.16 into main source tree
-	- many 3Dfx driver updates
-	- cygnus Makefiles now included
-	- updated DOS driver
-	- made a few changes to dosmesa.c and wmesa.c (VB->Unclipped)
-	- internally, colors now stored in GLubytes, not GLfixed
-	- optimized changing of GL_SHININESS parameter
-
-2.4  September 18, 1997
-    New:
-	- updated 3Dfx Glide driver
-	- hacks for 3Dfx rendering into an X window or fullscreen
-	- added depth buffer access functions to X/Mesa and OS/Mesa interfaces
-    Bug fixes:
-	- pixel buffer could overflow with long, wide lines
-	- fixed FP underflow problems in lighting
-	- glTexSubImage1D() had an unitialized variable
-	- incomplete texture objects could cause a segfault
-	- glDrawPixels with GL_COMPILE_AND_EXECUTE caused infinite loop
-	- flat-shaded quads in a strip were miscolored if clipped
-	- mipmapped triangle lod computation now works correctly
-	- fixed a few under/overflow bugs in triangle rasterizer
-	- glArrayElement() assigned bad normal if normal array disabled
-	- changed argument to glXReleaseBuffersMESA()
-	- fixed small triangle underflow bugs in tritemp.h (hopefully)
-	- glBindTexture(target, 0) caused a crash
-	- glTexImage[123]D() with NULL image pointer caused crash
-	- glPixelStore parameters are now ignored during display list execution
-	- fixed a two-sided lighting w/ clipping bug (black vertices)
-	- textures with width!=height were sometimes mis-rendered
-	- "weird" projection matrices could cause div by 0, other fp errors
-    Changes:
-	- changed precompiled header symbol from PCH to PC_HEADER
-	- split api.c into api1.c and api2.c
-	- added hash.c source file (but not used yet)
-	- a few Sun and HP configuration file changes
-	- MESA_GLX_FX env var replaces MESA_FX_WINDOW and MESA_FX_FULLSCREEN
-	- fixed a few cygnus build problems (src/Makefile.cygnus, src/wmesa.c)
-
-2.5  November 20, 1997
-    New:
-	- updated 3Dfx driver (v20) for GLQuake
-	- added GL_EXT_paletted_texture extension
-	- added GL_EXT_shared_texture_palette extension
-	- added GL_EXT_point_parameters extension
-	- now including Mark Kilgard's GLUT library v3.6
-	- new GLUT-based demos in gdemos/
-	- added a few more Unix config targets
-	- added Intel X86 assembly language vertex transformation code
-	- 3Dfx/Glide driver for Mesa now recognizes SST_SCREENREFRESH env var
-	- Windows 95 S3 Virge driver
-    Bug fixes:
-	- glCopyTexImage?D would crash due to uninitialized variable
-	- glColor w/ glColorMaterial in a display list caused a bug
-	- fixed several glDrawPixels() and ReadPixels() bugs in 3Dfx driver
-	- glVertex4*() vertices weren't always projected correctly
-	- trying to use mipmapped textured points or lines caused crash
-	- glColor[34][fd]() values now clamped to [0,1] before int conversion
-    Changes:
-	- new device driver functions for texture mapping
-	- hash tables used for display list and texture object lookup
-	- fixed GLX visual handling code to avoid saving redundant visuals
-	- 3Dfx Glide libraries automatically linked to libMesaGL.so
-	- dropped the Cirrus Logic Mondello code since it's obsolete
-	- updated Cygnus Makefiles (Stephane Rehel)
-	- updated Windows MSVC++ Makefiles (Oleg Letsinsky)
-	- procedure for making library files has changed: scripts now take
-	  a major and minor version arguments.  Make-config changed a lot.
-	- new implementation of glTexSubImage2D()
-	- updated widgets-mesa directory to create libMesaGLwM.a (Motif widget)
-	- separate linux-glide and linux-386-glide configurations
-
-2.6  February 12, 1998
-    New:
-	- Windows WGL functions
-	- updated VMS, DOS, Windows, Cygnus, BeOS, Amiga compilation support
-	- v0.22 of 3Dfx Glide driver
-	- more X86 assembly language optimizations
-	- faster blending for some modes
-	- XMesaSetFXmode() to switch between 3Dfx window and full-screen mode 
-	- added preliminary thread support
-	- added GLX_MESA_copy_sub_buffer extension
-	- some clipping optimizations
-    Bug fixes:
-	- fixed shading/material bug when drawing long primitive strips
-	- fixed clipping problem in long primitive strips
-	- fixed clipping bug when using 3Dfx driver
-	- fixed a problem when trying to use X fonts w/ 3Dfx driver
-	- fixed a texture filter bug in 3Dfx/Glide driver
-	- fixed bug in 3Dfx/Glide driver involving depth mask & clearing
-	- glLoadMatrix to set projection matrix confused the 3Dfx driver
-	- non-identity texture matrices didn't work with linux-386 configs
-	- glGenTextures() didn't reserve the returned texture IDs
-	- NULL proxy image sent to glTexImageXD() caused crash
-	- added texture state validation optimization (Henk Kok)
-	- fixed colormap reuse problem when using both RGB and CI windows
-	- 32bpp True/DirectColor X visuals weren't recognized
-	- fixed potential problem in evaluators memory allocation
-	- fixed assorted demo compilation bugs
-    Changes:
-	- replaced old Mesa/windows/ directory with Mesa/WIN32/ directory
-	- converted a few old glaux/gltk demos to GLUT
-	- renamed directories:  demos -> xdemos,  gdemos -> demos
-
-
-3.0  September 17, 1998
-    New:
-	- OpenGL 1.2 API
-	- GL_EXT_abgr pixel format extension
-	- GL_SGIS_texture_edge_clamp extension
-	- GL_SGIS_multitexture extension (to be replaced by GL_ARB_multitex)
-	- GL_EXT_multitexture extension (to be replaced by GL_ARB_multitex)
-	- GL_EXT_rescale_normal extension and renormal.c demo
-	- GLX_SGI_video_sync extension (a no-op)
-	- antialiased lines
-	- glGetTexImage() now implemented
-	- glDraw/Copy/ReadPixels() optimizations
-	- optimized textured triangle code (Marten Stromberg)
-	- more optimization of dithered TrueColor triangles in X driver
-	- Linux GGI driver
-	- updated MGL driver
-    Bug fixes:
-	- lots of assorted compilation fixes
-	- glInitNames didn't write initial hit record
-	- glBitmap didn't always check for invalid raster position
-	- switching between GLX and OSMesa contexts caused a crash
-	- fixed uninitialized variable in Mesa widget code
-	- fixed typo in texture code which caused book/texgen to crash
-	- fixed texture sampling bug when filter=GL_LINEAR and wrap=GL_CLAMP
-	- gluDisk() in POINT or LINE mode sometimes failed
-	- fixed texture + fog bug
-	- GL_COMPILE_AND_EXECUTE mode didn't work reliably
-	- glMultMatrix in projection matrix mode w/ 3Dfx driver could fail
-	- glDrawPixels(color index pixels) weren't converted to RGBA
-	- fixed possible getenv() buffer overflow security bug
-	- glBitmap in feedback mode was offset by xOrig, yOrig params
-	- device driver's DrawPixels hook was never used
-	- glDrawPixels with zoomY!=1 and top/bottom clipping didn't work
-	- glDrawPixels optimized for GL_LUMINANCE, GL_LUMINANCE_ALPHA, GLubyte
-	- fixed MakeCurrent bug in GLwRedrawObjects() in MesaWorkstation.c
-	- glCopyTexSubImage2D() didn't work with 3Dfx driver
-	- lines with width = 2 could cause crash
-	- glClear with scissor rect sometimes cleared whole buffer
-	- glTexSubImage2D( .. GL_COLOR_INDEX .. ) didn't work
-	- glTexImageXD( .. GL_ABGR_EXT .. ) didn't work
-	- computation of inverse modelview matrix sometimes failed
-	- fixed GL_CLAMP mode texture sampling bug
-	- textured line interpolation was somewhat broken
-	- textured triangle interpolation was also somewhat broken
-	- glGet(MODELVIEW/PROJECTION/TEXTURE_MATRIX_STACK_DEPTH) off by one
-	- evaluator state wasn't fully initialized
-	- texture coordinate clipping was buggy
-	- evaluator surfaces could be mis-colored
-	- glAccum(GL_RETURN, s) didn't obey glColorMask() settings
-	- zero area polygons shouldn't be culled if polygon mode is point/line
-	- clipped width and height of glReadPixels was sometimes off by one
-	- blending with alpha = 0 or 1.0 wasn't always exact
-	- reading of pixels from clipped region was buggy
-	- minor tweaking of X visual management in GLX emulator
-	- glPolygonStipple now obeys pixel unpacking parameters
-	- glGetPolygonStipple now obeys pixel packing parameters
-	- interleaved vertex array texture coordinates were broken
-	- query of proxy texture internal format was broken
-	- alpha channel wasn't reliably cleared
-	- fixed divide by zero error in gluScaleImage if dest size = 1 x 1
-    Conformance bug fixes:
-	- GL_SELECTION_BUFFER_POINTER and GL_SELECTION_BUFFER_SIZE were missing
-	- GL_TEXTURE_INTERNAL_FORMAT was missing
-	- glGet*(GL_POLYGON_STIPPLE) was broken
-	- glPush/PopAttrib() didn't save/restore all texture state
-	- glBitmap in feedback mode didn't work
-	- feedback of texture coords didn't always work
-	- glDrawPixels w/ format=GL_DEPTH_COMPONENT, type=GLbyte was broke
-	- glDrawPixels w/ format=GL_DEPTH_COMPONENT, type=GLubyte was broke
-	- glDrawPixels w/ format=GL_STENCIL_INDEX, type=GL_BITMAP was broke
-    Changes:
-	- upgraded GLUT to version 3.7
-	- only GL and GLU library code included in MesaLib.tar.gz
-	- GLUT and all demos now in MesaDemos.tar.gz
-	- glaux and gltk libraries removed
-	- IRIX -n32 and -64 libs go in lib32/ and lib64/ directories
-
-
-3.1 beta 1  November 19, 1998
-    New:
-	- GL_EXT_stencil_wrap extension
-	- GL_INGR_blend_func_separate extension
-	- GL_ARB_multitexture extension
-	- GL_NV_texgen_reflection extension
-	- newly optimized vertex transformation code
-	- updated GLUT 3.7 code
-	- better precision when using 32-bit Z buffer
-	- Allegro DJGPP driver
-    Bug fixes:
-	- glCopyPixels between front/back buffers didn't copy alpha correctly
-	- fixed out-of-bounds memory access in optimized 2-D texture code
-	- glPixelStorei didn't accept GL_PACK/UNPACK_IMAGE_HEIGHT parameter
-	- glGet*() didn't accept GL_MAX_3D_TEXTURE_SIZE parameter
-	- clipping of texture coordinates sometimes had bad R,Q values
-	- GL_CLAMP_TO_EDGE texture sampling was off by 0.5 texels
-	- glEdgeFlagPointer() now takes a GLvoid * instead of GLboolean *
-	- texture was sometimes applied twice with 3Dfx driver
-	- glPush/PopAttrib() fouled up texture object reference counts
-	- glDeleteLists(0, n) caused assertion failure
-	- bilinear texture sampling wasn't accurate enough
-	- glClear w/ glDepthMask(GL_FALSE) didn't work right on 3Dfx
-	- color components were reversed on big endian 32 bpp X visuals
-    Changes:
-	- removed GL_EXT_multitexture extension
-
-
-3.1 beta 2  May 24, 1999
-    New:
-	- multi-textured points and lines (mjk@nvidia.com)
-	- optimized 24bpp X rendering (bernd.paysan@gmx.de)
-	- added allegro support (bernie-t@geocities.com)
-	- cleaned-up Windows-related stuff (Ted Jump)
-	- minor stereo changes (KendallB@scitechsoft.com)
-	- new BeOS driver which implements BGLView class
-	- new Direct3D driver (see src/D3D)
-	- more efficient filled gluCylinder() function
-	- utilities:  util/showbuffer.[ch] and util/glstate.[ch]
-	- fixed some IRIX compiler warnings
-	- added support for building Mesa in XFree86 with
-	    SGI's GLX (kevin@precisioninsight.com)
-    Bug fixes:
-	- a variety of Windows/Mesa bug fixes (mjk@nvidia.com)
-	- packed pixel images weren't unpacked correctly
-	- patches some win32 files in GLUT (mjk@nvidia.com)
-	- glTexImage[123]D() didn't accept internalFormat == GL_COLOR_INDEX
-	- fixed lighting bug in Keith's new shading code
-	- fixed texture segfault seen in Lament screensaver
-	- fixed miscellaneous low-memory bugs
-	- glClear(GL_COLOR_BUFFER_BIT) with RGBA or CI masking was broken
-	- GL_LINEAR sampling of 3D textures was broken
-	- fixed SVR4 'cc' compiler macro problem (dawes@xfree86.org)
-	- added GL_TEXTURE_PRIORITY fix (keithh@netcomuk.co.uk)
-	- fixed wide point and wide line conformance bugs (brianp)
-    Changes:
-	- some device driver changes (see src/dd.h)
-	- new copyright on core Mesa code
-
-
-3.1 beta 3  September 17, 1999
-    New:
-	- optimized glAccum function
-	- optimized 24bpp rendering in XMesa driver
-	- GLU 1.2 polygon tessellator
-    Bug Fixes:
-	- glGetTexLevelParameter wasn't fully implemented
-	- glXUseXFont now handles multi-byte fonts
-	- glIsEnabled(GL_TEXTURE_2D / 3D) returned wrong result
-	- alpha channel of blending points, lines was sometimes incorrect
-    Changes:
-	- New library names:  "libGL" instead of "libMesaGL"
-	- New library numbering: libGL.so.1.2.310
-	- New subdirectories:  docs/ and bin/
-	- New Makefile-system (autoconf,automake,libtool)
-
-
-3.1 final  December 14, 1999
-    New:
-	- added demos/gloss.c
-	- added xdemos/glxdpyinfo.c
-	- added GLX_ARB_get_proc_address extension
-	- rewritten glTexImage code paths (faster, less memory, bug fixes)
-    Bug Fixes:
-	- several vertex array bug fixes
-	- overlapping glCopyPixels with pixel zooming now works
-	- glXUseXFont() bitmaps were vertically shifted by one pixel
-	- glCopyPixels with pixel zooming now works
-
-
-3.2 final  April 24, 2000
-    Bug fixes:
-	- fixed memcpy bugs in span.c
-	- fixed missing glEnd problem in demos/tessdemo.c
-	- fixed bug when clearing 24bpp Ximages
-	- fixed clipping problem found in Unreal Tournament
-	- fixed Loki's "ice bug" and "crazy triangles" seen in Heretic2
-	- fixed Loki's 3dfx RGB vs BGR bug
-	- fixed Loki's 3dfx smooth/flat shading bug in SoF
-    Changes:
-	- updated docs/README file
-	- use bcopy() optimizations on FreeBSD
-	- re-enabled the optimized persp_textured_triangle() function
-
-
-3.2.1  July 19, 2000
-    Bug fixes:
-	- gluBuild2DMipmaps() didn't accept GL_BGRA
-	- Fixed compile/makefile problems on IRIX
-	- fixed segfault in 3dfx driver when using GL selection/feedback
-	- no longer cull very, very tiny triangles
-	- blending w/ drawbuffer==GL_FRONT_BACK caused segfault (sw rendering)
-	- fixed Motif detection code in widgets-mesa/configure.in
-	- glColorMaterial and glMaterial updates to emissive and ambient
-	  didn't always work right
-	- Specular highlights weren't always in the right place
-	- clipped GL_LINE mode polygons had interior lines appear
-	- blend term GL_ONE_MINUS_CONSTANT_ALPHA was broken
-	- GL_NICEST fog didn't always work with flat shading
-	- glRect commands in display lists were sometimes miscolored
-	- Line Z offset didn't always work
-	- fixed texgen normal vector problem (gloss's teapot)
-	- numerous GL conformance bugs fixed
-    Changes:
-	- glColorMask(false, false, false, false) handled better/faster
-	- reverted to old GLU polygon tessellator, GLU 1.1
-	- updated Win32 build files
-
-
-3.3  July 21, 2000
-    New:
-	- antialiased triangles now implemented
-	- GL_EXT_texture_env_add texture mode extension
-	- GLX 1.3 API
-	- support for separate draw/read buffers (ie GL_SGI_make_current_read)
-	- thread-safe API dispath
-	- improved glxinfo program
-	- demos/texdown program to measure texture download performance
-	- glext.h header file
-	- demos/geartrain program
-	- GL_EXT_texture_lod_bias extension
-	- demos/lodbias program
-	- further optimized glRead/DrawPixels for 16-bit TrueColor X visuals
-	- GLX_EXT_visual_rating extension (a no-op, however)
-	- GL_HP_occlusion_test extension (for X and OS/Mesa drivers)
-	- demos/occlude program
-	- GL_SGIS_pixel_texture and GL_SGIX_pixel_texture extensions
-	- demos/pixeltex program
-	- GL_SGI_color_matrix extension
-	- GL_SGI_color_table extension
-	- GL_EXT_histogram extension
-	- GL_ARB_texture_cube_map extension
-	- added xdemos/glxheads and xdemos/manywin
-	- demos/texenv.c demo
-	- GL_EXT_texture_env_combine extension (by Holger Waechtler)
-	- Xlib driver is now thread-safe (see xdemos/glthreads)
-    Bug Fixes:
-	- various GL conformance failures fixed since 3.2.1
-    Changes:
-	- gl.h now uses #defines instead of C enums for all tokens
-	- glu.h now uses #defines instead of C enums for all tokens
-	- moved programs from 3Dfx/demos/ into demos/ directory
-
-
-3.4  November 3, 2000
-    New:
-	- optimized glDrawPixels for glPixelZoom(1,-1)
-    Bug Fixes:
-	- widgets-mesa/src/*.c files were missing from 3.3 distro
-	- include/GL/mesa_wgl.h file was missing from 3.3 distro
-	- fixed some Win32 compile problems
-	- texture object priorities weren't getting initialized to 1.0
-	- glAreTexturesResident return value was wrong when using hardware
-	- glXUseXFont segfaulted when using 3dfx driver (via MESA_GLX_FX)
-	- glReadPixels with GLushort packed types was broken
-	- fixed a few bugs in the GL_EXT_texture_env_combine texture code
-	- glPush/PopAttrib(GL_ENABLE_BIT) mishandled multi-texture enables
-	- fixed some typos/bugs in the VB code
-	- glDrawPixels(GL_COLOR_INDEX) to RGB window didn't work
-	- optimized glDrawPixels paths weren't being used
-	- per-fragment fog calculation didn't work without a Z buffer
-	- improved blending accuracy, fixes Glean blendFunc test failures
-	- glPixelStore(GL_PACK/UNPACK_SKIP_IMAGES) wasn't handled correctly
-	- glXGetProcAddressARB() didn't always return the right address
-	- gluBuild[12]DMipmaps() didn't grok the GL_BGR pixel format
-	- texture matrix changes weren't always detected (GLUT projtex demo)
-	- fixed random color problem in vertex fog code
-	- fixed Glide-related bug that let Quake get a 24-bit Z buffer
-    Changes:
-	- finished internal support for compressed textures for DRI
-
-
-3.4.1  February 14, 2001
-    New:
-	- fixed some Linux build problems
-	- fixed some Windows build problems
-	- GL_EXT_texture_env_dot3 extension (Gareth Hughes)
-    Bug fixes:
-	- added RENDER_START/RENDER_FINISH macros for glCopyTexImage in DRI
-	- various state-update code changes needed for DRI bugs
-	- disabled pixel transfer ops in glColorTable commands, not needed
-	- fixed bugs in glCopyConvolutionFilter1D/2D, glGetConvolutionFilter
-	- updated sources and fixed compile problems in widgets-mesa/
-	- GLX_PBUFFER enum value was wrong in glx.h
-	- fixed a glColorMaterial lighting bug
-	- fixed bad args to Read/WriteStencilSpan in h/w stencil clear function
-	- glXCopySubBufferMESA() Y position was off by one
-	- Error checking of glTexSubImage3D() was broken (bug 128775)
-	- glPopAttrib() didn't restore all derived Mesa state correctly
-	- Better glReadPixels accuracy for 16bpp color - fixes lots of OpenGL
-	  conformance problems at 16bpp.
-	- clearing depth buffer with scissoring was broken, would segfault
-	- OSMesaGetDepthBuffer() returned bad bytesPerValue value
-	- fixed a line clipping bug (reported by Craig McDaniel)
-	- fixed RGB color over/underflow bug for very tiny triangles
-    Known problems:
-	- NURBS or evaluator surfaces inside display lists don't always work
-
-
-3.4.2  May 17, 2001
-    Bug fixes:
-	- deleting the currently bound texture could cause bad problems
-	- using fog could result in random vertex alpha values
-	- AA triangle rendering could touch pixels outside right window bound
-	- fixed byteswapping problem in clear_32bit_ximage() function
-	- fixed bugs in wglUseFontBitmapsA(), by Frank Warmerdam
-	- fixed memory leak in glXUseXFont()
-	- fragment sampling in AA triangle function was off by 1/2 pixel
-	- Windows: reading pixels from framebuffer didn't always work
-	- glConvolutionFilter2D could segfault or cause FP exception
-	- fixed segfaults in FX and X drivers when using tex unit 1 but not 0
-	- GL_NAND logicop didn't work right in RGBA mode
-	- fixed a memory corruption bug in vertex buffer reset code
-	- clearing the softwara alpha buffer with scissoring was broken
-	- fixed a few color index mode fog bugs
-	- fixed some bad assertions in color index mode
-	- fixed FX line 'stipple' bug #420091
-	- fixed stencil buffer clear width/height typo
-	- fixed GL error glitches in gl[Client]ActiveTextureARB()
-	- fixed Windows compilation problem in texutil.c
-	- fixed 1/8-pixel AA triangle sampling error
-    Changes:
-	- optimized writing mono-colored pixel spans to X pixmaps
-	- increased max viewport size to 2048 x 2048
-
-
-3.5  June 21, 2001
-    New:
-	- internals of Mesa divided into modular pieces (Keith Whitwell)
-	- 100% OpenGL 1.2 conformance (passes all conformance tests)
-	- new AA line algorithm
-	- GL_EXT_convolution extension
-	- GL_ARB_imaging subset
-	- OSMesaCreateContextExt() function
-	- GL_ARB_texture_env_add extension (same as GL_EXT_texture_env_add)
-	- GL_MAX_TEXTURE_UNITS_ARB now defaults to eight
-	- GL_EXT_fog_coord extension (Keith Whitwell)
-	- GL_EXT_secondary_color extension (Keith Whitwell)
-	- GL_ARB_texture_env_add extension (same as GL_EXT_texture_env_add)
-	- GL_SGIX_depth_texture extension
-	- GL_SGIX_shadow and GL_SGIX_shadow_ambient extensions
-	- demos/shadowtex.c demo of GL_SGIX_depth_texture and GL_SGIX_shadow
-	- GL_ARB_texture_env_combine extension
-	- GL_ARB_texture_env_dot3 extension
-	- GL_ARB_texture_border_clamp (aka GL_SGIS_texture_border_clamp)
-	- OSMesaCreateContextExt() function
-	- libOSMesa.so library, contains the OSMesa driver interface
-	- GL/glxext.h header file for GLX extensions
-	- somewhat faster software texturing, fogging, depth testing
-	- all color-index conformance tests now pass (only 8bpp tested)
-	- SPARC assembly language TCL optimizations (David Miller)
-	- GL_SGIS_generate_mipmap extension
-    Bug Fixes:
-	- fbiRev and tmuRev were unitialized when using Glide3
-	- fixed a few color index mode conformance failures; all pass now
-	- now appling antialiasing coverage to alpha after texturing
-	- colors weren't getting clamped to [0,1] before color table lookup
-	- fixed RISC alignment errors caused by COPY_4UBV macro
-	- drawing wide, flat-shaded lines could cause a segfault
-	- vertices now snapped to 1/16 pixel to fix rendering of tiny triangles
-    Changes:
-	- SGI's Sample Implementation (SI) 1.3 GLU library replaces Mesa GLU
-	- new libOSMesa.so library, contains the OSMesa driver interface
-
-
-4.0  October 22, 2001
-    New:
-	- Mesa 4.0 implements the OpenGL 1.3 specification
-	- GL_IBM_rasterpos_clip extension
-	- GL_EXT_texture_edge_clamp extension (aka GL_SGIS_texture_edge_clamp)
-	- GL_ARB_texture_mirrored_repeat extension
-	- WindML UGL driver (Stephane Raimbault)
-	- added OSMESA_MAX_WIDTH/HEIGHT queries
-	- attempted compiliation fixes for Solaris 5, 7 and 8
-	- updated glext.h and glxext.h files
-	- updated Windows driver (Karl Schultz)
-    Bug fixes:
-	- added some missing GLX 1.3 tokens to include/GL/glx.h
-	- GL_COLOR_MATRIX changes weren't recognized by teximage functions
-	- glCopyPixels with scale and bias was broken
-	- glRasterPos with lighting could segfault
-	- glDeleteTextures could leave a dangling pointer
-	- Proxy textures for cube maps didn't work
-	- fixed a number of 16-bit color channel bugs
-	- fixed a few minor memory leaks
-	- GLX context sharing was broken in 3.5
-	- fixed state-update bugs in glPopClientAttrib()
-	- fixed glDrawRangeElements() bug
-	- fixed a glPush/PopAttrib() bug related to texture binding
-	- flat-shaded, textured lines were broken
-	- fixed a dangling pointer problem in the XMesa code (Chris Burghart)
-	- lighting didn't always produce the correct alpha value
-	- fixed 3DNow! code to not read past end of arrays (Andrew Lewycky)
-
-
-4.0.1 December 17, 2001
-    New:
-	- better sub-pixel sample positions for AA triangles (Ray Tice)
-	- slightly faster blending for (GL_ZERO, GL_ONE) and (GL_ONE, GL_ZERO)
-    Bug fixes:
-	- added missing break statements in glGet*() for multisample cases
-	- fixed uninitialized hash table mutex bug (display lists / texobjs)
-	- fixed bad teximage error check conditional (bug 476846)
-	- fixed demos readtex.c compilation problem on Windows (Karl Schultz)
-	- added missing glGet() query for GL_MAX_TEXTURE_LOD_BIAS_EXT
-	- silence some compiler warnings (gcc 2.96)
-	- enable the #define GL_VERSION_1_3 in GL/gl.h
-	- added GL 1.3 and GLX 1.4 entries to gl_mangle.h and glx_mangle.h
-	- fixed glu.h typedef problem found with MSDev 6.0
-	- build libGL.so with -Bsymbolic (fixes bug found with Chromium)
-	- added missing 'const' to glXGetContextIDEXT() in glxext.h
-	- fixed a few glXGetProcAddress() errors (texture compression, etc)
-	- fixed start index bug in compiled vertex arrays (Keith)
-	- fixed compilation problems in src/SPARC/glapi_sparc.S
-	- fixed triangle strip "parity" bug found in VTK medical1 demo (Keith)
-	- use glXGetProcAddressARB in GLUT to avoid extension linking problems
-	- provoking vertex of flat-shaded, color-index triangles was wrong
-	- fixed a few display list bugs (GLUT walker, molecule, etc) (Keith)
-	- glTexParameter didn't flush the vertex buffer (Ray Tice)
-	- feedback attributes for glDraw/CopyPixels and glBitmap were wrong
-	- fixed bug in normal length caching (ParaView lighting bug)
-	- fixed separate_specular color bug found in Chimera (18 Dec 2001)
-
-
-4.0.2  April 2, 2002
-    New:
-	- New DOS (DJGPP) driver written by Daniel Borca
-	- New driver interface functions for TCL drivers (such as Radeon DRI)
-	- GL_RENDERER string returns "Mesa Offscreen16" or "Mesa Offscreen32"
-	  if using deep color channels
-	- latest GL/glext.h and GL/glxext.h headers from SGI
-    Bug fixes:
-	- GL_BLEND with non-black texture env color wasn't always correct
-	- GL_REPLACE with GL_RGB texture format wasn't always correct (alpha)
-	- glTexEnviv( pname != GL_TEXTURE_ENV_COLOR ) was broken
-	- glReadPixels was sometimes mistakenly clipped by the scissor box
-	- glDraw/ReadPixels didn't catch all the errors that they should have
-	- Fixed 24bpp rendering problem in Windows driver (Karl Schultz)
-	- 16-bit GLchan mode fixes (m_trans_tmp.h, s_triangle.c)
-	- Fixed 1-bit float->int conversion bug in glDrawPixels(GL_DEPTH_COMP)
-	- glColorMask as sometimes effecting glXSwapBuffers()
-	- fixed a potential bug in XMesaGarbageCollect()
-	- N threads rendering into one window didn't work reliably
-	- glCopyPixels didn't work for deep color channels
-	- improved 8 -> 16bit/channel texture image conversion (Gerk Huisma)
-	- glPopAttrib() didn't correctly restore user clip planes
-	- user clip planes failed for some perspective projections (Chromium)
-    Known bugs:
-	- mipmap LOD computation
-
-
-4.0.3  June 25, 2002
-    New:
-	- updated GL/glext.h file (version 15)
-	- corrected MMX blend code (Jose Fonseca)
-	- support for software-based alpha planes in Windows driver
-	- updated GGI driver (Filip Spacek)
-    Bug fixes:
-	- glext.h had wrong values for GL_DOT3_RGB[A]_EXT tokens
-	- OSMesaMakeCurrent() didn't recognize buffer size changes
-	- assorted conformance fixes for 16-bit/channel rendering
-	- texcombine alpha subtraction mode was broken
-	- fixed lighting bug with non-uniform scaling and display lists
-	- fixed bug when deleting shared display lists
-	- disabled SPARC cliptest assembly code (Mesa bug 544665)
-	- fixed a couple Solaris compilation/link problems
-	- blending clipped glDrawPixels didn't always work
-	- glGetTexImage() didn't accept packed pixel types
-	- glPixelMapu[is]v() could explode given too large of pixelmap
-	- glGetTexParameter[if]v() didn't accept GL_TEXTURE_MAX_ANISOTROPY_EXT
-	- glXCopyContext() could lead to segfaults
-	- glCullFace(GL_FRONT_AND_BACK) didn't work (bug 572665)
-    Changes:
-	- lots of C++ (g++) code clean-ups
-	- lots of T&L updates for the Radeon DRI driver
-    Known bugs:
-	- mipmap LOD computation (fixed for Mesa 4.1)
-
-
-4.0.4  October 3, 2002
-    New:
-	- GL_NV_texture_rectangle extension
-	- updated glext.h header (version 17)
-	- updated DOS driver (Daniel Borca)
-	- updated BeOS R5 driver (Philippe Houdoin)
-	- added GL_IBM_texture_mirror_repeat
-	- glxinfo now takes -l option to print interesting OpenGL limits info
-	- GL_MESA_ycbcr_texture extension
-	- GL_APPLE_client_storage extension (for some DRI drivers only)
-	- GL_MESA_pack_invert extension
-    Bug fixes:
-	- fixed GL_LINEAR fog bug by adding clamping
-	- fixed FP exceptions found using Alpha CPU
-	- 3dfx MESA_GLX_FX=window (render to window) didn't work
-	- fixed memory leak in wglCreateContest (Karl Schultz)
-	- define GLAPIENTRY and GLAPI if undefined in glu.h
-	- wglGetProcAddress didn't handle all API functions
-	- when testing for OpenGL 1.2 vs 1.3, check for GL_ARB_texture_cube_map
-	- removed GL_MAX_CONVOLUTION_WIDTH/HEIGHT from glGetInteger/Float/etc()
-	- error checking in compressed tex image functions had some glitches
-	- fixed AIX compile problem in src/config.c
-	- glGetTexImage was using pixel unpacking instead of packing params
-	- auto-mipmap generation for cube maps was incorrect
-    Changes:
-	- max texture units reduced to six to accomodate texture rectangles
-	- removed unfinished GL_MESA_sprite_point extension code
-
-
-4.1  October 29, 2002
-    New:
-	- GL_NV_vertex_program extension
-	- GL_NV_vertex_program1_1 extension
-	- GL_ARB_window_pos extension
-	- GL_ARB_depth_texture extension
-	- GL_ARB_shadow extension
-	- GL_ARB_shadow_ambient extension
-	- GL_EXT_shadow_funcs extension
-	- GL_ARB_point_parameters extension
-	- GL_ARB_texture_env_crossbar
-	- GL_NV_point_sprite extension
-	- GL_NV_texture_rectangle extension
-	- GL_EXT_multi_draw_arrays extension
-	- GL_EXT_stencil_two_side extension
-	- GLX_SGIX_fbconfig and GLX_SGIX_pbuffer extensions
-	- GL_ATI_texture_mirror_once extension (Ian Romanick)
-	- massive overhaul/simplification of software rasterizer module,
-	  many contributions from Klaus Niederkrueger
-	- faster software texturing in some cases (i.e. trilinear filtering)
-	- new OSMesaGetProcAddress() function
-	- more blend modes implemented with MMX code (Jose Fonseca)
-	- added glutGetProcAddress() to GLUT
-	- added GLUT_FPS env var to compute frames/second in glutSwapBuffers()
-	- pbinfo and pbdemo PBuffer programs
-	- glxinfo -v prints transprent pixel info (Gerd Sussner)
-    Bug fixes:
-	- better mipmap LOD computation (prevents excessive blurriness)
-	- OSMesaMakeCurrent() didn't recognize buffer size changes
-	- assorted conformance fixes for 16-bit/channel rendering
-	- texcombine alpha subtraction mode was broken
-	- fixed some blend problems when GLchan==GLfloat (Gerk Huisma)
-	- clamp colors to [0,inf] in OSMesa if GLchan==GLfloat (Gerk Huisma)
-	- fixed divide by zero error in NURBS tessellator (Jon Perry)
-	- fixed GL_LINEAR fog bug by adding clamping
-	- fixed FP exceptions found using Alpha CPU
-	- 3dfx/glide driver render-to-window feature was broken
-	- added missing GLX_TRANSPARENT_RGB token to glx.h
-	- fixed error checking related to paletted textures
-	- fixed reference count error in glDeleteTextures (Randy Fayan)
-    Changes:
-	- New spec file and Python code to generate some GL dispatch files
-	- Glide driver defaults to "no" with autoconf/automake
-	- updated demos/stex3d with new options
-
-
-5.0  November 13, 2002
-    New:
-	- OpenGL 1.4 support (glGetString(GL_VERSION) returns "1.4")
-	- removed some overlooked debugging code
-	- glxinfo updated to support GLX_ARB_multisample
-	- GLUT now support GLX_ARB_multisample
-	- updated DOS driver (Daniel Borca)
-    Bug fixes:
-	- GL_POINT and GL_LINE-mode polygons didn't obey cull state
-	- fixed potential bug in _mesa_align_malloc/calloc()
-	- fixed missing triangle bug when running vertex programs
-	- fixed a few HPUX compilation problems
-	- FX (Glide) driver didn't compile
-	- setting GL_TEXTURE_BORDER_COLOR with glTexParameteriv() didn't work
-	- a few EXT functions, like glGenTexturesEXT, were no-ops
-	- a few OpenGL 1.4 functions like glFogCoord*, glBlendFuncSeparate,
-	  glMultiDrawArrays and glMultiDrawElements were missing
-	- glGet*(GL_ACTIVE_STENCIL_FACE_EXT) was broken
-	- Pentium 4 Mobile was mistakenly identified as having 3DNow!
-	- fixed one-bit error in point/line fragment Z calculation
-	- fixed potential segfault in fakeglx code
-	- fixed color overflow problem in DOT3 texture env mode
-
-
-5.0.1  March 30, 2003
-    New:
-	- DOS driver updates from Daniel Borca
-	- updated GL/gl_mangle.h file (Bill Hoffman)
-    Bug fixes:
-	- auto mipmap generation for cube maps was broken (bug 641363)
-	- writing/clearing software alpha channels was unreliable
-	- minor compilation fixes for OS/2 (Evgeny Kotsuba)
-	- fixed some bad assertions found with shadowtex demo
-	- fixed error checking bug in glCopyTexSubImage2D (bug 659020)
-	- glRotate(angle, -x, 0, 0) was incorrect (bug 659677)
-	- fixed potential segfault in texture object validation (bug 659012)
-	- fixed some bogus code in _mesa_test_os_sse_exception_support (Linus)
-	- fix fog stride bug in tnl code for h/w drivers (Michel Danzer)
-	- fixed glActiveTexture / glMatrixMode(GL_TEXTURE) bug (#669080)
-	- glGet(GL_CURRENT_SECONDARY_COLOR) should return 4 values, not 3
-	- fixed compilation problem on Solaris7/x86 (bug 536406)
-	- fixed prefetch bug in 3DNow! code (Felix Kuhling)
-	- fixed NeXT build problem (FABSF macro)
-	- glDrawPixels Z values when glPixelZoom!=1 were invalid (bug 687811)
-	- zoomed glDraw/CopyPixels with clipping sometimes failed (bug 689964)
-	- AA line and triangle Z values are now rounded, not truncated
-	- fixed color interpolation bug when GLchan==GLfloat (bug 694461)
-	- glArePrograms/TexturesResident() wasn't 100% correct (Jose Fonseca)
-	- fixed a minor GL_COLOR_MATERIAL bug
-	- NV vertex program EXP instruction was broken
-	- glColorMask misbehaved with X window / pixmap rendering
-	- fix autoconf/libtool GLU C++ linker problem on Linux (a total hack)
-	- attempt to fix GGI compilation problem when MesaDemos not present
-	- NV vertex program ARL-relative fetches didn't work
-    Changes:
-	- use glPolygonOffset in gloss demo to avoid z-fighting artifacts
-	- updated winpos and pointblast demos to use ARB extensions
-	- disable SPARC normal transformation code (bug 673938)
-	- GLU fixes for OS/2 (Evgeny Kotsuba)
-
-
-5.0.2  September 5, 2003
-    Bug fixes:
-	- fixed texgen problem causing texcoord's Q to be zero (stex3d)
-	- default GL_TEXTURE_COMPARE_MODE_ARB was wrong
-	- GL_CURRENT_MATRIX_NV query was wrong
-	- GL_CURRENT_MATRIX_STACK_DEPTH_NV query was off by one
-	- GL_LIST_MODE query wasn't correct
-	- GL_FOG_COORDINATE_SOURCE_EXT query wasn't supported
-	- GL_SECONDARY_COLOR_ARRAY_SIZE_EXT query returned wrong value
-	- blended, wide lines didn't always work correctly (bug 711595)
-	- glVertexAttrib4svNV w component was always 1
-	- fixed bug in GL_IBM_rasterpos_clip (missing return)
-	- GL_DEPTH_TEXTURE_MODE = GL_ALPHA didn't work correctly
-	- a few Solaris compilation fixes
-	- fixed glClear() problem for DRI drivers (non-existant stencil, etc)
-	- fixed int/REAL mixup in GLU NURBS curve evaluator (Eric Cazeaux)
-	- fixed delete [] bug in SI GLU (bug 721765) (Diego Santa Cruz)
-	- glFog() didn't clamp fog colors
-	- fixed bad float/int conversion for GL_TEXTURE_PRIORITY in the
-	  gl[Get]TexParameteri[v] functions
-	- fixed invalid memory references in glTexGen functions (bug 781602)
-	- integer-valued color arrays weren't handled correctly
-	- glDrawPixels(GL_DEPTH_COMPONENT) with glPixelZoom didn't work
-	- GL_EXT_texture_lod_bias is part of 1.4, overlooked in 5.0.1
-    Changes:
-	- build GLUT with -fexceptions so C++ apps propogate exceptions
-
-
-5.1  December 17, 2003
-    New:
-	- reorganized directory tree
-	- GL_ARB_vertex/fragment_program extensions (Michal Krol & Karl Rasche)
-	- GL_ATI_texture_env_combine3 extension (Ian Romanick)
-	- GL_SGI_texture_color_table extension (Eric Plante)
-	- GL_NV_fragment_program extension
-	- GL_NV_light_max_exponent extension
-	- GL_EXT_texture_rectangle (identical to GL_NV_texture_rectangle)
-	- GL_ARB_occlusion_query extension
-	- GL_ARB_point_sprite extension
-	- GL_ARB_texture_non_power_of_two extension
-	- GL_IBM_multimode_draw_arrays extension
-	- GL_EXT_texture_mirror_clamp extension (Ian Romanick)
-	- GL_ARB_vertex_buffer_object extension
-	- new X86 feature detection code (Petr Sebor)
-	- less memory used for display lists and vertex buffers
-	- demo of per-pixel lighting with a fragment program (demos/fplight.c)
-	- new version (18) of glext.h header
-	- new spriteblast.c demo of GL_ARB_point_sprite
-	- faster glDrawPixels in X11 driver in some cases (see RELNOTES-5.1)
-	- faster glCopyPixels in X11 driver in some cases (see RELNOTES-5.1)
-    Bug fixes:
-	- really enable OpenGL 1.4 features in DOS driver.
-	- fixed issues in glDrawPixels and glCopyPixels for very wide images
-	- glPixelMapf/ui/usv()'s size parameter is GLsizei, not GLint
-	- fixed some texgen bugs reported by Daniel Borca
-	- fixed wglMakeCurrent(NULL, NULL) bug (#835861)
-	- fixed glTexSubImage3D z-offset bug (Cedric Gautier)
-	- fixed RGBA blend enable bug (Ville Syrjala)
-	- glAccum is supposed to be a no-op in selection/feedback mode
-	- fixed texgen bug #597589 (John Popplewell)
-    Changes:
-	- dropped API trace feature (src/Trace/)
-	- documentation overhaul.  merged with website content.  more html.
-	- glxgears.c demo updated to use GLX swap rate extensions
-	- glTexImage1/2/3D now allows width/height/depth = 0
-	- disable SPARC asm code on Linux (bug 852204)
-
-
-6.0  January 16, 2004
-    New:
-	- full OpenGL 1.5 support
-	- updated GL/glext.h file to version 21
-    Changes:
-	- changed max framebuffer size to 4Kx4K (MAX_WIDTH/HEIGHT in config.h)
-    Bug fixes:
-	- fixed bug in UNCLAMPED_FLOAT_TO_UBYTE macro; solves a color
-	  clamping issue
-	- updated suno5-gcc configs
-	- glColor3 functions sometimes resulted in undefined alpha values
-	- fixed FP divide by zero error seen on VMS with xlockmore, others
-	- fixed vertex/fragment program debug problem (bug 873011)
-	- building on AIX with gcc works now
-	- glDeleteProgramsARB failed for ARB fragment programs (bug 876160)
-	- glDrawRangeElements tried to modify potentially read-only storage
-	- updated files for building on Windows
-
-
-6.0.1  April 2, 2004
-    New:
-	- upgraded glext.h to version 22
-	- new build targets (Dan Schikore)
-	- new linux-x86-opteron build target (Heath Feather)
-    Bug fixes:
-	- glBindProgramARB didn't update all necessary state
-	- fixed build problems on OpenBSD
-	- omit CVS directories from tarballs
-	- glGetTexImage(GL_COLOR_INDEX) was broken
-	- fixed an infinite loop in t&l module
-	- silenced some valgrind warnings about using unitialized memory
-	- fixed some compilation/link glitches on IRIX (Mike Stephens)
-	- glBindProgram wasn't getting compiled into display lists
-	- GLX_FBCONFIG_ID wasn't recognized in glXChooseFBConfig() (bug 888079)
-	- two-sided lighting and vertex program didn't work (bug 887330)
-	- stores to program parameter registers in vertex state programs
-	  didn't work.
-	- fixed glOrtho bug found with gcc 3.2.2 (RH9)
-	- glXCreateWindow() wasn't fully implemented (bug 890894)
-	- generic vertex attribute arrays didn't work in display lists
-	- vertex buffer objects' default usage and access fields were wrong
-	- glDrawArrays with start!=0 was broken
-	- fragment program PK2H, UP2H, UP4B and UP4UB instructions were broken
-	- linux-osmesa16-static config didn't work
-	- fixed a few color index rendering problems (bug 910687)
-	- glInterleavedArrays didn't respect GL_CLIENT_ACTIVE_TEXTURE
-	- OSMesa RGB and BGR modes were broken
-	- glProgramStringARB mistakenly required a null-terminated string
-	- fragment program XPD instruction was incorrect
-	- glGetMaterial() didn't work reliably
-	- ARB_fragment_program KIL instruction was incorrect
-
-
-6.1  August 18, 2004
-    New:
-	- Revamped Makefile system
-	- glXUseRotatedXFont() utility (see xdemos/xuserotfont.c)
-	- internal driver interface changes related to texture object
-	  allocation, vertex/fragment programs, BlendEquationSeparate, etc.
-	- option to walk triangle edges with double-precision floats
-	  (Justin Novosad of Discreet) (see config.h file)
-	- support for AUX buffers in software GLX driver
-	- updated glext.h to version 24 and glxext.h to version 6
-	- new MESA_GLX_FORCE_ALPHA and MESA_GLX_DEPTH_BITS env vars
-	- updated BeOS support (Philippe Houdoin)
-    Changes:
-	- fragment fog interpolation is perspective corrected now
-	- new glTexImage code, much cleaner, may be a bit faster
-    Bug fixes:
-	- glArrayElement in display lists didn't handle generic vertex attribs
-	- glFogCoord didn't always work properly
-	- ARB_fragment_program fog options didn't work
-	- frag prog TEX instruction no longer incorrectly divides s,t,r by q
-	- ARB frag prog TEX and TEXP instructions now use LOD=0
-	- glTexEnviv in display lists didn't work
-	- glRasterPos didn't do texgen or apply texture matrix
-	- GL_DOUBLE-valued vertex arrays were broken in some cases
-	- fixed texture rectangle edge/border sampling bugs
-	- sampling an incomplete texture in a fragment program would segfault
-	- glTexImage was missing a few error checks
-	- fixed some minor glGetTexParameter glitches
-	- GL_INTENSITY was mistakenly accepted as a <format> to glTexImage
-	- fragment program writes to RC/HC register were broken
-	- fixed a few glitches in GL_HP_occlusion_test extension
-	- glBeginQueryARB and glEndQueryARB didn't work inside display lists
-	- vertex program state references were broken
-	- fixed triangle color interpolation bug on AIX (Shane Blackett)
-	- fixed a number of minor memory leaks (bug #1002030)
-
-
-6.2  October 2, 2004
-    New:
-	- enabled GL_ARB_texture_rectangle (same as GL_NV_texture_rectangle)
-	- updated Doxygen support (Jose Fonseca)
-    Changes:
-	- some GGI driver updates (Christoph Egger, bug 1025977)
-    Bug fixes:
-	- Omit GL_ARB_texture_non_power_of_two from list of OpenGL 1.5 features
-	- fixed a few compilation issues on IRIX
-	- fixed a matrix classification bug (reported by Wes Bethel)
-	- we weren't reseting the vertex/fragment program error state
-	  before parsing (Dave Reveman)
-	- adjust texcoords for sampling texture rectangles (Dave Reveman)
-	- glGet*(GL_MAX_VERTEX_ATTRIBS_ARB) wasn't implemented
-	- repeated calls to glDeleteTexture(t) could lead to a crash
-	- fixed potential ref count bugs in VBOs and vertex/fragment programs
-	- spriteblast demo didn't handle window size changes correctly
-	- glTexSubImage didn't handle pixels=NULL correctly for PBOs
-	- fixed color index mode glDrawPixels bug (Karl Schultz)
-
-
-6.2.1  December 9, 2004
-    Bug fixes:
-	- don't apply regular fog or color sum when using a fragment program
-	- glProgramEnvParameter4fARB always generated an error on
-	  GL_FRAGMENT_PROGRAM_ARB (fdo bug 1645)
-	- glVertexAttrib3svNV and glVertexAttrib3svARB were broken
-	- fixed width/height mix-up in glSeparableFilter2D()
-	- fixed regression in glCopyPixels + convolution
-	- glReadPixels from a clipped front color buffer didn't always work
-	- glTexImage didn't accept GL_RED/GREEN/BLUE as the format
-	- Attempting queries/accesses of VBO 0 weren't detected as errors
-	- paletted textures failed if the palette had fewer than 256 entries
-    Changes:
-	- fixed a bunch of compiler warnings found with gcc 3.4
-	- bug reports should to go bugzilla.freedesktop.org
-
-
-6.3  July 20, 2005
-    New:
-	- GL_EXT_framebuffer_object extension
-	- GL_ARB_draw_buffers extension
-	- GL_ARB_pixel_buffer_object extension
-	- GL_OES_read_format extension (Ian Romanick)
-	- DirectFB driver (Claudio Ciccani)
-	- x86_64 vertex transformation code (Mikko T.)
-	- Updated GL/glext.h to version 29
-    Changes:
-	- added -stereo option for glxgears demo (Jacek Rosik)
-	- updated the PBuffer demo code in xdemos/ directory
-	- glDeleteTextures/Programs/Buffers() now makes the object ID
-	  available for immediate re-use
-	- assorted 64-bit clean-ups fixes (x86_64 and Win64)
-	- lots of internal changes for GL_EXT_framebuffer_object
-    Bug fixes:
-	- some functions didn't support PBO functionality
-	- glGetTexImage didn't convert color index images to RGBA as required
-	- fragment program texcoords were sometimes wrong for points and lines
-	- fixed problem with negative dot product in arbfplight, fplight demos
-	- fixed bug in perspective correction of antialiased, textured lines
-	- querying GL_POST_CONVOLUTION_ALPHA_BIAS_EXT returned wrong value
-	- fixed a couple per-pixel fog bugs (Soju Matsumoto)
-	- glGetBooleanv(GL_FRAGMENT_PROGRAM_BINDING_NV) was broken
-	- fixed float parsing bug in ARB frag/vert programs (bug 2520)
-	- XMesaGetDepthBuffer() returned incorrect value for bytesPerValue
-	- GL_COLOR_MATERIAL with glColor3 didn't properly set diffuse alpha
-	- glXChooseFBConfig() crashed if attribList pointer was NULL
-	- program state.light[n].spot.direction.w was wrong value (bug 3083)
-	- fragment program fog option required glEnable(GL_FOG) - wrong.
-	- glColorTable() could produce a Mesa implementation error (bug 3135)
-	- RasterPos could get corrupted by color index rendering path
-	- Removed bad XTranslateCoordinates call when rendering to Pixmaps
-	- glPopAttrib() didn't properly restore GL_TEXTURE_GEN enable state
-	- fixed a few Darwin compilation problems
-
-
-6.3.1
-    This was an intermediate release for X.org which wasn't otherwise released.
-
-
-6.3.2  August 19, 2005
-    New:
-	- The distribution now includes the DRI drivers and GLX code
-    Changes:
-	- Made the DRI "new" driver interface standard, remove old code
-    Bug fixes:
-	- GL_ARB_vertex/fragment_shader were mistakenly listed in the
-	  extensions string
-	- negative relative addressing in vertex programs was broken
-	- update/fix SPARC assembly code for vertex transformation
-	- fixed memory leak when freeing GLX drawables/renderbuffers
-	- fixed display list memory leak
-	- the GL_PIXEL_MAP_I_TO_I table is now floating point, not integer
-	- wglGetProcAddress() didn't handle wgl-functions
-	- fixed glxext.h cross-compile issue (Colin Harrison)
-	- assorted DRI driver fixes
-
-
-6.4  October 24, 2005
-    New:
-	- Added a fast XOR line drawing function in Xlib driver
-	- Added support for GL_ARB_texture_mirrored_repeat to savage
-	  driver (supported only on Savage4 hardware).
-    Changes:
-	- Mesa now packaged in three parts: Library, Demos and GLUT
-    Bug fixes:
-	- GLX_X_RENDERABLE token wasn't accepted by glXChooseFBConfig
-	- Some files were present multiple times in the 6.3.2 tarballs
-	- r200_vtxtmp_x86.S file was missing from 6.3.2 tarball (bug 4207)
-	- glxgears_fbconfig demo didn't work (bug 4237)
-	- fixed bug when bilinear sampling 2d textures with borders
-	- glXCreatePbuffer() could segfault instead of returning 0 (bug 4235)
-	- fixed undefined frexp and rand in X.org libGLcore.a (bug 4242)
-	- fixed a few problems with proxy color tables (bug 4270)
-	- fixed precision problem in Z clearing (bug 4395)
-	- glBitmap, glDraw/CopyPixels mistakenly generated selection hits
-	- fixed potential segfault caused by reading pixels outside
-	  of renderbuffer bounds
-	- glGetTexLevelParameter didn't accept GL_TEXTURE_DEPTH_SIZE_ARB
-	- fixed memory corruption bug involving software alpha buffers
-	- glReadPixels clipped by window bounds was sometimes broken
-	- glDraw/CopyPixels of stencil data ignored the stencil write mask
-	- glReadPixels from a texture bound to a framebuffer object didn't work
-	- glIsRender/FramebufferEXT weren't totally correct
-	- fixed a number of point size attenuation/fade bugs
-	- fixed glFogCoord bug 4729
-	- GLX encoding for transpose matrix functions was broken
-	- fixed broken fragment program KIL and SWZ instructions
-	- fragment programs that wrote result.depth.z didn't work
-
-
-6.4.1  November 30, 2005
-    Bug fixes:
-	- redefining a vertex program string didn't take effect in TNL module
-	- fixed occasional segfault upon vertex/fragment parsing error
-	- vertex program LIT instruction didn't handle 0^0=1 correctly
-	- fragment program fog option didn't work with glDrawPixels, glBitmap
-	- USE_MGL_NAMESPACE didn't work for x86-64
-	- OSMesa demos were missing from previous release tarballs
-	- fixed problem with float->ushort conversion in glClear (bug 4992)
-	- popping of GL_EYE_PLANE texgen state was broken (bug 4996)
-	- popping of GL_SPOT_DIRECTION light state was broken (bug 5005)
-	- fixed occasional triangle color interpolation problem on VMS
-	- work around invalid free() call (bug 5131)
-	- fixed BSD X server compilation problem by including stdint.h
-
-
-6.4.2  February 2, 2006
-    New:
-	- added OSMesaColorClamp() function/feature
-	- added wglGetExtensionStringARB() function
-    Bug fixes:
-	- fixed some problems when building on Windows
-	- GLw header files weren't installed by installmesa script (bug 5396)
-	- GL/glfbdev.h file was missing from tarballs
-	- fixed TNL initialization bug which could lead to crash (bug 5791)
-
-
-6.5  March 31, 2006
-    New:
-	- OpenGL Shading Language support through GL_ARB_shader_objects,
-	  GL_ARB_shading_language_100, GL_ARB_vertex_shader and
-	  GL_ARB_fragment_shader (done by Michal Krol)
-	- GL_EXT_packed_depth_stencil extension
-	- GL_EXT_timer_query extension
-	- GL_EXT_framebuffer_blit extension
-	- GL_ARB_half_float_pixel
-	- reflect demo improved to support multiple windows
-	- singlebuffer demo (shows no/little-flicker single-buffered rendering)
-	- r200: enable GL_ARB_texture_env_crossbar, separate the texture
-	  sampling unit bits from the texture env combine enable bits
-	- r200: add support for GL_ATI_fragment_shader
-	- added fast XOR-mode line drawing optimization
-	- radeon: add support for all 3 tmus, GL_ARB_texture_cube_map
-	  and GL_EXT_fog_coord
-	- MESA_GLX_ALPHA_BITS env var for xlib driver
-	- many DRI driver updates (including screen rotation support
-	  for the Intel DRI driver)
-    Changes:
-	- removed GL_HP_occlusion_test (use GL_ARB_occlusion_query instead)
-	- removed GL_SGIX/SGIS_pixel_texture extensions
-    Bug fixes:
-	- fixed glxcontextmodes.c datatype problem (bug 5835)
-	- fixed aix-gcc build/install bugs (bug 5874)
-	- fixed some bugs in texture env program generation
-	- glXCopyContext() didn't handle texture object bindings properly
-	- glXCopyContext() didn't copy all lighting state
-	- fixed FreeBSD config (Pedro Giffuni)
-	- fixed some minor framebuffer object bugs
-	- replaced dprintf() with _glu_printf() in GLU (bug 6244)
-	- fixed a number of thread safety bugs/regressions
-	- fixed a number of GLU tesselator bugs (John Shell, bug 6339)
-	- paletted texturing was broken w/ floating point palettes (K. Schultz)
-	- lots of assorted framebuffer object bug fixes
-
-6.5.1  August 31, 2006
-    New:
-	- Intel i965 DRI driver
-	- GL_APPLE_vertex_array_object extension (Ian Romanick)
-	- GL_EXT_texture_sRGB extension
-	- GL_EXT_gpu_program_parameters (Ian Romanick)
-	- "engine" demo
-	- updated fbdev driver and GLUT for fbdev (Sean D'Epagnier)
-	- many updates to the DRI drivers
-    Changes:
-	- The glVertexAttribARB functions no longer alias the conventional
-	  vertex attributes.
-	- glxinfo program prints more info with -l option
-	- GL_FRAGMENT_PROGRAM_NV and GL_FRAGMENT_PROGRAM_ARB are now
-	  compatible, in terms of glBindProgramARB()
-    Bug fixes:
-	- fixed broken texture border handling for depth textures (bug 6498)
-	- removed the test for duplicated framebuffer attachments, per
-	  version 117 of the GL_EXT_framebuffer_object specification
-	- fixed a few render-to-texture bugs, including render to depth texture
-	- clipping of lines against user-defined clip planes was broken (6512)
-	- assembly language dispatch for SPARC was broken (bug 6484)
-	- assorted compilation fixes on various Unix platforms (Dan Schikore)
-	- glPopAttrib could restore an invalid value for GL_DRAW_BUFFER
-	- assorted minor fixes for 16 and 32 bit/channel modes
-	- fixed assorted bugs in texture compression paths
-	- fixed indirect rendering vertex array crashes (bug 6863)
-	- glDrawPixels GL_INDEX_OFFSET didn't always work
-	- fixed convolution memory leak (bug 7077)
-	- rectangular depth textures didn't work
-	- invalid mode to glBegin didn't generate an error (bug 7142)
-	- 'normalized' parameter to glVertexAttribPointerARB didn't work
-	- disable bogus GLX_SGI_video_sync extension in xlib driver
-	- fixed R128 driver locking bug (Martijn van Oosterhout)
-	- using evaluators with vertex programs caused crashes (bug 7564)
-	- fragment.position wasn't set correctly for point/line primitives
-	- fixed parser bug for scalar sources for GL_NV_fragment_program
-	- max fragment program length was incorrectly 128, now 1024
-	- writes to result.depth in fragment programs weren't clamped to [0,1]
-	- fixed potential dangling pointer bug in glBindProgram()
-	- fixed some memory leaks (and potential crashes) in Xlib driver
+
+
+Mesa Version History
+====================
+
+1.0 beta   February 1995
+    Initial release
+
+1.1 beta   March 4, 1995
+    Changes:
+	faster point and line drawing (2x faster)
+	more systems supported, better Makefiles
+	Renamed lib*.a files to avoid collisions
+	many small bug fixes
+    New:
+	pseudo-GLX functions added
+	new implementation of evaluators (eval2.c)
+	GLUT support
+
+1.1.1 beta   March 7, 1995
+    Changes:
+	Reverted from eval2.c to eval.c due to FPE on Linux
+	more speed improvements
+	more Makefile changes
+
+1.1.2 beta   March 14, 1995
+    New:
+	implementation of SGI's blending extensions
+	glXUseXFont implemented
+	added MESA_DEBUG environment variable support
+    Changes:
+	Using eval2.c again
+	more FPE-prevention checks (0-length normals are OK)
+	a few small bug fixes
+	much faster pixel logic ops!
+	faster transformation arithmetic
+
+1.1.3 beta   March 31, 1995
+    New:
+	gluScaleImage() and gluBuild2DMipMaps() implemented
+	Mesa widgets for Xt/Motif
+	blendEXT demos
+	added environment variables for selecting visuals
+    Changes:
+	almost all GLUT demos work correctly now
+	faster X device driver functions
+	more bug fixes
+
+1.1.4 beta   April 20, 1995
+    Bug fixes:
+	- missing #define SEEK_SET in src-tk/image.c
+	- compile glShadeModel into display lists
+	- fixed pow() domain error in src/light.c
+	- fixed "flickering bitmaps" in double buffer mode
+	- fixed tk.h and aux.h for C++
+	- state of LIGHT_MODEL_LOCAL_VIEWER was inverted
+    New features:
+	- MUCH, MUCH nicer dithering in 8-bit RGB mode
+	- updated widgets and widget demos
+	- Implemented GLXPixmap functions
+	- Added GLU 1.1 and GLX 1.1 functions
+	- Changed the X/Mesa interface API, more versatile
+	- Implemented gluPartialDisk()
+
+1.2   May 22, 1995
+    Bug fixes:
+	- IRIX 4.x makefile problem
+	- modified tk to share root colormap as needed
+	- gluLookAt normalization problem
+	- suppress Expose, NoExpose events in swapbuffers
+	- glBitmap() and glDrawPixels() clipping
+    New features:
+	- GL_BLEND, GL_MODULATE, GL_DECAL, and GL_REPLACE_EXT texture
+	      modes implemented
+	- texture maps stored more efficiently
+	- texture maps can be compiled into display lists
+	- Bogdan Sikorski's GLU polygon tesselation code
+	- Linas Vepstas's sweep and extrusion library
+	- glXCreateContext()'s shareList parameter works as it's supposed to.
+	      XMesaCreateContext() updated to accept a shareList parameter too.
+	- Mesa can be compiled with real OpenGL .h files
+	- MESA_BACK_BUFFER environment variable
+	- better GLX error checking
+
+1.2.1   June 22, 1995
+    Bug fixes:
+	- X/Mesa double buffer window resize crash
+	- widgets now pass PointerMotion events
+	- X/Mesa incorrect default clear color and drawing color
+	- more robust X MIT-SHM support in X/Mesa
+	- glTexImage( format=GL_LUMINANCE ) didn't work
+	- GL_LINE mode polygons with line width > 1.0 could cause a crash
+	- numerous feedback bugs
+	- glReadPixels() from depth buffer was wrong
+	- error prone depth and stencil buffer allocation
+    New features:
+	- Preliminary Microsoft Windows driver
+	- Implemented a number of missing functions: glEvalCoord[12][df]v(),
+	    glGet...(), etc.
+	- Added a few missing symbols to gl.h and glu.h
+	- Faster rendering of smooth-shaded, RGBA, depth-buffered polygons.
+	- Faster rendering of lines when width=2.0
+	- Stencil-related functions now work in display lists
+    Changes:
+	- renamed aux.h as glaux.h (MS-DOS names can't start with aux)
+	- most filenames are in 8.3 format to accomodate MS-DOS
+	- use GLubytes to store arrays of colors instead of GLints
+
+1.2.2   August 2, 1995
+    New features:
+	- texture mapped points and lines
+	- NURBS! (but not 100% complete)
+	- viewports may safely extend beyond window boundaries
+	- MESA_PRIVATE_CMAP environment variable
+	- Grayscale X display support
+	- two new demos:  demos/gears.c and demos/shadow.c
+	- MachTen for Macintosh configuration
+    Bug fixes:
+	- glGet*(GL_DEPTH_BITS) returned bytes, not bits
+	- point, line, and bitmap rasterization suffered from roundoff errors
+	- fixed a division by zero error in line clippping
+	- occasional wrong default background color really fixed!
+	- glDepthFunc(GL_ALWAYS) with glDepthMask(GL_FALSE) didn't work
+	- gluBuild2DMipmaps malloc problem fixed
+	- view volume clipping of smooth shaded lines resulted in bad colors
+    Changes:
+	- new visual selection method in glXChooseVisual()
+	- improved GLU quadric functions
+	- call XSync for glFinish and XFlush for glFlush
+	- glVertex() calls now use a function pointer to avoid conditionals
+	- removed contrib directory from Mesa tar file (available on ftp site)
+	- AIX shared library support
+	- Removed GLUenum type as it's not in OpenGL
+
+1.2.3   September 26, 1995
+    New features:
+	- Mesa header files now equivalent to SGI OpenGL headers
+	- Support for HP's Color Recovery dithering displays
+	- Faster vertex transformation
+	- Faster raster operations into X windows under certain conditions
+	- New configurations:  HP w/ shared libs, Ultrix w/ GCC, Data General
+	- 4-bit visuals now supported
+    Bug fixes:
+	- glScissor bug fixed
+	- round-off errors in clipping lines against clip planes fixed
+	- byte swapping between hosts and display servers implemented
+	- glGetError() can be called without a current rendering context
+	- problem with accidentally culled polygons is fixed
+	- fixed some widget compilation problems
+
+1.2.4   November 17, 1995
+    New features:
+	- More speed improvements (lighting, fogging, polygon drawing)
+	- Window system and OS-independent off-screen rendering
+	- Preliminary Fortran bindings
+	- glPolygonOffsetEXT implemented
+	- glColorMask and glIndexMask now fully implemented
+	- glPixelZoom implemented
+	- display lists fully implemented
+	- gamma correction
+	- dithering in 8-bit TrueColor/DirectColor visuals
+    Changes:
+	- Improved device driver interface
+	- tk.h renamed to gltk.h to avoid conflicts with Tcl's Tk
+	- Dithering support moved from core into device driver
+    Bug fixes:
+	- glEnable/Disable( GL_LIGHTING ) didn't always take effect
+	- glReadPixels byte swapping was broken
+	- glMaterial with pname==GL_AMBIENT_AND_DIFFUSE was broken
+	- duplicate glColor4b() prototype in GL/gl.h removed
+	- stripes in wave -ci demo fixed
+	- GL_LINEAR_MIPMAP_NEAREST had wrong value
+	- bugs in HP Color Recovery support fixed
+	- fixed bug when blending lines, points, bitmaps outside of window
+
+1.2.5   November 30, 1995
+    New Features:
+	- updated MS Windows driver
+	- new implementation of StaticGray/GrayScale visual support
+    Bug fixes:
+	- pixelzooming with gamma correction or blending didn't work
+	- HP color recovery visual wasn't being picked by glXChooseVisual
+	- glClear didn't always observe glColorMask changes
+	- olympic and offset demos didn't compile on some Suns
+	- texcoord clamping wasn't correct
+	- a polygon optimization introduced an occasional sampling problem
+
+1.2.6   January 26, 1996
+    New Features:
+	- faster line and polygon rendering under certain conditions.  See
+	  Performance Tips 9 and 10 in README
+	- profiling
+	- lighting is a bit faster
+	- better perspective corrected texture mapping
+	- Amiga AmiWin (X11) support
+	- preliminary Linux SVGA driver
+    Changes:
+	- now using a 16-bit depth buffer, faster, smaller
+	- GL_NORMALIZE is disabled by default
+    Bug fixes:
+	- projective texture mapping
+	- fixed a memory leak in the context destroy function
+	- GL_POLYGON with less than 3 vertices caused a crash
+	- glGet*() returned wrong result for GL_INDEX_MODE
+	- reading pixels from an unmapped X window caused a BadMatch error
+
+1.2.7   March 5, 1996
+    New:
+	- faster lighting
+	- faster 16-bit TrueColor rendering on Linux
+	- faster 32-bit TrueColor rendering on Linux, HP, IBM
+	- non-depth-buffered XImage polygons are faster
+	- vertex array extension
+	- software alpha planes
+	- updated Macintosh driver
+	- new NeXT driver
+	- GLU quadric functions generate texture coordinates
+	- reflect.c demo - reflective, textured surface demo
+    Changes:
+	- gamma correction code moved into the X driver for better performance
+    Bug fixes:
+	- multiple glClipPlane()'s didn't work reliably
+	- glPolygonMode() didn't always work
+	- glCullFace( GL_FRONT_AND_BACK ) didn't work
+	- texture mapping with gamma correction was buggy
+	- floating point exceptions in texture coordinate interpolation
+	- XImage byte swapping didn't always work
+	- polygon edge flags weren't always used correctly
+
+1.2.8   May 22, 1996
+    New:
+	- overlay planes on X servers with the SERVER_OVERLAY_VISUALS property
+	- better monochrome output
+	- more IRIX 6.x configurations
+	- more robust RGB mode color allocation
+	- added MESA_XSYNC environment variable
+	- GLX_MESA_pixmap_colormap and GLX_EXT_visual_info extensions
+	- GL_MESA_window_pos extension
+	- faster glReadPixels/glDrawPixels for GL_DEPTH and GL_UNSIGNED_SHORT
+	  and GL_UNSIGNED_INT
+        - driver for prototype Cirrus Mondello 3-D board
+	- updated AmigaDOS driver
+        - a few small speed optimizations in polygon rendering
+    Changes:
+	- internal device driver interface modified to simplify device
+	  driver implementations and to support hardware Z buffers
+	- several changes to the X/Mesa interface (xmesa.h)
+    Bug fixes:
+	- fixed pow(0,0) domain error triggered on some systems
+	- glStencilClear() in a display list caused an infinite loop
+	- glRasterPos*() was sometimes off by +/-0.5 in X and Y
+	- color masking and blending were performed in wrong order
+	- auxSolidCylinder() sometimes drew a wire-frame cylinder
+	- fixed file writing bug in osdemo.c
+	- pixel mapping didn't always work
+	- the GL_GEQUAL stencil func didn't work
+	- the GL_INVERT stencil op didn't work
+	- the stencil write mask didn't work
+	- glPush/PopAttrib() didn't do enough error checking
+	- glIsList() didn't always work correctly
+
+2.0  October 10, 1996
+    New:
+	- Implements OpenGL 1.1 API functions
+	- all texture filtering modes supported (mipmapping)
+	- faster texture mapping, see Performance Tip 11 in README
+	- antialiased RGB points
+	- X support for line and polygon stippling
+	- glDrawBuffer( GL_FRONT_AND_BACK ) works
+	- util/ directory of useful stuff
+	- demos/texobj demo of texture objects
+    Changes:
+	- major internal changes for thread-safeness
+	- new device driver interface
+	- MESA_ALPHA env variable removed
+	- triangle rasterizer replaces polygon rasterizer
+    Bug fixes:
+	- glPopAttrib() bug
+	- glDrawBuffer(GL_NONE) works now
+
+2.1  December 14, 1996
+    New:
+	- VMS support
+	- MS-DOS driver
+	- OpenStep support
+	- updated, combined Windows 95/NT driver
+	- implemented glGetLighti() and glGetTexGen*()
+	- GLX does garbage collection of ancillary buffers
+    Bug fixes:
+	- removed unused _EXT constants from gl.h
+	- fixed polygon offset bugs
+	- Z coordinates of clipped lines were incorrect
+	- glEdgeFlag() in display lists didn't always work
+	- glLight*() in display lists didn't work
+	- fixed X line stipple bugs (Michael Pichler)
+	- glXUseXfonts XFreeFont/XFreeFontInfo bug fixed
+	- fixed a feedback bug
+	- glTexGen*() now transforms GL_EYE_PLANE by inverse modelview matrix
+	- polygons were sometimes culled instead of clipped
+	- triangle rasterizer suffered from float/int overflow exceptions
+	- fixed FP underflow exception in lighting (specular exponent)
+	- glEnable/glDisable of GL_EXT_vertex_array enums didn't work
+	- fixed free(NULL) in GLU tesselator code
+	- using 24-bit color on some X servers resulted in garbage rendering
+	- 32-bit per pixel mode for XFree86 now works
+	- glRotate(a,0,0,0) gave unpredictable results
+	- GL_LINE_STRIP with > 480 vertices had occasional clipping problems
+	- 8-bit TrueColor GLXPixmap rendering incorrectly required a colormap
+	- glMaterial() wasn't ignored when GL_COLOR_MATERIAL was enabled
+	- glEnable(GL_COLOR_MATERIAL) followed by glColor() didn't work right
+	- accumulation buffer was limited to positive values
+	- projective textures didn't work
+	- selection buffer overflows weren't handled correctly
+    Changes:
+	- restored the GL_EXT_polygon_offset extension
+	- slightly faster RGB dithering
+	- the SVGA driver works again
+	- Amiga driver now distributed separately
+	- NeXT driver updated for Mesa 2.x
+
+2.2  March 14, 1997
+    New:
+	- better color selection when dithering
+	- added GL_EXT_texture_object extension
+	- updated MS-DOS driver for DJGPP
+	- added openbsd make configuration
+	- faster dithered flat-shaded triangles
+	- various compilation problems with Motif widgets fixed
+	- gl.h, glx.h and glu.h name mangling option
+	- BeOS driver
+	- 3D texture mapping extension
+	- GL_MESA_resize_buffers extension
+	- morph3d, stex3d and spectex demos
+	- 3Dfx support
+    Bug fixes:
+	- glColorMaterial should finally work right in all respects
+	- linear interpolation of mipmap levels was incorrectly weighted
+	- readpix.c didn't compile on Macintosh
+	- GL_INVERT and related logic ops didn't work right
+	- glTexImage[12]D() didn't check its parameters consistantly
+	- fixed a memory leak in glTexImage[12]D()
+	- kludged around a SunOS 5.x/GCC compiler bug in the feedback code
+	- glReadPixels aborted instead of normally catching some errors
+	- a few 1.1 constants were missing or misnamed in gl.h
+	- glBegin(p); glBegin(q); didn't generate an error
+	- fixed a memory leak in GLX code
+	- clipping of concave polygons could cause a core dump
+	- 1-component alpha texture maps didn't work
+	- fixed a GLU polygon tesselator bug
+	- polygons with colinear vertices were sometimes culled
+	- feedback triangle colors were wrong when using smooth shading
+	- textures with borders didn't work correctly
+	- colors returned in feedback mode were wrong when using lighting
+	- spotlights didn't effect ambient lighting correctly
+	- gluPartialDisk() had a few bugs
+    Changes:
+	- device driver interface expanded to support texture mapping
+	- faster matrix inversion subroutine
+	- commented out #include "wmesa_extend.h" from src/wmesa.c
+	- fixed many compiler warnings in the demo programs
+
+2.3  June 30, 1997
+    New:
+	- Mesa distribution divided into two pieces:  library code and demos
+	- faster vertex transformation, clip testing, lighting
+	- faster line drawing
+	- TrueColor visuals how have dithering (for depths < 24 bits)
+	- added MESA_NO_DITHER environment variable
+	- new device driver function: NearFar(), RenderVB(), RasterSetup()
+	- added LynxOS configuration
+	- added cygnus Win32 configuration
+	- added texcyl.c GLUT demo
+	- added XMesaDitherColor() to X/Mesa interface
+	- new NURBS code from Bogdan Sikorski
+	- added demos/shape.c (non-rectangular X window!)
+    Bug fixes:
+	- glEnable/DisableClientState() were missing from GL/gl.h
+	- GL_SPHERE_MAP texcoord generation didn't work correctly
+	- glXGetConfig() returned wrong number of depth, stencil, accum bits
+	- glDrawPixels feedback/selection didn't examine RasterPos valid bit
+	- black and white were reversed on some monochrome displays
+	- fixed potential image memory leak (wasn't setting reference counter)
+	- glDrawPixels sometimes didn't recognize some GL state changes
+	- gluProject/UnProject() didn't check for divide by zero
+	- stex3d demo called random() and srandom(), not portable
+	- fixed memory leaks in context.c and drawpix.c
+	- fixed NULL dereferencing problem in gl_update_texture_state()
+	- glReadPixels between glBegin/glEnd didn't generate an error.
+	- fixed memory leak in polygon tesselator (Randy Frank)
+	- fixed seg fault bug drawing flat-shaded, depth-tested lines
+	- clipped GL_TRIANGLE_STRIPs sometimes had wrong color when flat-shaded
+	- glBindTexture sometimes didn't work
+	- fixed a bug deep in glXReleaseBuffersMESA()
+	- fog was mistakenly applied to alpha
+	- glPopMatrix didn't set "dirty matrix" flag
+	- glPolygonStipple pattern was sometimes wrong
+	- glClear wasn't disabled during feedback and selection
+	- fixed memory leak in glTexSubImage[123]D
+    Changes:
+	- many library source files reorganized
+	- faster X color allocation, colors also freed when finished with them
+	- new texture sampling function pointer in texture objects
+	- incorporated 3Dfx VooDoo driver v0.16 into main source tree
+	- many 3Dfx driver updates
+	- cygnus Makefiles now included
+	- updated DOS driver
+	- made a few changes to dosmesa.c and wmesa.c (VB->Unclipped)
+	- internally, colors now stored in GLubytes, not GLfixed
+	- optimized changing of GL_SHININESS parameter
+
+2.4  September 18, 1997
+    New:
+	- updated 3Dfx Glide driver
+	- hacks for 3Dfx rendering into an X window or fullscreen
+	- added depth buffer access functions to X/Mesa and OS/Mesa interfaces
+    Bug fixes:
+	- pixel buffer could overflow with long, wide lines
+	- fixed FP underflow problems in lighting
+	- glTexSubImage1D() had an unitialized variable
+	- incomplete texture objects could cause a segfault
+	- glDrawPixels with GL_COMPILE_AND_EXECUTE caused infinite loop
+	- flat-shaded quads in a strip were miscolored if clipped
+	- mipmapped triangle lod computation now works correctly
+	- fixed a few under/overflow bugs in triangle rasterizer
+	- glArrayElement() assigned bad normal if normal array disabled
+	- changed argument to glXReleaseBuffersMESA()
+	- fixed small triangle underflow bugs in tritemp.h (hopefully)
+	- glBindTexture(target, 0) caused a crash
+	- glTexImage[123]D() with NULL image pointer caused crash
+	- glPixelStore parameters are now ignored during display list execution
+	- fixed a two-sided lighting w/ clipping bug (black vertices)
+	- textures with width!=height were sometimes mis-rendered
+	- "weird" projection matrices could cause div by 0, other fp errors
+    Changes:
+	- changed precompiled header symbol from PCH to PC_HEADER
+	- split api.c into api1.c and api2.c
+	- added hash.c source file (but not used yet)
+	- a few Sun and HP configuration file changes
+	- MESA_GLX_FX env var replaces MESA_FX_WINDOW and MESA_FX_FULLSCREEN
+	- fixed a few cygnus build problems (src/Makefile.cygnus, src/wmesa.c)
+
+2.5  November 20, 1997
+    New:
+	- updated 3Dfx driver (v20) for GLQuake
+	- added GL_EXT_paletted_texture extension
+	- added GL_EXT_shared_texture_palette extension
+	- added GL_EXT_point_parameters extension
+	- now including Mark Kilgard's GLUT library v3.6
+	- new GLUT-based demos in gdemos/
+	- added a few more Unix config targets
+	- added Intel X86 assembly language vertex transformation code
+	- 3Dfx/Glide driver for Mesa now recognizes SST_SCREENREFRESH env var
+	- Windows 95 S3 Virge driver
+    Bug fixes:
+	- glCopyTexImage?D would crash due to uninitialized variable
+	- glColor w/ glColorMaterial in a display list caused a bug
+	- fixed several glDrawPixels() and ReadPixels() bugs in 3Dfx driver
+	- glVertex4*() vertices weren't always projected correctly
+	- trying to use mipmapped textured points or lines caused crash
+	- glColor[34][fd]() values now clamped to [0,1] before int conversion
+    Changes:
+	- new device driver functions for texture mapping
+	- hash tables used for display list and texture object lookup
+	- fixed GLX visual handling code to avoid saving redundant visuals
+	- 3Dfx Glide libraries automatically linked to libMesaGL.so
+	- dropped the Cirrus Logic Mondello code since it's obsolete
+	- updated Cygnus Makefiles (Stephane Rehel)
+	- updated Windows MSVC++ Makefiles (Oleg Letsinsky)
+	- procedure for making library files has changed: scripts now take
+	  a major and minor version arguments.  Make-config changed a lot.
+	- new implementation of glTexSubImage2D()
+	- updated widgets-mesa directory to create libMesaGLwM.a (Motif widget)
+	- separate linux-glide and linux-386-glide configurations
+
+2.6  February 12, 1998
+    New:
+	- Windows WGL functions
+	- updated VMS, DOS, Windows, Cygnus, BeOS, Amiga compilation support
+	- v0.22 of 3Dfx Glide driver
+	- more X86 assembly language optimizations
+	- faster blending for some modes
+	- XMesaSetFXmode() to switch between 3Dfx window and full-screen mode 
+	- added preliminary thread support
+	- added GLX_MESA_copy_sub_buffer extension
+	- some clipping optimizations
+    Bug fixes:
+	- fixed shading/material bug when drawing long primitive strips
+	- fixed clipping problem in long primitive strips
+	- fixed clipping bug when using 3Dfx driver
+	- fixed a problem when trying to use X fonts w/ 3Dfx driver
+	- fixed a texture filter bug in 3Dfx/Glide driver
+	- fixed bug in 3Dfx/Glide driver involving depth mask & clearing
+	- glLoadMatrix to set projection matrix confused the 3Dfx driver
+	- non-identity texture matrices didn't work with linux-386 configs
+	- glGenTextures() didn't reserve the returned texture IDs
+	- NULL proxy image sent to glTexImageXD() caused crash
+	- added texture state validation optimization (Henk Kok)
+	- fixed colormap reuse problem when using both RGB and CI windows
+	- 32bpp True/DirectColor X visuals weren't recognized
+	- fixed potential problem in evaluators memory allocation
+	- fixed assorted demo compilation bugs
+    Changes:
+	- replaced old Mesa/windows/ directory with Mesa/WIN32/ directory
+	- converted a few old glaux/gltk demos to GLUT
+	- renamed directories:  demos -> xdemos,  gdemos -> demos
+
+
+3.0  September 17, 1998
+    New:
+	- OpenGL 1.2 API
+	- GL_EXT_abgr pixel format extension
+	- GL_SGIS_texture_edge_clamp extension
+	- GL_SGIS_multitexture extension (to be replaced by GL_ARB_multitex)
+	- GL_EXT_multitexture extension (to be replaced by GL_ARB_multitex)
+	- GL_EXT_rescale_normal extension and renormal.c demo
+	- GLX_SGI_video_sync extension (a no-op)
+	- antialiased lines
+	- glGetTexImage() now implemented
+	- glDraw/Copy/ReadPixels() optimizations
+	- optimized textured triangle code (Marten Stromberg)
+	- more optimization of dithered TrueColor triangles in X driver
+	- Linux GGI driver
+	- updated MGL driver
+    Bug fixes:
+	- lots of assorted compilation fixes
+	- glInitNames didn't write initial hit record
+	- glBitmap didn't always check for invalid raster position
+	- switching between GLX and OSMesa contexts caused a crash
+	- fixed uninitialized variable in Mesa widget code
+	- fixed typo in texture code which caused book/texgen to crash
+	- fixed texture sampling bug when filter=GL_LINEAR and wrap=GL_CLAMP
+	- gluDisk() in POINT or LINE mode sometimes failed
+	- fixed texture + fog bug
+	- GL_COMPILE_AND_EXECUTE mode didn't work reliably
+	- glMultMatrix in projection matrix mode w/ 3Dfx driver could fail
+	- glDrawPixels(color index pixels) weren't converted to RGBA
+	- fixed possible getenv() buffer overflow security bug
+	- glBitmap in feedback mode was offset by xOrig, yOrig params
+	- device driver's DrawPixels hook was never used
+	- glDrawPixels with zoomY!=1 and top/bottom clipping didn't work
+	- glDrawPixels optimized for GL_LUMINANCE, GL_LUMINANCE_ALPHA, GLubyte
+	- fixed MakeCurrent bug in GLwRedrawObjects() in MesaWorkstation.c
+	- glCopyTexSubImage2D() didn't work with 3Dfx driver
+	- lines with width = 2 could cause crash
+	- glClear with scissor rect sometimes cleared whole buffer
+	- glTexSubImage2D( .. GL_COLOR_INDEX .. ) didn't work
+	- glTexImageXD( .. GL_ABGR_EXT .. ) didn't work
+	- computation of inverse modelview matrix sometimes failed
+	- fixed GL_CLAMP mode texture sampling bug
+	- textured line interpolation was somewhat broken
+	- textured triangle interpolation was also somewhat broken
+	- glGet(MODELVIEW/PROJECTION/TEXTURE_MATRIX_STACK_DEPTH) off by one
+	- evaluator state wasn't fully initialized
+	- texture coordinate clipping was buggy
+	- evaluator surfaces could be mis-colored
+	- glAccum(GL_RETURN, s) didn't obey glColorMask() settings
+	- zero area polygons shouldn't be culled if polygon mode is point/line
+	- clipped width and height of glReadPixels was sometimes off by one
+	- blending with alpha = 0 or 1.0 wasn't always exact
+	- reading of pixels from clipped region was buggy
+	- minor tweaking of X visual management in GLX emulator
+	- glPolygonStipple now obeys pixel unpacking parameters
+	- glGetPolygonStipple now obeys pixel packing parameters
+	- interleaved vertex array texture coordinates were broken
+	- query of proxy texture internal format was broken
+	- alpha channel wasn't reliably cleared
+	- fixed divide by zero error in gluScaleImage if dest size = 1 x 1
+    Conformance bug fixes:
+	- GL_SELECTION_BUFFER_POINTER and GL_SELECTION_BUFFER_SIZE were missing
+	- GL_TEXTURE_INTERNAL_FORMAT was missing
+	- glGet*(GL_POLYGON_STIPPLE) was broken
+	- glPush/PopAttrib() didn't save/restore all texture state
+	- glBitmap in feedback mode didn't work
+	- feedback of texture coords didn't always work
+	- glDrawPixels w/ format=GL_DEPTH_COMPONENT, type=GLbyte was broke
+	- glDrawPixels w/ format=GL_DEPTH_COMPONENT, type=GLubyte was broke
+	- glDrawPixels w/ format=GL_STENCIL_INDEX, type=GL_BITMAP was broke
+    Changes:
+	- upgraded GLUT to version 3.7
+	- only GL and GLU library code included in MesaLib.tar.gz
+	- GLUT and all demos now in MesaDemos.tar.gz
+	- glaux and gltk libraries removed
+	- IRIX -n32 and -64 libs go in lib32/ and lib64/ directories
+
+
+3.1 beta 1  November 19, 1998
+    New:
+	- GL_EXT_stencil_wrap extension
+	- GL_INGR_blend_func_separate extension
+	- GL_ARB_multitexture extension
+	- GL_NV_texgen_reflection extension
+	- newly optimized vertex transformation code
+	- updated GLUT 3.7 code
+	- better precision when using 32-bit Z buffer
+	- Allegro DJGPP driver
+    Bug fixes:
+	- glCopyPixels between front/back buffers didn't copy alpha correctly
+	- fixed out-of-bounds memory access in optimized 2-D texture code
+	- glPixelStorei didn't accept GL_PACK/UNPACK_IMAGE_HEIGHT parameter
+	- glGet*() didn't accept GL_MAX_3D_TEXTURE_SIZE parameter
+	- clipping of texture coordinates sometimes had bad R,Q values
+	- GL_CLAMP_TO_EDGE texture sampling was off by 0.5 texels
+	- glEdgeFlagPointer() now takes a GLvoid * instead of GLboolean *
+	- texture was sometimes applied twice with 3Dfx driver
+	- glPush/PopAttrib() fouled up texture object reference counts
+	- glDeleteLists(0, n) caused assertion failure
+	- bilinear texture sampling wasn't accurate enough
+	- glClear w/ glDepthMask(GL_FALSE) didn't work right on 3Dfx
+	- color components were reversed on big endian 32 bpp X visuals
+    Changes:
+	- removed GL_EXT_multitexture extension
+
+
+3.1 beta 2  May 24, 1999
+    New:
+	- multi-textured points and lines (mjk@nvidia.com)
+	- optimized 24bpp X rendering (bernd.paysan@gmx.de)
+	- added allegro support (bernie-t@geocities.com)
+	- cleaned-up Windows-related stuff (Ted Jump)
+	- minor stereo changes (KendallB@scitechsoft.com)
+	- new BeOS driver which implements BGLView class
+	- new Direct3D driver (see src/D3D)
+	- more efficient filled gluCylinder() function
+	- utilities:  util/showbuffer.[ch] and util/glstate.[ch]
+	- fixed some IRIX compiler warnings
+	- added support for building Mesa in XFree86 with
+	    SGI's GLX (kevin@precisioninsight.com)
+    Bug fixes:
+	- a variety of Windows/Mesa bug fixes (mjk@nvidia.com)
+	- packed pixel images weren't unpacked correctly
+	- patches some win32 files in GLUT (mjk@nvidia.com)
+	- glTexImage[123]D() didn't accept internalFormat == GL_COLOR_INDEX
+	- fixed lighting bug in Keith's new shading code
+	- fixed texture segfault seen in Lament screensaver
+	- fixed miscellaneous low-memory bugs
+	- glClear(GL_COLOR_BUFFER_BIT) with RGBA or CI masking was broken
+	- GL_LINEAR sampling of 3D textures was broken
+	- fixed SVR4 'cc' compiler macro problem (dawes@xfree86.org)
+	- added GL_TEXTURE_PRIORITY fix (keithh@netcomuk.co.uk)
+	- fixed wide point and wide line conformance bugs (brianp)
+    Changes:
+	- some device driver changes (see src/dd.h)
+	- new copyright on core Mesa code
+
+
+3.1 beta 3  September 17, 1999
+    New:
+	- optimized glAccum function
+	- optimized 24bpp rendering in XMesa driver
+	- GLU 1.2 polygon tessellator
+    Bug Fixes:
+	- glGetTexLevelParameter wasn't fully implemented
+	- glXUseXFont now handles multi-byte fonts
+	- glIsEnabled(GL_TEXTURE_2D / 3D) returned wrong result
+	- alpha channel of blending points, lines was sometimes incorrect
+    Changes:
+	- New library names:  "libGL" instead of "libMesaGL"
+	- New library numbering: libGL.so.1.2.310
+	- New subdirectories:  docs/ and bin/
+	- New Makefile-system (autoconf,automake,libtool)
+
+
+3.1 final  December 14, 1999
+    New:
+	- added demos/gloss.c
+	- added xdemos/glxdpyinfo.c
+	- added GLX_ARB_get_proc_address extension
+	- rewritten glTexImage code paths (faster, less memory, bug fixes)
+    Bug Fixes:
+	- several vertex array bug fixes
+	- overlapping glCopyPixels with pixel zooming now works
+	- glXUseXFont() bitmaps were vertically shifted by one pixel
+	- glCopyPixels with pixel zooming now works
+
+
+3.2 final  April 24, 2000
+    Bug fixes:
+	- fixed memcpy bugs in span.c
+	- fixed missing glEnd problem in demos/tessdemo.c
+	- fixed bug when clearing 24bpp Ximages
+	- fixed clipping problem found in Unreal Tournament
+	- fixed Loki's "ice bug" and "crazy triangles" seen in Heretic2
+	- fixed Loki's 3dfx RGB vs BGR bug
+	- fixed Loki's 3dfx smooth/flat shading bug in SoF
+    Changes:
+	- updated docs/README file
+	- use bcopy() optimizations on FreeBSD
+	- re-enabled the optimized persp_textured_triangle() function
+
+
+3.2.1  July 19, 2000
+    Bug fixes:
+	- gluBuild2DMipmaps() didn't accept GL_BGRA
+	- Fixed compile/makefile problems on IRIX
+	- fixed segfault in 3dfx driver when using GL selection/feedback
+	- no longer cull very, very tiny triangles
+	- blending w/ drawbuffer==GL_FRONT_BACK caused segfault (sw rendering)
+	- fixed Motif detection code in widgets-mesa/configure.in
+	- glColorMaterial and glMaterial updates to emissive and ambient
+	  didn't always work right
+	- Specular highlights weren't always in the right place
+	- clipped GL_LINE mode polygons had interior lines appear
+	- blend term GL_ONE_MINUS_CONSTANT_ALPHA was broken
+	- GL_NICEST fog didn't always work with flat shading
+	- glRect commands in display lists were sometimes miscolored
+	- Line Z offset didn't always work
+	- fixed texgen normal vector problem (gloss's teapot)
+	- numerous GL conformance bugs fixed
+    Changes:
+	- glColorMask(false, false, false, false) handled better/faster
+	- reverted to old GLU polygon tessellator, GLU 1.1
+	- updated Win32 build files
+
+
+3.3  July 21, 2000
+    New:
+	- antialiased triangles now implemented
+	- GL_EXT_texture_env_add texture mode extension
+	- GLX 1.3 API
+	- support for separate draw/read buffers (ie GL_SGI_make_current_read)
+	- thread-safe API dispath
+	- improved glxinfo program
+	- demos/texdown program to measure texture download performance
+	- glext.h header file
+	- demos/geartrain program
+	- GL_EXT_texture_lod_bias extension
+	- demos/lodbias program
+	- further optimized glRead/DrawPixels for 16-bit TrueColor X visuals
+	- GLX_EXT_visual_rating extension (a no-op, however)
+	- GL_HP_occlusion_test extension (for X and OS/Mesa drivers)
+	- demos/occlude program
+	- GL_SGIS_pixel_texture and GL_SGIX_pixel_texture extensions
+	- demos/pixeltex program
+	- GL_SGI_color_matrix extension
+	- GL_SGI_color_table extension
+	- GL_EXT_histogram extension
+	- GL_ARB_texture_cube_map extension
+	- added xdemos/glxheads and xdemos/manywin
+	- demos/texenv.c demo
+	- GL_EXT_texture_env_combine extension (by Holger Waechtler)
+	- Xlib driver is now thread-safe (see xdemos/glthreads)
+    Bug Fixes:
+	- various GL conformance failures fixed since 3.2.1
+    Changes:
+	- gl.h now uses #defines instead of C enums for all tokens
+	- glu.h now uses #defines instead of C enums for all tokens
+	- moved programs from 3Dfx/demos/ into demos/ directory
+
+
+3.4  November 3, 2000
+    New:
+	- optimized glDrawPixels for glPixelZoom(1,-1)
+    Bug Fixes:
+	- widgets-mesa/src/*.c files were missing from 3.3 distro
+	- include/GL/mesa_wgl.h file was missing from 3.3 distro
+	- fixed some Win32 compile problems
+	- texture object priorities weren't getting initialized to 1.0
+	- glAreTexturesResident return value was wrong when using hardware
+	- glXUseXFont segfaulted when using 3dfx driver (via MESA_GLX_FX)
+	- glReadPixels with GLushort packed types was broken
+	- fixed a few bugs in the GL_EXT_texture_env_combine texture code
+	- glPush/PopAttrib(GL_ENABLE_BIT) mishandled multi-texture enables
+	- fixed some typos/bugs in the VB code
+	- glDrawPixels(GL_COLOR_INDEX) to RGB window didn't work
+	- optimized glDrawPixels paths weren't being used
+	- per-fragment fog calculation didn't work without a Z buffer
+	- improved blending accuracy, fixes Glean blendFunc test failures
+	- glPixelStore(GL_PACK/UNPACK_SKIP_IMAGES) wasn't handled correctly
+	- glXGetProcAddressARB() didn't always return the right address
+	- gluBuild[12]DMipmaps() didn't grok the GL_BGR pixel format
+	- texture matrix changes weren't always detected (GLUT projtex demo)
+	- fixed random color problem in vertex fog code
+	- fixed Glide-related bug that let Quake get a 24-bit Z buffer
+    Changes:
+	- finished internal support for compressed textures for DRI
+
+
+3.4.1  February 14, 2001
+    New:
+	- fixed some Linux build problems
+	- fixed some Windows build problems
+	- GL_EXT_texture_env_dot3 extension (Gareth Hughes)
+    Bug fixes:
+	- added RENDER_START/RENDER_FINISH macros for glCopyTexImage in DRI
+	- various state-update code changes needed for DRI bugs
+	- disabled pixel transfer ops in glColorTable commands, not needed
+	- fixed bugs in glCopyConvolutionFilter1D/2D, glGetConvolutionFilter
+	- updated sources and fixed compile problems in widgets-mesa/
+	- GLX_PBUFFER enum value was wrong in glx.h
+	- fixed a glColorMaterial lighting bug
+	- fixed bad args to Read/WriteStencilSpan in h/w stencil clear function
+	- glXCopySubBufferMESA() Y position was off by one
+	- Error checking of glTexSubImage3D() was broken (bug 128775)
+	- glPopAttrib() didn't restore all derived Mesa state correctly
+	- Better glReadPixels accuracy for 16bpp color - fixes lots of OpenGL
+	  conformance problems at 16bpp.
+	- clearing depth buffer with scissoring was broken, would segfault
+	- OSMesaGetDepthBuffer() returned bad bytesPerValue value
+	- fixed a line clipping bug (reported by Craig McDaniel)
+	- fixed RGB color over/underflow bug for very tiny triangles
+    Known problems:
+	- NURBS or evaluator surfaces inside display lists don't always work
+
+
+3.4.2  May 17, 2001
+    Bug fixes:
+	- deleting the currently bound texture could cause bad problems
+	- using fog could result in random vertex alpha values
+	- AA triangle rendering could touch pixels outside right window bound
+	- fixed byteswapping problem in clear_32bit_ximage() function
+	- fixed bugs in wglUseFontBitmapsA(), by Frank Warmerdam
+	- fixed memory leak in glXUseXFont()
+	- fragment sampling in AA triangle function was off by 1/2 pixel
+	- Windows: reading pixels from framebuffer didn't always work
+	- glConvolutionFilter2D could segfault or cause FP exception
+	- fixed segfaults in FX and X drivers when using tex unit 1 but not 0
+	- GL_NAND logicop didn't work right in RGBA mode
+	- fixed a memory corruption bug in vertex buffer reset code
+	- clearing the softwara alpha buffer with scissoring was broken
+	- fixed a few color index mode fog bugs
+	- fixed some bad assertions in color index mode
+	- fixed FX line 'stipple' bug #420091
+	- fixed stencil buffer clear width/height typo
+	- fixed GL error glitches in gl[Client]ActiveTextureARB()
+	- fixed Windows compilation problem in texutil.c
+	- fixed 1/8-pixel AA triangle sampling error
+    Changes:
+	- optimized writing mono-colored pixel spans to X pixmaps
+	- increased max viewport size to 2048 x 2048
+
+
+3.5  June 21, 2001
+    New:
+	- internals of Mesa divided into modular pieces (Keith Whitwell)
+	- 100% OpenGL 1.2 conformance (passes all conformance tests)
+	- new AA line algorithm
+	- GL_EXT_convolution extension
+	- GL_ARB_imaging subset
+	- OSMesaCreateContextExt() function
+	- GL_ARB_texture_env_add extension (same as GL_EXT_texture_env_add)
+	- GL_MAX_TEXTURE_UNITS_ARB now defaults to eight
+	- GL_EXT_fog_coord extension (Keith Whitwell)
+	- GL_EXT_secondary_color extension (Keith Whitwell)
+	- GL_ARB_texture_env_add extension (same as GL_EXT_texture_env_add)
+	- GL_SGIX_depth_texture extension
+	- GL_SGIX_shadow and GL_SGIX_shadow_ambient extensions
+	- demos/shadowtex.c demo of GL_SGIX_depth_texture and GL_SGIX_shadow
+	- GL_ARB_texture_env_combine extension
+	- GL_ARB_texture_env_dot3 extension
+	- GL_ARB_texture_border_clamp (aka GL_SGIS_texture_border_clamp)
+	- OSMesaCreateContextExt() function
+	- libOSMesa.so library, contains the OSMesa driver interface
+	- GL/glxext.h header file for GLX extensions
+	- somewhat faster software texturing, fogging, depth testing
+	- all color-index conformance tests now pass (only 8bpp tested)
+	- SPARC assembly language TCL optimizations (David Miller)
+	- GL_SGIS_generate_mipmap extension
+    Bug Fixes:
+	- fbiRev and tmuRev were unitialized when using Glide3
+	- fixed a few color index mode conformance failures; all pass now
+	- now appling antialiasing coverage to alpha after texturing
+	- colors weren't getting clamped to [0,1] before color table lookup
+	- fixed RISC alignment errors caused by COPY_4UBV macro
+	- drawing wide, flat-shaded lines could cause a segfault
+	- vertices now snapped to 1/16 pixel to fix rendering of tiny triangles
+    Changes:
+	- SGI's Sample Implementation (SI) 1.3 GLU library replaces Mesa GLU
+	- new libOSMesa.so library, contains the OSMesa driver interface
+
+
+4.0  October 22, 2001
+    New:
+	- Mesa 4.0 implements the OpenGL 1.3 specification
+	- GL_IBM_rasterpos_clip extension
+	- GL_EXT_texture_edge_clamp extension (aka GL_SGIS_texture_edge_clamp)
+	- GL_ARB_texture_mirrored_repeat extension
+	- WindML UGL driver (Stephane Raimbault)
+	- added OSMESA_MAX_WIDTH/HEIGHT queries
+	- attempted compiliation fixes for Solaris 5, 7 and 8
+	- updated glext.h and glxext.h files
+	- updated Windows driver (Karl Schultz)
+    Bug fixes:
+	- added some missing GLX 1.3 tokens to include/GL/glx.h
+	- GL_COLOR_MATRIX changes weren't recognized by teximage functions
+	- glCopyPixels with scale and bias was broken
+	- glRasterPos with lighting could segfault
+	- glDeleteTextures could leave a dangling pointer
+	- Proxy textures for cube maps didn't work
+	- fixed a number of 16-bit color channel bugs
+	- fixed a few minor memory leaks
+	- GLX context sharing was broken in 3.5
+	- fixed state-update bugs in glPopClientAttrib()
+	- fixed glDrawRangeElements() bug
+	- fixed a glPush/PopAttrib() bug related to texture binding
+	- flat-shaded, textured lines were broken
+	- fixed a dangling pointer problem in the XMesa code (Chris Burghart)
+	- lighting didn't always produce the correct alpha value
+	- fixed 3DNow! code to not read past end of arrays (Andrew Lewycky)
+
+
+4.0.1 December 17, 2001
+    New:
+	- better sub-pixel sample positions for AA triangles (Ray Tice)
+	- slightly faster blending for (GL_ZERO, GL_ONE) and (GL_ONE, GL_ZERO)
+    Bug fixes:
+	- added missing break statements in glGet*() for multisample cases
+	- fixed uninitialized hash table mutex bug (display lists / texobjs)
+	- fixed bad teximage error check conditional (bug 476846)
+	- fixed demos readtex.c compilation problem on Windows (Karl Schultz)
+	- added missing glGet() query for GL_MAX_TEXTURE_LOD_BIAS_EXT
+	- silence some compiler warnings (gcc 2.96)
+	- enable the #define GL_VERSION_1_3 in GL/gl.h
+	- added GL 1.3 and GLX 1.4 entries to gl_mangle.h and glx_mangle.h
+	- fixed glu.h typedef problem found with MSDev 6.0
+	- build libGL.so with -Bsymbolic (fixes bug found with Chromium)
+	- added missing 'const' to glXGetContextIDEXT() in glxext.h
+	- fixed a few glXGetProcAddress() errors (texture compression, etc)
+	- fixed start index bug in compiled vertex arrays (Keith)
+	- fixed compilation problems in src/SPARC/glapi_sparc.S
+	- fixed triangle strip "parity" bug found in VTK medical1 demo (Keith)
+	- use glXGetProcAddressARB in GLUT to avoid extension linking problems
+	- provoking vertex of flat-shaded, color-index triangles was wrong
+	- fixed a few display list bugs (GLUT walker, molecule, etc) (Keith)
+	- glTexParameter didn't flush the vertex buffer (Ray Tice)
+	- feedback attributes for glDraw/CopyPixels and glBitmap were wrong
+	- fixed bug in normal length caching (ParaView lighting bug)
+	- fixed separate_specular color bug found in Chimera (18 Dec 2001)
+
+
+4.0.2  April 2, 2002
+    New:
+	- New DOS (DJGPP) driver written by Daniel Borca
+	- New driver interface functions for TCL drivers (such as Radeon DRI)
+	- GL_RENDERER string returns "Mesa Offscreen16" or "Mesa Offscreen32"
+	  if using deep color channels
+	- latest GL/glext.h and GL/glxext.h headers from SGI
+    Bug fixes:
+	- GL_BLEND with non-black texture env color wasn't always correct
+	- GL_REPLACE with GL_RGB texture format wasn't always correct (alpha)
+	- glTexEnviv( pname != GL_TEXTURE_ENV_COLOR ) was broken
+	- glReadPixels was sometimes mistakenly clipped by the scissor box
+	- glDraw/ReadPixels didn't catch all the errors that they should have
+	- Fixed 24bpp rendering problem in Windows driver (Karl Schultz)
+	- 16-bit GLchan mode fixes (m_trans_tmp.h, s_triangle.c)
+	- Fixed 1-bit float->int conversion bug in glDrawPixels(GL_DEPTH_COMP)
+	- glColorMask as sometimes effecting glXSwapBuffers()
+	- fixed a potential bug in XMesaGarbageCollect()
+	- N threads rendering into one window didn't work reliably
+	- glCopyPixels didn't work for deep color channels
+	- improved 8 -> 16bit/channel texture image conversion (Gerk Huisma)
+	- glPopAttrib() didn't correctly restore user clip planes
+	- user clip planes failed for some perspective projections (Chromium)
+    Known bugs:
+	- mipmap LOD computation
+
+
+4.0.3  June 25, 2002
+    New:
+	- updated GL/glext.h file (version 15)
+	- corrected MMX blend code (Jose Fonseca)
+	- support for software-based alpha planes in Windows driver
+	- updated GGI driver (Filip Spacek)
+    Bug fixes:
+	- glext.h had wrong values for GL_DOT3_RGB[A]_EXT tokens
+	- OSMesaMakeCurrent() didn't recognize buffer size changes
+	- assorted conformance fixes for 16-bit/channel rendering
+	- texcombine alpha subtraction mode was broken
+	- fixed lighting bug with non-uniform scaling and display lists
+	- fixed bug when deleting shared display lists
+	- disabled SPARC cliptest assembly code (Mesa bug 544665)
+	- fixed a couple Solaris compilation/link problems
+	- blending clipped glDrawPixels didn't always work
+	- glGetTexImage() didn't accept packed pixel types
+	- glPixelMapu[is]v() could explode given too large of pixelmap
+	- glGetTexParameter[if]v() didn't accept GL_TEXTURE_MAX_ANISOTROPY_EXT
+	- glXCopyContext() could lead to segfaults
+	- glCullFace(GL_FRONT_AND_BACK) didn't work (bug 572665)
+    Changes:
+	- lots of C++ (g++) code clean-ups
+	- lots of T&L updates for the Radeon DRI driver
+    Known bugs:
+	- mipmap LOD computation (fixed for Mesa 4.1)
+
+
+4.0.4  October 3, 2002
+    New:
+	- GL_NV_texture_rectangle extension
+	- updated glext.h header (version 17)
+	- updated DOS driver (Daniel Borca)
+	- updated BeOS R5 driver (Philippe Houdoin)
+	- added GL_IBM_texture_mirror_repeat
+	- glxinfo now takes -l option to print interesting OpenGL limits info
+	- GL_MESA_ycbcr_texture extension
+	- GL_APPLE_client_storage extension (for some DRI drivers only)
+	- GL_MESA_pack_invert extension
+    Bug fixes:
+	- fixed GL_LINEAR fog bug by adding clamping
+	- fixed FP exceptions found using Alpha CPU
+	- 3dfx MESA_GLX_FX=window (render to window) didn't work
+	- fixed memory leak in wglCreateContest (Karl Schultz)
+	- define GLAPIENTRY and GLAPI if undefined in glu.h
+	- wglGetProcAddress didn't handle all API functions
+	- when testing for OpenGL 1.2 vs 1.3, check for GL_ARB_texture_cube_map
+	- removed GL_MAX_CONVOLUTION_WIDTH/HEIGHT from glGetInteger/Float/etc()
+	- error checking in compressed tex image functions had some glitches
+	- fixed AIX compile problem in src/config.c
+	- glGetTexImage was using pixel unpacking instead of packing params
+	- auto-mipmap generation for cube maps was incorrect
+    Changes:
+	- max texture units reduced to six to accomodate texture rectangles
+	- removed unfinished GL_MESA_sprite_point extension code
+
+
+4.1  October 29, 2002
+    New:
+	- GL_NV_vertex_program extension
+	- GL_NV_vertex_program1_1 extension
+	- GL_ARB_window_pos extension
+	- GL_ARB_depth_texture extension
+	- GL_ARB_shadow extension
+	- GL_ARB_shadow_ambient extension
+	- GL_EXT_shadow_funcs extension
+	- GL_ARB_point_parameters extension
+	- GL_ARB_texture_env_crossbar
+	- GL_NV_point_sprite extension
+	- GL_NV_texture_rectangle extension
+	- GL_EXT_multi_draw_arrays extension
+	- GL_EXT_stencil_two_side extension
+	- GLX_SGIX_fbconfig and GLX_SGIX_pbuffer extensions
+	- GL_ATI_texture_mirror_once extension (Ian Romanick)
+	- massive overhaul/simplification of software rasterizer module,
+	  many contributions from Klaus Niederkrueger
+	- faster software texturing in some cases (i.e. trilinear filtering)
+	- new OSMesaGetProcAddress() function
+	- more blend modes implemented with MMX code (Jose Fonseca)
+	- added glutGetProcAddress() to GLUT
+	- added GLUT_FPS env var to compute frames/second in glutSwapBuffers()
+	- pbinfo and pbdemo PBuffer programs
+	- glxinfo -v prints transprent pixel info (Gerd Sussner)
+    Bug fixes:
+	- better mipmap LOD computation (prevents excessive blurriness)
+	- OSMesaMakeCurrent() didn't recognize buffer size changes
+	- assorted conformance fixes for 16-bit/channel rendering
+	- texcombine alpha subtraction mode was broken
+	- fixed some blend problems when GLchan==GLfloat (Gerk Huisma)
+	- clamp colors to [0,inf] in OSMesa if GLchan==GLfloat (Gerk Huisma)
+	- fixed divide by zero error in NURBS tessellator (Jon Perry)
+	- fixed GL_LINEAR fog bug by adding clamping
+	- fixed FP exceptions found using Alpha CPU
+	- 3dfx/glide driver render-to-window feature was broken
+	- added missing GLX_TRANSPARENT_RGB token to glx.h
+	- fixed error checking related to paletted textures
+	- fixed reference count error in glDeleteTextures (Randy Fayan)
+    Changes:
+	- New spec file and Python code to generate some GL dispatch files
+	- Glide driver defaults to "no" with autoconf/automake
+	- updated demos/stex3d with new options
+
+
+5.0  November 13, 2002
+    New:
+	- OpenGL 1.4 support (glGetString(GL_VERSION) returns "1.4")
+	- removed some overlooked debugging code
+	- glxinfo updated to support GLX_ARB_multisample
+	- GLUT now support GLX_ARB_multisample
+	- updated DOS driver (Daniel Borca)
+    Bug fixes:
+	- GL_POINT and GL_LINE-mode polygons didn't obey cull state
+	- fixed potential bug in _mesa_align_malloc/calloc()
+	- fixed missing triangle bug when running vertex programs
+	- fixed a few HPUX compilation problems
+	- FX (Glide) driver didn't compile
+	- setting GL_TEXTURE_BORDER_COLOR with glTexParameteriv() didn't work
+	- a few EXT functions, like glGenTexturesEXT, were no-ops
+	- a few OpenGL 1.4 functions like glFogCoord*, glBlendFuncSeparate,
+	  glMultiDrawArrays and glMultiDrawElements were missing
+	- glGet*(GL_ACTIVE_STENCIL_FACE_EXT) was broken
+	- Pentium 4 Mobile was mistakenly identified as having 3DNow!
+	- fixed one-bit error in point/line fragment Z calculation
+	- fixed potential segfault in fakeglx code
+	- fixed color overflow problem in DOT3 texture env mode
+
+
+5.0.1  March 30, 2003
+    New:
+	- DOS driver updates from Daniel Borca
+	- updated GL/gl_mangle.h file (Bill Hoffman)
+    Bug fixes:
+	- auto mipmap generation for cube maps was broken (bug 641363)
+	- writing/clearing software alpha channels was unreliable
+	- minor compilation fixes for OS/2 (Evgeny Kotsuba)
+	- fixed some bad assertions found with shadowtex demo
+	- fixed error checking bug in glCopyTexSubImage2D (bug 659020)
+	- glRotate(angle, -x, 0, 0) was incorrect (bug 659677)
+	- fixed potential segfault in texture object validation (bug 659012)
+	- fixed some bogus code in _mesa_test_os_sse_exception_support (Linus)
+	- fix fog stride bug in tnl code for h/w drivers (Michel Danzer)
+	- fixed glActiveTexture / glMatrixMode(GL_TEXTURE) bug (#669080)
+	- glGet(GL_CURRENT_SECONDARY_COLOR) should return 4 values, not 3
+	- fixed compilation problem on Solaris7/x86 (bug 536406)
+	- fixed prefetch bug in 3DNow! code (Felix Kuhling)
+	- fixed NeXT build problem (FABSF macro)
+	- glDrawPixels Z values when glPixelZoom!=1 were invalid (bug 687811)
+	- zoomed glDraw/CopyPixels with clipping sometimes failed (bug 689964)
+	- AA line and triangle Z values are now rounded, not truncated
+	- fixed color interpolation bug when GLchan==GLfloat (bug 694461)
+	- glArePrograms/TexturesResident() wasn't 100% correct (Jose Fonseca)
+	- fixed a minor GL_COLOR_MATERIAL bug
+	- NV vertex program EXP instruction was broken
+	- glColorMask misbehaved with X window / pixmap rendering
+	- fix autoconf/libtool GLU C++ linker problem on Linux (a total hack)
+	- attempt to fix GGI compilation problem when MesaDemos not present
+	- NV vertex program ARL-relative fetches didn't work
+    Changes:
+	- use glPolygonOffset in gloss demo to avoid z-fighting artifacts
+	- updated winpos and pointblast demos to use ARB extensions
+	- disable SPARC normal transformation code (bug 673938)
+	- GLU fixes for OS/2 (Evgeny Kotsuba)
+
+
+5.0.2  September 5, 2003
+    Bug fixes:
+	- fixed texgen problem causing texcoord's Q to be zero (stex3d)
+	- default GL_TEXTURE_COMPARE_MODE_ARB was wrong
+	- GL_CURRENT_MATRIX_NV query was wrong
+	- GL_CURRENT_MATRIX_STACK_DEPTH_NV query was off by one
+	- GL_LIST_MODE query wasn't correct
+	- GL_FOG_COORDINATE_SOURCE_EXT query wasn't supported
+	- GL_SECONDARY_COLOR_ARRAY_SIZE_EXT query returned wrong value
+	- blended, wide lines didn't always work correctly (bug 711595)
+	- glVertexAttrib4svNV w component was always 1
+	- fixed bug in GL_IBM_rasterpos_clip (missing return)
+	- GL_DEPTH_TEXTURE_MODE = GL_ALPHA didn't work correctly
+	- a few Solaris compilation fixes
+	- fixed glClear() problem for DRI drivers (non-existant stencil, etc)
+	- fixed int/REAL mixup in GLU NURBS curve evaluator (Eric Cazeaux)
+	- fixed delete [] bug in SI GLU (bug 721765) (Diego Santa Cruz)
+	- glFog() didn't clamp fog colors
+	- fixed bad float/int conversion for GL_TEXTURE_PRIORITY in the
+	  gl[Get]TexParameteri[v] functions
+	- fixed invalid memory references in glTexGen functions (bug 781602)
+	- integer-valued color arrays weren't handled correctly
+	- glDrawPixels(GL_DEPTH_COMPONENT) with glPixelZoom didn't work
+	- GL_EXT_texture_lod_bias is part of 1.4, overlooked in 5.0.1
+    Changes:
+	- build GLUT with -fexceptions so C++ apps propogate exceptions
+
+
+5.1  December 17, 2003
+    New:
+	- reorganized directory tree
+	- GL_ARB_vertex/fragment_program extensions (Michal Krol & Karl Rasche)
+	- GL_ATI_texture_env_combine3 extension (Ian Romanick)
+	- GL_SGI_texture_color_table extension (Eric Plante)
+	- GL_NV_fragment_program extension
+	- GL_NV_light_max_exponent extension
+	- GL_EXT_texture_rectangle (identical to GL_NV_texture_rectangle)
+	- GL_ARB_occlusion_query extension
+	- GL_ARB_point_sprite extension
+	- GL_ARB_texture_non_power_of_two extension
+	- GL_IBM_multimode_draw_arrays extension
+	- GL_EXT_texture_mirror_clamp extension (Ian Romanick)
+	- GL_ARB_vertex_buffer_object extension
+	- new X86 feature detection code (Petr Sebor)
+	- less memory used for display lists and vertex buffers
+	- demo of per-pixel lighting with a fragment program (demos/fplight.c)
+	- new version (18) of glext.h header
+	- new spriteblast.c demo of GL_ARB_point_sprite
+	- faster glDrawPixels in X11 driver in some cases (see RELNOTES-5.1)
+	- faster glCopyPixels in X11 driver in some cases (see RELNOTES-5.1)
+    Bug fixes:
+	- really enable OpenGL 1.4 features in DOS driver.
+	- fixed issues in glDrawPixels and glCopyPixels for very wide images
+	- glPixelMapf/ui/usv()'s size parameter is GLsizei, not GLint
+	- fixed some texgen bugs reported by Daniel Borca
+	- fixed wglMakeCurrent(NULL, NULL) bug (#835861)
+	- fixed glTexSubImage3D z-offset bug (Cedric Gautier)
+	- fixed RGBA blend enable bug (Ville Syrjala)
+	- glAccum is supposed to be a no-op in selection/feedback mode
+	- fixed texgen bug #597589 (John Popplewell)
+    Changes:
+	- dropped API trace feature (src/Trace/)
+	- documentation overhaul.  merged with website content.  more html.
+	- glxgears.c demo updated to use GLX swap rate extensions
+	- glTexImage1/2/3D now allows width/height/depth = 0
+	- disable SPARC asm code on Linux (bug 852204)
+
+
+6.0  January 16, 2004
+    New:
+	- full OpenGL 1.5 support
+	- updated GL/glext.h file to version 21
+    Changes:
+	- changed max framebuffer size to 4Kx4K (MAX_WIDTH/HEIGHT in config.h)
+    Bug fixes:
+	- fixed bug in UNCLAMPED_FLOAT_TO_UBYTE macro; solves a color
+	  clamping issue
+	- updated suno5-gcc configs
+	- glColor3 functions sometimes resulted in undefined alpha values
+	- fixed FP divide by zero error seen on VMS with xlockmore, others
+	- fixed vertex/fragment program debug problem (bug 873011)
+	- building on AIX with gcc works now
+	- glDeleteProgramsARB failed for ARB fragment programs (bug 876160)
+	- glDrawRangeElements tried to modify potentially read-only storage
+	- updated files for building on Windows
+
+
+6.0.1  April 2, 2004
+    New:
+	- upgraded glext.h to version 22
+	- new build targets (Dan Schikore)
+	- new linux-x86-opteron build target (Heath Feather)
+    Bug fixes:
+	- glBindProgramARB didn't update all necessary state
+	- fixed build problems on OpenBSD
+	- omit CVS directories from tarballs
+	- glGetTexImage(GL_COLOR_INDEX) was broken
+	- fixed an infinite loop in t&l module
+	- silenced some valgrind warnings about using unitialized memory
+	- fixed some compilation/link glitches on IRIX (Mike Stephens)
+	- glBindProgram wasn't getting compiled into display lists
+	- GLX_FBCONFIG_ID wasn't recognized in glXChooseFBConfig() (bug 888079)
+	- two-sided lighting and vertex program didn't work (bug 887330)
+	- stores to program parameter registers in vertex state programs
+	  didn't work.
+	- fixed glOrtho bug found with gcc 3.2.2 (RH9)
+	- glXCreateWindow() wasn't fully implemented (bug 890894)
+	- generic vertex attribute arrays didn't work in display lists
+	- vertex buffer objects' default usage and access fields were wrong
+	- glDrawArrays with start!=0 was broken
+	- fragment program PK2H, UP2H, UP4B and UP4UB instructions were broken
+	- linux-osmesa16-static config didn't work
+	- fixed a few color index rendering problems (bug 910687)
+	- glInterleavedArrays didn't respect GL_CLIENT_ACTIVE_TEXTURE
+	- OSMesa RGB and BGR modes were broken
+	- glProgramStringARB mistakenly required a null-terminated string
+	- fragment program XPD instruction was incorrect
+	- glGetMaterial() didn't work reliably
+	- ARB_fragment_program KIL instruction was incorrect
+
+
+6.1  August 18, 2004
+    New:
+	- Revamped Makefile system
+	- glXUseRotatedXFont() utility (see xdemos/xuserotfont.c)
+	- internal driver interface changes related to texture object
+	  allocation, vertex/fragment programs, BlendEquationSeparate, etc.
+	- option to walk triangle edges with double-precision floats
+	  (Justin Novosad of Discreet) (see config.h file)
+	- support for AUX buffers in software GLX driver
+	- updated glext.h to version 24 and glxext.h to version 6
+	- new MESA_GLX_FORCE_ALPHA and MESA_GLX_DEPTH_BITS env vars
+	- updated BeOS support (Philippe Houdoin)
+    Changes:
+	- fragment fog interpolation is perspective corrected now
+	- new glTexImage code, much cleaner, may be a bit faster
+    Bug fixes:
+	- glArrayElement in display lists didn't handle generic vertex attribs
+	- glFogCoord didn't always work properly
+	- ARB_fragment_program fog options didn't work
+	- frag prog TEX instruction no longer incorrectly divides s,t,r by q
+	- ARB frag prog TEX and TEXP instructions now use LOD=0
+	- glTexEnviv in display lists didn't work
+	- glRasterPos didn't do texgen or apply texture matrix
+	- GL_DOUBLE-valued vertex arrays were broken in some cases
+	- fixed texture rectangle edge/border sampling bugs
+	- sampling an incomplete texture in a fragment program would segfault
+	- glTexImage was missing a few error checks
+	- fixed some minor glGetTexParameter glitches
+	- GL_INTENSITY was mistakenly accepted as a <format> to glTexImage
+	- fragment program writes to RC/HC register were broken
+	- fixed a few glitches in GL_HP_occlusion_test extension
+	- glBeginQueryARB and glEndQueryARB didn't work inside display lists
+	- vertex program state references were broken
+	- fixed triangle color interpolation bug on AIX (Shane Blackett)
+	- fixed a number of minor memory leaks (bug #1002030)
+
+
+6.2  October 2, 2004
+    New:
+	- enabled GL_ARB_texture_rectangle (same as GL_NV_texture_rectangle)
+	- updated Doxygen support (Jose Fonseca)
+    Changes:
+	- some GGI driver updates (Christoph Egger, bug 1025977)
+    Bug fixes:
+	- Omit GL_ARB_texture_non_power_of_two from list of OpenGL 1.5 features
+	- fixed a few compilation issues on IRIX
+	- fixed a matrix classification bug (reported by Wes Bethel)
+	- we weren't reseting the vertex/fragment program error state
+	  before parsing (Dave Reveman)
+	- adjust texcoords for sampling texture rectangles (Dave Reveman)
+	- glGet*(GL_MAX_VERTEX_ATTRIBS_ARB) wasn't implemented
+	- repeated calls to glDeleteTexture(t) could lead to a crash
+	- fixed potential ref count bugs in VBOs and vertex/fragment programs
+	- spriteblast demo didn't handle window size changes correctly
+	- glTexSubImage didn't handle pixels=NULL correctly for PBOs
+	- fixed color index mode glDrawPixels bug (Karl Schultz)
+
+
+6.2.1  December 9, 2004
+    Bug fixes:
+	- don't apply regular fog or color sum when using a fragment program
+	- glProgramEnvParameter4fARB always generated an error on
+	  GL_FRAGMENT_PROGRAM_ARB (fdo bug 1645)
+	- glVertexAttrib3svNV and glVertexAttrib3svARB were broken
+	- fixed width/height mix-up in glSeparableFilter2D()
+	- fixed regression in glCopyPixels + convolution
+	- glReadPixels from a clipped front color buffer didn't always work
+	- glTexImage didn't accept GL_RED/GREEN/BLUE as the format
+	- Attempting queries/accesses of VBO 0 weren't detected as errors
+	- paletted textures failed if the palette had fewer than 256 entries
+    Changes:
+	- fixed a bunch of compiler warnings found with gcc 3.4
+	- bug reports should to go bugzilla.freedesktop.org
+
+
+6.3  July 20, 2005
+    New:
+	- GL_EXT_framebuffer_object extension
+	- GL_ARB_draw_buffers extension
+	- GL_ARB_pixel_buffer_object extension
+	- GL_OES_read_format extension (Ian Romanick)
+	- DirectFB driver (Claudio Ciccani)
+	- x86_64 vertex transformation code (Mikko T.)
+	- Updated GL/glext.h to version 29
+    Changes:
+	- added -stereo option for glxgears demo (Jacek Rosik)
+	- updated the PBuffer demo code in xdemos/ directory
+	- glDeleteTextures/Programs/Buffers() now makes the object ID
+	  available for immediate re-use
+	- assorted 64-bit clean-ups fixes (x86_64 and Win64)
+	- lots of internal changes for GL_EXT_framebuffer_object
+    Bug fixes:
+	- some functions didn't support PBO functionality
+	- glGetTexImage didn't convert color index images to RGBA as required
+	- fragment program texcoords were sometimes wrong for points and lines
+	- fixed problem with negative dot product in arbfplight, fplight demos
+	- fixed bug in perspective correction of antialiased, textured lines
+	- querying GL_POST_CONVOLUTION_ALPHA_BIAS_EXT returned wrong value
+	- fixed a couple per-pixel fog bugs (Soju Matsumoto)
+	- glGetBooleanv(GL_FRAGMENT_PROGRAM_BINDING_NV) was broken
+	- fixed float parsing bug in ARB frag/vert programs (bug 2520)
+	- XMesaGetDepthBuffer() returned incorrect value for bytesPerValue
+	- GL_COLOR_MATERIAL with glColor3 didn't properly set diffuse alpha
+	- glXChooseFBConfig() crashed if attribList pointer was NULL
+	- program state.light[n].spot.direction.w was wrong value (bug 3083)
+	- fragment program fog option required glEnable(GL_FOG) - wrong.
+	- glColorTable() could produce a Mesa implementation error (bug 3135)
+	- RasterPos could get corrupted by color index rendering path
+	- Removed bad XTranslateCoordinates call when rendering to Pixmaps
+	- glPopAttrib() didn't properly restore GL_TEXTURE_GEN enable state
+	- fixed a few Darwin compilation problems
+
+
+6.3.1
+    This was an intermediate release for X.org which wasn't otherwise released.
+
+
+6.3.2  August 19, 2005
+    New:
+	- The distribution now includes the DRI drivers and GLX code
+    Changes:
+	- Made the DRI "new" driver interface standard, remove old code
+    Bug fixes:
+	- GL_ARB_vertex/fragment_shader were mistakenly listed in the
+	  extensions string
+	- negative relative addressing in vertex programs was broken
+	- update/fix SPARC assembly code for vertex transformation
+	- fixed memory leak when freeing GLX drawables/renderbuffers
+	- fixed display list memory leak
+	- the GL_PIXEL_MAP_I_TO_I table is now floating point, not integer
+	- wglGetProcAddress() didn't handle wgl-functions
+	- fixed glxext.h cross-compile issue (Colin Harrison)
+	- assorted DRI driver fixes
+
+
+6.4  October 24, 2005
+    New:
+	- Added a fast XOR line drawing function in Xlib driver
+	- Added support for GL_ARB_texture_mirrored_repeat to savage
+	  driver (supported only on Savage4 hardware).
+    Changes:
+	- Mesa now packaged in three parts: Library, Demos and GLUT
+    Bug fixes:
+	- GLX_X_RENDERABLE token wasn't accepted by glXChooseFBConfig
+	- Some files were present multiple times in the 6.3.2 tarballs
+	- r200_vtxtmp_x86.S file was missing from 6.3.2 tarball (bug 4207)
+	- glxgears_fbconfig demo didn't work (bug 4237)
+	- fixed bug when bilinear sampling 2d textures with borders
+	- glXCreatePbuffer() could segfault instead of returning 0 (bug 4235)
+	- fixed undefined frexp and rand in X.org libGLcore.a (bug 4242)
+	- fixed a few problems with proxy color tables (bug 4270)
+	- fixed precision problem in Z clearing (bug 4395)
+	- glBitmap, glDraw/CopyPixels mistakenly generated selection hits
+	- fixed potential segfault caused by reading pixels outside
+	  of renderbuffer bounds
+	- glGetTexLevelParameter didn't accept GL_TEXTURE_DEPTH_SIZE_ARB
+	- fixed memory corruption bug involving software alpha buffers
+	- glReadPixels clipped by window bounds was sometimes broken
+	- glDraw/CopyPixels of stencil data ignored the stencil write mask
+	- glReadPixels from a texture bound to a framebuffer object didn't work
+	- glIsRender/FramebufferEXT weren't totally correct
+	- fixed a number of point size attenuation/fade bugs
+	- fixed glFogCoord bug 4729
+	- GLX encoding for transpose matrix functions was broken
+	- fixed broken fragment program KIL and SWZ instructions
+	- fragment programs that wrote result.depth.z didn't work
+
+
+6.4.1  November 30, 2005
+    Bug fixes:
+	- redefining a vertex program string didn't take effect in TNL module
+	- fixed occasional segfault upon vertex/fragment parsing error
+	- vertex program LIT instruction didn't handle 0^0=1 correctly
+	- fragment program fog option didn't work with glDrawPixels, glBitmap
+	- USE_MGL_NAMESPACE didn't work for x86-64
+	- OSMesa demos were missing from previous release tarballs
+	- fixed problem with float->ushort conversion in glClear (bug 4992)
+	- popping of GL_EYE_PLANE texgen state was broken (bug 4996)
+	- popping of GL_SPOT_DIRECTION light state was broken (bug 5005)
+	- fixed occasional triangle color interpolation problem on VMS
+	- work around invalid free() call (bug 5131)
+	- fixed BSD X server compilation problem by including stdint.h
+
+
+6.4.2  February 2, 2006
+    New:
+	- added OSMesaColorClamp() function/feature
+	- added wglGetExtensionStringARB() function
+    Bug fixes:
+	- fixed some problems when building on Windows
+	- GLw header files weren't installed by installmesa script (bug 5396)
+	- GL/glfbdev.h file was missing from tarballs
+	- fixed TNL initialization bug which could lead to crash (bug 5791)
+
+
+6.5  March 31, 2006
+    New:
+	- OpenGL Shading Language support through GL_ARB_shader_objects,
+	  GL_ARB_shading_language_100, GL_ARB_vertex_shader and
+	  GL_ARB_fragment_shader (done by Michal Krol)
+	- GL_EXT_packed_depth_stencil extension
+	- GL_EXT_timer_query extension
+	- GL_EXT_framebuffer_blit extension
+	- GL_ARB_half_float_pixel
+	- reflect demo improved to support multiple windows
+	- singlebuffer demo (shows no/little-flicker single-buffered rendering)
+	- r200: enable GL_ARB_texture_env_crossbar, separate the texture
+	  sampling unit bits from the texture env combine enable bits
+	- r200: add support for GL_ATI_fragment_shader
+	- added fast XOR-mode line drawing optimization
+	- radeon: add support for all 3 tmus, GL_ARB_texture_cube_map
+	  and GL_EXT_fog_coord
+	- MESA_GLX_ALPHA_BITS env var for xlib driver
+	- many DRI driver updates (including screen rotation support
+	  for the Intel DRI driver)
+    Changes:
+	- removed GL_HP_occlusion_test (use GL_ARB_occlusion_query instead)
+	- removed GL_SGIX/SGIS_pixel_texture extensions
+    Bug fixes:
+	- fixed glxcontextmodes.c datatype problem (bug 5835)
+	- fixed aix-gcc build/install bugs (bug 5874)
+	- fixed some bugs in texture env program generation
+	- glXCopyContext() didn't handle texture object bindings properly
+	- glXCopyContext() didn't copy all lighting state
+	- fixed FreeBSD config (Pedro Giffuni)
+	- fixed some minor framebuffer object bugs
+	- replaced dprintf() with _glu_printf() in GLU (bug 6244)
+	- fixed a number of thread safety bugs/regressions
+	- fixed a number of GLU tesselator bugs (John Shell, bug 6339)
+	- paletted texturing was broken w/ floating point palettes (K. Schultz)
+	- lots of assorted framebuffer object bug fixes
+
+6.5.1  August 31, 2006
+    New:
+	- Intel i965 DRI driver
+	- GL_APPLE_vertex_array_object extension (Ian Romanick)
+	- GL_EXT_texture_sRGB extension
+	- GL_EXT_gpu_program_parameters (Ian Romanick)
+	- "engine" demo
+	- updated fbdev driver and GLUT for fbdev (Sean D'Epagnier)
+	- many updates to the DRI drivers
+    Changes:
+	- The glVertexAttribARB functions no longer alias the conventional
+	  vertex attributes.
+	- glxinfo program prints more info with -l option
+	- GL_FRAGMENT_PROGRAM_NV and GL_FRAGMENT_PROGRAM_ARB are now
+	  compatible, in terms of glBindProgramARB()
+    Bug fixes:
+	- fixed broken texture border handling for depth textures (bug 6498)
+	- removed the test for duplicated framebuffer attachments, per
+	  version 117 of the GL_EXT_framebuffer_object specification
+	- fixed a few render-to-texture bugs, including render to depth texture
+	- clipping of lines against user-defined clip planes was broken (6512)
+	- assembly language dispatch for SPARC was broken (bug 6484)
+	- assorted compilation fixes on various Unix platforms (Dan Schikore)
+	- glPopAttrib could restore an invalid value for GL_DRAW_BUFFER
+	- assorted minor fixes for 16 and 32 bit/channel modes
+	- fixed assorted bugs in texture compression paths
+	- fixed indirect rendering vertex array crashes (bug 6863)
+	- glDrawPixels GL_INDEX_OFFSET didn't always work
+	- fixed convolution memory leak (bug 7077)
+	- rectangular depth textures didn't work
+	- invalid mode to glBegin didn't generate an error (bug 7142)
+	- 'normalized' parameter to glVertexAttribPointerARB didn't work
+	- disable bogus GLX_SGI_video_sync extension in xlib driver
+	- fixed R128 driver locking bug (Martijn van Oosterhout)
+	- using evaluators with vertex programs caused crashes (bug 7564)
+	- fragment.position wasn't set correctly for point/line primitives
+	- fixed parser bug for scalar sources for GL_NV_fragment_program
+	- max fragment program length was incorrectly 128, now 1024
+	- writes to result.depth in fragment programs weren't clamped to [0,1]
+	- fixed potential dangling pointer bug in glBindProgram()
+	- fixed some memory leaks (and potential crashes) in Xlib driver
diff --git a/mesalib/docs/enums.txt b/mesalib/docs/enums.txt
index cb97b5a31..b37768e20 100644
--- a/mesalib/docs/enums.txt
+++ b/mesalib/docs/enums.txt
@@ -1,57 +1,57 @@
-
-See the OpenGL ARB enum registry at http://www.opengl.org/registry/api/enum.spec
-
-Blocks allocated to Mesa:
-       0x8750-0x875F
-       0x8BB0-0x8BBF
-
-
-GL_MESA_packed_depth_stencil
-	GL_DEPTH_STENCIL_MESA            0x8750
-	GL_UNSIGNED_INT_24_8_MESA        0x8751
-	GL_UNSIGNED_INT_8_24_REV_MESA    0x8752
-	GL_UNSIGNED_SHORT_15_1_MESA      0x8753
-	GL_UNSIGNED_SHORT_1_15_REV_MESA  0x8754
-
-GL_MESA_trace.spec:
-	GL_TRACE_ALL_BITS_MESA           0xFFFF
-	GL_TRACE_OPERATIONS_BIT_MESA     0x0001
-	GL_TRACE_PRIMITIVES_BIT_MESA     0x0002
-	GL_TRACE_ARRAYS_BIT_MESA         0x0004
-	GL_TRACE_TEXTURES_BIT_MESA       0x0008
-	GL_TRACE_PIXELS_BIT_MESA         0x0010
-	GL_TRACE_ERRORS_BIT_MESA         0x0020
-	GL_TRACE_MASK_MESA               0x8755
-	GL_TRACE_NAME_MESA               0x8756
-
-MESA_ycbcr_texture.spec:
-	GL_YCBCR_MESA                    0x8757
-	GL_UNSIGNED_SHORT_8_8_MESA       0x85BA /* same as Apple's */
-	GL_UNSIGNED_SHORT_8_8_REV_MESA   0x85BB /* same as Apple's */
-
-GL_MESA_pack_invert.spec
-	GL_PACK_INVERT_MESA              0x8758
-
-GL_MESA_shader_debug.spec: (obsolete)
-        GL_DEBUG_OBJECT_MESA             0x8759
-        GL_DEBUG_PRINT_MESA              0x875A
-        GL_DEBUG_ASSERT_MESA             0x875B
-
-GL_MESA_program_debug.spec: (obsolete)
-	GL_FRAGMENT_PROGRAM_CALLBACK_MESA      0x????
-	GL_VERTEX_PROGRAM_CALLBACK_MESA        0x????
-	GL_FRAGMENT_PROGRAM_POSITION_MESA      0x????
-	GL_VERTEX_PROGRAM_POSITION_MESA        0x????
-	GL_FRAGMENT_PROGRAM_CALLBACK_FUNC_MESA 0x????
-	GL_FRAGMENT_PROGRAM_CALLBACK_DATA_MESA 0x????
-	GL_VERTEX_PROGRAM_CALLBACK_FUNC_MESA   0x????
-	GL_VERTEX_PROGRAM_CALLBACK_DATA_MESA   0x????
-
-GL_MESAX_texture_stack:
-	GL_TEXTURE_1D_STACK_MESAX            0x8759
-	GL_TEXTURE_2D_STACK_MESAX            0x875A
-	GL_PROXY_TEXTURE_1D_STACK_MESAX      0x875B
-	GL_PROXY_TEXTURE_2D_STACK_MESAX      0x875C
-	GL_TEXTURE_1D_STACK_BINDING_MESAX    0x875D
-	GL_TEXTURE_2D_STACK_BINDING_MESAX    0x875E
-
+
+See the OpenGL ARB enum registry at http://www.opengl.org/registry/api/enum.spec
+
+Blocks allocated to Mesa:
+       0x8750-0x875F
+       0x8BB0-0x8BBF
+
+
+GL_MESA_packed_depth_stencil
+	GL_DEPTH_STENCIL_MESA            0x8750
+	GL_UNSIGNED_INT_24_8_MESA        0x8751
+	GL_UNSIGNED_INT_8_24_REV_MESA    0x8752
+	GL_UNSIGNED_SHORT_15_1_MESA      0x8753
+	GL_UNSIGNED_SHORT_1_15_REV_MESA  0x8754
+
+GL_MESA_trace.spec:
+	GL_TRACE_ALL_BITS_MESA           0xFFFF
+	GL_TRACE_OPERATIONS_BIT_MESA     0x0001
+	GL_TRACE_PRIMITIVES_BIT_MESA     0x0002
+	GL_TRACE_ARRAYS_BIT_MESA         0x0004
+	GL_TRACE_TEXTURES_BIT_MESA       0x0008
+	GL_TRACE_PIXELS_BIT_MESA         0x0010
+	GL_TRACE_ERRORS_BIT_MESA         0x0020
+	GL_TRACE_MASK_MESA               0x8755
+	GL_TRACE_NAME_MESA               0x8756
+
+MESA_ycbcr_texture.spec:
+	GL_YCBCR_MESA                    0x8757
+	GL_UNSIGNED_SHORT_8_8_MESA       0x85BA /* same as Apple's */
+	GL_UNSIGNED_SHORT_8_8_REV_MESA   0x85BB /* same as Apple's */
+
+GL_MESA_pack_invert.spec
+	GL_PACK_INVERT_MESA              0x8758
+
+GL_MESA_shader_debug.spec: (obsolete)
+        GL_DEBUG_OBJECT_MESA             0x8759
+        GL_DEBUG_PRINT_MESA              0x875A
+        GL_DEBUG_ASSERT_MESA             0x875B
+
+GL_MESA_program_debug.spec: (obsolete)
+	GL_FRAGMENT_PROGRAM_CALLBACK_MESA      0x????
+	GL_VERTEX_PROGRAM_CALLBACK_MESA        0x????
+	GL_FRAGMENT_PROGRAM_POSITION_MESA      0x????
+	GL_VERTEX_PROGRAM_POSITION_MESA        0x????
+	GL_FRAGMENT_PROGRAM_CALLBACK_FUNC_MESA 0x????
+	GL_FRAGMENT_PROGRAM_CALLBACK_DATA_MESA 0x????
+	GL_VERTEX_PROGRAM_CALLBACK_FUNC_MESA   0x????
+	GL_VERTEX_PROGRAM_CALLBACK_DATA_MESA   0x????
+
+GL_MESAX_texture_stack:
+	GL_TEXTURE_1D_STACK_MESAX            0x8759
+	GL_TEXTURE_2D_STACK_MESAX            0x875A
+	GL_PROXY_TEXTURE_1D_STACK_MESAX      0x875B
+	GL_PROXY_TEXTURE_2D_STACK_MESAX      0x875C
+	GL_TEXTURE_1D_STACK_BINDING_MESAX    0x875D
+	GL_TEXTURE_2D_STACK_BINDING_MESAX    0x875E
+
diff --git a/mesalib/docs/install.html b/mesalib/docs/install.html
index 2ba7cb867..a344c2ca7 100644
--- a/mesalib/docs/install.html
+++ b/mesalib/docs/install.html
@@ -16,8 +16,7 @@
   <li><a href="#prereq-dri">For DRI and hardware acceleration</a>
   </ul>
 <li><a href="#autoconf">Building with autoconf (Linux/Unix/X11)</a>
-<li><a href="#scons">Building with SCons (Windows)</a>
-<li><a href="#legacy">Building with legacy Makefiles (deprecated)</a>
+<li><a href="#scons">Building with SCons (Windows/Linux)</a>
 <li><a href="#other">Building for other systems</a>
 <li><a href="#libs">Library Information</a>
 <li><a href="#pkg-config">Building OpenGL programs with pkg-config
@@ -95,7 +94,7 @@ for more details.
 
 
 <a name="scons">
-<H1>3. Building with SCons (Windows)</H1>
+<H1>3. Building with SCons (Windows/Linux)</H1>
 
 <p>
 To build Mesa with SCons on Linux or Windows do
@@ -129,118 +128,8 @@ Put them all in the same directory to test them.
 
 
 
-<a name="legacy">
-<h1>4. Building with legacy Makefiles (deprecated)</h1>
-
-<p>
-The legacy Mesa build system is based on a collection of pre-defined
-system configurations.
-Some of these might work for older systems not supported by autoconf.
-</p>
-<p>
-To see the list of configurations, just type <code>make</code>.
-Then choose a configuration from the list and type <code>make</code>
-<em>configname</em>.
-</p>
-
-<p>
-Mesa may be built in several different ways using the predefined configurations:
-</p>
-<ul>
-<li><b><em>Stand-alone/Xlib mode</em></b> - Mesa will be compiled as
-a software renderer using Xlib to do all rendering.
-The libGL.so library will be a self-contained rendering library that will
-allow you to run OpenGL/GLX applications on any X server (regardless of
-whether it supports the GLX X server extension).
-You will <em>not</em> be able to use hardware 3D acceleration.
-<p>
-To compile stand-alone Mesa type <code>make</code> in the top-level directory.
-You'll see a list of supported system configurations.
-Choose one from the list (such as linux-x86), and type:
-</p>
-<pre>
-    make linux-x86
-</pre>
-<p>This will produce libGL.so and several other libraries</p>
-</li>
-
-<li><b><em>DRI/accelerated</em></b> - The DRI hardware drivers for
-accelerated OpenGL rendering (for ATI, Intel, Matrox, etc) will be built.
-The libGL.so library will support the GLX extension and will load/use
-the DRI hardware drivers.
-
-
-<p>
-Build Mesa and the DRI hardware drivers by running
-</p>
-<pre>
-   make linux-dri
-</pre>
-<p>
-There are also <code>linux-dri-x86</code>, <code>linux-dri-x86-64</code>, 
-and <code>linux-ppc</code> configurations which are optimized for those
-architectures.
-</p>
-<p>
-Make sure you have the prerequisite versions of DRM and Xserver mentioned
-above.
-</p>
-
-</ul>
-
-
-<p>
-Later, if you want to rebuild for a different configuration run
-<code>make realclean</code> before rebuilding.
-</p>
-
-
-
-<a name="install">
-<H2>Installing the header and library files</H2>
-
-<p>
-The standard location for the OpenGL header files on Unix-type systems is
-in <code>/usr/include/GL/</code>.
-The standard location for the libraries is <code>/usr/lib/</code>.
-For more information see, the
-<a href="http://oss.sgi.com/projects/ogl-sample/ABI/" target="_parent">
-Linux/OpenGL ABI specification</a>.
-</p>
-
-<p>
-If you'd like Mesa to co-exist with another implementation of OpenGL that's
-already installed, you'll have to choose different directories, like
-<code>/usr/local/include/GL/</code> and <code>/usr/local/lib/</code>.
-</p>
-
-<p>
-To install Mesa's headers and libraries, run <code>make install</code>.
-But first, check the Mesa/configs/default file and examine the values
-of the <b>INSTALL_DIR</b> and <b>DRI_DRIVER_INSTALL_DIR</b> variables.
-Change them if needed, then run <code>make install</code>.
-</p>
-
-<p>
-The variable
-<b>DESTDIR</b> may also be used to install the contents to a temporary
-staging directory.
-This can be useful for package management.
-For example: <code>make install DESTDIR=/somepath/</code>
-</p>
-
-<p>
-Note: at runtime you can use the LD_LIBRARY_PATH environment variable
-(on Linux at least) to switch
-between the Mesa libraries and other vendor's libraries whenever you want.
-This is a handy way to compare multiple OpenGL implementations.
-</p>
-
-
-
-
 <a name="other">
-<H1>5. Building for other systems</H1>
+<H1>4. Building for other systems</H1>
 
 <p>
 Documentation for other environments (some may be very out of date):
@@ -255,7 +144,7 @@ Documentation for other environments (some may be very out of date):
 
 
 <a name="libs">
-<H1>6. Library Information</H1>
+<H1>5. Library Information</H1>
 
 <p>
 When compilation has finished, look in the top-level <code>lib/</code>
@@ -300,7 +189,7 @@ versions of libGL and device drivers.
 
 
 <a name="pkg-config">
-<H1>7. Building OpenGL programs with pkg-config</H1>
+<H1>6. Building OpenGL programs with pkg-config</H1>
 
 <p>
 Running <code>make install</code> will install package configuration files
diff --git a/mesalib/docs/mesa.css b/mesalib/docs/mesa.css
index 8e4944e44..a53a9df8b 100644
--- a/mesalib/docs/mesa.css
+++ b/mesalib/docs/mesa.css
@@ -1,33 +1,33 @@
-/* Mesa CSS */
-body {
-	background-color: #ffffff;
-	font: 14px 'Lucida Grande', Geneva, Arial, Verdana, sans-serif;
-	color: black;
- 	link: #111188;
-}
-
-h1 {
-	font: 24px 'Lucida Grande', Geneva, Arial, Verdana, sans-serif;
-	font-weight: bold;
-	color: black;
-}
-
-h2 {
-	font: 18px 'Lucida Grande', Geneva, Arial, Verdana, sans-serif, bold;
-	font-weight: bold;
-	color: black;
-}
-
-code {
-	font-family: monospace;
-	font-size: 10pt;
-	color: black;
-}
-
-
-pre {
-	/*font-family: monospace;*/
-	font-size: 10pt;
-	/*color: black;*/
-}
-
+/* Mesa CSS */
+body {
+	background-color: #ffffff;
+	font: 14px 'Lucida Grande', Geneva, Arial, Verdana, sans-serif;
+	color: black;
+ 	link: #111188;
+}
+
+h1 {
+	font: 24px 'Lucida Grande', Geneva, Arial, Verdana, sans-serif;
+	font-weight: bold;
+	color: black;
+}
+
+h2 {
+	font: 18px 'Lucida Grande', Geneva, Arial, Verdana, sans-serif, bold;
+	font-weight: bold;
+	color: black;
+}
+
+code {
+	font-family: monospace;
+	font-size: 10pt;
+	color: black;
+}
+
+
+pre {
+	/*font-family: monospace;*/
+	font-size: 10pt;
+	/*color: black;*/
+}
+
diff --git a/mesalib/docs/relnotes-7.10.html b/mesalib/docs/relnotes-7.10.html
index bb73975b7..3f4f229b0 100644
--- a/mesalib/docs/relnotes-7.10.html
+++ b/mesalib/docs/relnotes-7.10.html
@@ -1,2795 +1,2795 @@
-<HTML>
-
-<head>
-<TITLE>Mesa Release Notes</TITLE>
-<link rel="stylesheet" type="text/css" href="mesa.css">
-<meta http-equiv="content-type" content="text/html; charset=utf-8" />
-</head>
-
-<BODY>
-
-<body bgcolor="#eeeeee">
-
-<H1>Mesa 7.10 Release Notes / January 7, 2011</H1>
-
-<p>
-Mesa 7.10 is a new development release.
-People who are concerned with stability and reliability should stick
-with a previous release or wait for Mesa 7.10.1.
-</p>
-<p>
-Mesa 7.10 implements the OpenGL 2.1 API, but the version reported by
-glGetString(GL_VERSION) depends on the particular driver being used.
-Some drivers don't support all the features required in OpenGL 2.1.
-</p>
-<p>
-See the <a href="install.html">Compiling/Installing page</a> for prerequisites
-for DRI hardware acceleration.
-</p>
-
-
-<h2>MD5 checksums</h2>
-<pre>
-0a70c15c135561824bdcae92bf232e43  MesaLib-7.10.tar.gz
-33fb94eccc02cbb4d8d1365615e38e46  MesaLib-7.10.tar.bz2
-5cafdc0eda0f9bf370b95c98df3338fa  MesaLib-7.10.zip
-bc644be551ed585fc4f66c16b64a91c9  MesaGLUT-7.10.tar.gz
-5c2677a155672352d62b177e4f0f92e8  MesaGLUT-7.10.tar.bz2
-2ce5001f74496d1ba719ef74d910a5cf  MesaGLUT-7.10.zip
-</pre>
-
-
-<h2>New features</h2>
-<ul>
-<li>GL_ARB_explicit_attrib_location extension (Intel and software drivers).
-<li>GL_ARB_texture_rg (Intel, software drivers, gallium drivers).
-<li>GL_EXT_separate_shader_objects extension (Intel and software drivers).
-<li>GL_NV_primitive_restart extension (Gallium softpipe, llvmpipe).
-<li>New fragment shader back-end for i965-class hardware.
-<li>Support for Sandybridge chipset in i965 DRI driver.
-</ul>
-
-
-<h2>Bug fixes</h2>
-<p>This list is likely incomplete.</p>
-<ul>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28800">Bug 28800</a> - [r300c, r300g] Texture corruption with World of Warcraft</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29420">Bug 29420</a> - Amnesia / HPL2 RendererFeatTest - not rendering correctly</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29946">Bug 29946</a> - [swrast] piglit valgrind glsl-array-bounds-04 fails</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30261">Bug 30261</a> - [GLSL 1.20] allowing inconsistent invariant declaration between two vertex shaders</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30632">Bug 30632</a> - [softpipe] state_tracker/st_manager.c:489: st_context_notify_invalid_framebuffer: Assertion `stfb &amp;&amp; stfb-&gt;iface == stfbi' failed.</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30694">Bug 30694</a> - wincopy will crash on Gallium drivers when going to front buffer</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30771">Bug 30771</a> -  [r600g] vert-tex glsl demo</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30787">Bug 30787</a> - Invalid asm shader does not generate draw-time error when used with GLSL shader</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30974">Bug 30974</a> - [llvmpipe] SIGABRT src/gallium/drivers/llvmpipe/lp_state_fs.c:779</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30993">Bug 30993</a> - getFramebufferAttachmentParameteriv wrongly generates error</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31101">Bug 31101</a> -  [glsl2] abort() in ir_validate::visit_enter(ir_assignment *ir)</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31193">Bug 31193</a> -  [regression] aa43176e break water reflections</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31194">Bug 31194</a> - The mesa meta save/restore code doesn't ref the current GLSL program</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31371">Bug 31371</a> - glslparsertest: ir.cpp:358: ir_constant::ir_constant(const glsl_type*, const ir_constant_data*): Assertion `(type->base_type &gt;= 0) &amp;&amp; (type->base_type &lt;= 3)' failed.</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31439">Bug 31439</a> - Crash in glBufferSubData() with size == 0</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31495">Bug 31495</a> - [i965 gles2c bisected] OpenGL ES 2.0 conformance GL2Tests_GetBIFD_input.run regressed</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31514">Bug 31514</a> - isBuffer returns true for unbound buffers</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31560">Bug 31560</a> - [tdfx] tdfx_tex.c:702: error: ‘const struct gl_color_table’ has no member named ‘Format’</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31617">Bug 31617</a> - Radeon/Compiz: 'failed to attach dri2 front buffer', error case not handled</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31648">Bug 31648</a> -  [GLSL] array-struct-array gets assertion: `(size &gt;= 1) && (size &lt;= 4)' failed.</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31650">Bug 31650</a> - [GLSL] varying gl_TexCoord fails to be re-declared to different size in the second shader</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31673">Bug 31673</a> - GL_FRAGMENT_PRECISION_HIGH preprocessor macro undefined in GLSL ES</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31690">Bug 31690</a> -  i915 shader compiler fails to flatten if in Aquarium webgl demo.</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31832">Bug 31832</a> - [i915] Bad renderbuffer format: 21</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31841">Bug 31841</a> - [drm:radeon_cs_ioctl] *ERROR* Invalid command stream !</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31894">Bug 31894</a> - Writing to gl_PointSize with GLES2 corrupts other varyings</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31909">Bug 31909</a> - [i965] brw_fs.cpp:1461: void fs_visitor::emit_bool_to_cond_code(ir_rvalue*): Assertion `expr-&gt;operands[i]-&gt;type-&gt;is_scalar()' failed.</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31934">Bug 31934</a> - [gallium] Mapping empty buffer object causes SIGSEGV</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31983">Bug 31983</a> -  [i915 gles2] "if (expression with builtin/varying variables) discard" breaks linkage</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31985">Bug 31985</a> - [GLSL 1.20] initialized uniform array considered as "unsized"</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31987">Bug 31987</a> - [gles2] if input a wrong pname(GL_NONE) to glGetBoolean, it will not case GL_INVALID_ENUM</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32035">Bug 32035</a> - [GLSL bisected] comparing unsized array gets segfault</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32070">Bug 32070</a> - llvmpipe renders stencil demo incorrectly</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32273">Bug 32273</a> - assertion fails when starting vdrift 2010 release with shaders enabled</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32287">Bug 32287</a> - [bisected GLSL] float-int failure</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32311">Bug 32311</a> - [965 bisected] Array look-ups broken on GM45</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32520">Bug 32520</a> -  [gles2] glBlendFunc(GL_ZERO, GL_DST_COLOR) will result in GL_INVALID_ENUM</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32825">Bug 32825</a> - egl_glx driver completely broken in 7.9 branch [fix in master]</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=">Bug </a> - </li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=">Bug </a> - </li>
-
-</ul>
-
-
-<h2>Changes</h2>
-<p>Adam Jackson (2):
-<ul>
-  <li>i965: Update renderer strings for sandybridge</li>
-  <li>drivers/x11: unifdef XFree86Server</li>
-</ul></p>
-
-<p>Alex Deucher (30):
-<ul>
-  <li>r600c: fix mipmap stride on evergreen</li>
-  <li>r600c: add reloc for CB_COLOR0_ATTRIB</li>
-  <li>r600c: pull over 6xx/7xx vertex fixes for evergreen</li>
-  <li>r600c: fix segfault in evergreen stencil code</li>
-  <li>r100: revalidate after radeon_update_renderbuffers</li>
-  <li>r600c: add missing radeon_prepare_render() call on evergreen</li>
-  <li>r600c: properly align mipmaps to group size</li>
-  <li>egl_dri2: Add radeon chip ids</li>
-  <li>r600c/evergreen: texture align is group_bytes just like 6xx/7xx</li>
-  <li>r600g: fix buffer alignment</li>
-  <li>r600g: All EVENT_WRITE packets need the EVENT_INDEX field</li>
-  <li>r600g: translate ARR instruction for evergreen</li>
-  <li>r600g: use meaningful defines for chiprev</li>
-  <li>r600g: use full range of VS resources for vertex samplers</li>
-  <li>r600g: fix additional EVENT_WRITE packet</li>
-  <li>r600g: fix some winsys functions to deal properly with evergreen</li>
-  <li>r600c: add Ontario Fusion APU support</li>
-  <li>r600g: add support for ontario APUs</li>
-  <li>r600c: fix VC flush on cedar and palm</li>
-  <li>gallium/egl: fix r300 vs r600 loading</li>
-  <li>r600c: fix some opcodes on evergreen</li>
-  <li>r600c: bump texture limits to hw limits</li>
-  <li>r600g: bump texture/cb limits appropriately for evergreen</li>
-  <li>radeon: bump mip tree levels to 15</li>
-  <li>r600g: fix rendering with a vertex attrib having a zero stride</li>
-  <li>r600g: remove useless switch statements</li>
-  <li>r600g: add support for NI (northern islands) asics</li>
-  <li>r600c: add support for NI asics</li>
-  <li>r600g: support up to 64 shader constants</li>
-  <li>r600c: fix up SQ setup in blit code for Ontario/NI</li>
-</ul></p>
-
-<p>Andre Maasikas (3):
-<ul>
-  <li>r600c: fix buffer height setting in dri2 case</li>
-  <li>r600g: break alu clause earlier</li>
-  <li>r600g: fix evergreen interpolation setup</li>
-</ul></p>
-
-<p>Andrew Randrianasulu (2):
-<ul>
-  <li>dri/nv04: Don't expose ARB_texture_env_combine/dot3.</li>
-  <li>dri/nv04: Enable eng3dm for A8/L8 textures.</li>
-</ul></p>
-
-<p>Aras Pranckevicius (2):
-<ul>
-  <li>glsl: fix crash in loop analysis when some controls can't be determined</li>
-  <li>glsl: fix matrix type check in ir_algebraic</li>
-</ul></p>
-
-<p>Bas Nieuwenhuizen (3):
-<ul>
-  <li>r600g: set ENABLE_KILL in the shader state in the new design</li>
-  <li>r600g: set ENABLE_KILL on evergreen too</li>
-  <li>r600g: use dirty list to track dirty blocks</li>
-</ul></p>
-
-<p>Ben Skeggs (3):
-<ul>
-  <li>nv50: DST</li>
-  <li>nv50: DPH</li>
-  <li>nv50: silence some unknown get_param warnings</li>
-</ul></p>
-
-<p>Benjamin Franzke (2):
-<ul>
-  <li>st/egl image: multiply drm buf-stride with blocksize</li>
-  <li>r600g: implement texture_get_handle (needed for eglExportDRMImageMESA)</li>
-</ul></p>
-
-<p>Brian Paul (296):
-<ul>
-  <li>glx: add const qualifiers to __indirect_glMultiDrawArraysEXT()</li>
-  <li>glsl2: fix signed/unsigned comparison warning</li>
-  <li>llvmpipe: cast to silence warning</li>
-  <li>llvmpipe: s/boolean/unsigned/ in bitfield to silence warning</li>
-  <li>nv50: use unsigned int for bitfields to silence warnings</li>
-  <li>tgsi: fix incorrect usage_mask for shadow tex instructions</li>
-  <li>gallivm: expand AoS sampling to cover all filtering modes</li>
-  <li>gallivm: fix incorrect vector shuffle datatype</li>
-  <li>gallivm: move i32_vec_type inside the #ifdef</li>
-  <li>mesa: include mfeatures.h in formats.c</li>
-  <li>gallivm: fix wrong return value in bitwise functions</li>
-  <li>tgsi/sse: fix aos_to_soa() loop to handle num_inputs==0</li>
-  <li>gallivm: added missing case for PIPE_TEXTURE_RECT</li>
-  <li>gallium: better docs for pipe_rasterizer_state::sprite_coord_enable</li>
-  <li>gallium: rework handling of sprite_coord_enable state</li>
-  <li>gallium/docs: added new pipeline.txt diagram</li>
-  <li>mesa: don't call valid_texture_object() in non-debug builds</li>
-  <li>glsl2: silence compiler warnings in printf() calls</li>
-  <li>docs: remove old broken link</li>
-  <li>docs: mark as obsolete, remove dead links</li>
-  <li>llvmpipe: fix query bug when no there's no scene</li>
-  <li>gallivm: remove debug code</li>
-  <li>llvmpipe: maintain fragment shader state for draw module</li>
-  <li>llvmpipe: indentation fix</li>
-  <li>llvmpipe: reformatting, remove trailing whitespace, etc</li>
-  <li>llvmpipe: clean-up, comments in setup_point_coefficient()</li>
-  <li>llvmpipe: rename sprite field, add sprite_coord_origin</li>
-  <li>llvmpipe: implement sprite coord origin modes</li>
-  <li>draw: fix test for using the wide-point stage</li>
-  <li>llvmpipe: check bitshift against PIPE_MAX_SHADER_OUTPUTS</li>
-  <li>draw: check bitshift against PIPE_MAX_SHADER_OUTPUS</li>
-  <li>Merge branch 'sprite-coord'</li>
-  <li>draw: new draw_fs.[ch] files</li>
-  <li>glsl2: fix typo in error msg</li>
-  <li>gallivm: fix lp_build_sample_compare()</li>
-  <li>softpipe: add missing calls to set draw vertex samplers/views</li>
-  <li>mesa: don't advertise bogus GL_ARB_shading_language_120 extension</li>
-  <li>configs: remove egl-swrast target from linux-dri config</li>
-  <li>llvmpipe: fix sprite texcoord setup for non-projective texturing</li>
-  <li>mesa: fix assertions to handle srgb formats</li>
-  <li>st/mesa: add missing MESA_FORMAT_S8 case in st_mesa_format_to_pipe_format()</li>
-  <li>st/mesa: use the wrapped renderbuffer in CopyPixels()</li>
-  <li>llvmpipe: make min/max lod and lod bias dynamic state</li>
-  <li>llvmpipe: make texture border_color dynamic state</li>
-  <li>softpipe: fix repeat() function for NPOT textures</li>
-  <li>gallivm: fix repeat() function for NPOT textures</li>
-  <li>swrast: update comments for REMAINDER() macro</li>
-  <li>softpipe: rename sp_state_fs.c -&gt; sp_state_shader.c</li>
-  <li>softpipe: make shader-related functions static</li>
-  <li>softpipe: make blend/stencil/depth functions static</li>
-  <li>softpipe: make sampler state functions static</li>
-  <li>softpipe: make vertex state functions static</li>
-  <li>softpipe: make rasterizer state functions static</li>
-  <li>softpipe: make stream out state functions static</li>
-  <li>softpipe: make clip state functions static</li>
-  <li>softpipe: minor asst. clean-ups</li>
-  <li>softpipe: allocate tile data on demand</li>
-  <li>llvmpipe: fix swizzling of texture border color</li>
-  <li>softpipe: fix swizzling of texture border color</li>
-  <li>draw: pass sampler state down to llvm jit state</li>
-  <li>gallivm: check for level=0 case in lp_build_minify()</li>
-  <li>gallivm: added some comments</li>
-  <li>draw: check for null sampler pointers</li>
-  <li>swrast: fix choose_depth_texture_level() to respect mipmap filtering state</li>
-  <li>st/mesa: replace assertion w/ conditional in framebuffer invalidation</li>
-  <li>glsl2: fix signed/unsigned comparison warning</li>
-  <li>st/xlib: add some comments</li>
-  <li>ir_to_mesa: assorted clean-ups, const qualifiers, new comments</li>
-  <li>mesa: remove assertion w/ undeclared variable texelBytes</li>
-  <li>gallivm: remove newlines</li>
-  <li>draw/llvmpipe: replace DRAW_MAX_TEXTURE_LEVELS with PIPE_MAX_TEXTURE_LEVELS</li>
-  <li>mesa: reformatting, comments, code movement</li>
-  <li>x11: fix breakage from gl_config::visualType removal</li>
-  <li>gallivm: work-around trilinear mipmap filtering regression with LLVM 2.8</li>
-  <li>mesa: remove post-convolution width/height vars</li>
-  <li>gallivm: add compile-time option to emit inst addrs and/or line numbers</li>
-  <li>llvmpipe: code to dump bytecode to file (disabled)</li>
-  <li>gallivm: added lp_build_print_vec4()</li>
-  <li>gallivm: added lp_build_load_volatile()</li>
-  <li>glsl: add ir_unop_round_even case to silence unhandled enum warning</li>
-  <li>st/mesa: fix regressions in glDrawPixels(GL_STENCIL_INDEX)</li>
-  <li>st/mesa: reformatting in st_cb_drawpixels.c</li>
-  <li>st/mesa: use GLuint to avoid problem w/ uint not defined on mingw32</li>
-  <li>st/mesa: update function name, comments</li>
-  <li>gallivm: use util_snprintf()</li>
-  <li>llvmpipe: remove lp_setup_coef*.c files from Makefile</li>
-  <li>mesa: fix mesa version string construction</li>
-  <li>gallivm: fix incorrect type for zero vector in emit_kilp()</li>
-  <li>llvmpipe/draw: always enable LLVMAddInstructionCombiningPass()</li>
-  <li>draw: use float version of LLVM Mul/Add instructions</li>
-  <li>draw: fix typo in comment</li>
-  <li>mesa: add GL_RG case to _mesa_source_buffer_exists()</li>
-  <li>mesa: add missing cases for packing red/green images</li>
-  <li>st/mesa: added cases for GL_COMPRESSED_RED/RG in st_choose_format()</li>
-  <li>docs: update texture red/green support in GL3.txt</li>
-  <li>docs: add GL_ARB_texture_rg to release notes</li>
-  <li>mesa: driver hook for primitive restart</li>
-  <li>mesa: set/get primitive restart state</li>
-  <li>mesa: API spec for primitive restart</li>
-  <li>mesa: regenerated files with primitive restart</li>
-  <li>mesa: plug in primitive restart function</li>
-  <li>vbo: support for primitive restart</li>
-  <li>gallium: new CAP, state for primitive restart</li>
-  <li>st/mesa: support for primitive restart</li>
-  <li>draw: implement primitive splitting for primitive restart</li>
-  <li>softpipe: enable primitive restart</li>
-  <li>llvmpipe: enable primitive restart</li>
-  <li>docs: added GL_NV_primitive_restart extension</li>
-  <li>Merge branch 'primitive-restart-cleanup'</li>
-  <li>winsys/xlib: formatting fixes</li>
-  <li>winsys/xlib: use Bool type for shm field</li>
-  <li>winsys/xlib: fix up allocation/dealloction of XImage</li>
-  <li>winsys/xlib: rename xm-&gt;xlib</li>
-  <li>galahad: silence warnings</li>
-  <li>mesa: move declaration before code</li>
-  <li>docs: updated GL3 status for primitive restart</li>
-  <li>mesa: 80-column wrapping</li>
-  <li>mesa: simplify fbo format checking code</li>
-  <li>mesa: split up the image.c file</li>
-  <li>mesa: add pixel packing for unscaled integer types</li>
-  <li>mesa: _mesa_ClearColorIuiEXT() and _mesa_ClearColorIiEXT()</li>
-  <li>mesa: _mesa_is_format_integer() function</li>
-  <li>mesa: minor reformatting, clean-ups</li>
-  <li>mesa: added _mesa_is_fragment_shader_active() helper</li>
-  <li>mesa: new glDrawPixels error check for integer formats</li>
-  <li>softpipe: added some texture sample debug code (disabled)</li>
-  <li>mesa: added new gl_extensions::EXT_gpu_shader4 field</li>
-  <li>mesa: added new gl_framebuffer::_IntegerColor field</li>
-  <li>mesa: added glGet query for GL_RGBA_INTEGER_MODE_EXT</li>
-  <li>mesa: compute _IntegerColor field in _mesa_test_framebuffer_completeness()</li>
-  <li>mesa: added cases for GL_EXT_texture_integer formats</li>
-  <li>mesa: added cases for GL_EXT_texture_integer</li>
-  <li>st/mesa: add format selection for signed/unsigned integer formats</li>
-  <li>mesa: simplify target_can_be_compressed() function</li>
-  <li>glapi: GL_EXT_texture_integer API</li>
-  <li>glapi: include/build EXT_texture_integer.xml</li>
-  <li>mesa: regenerated API files for GL_EXT_texture_integer</li>
-  <li>mesa: plug in GL_EXT_texture_integer functions</li>
-  <li>mesa: display list support for GL_EXT_texture_integer</li>
-  <li>st/mesa: be smarter choosing texture format for glDrawPixels()</li>
-  <li>softpipe: remove &gt;32bpp color restriction</li>
-  <li>mesa: silence enum comparison warning</li>
-  <li>mesa: fix uninitialized var warning</li>
-  <li>xlib: silence unused var warning</li>
-  <li>util: use pointer_to_func() to silence warning</li>
-  <li>rtasm: use pointer_to_func() to silence warning</li>
-  <li>translate: use function typedefs, casts to silence warnings</li>
-  <li>translate: remove unused prototypes</li>
-  <li>mesa: additional glReadPixels error checks for GL_EXT_texture_integer</li>
-  <li>mesa: additional switch cases for GL_EXT_texture_integer</li>
-  <li>mesa: additional teximage error checks for GL_EXT_texture_integer</li>
-  <li>mesa: do integer FB / shader validation check in _mesa_valid_to_render()</li>
-  <li>mesa: call _mesa_valid_to_render() in glDrawPixels, glCopyPixels, glBitmap</li>
-  <li>mesa: remove the unused _mesa_is_fragment_shader_active() function</li>
-  <li>mesa: fix bug in _mesa_is_format_integer()</li>
-  <li>mesa: rename function to _mesa_is_format_integer_color()</li>
-  <li>mesa: remove 'normalized' parameter from _mesa_VertexAttribIPointer()</li>
-  <li>vbo: re-indent file</li>
-  <li>glapi: xml spec file for GL_EXT_gpu_shader4</li>
-  <li>glapi: include EXT_gpu_shader4.xml</li>
-  <li>glapi: regenerated API files</li>
-  <li>mesa: plug in stubs for glBindFragDataLocation(), glGetFragDataLocation()</li>
-  <li>mesa: add glGetUniformuiv(), plug in uint glUniform funcs</li>
-  <li>mesa: plug in more GL_EXT_gpu_shader4 functions</li>
-  <li>mesa: add new GLvertexformat entries for integer-valued attributes</li>
-  <li>mesa: implement integer-valued vertex attribute functions</li>
-  <li>mesa: add gl_client_array::Integer field and related vertex array state code</li>
-  <li>mesa: consolidate glVertex/Color/etcPointer() code</li>
-  <li>mesa: state/queries for GL_MIN/MAX_PROGRAM_TEXEL_OFFSET_EXT</li>
-  <li>mesa: glArrayElement support for integer-valued arrays</li>
-  <li>mesa: clean-up array element code</li>
-  <li>mesa: add extension table entry for GL_EXT_gpu_shader4</li>
-  <li>mesa: remove obsolete comment</li>
-  <li>mesa: fix incorrect type in _mesa_texstore_rgba_int16()</li>
-  <li>mesa: fix integer cases in _mesa_is_legal_format_and_type()</li>
-  <li>mesa: add const qualifier to _mesa_is_legal_format_and_type()</li>
-  <li>mesa: additional integer formats in _mesa_bytes_per_pixel()</li>
-  <li>mesa: pixel transfer ops do not apply to integer-valued textures</li>
-  <li>mesa: remove dead code</li>
-  <li>osmesa: fix renderbuffer memleak in OSMesaMakeCurrent()</li>
-  <li>mesa: use GLubyte for edge flag arrays</li>
-  <li>mesa: move the gl_config struct declaration</li>
-  <li>dri/util: add a bunch of comments</li>
-  <li>mesa: remove always-false conditional in check_compatible()</li>
-  <li>mesa: fix aux/accum comment and error message mixups</li>
-  <li>llvmpipe: assign context's frag shader pointer before using it</li>
-  <li>llvmpipe: add a cast</li>
-  <li>mesa: silence new warnings in texobj.c</li>
-  <li>egl/gdi: fix typo: xsurf-&gt;gsurf</li>
-  <li>mesa: code to unpack RGBA as uints</li>
-  <li>gallivm: implement scatter stores into temp register file</li>
-  <li>gallivm: add some LLVM var labels</li>
-  <li>gallivm: added debug code to dump temp registers</li>
-  <li>gallivm: add pixel offsets in scatter stores</li>
-  <li>gallivm: added lp_elem_type()</li>
-  <li>gallivm: implement execution mask for scatter stores</li>
-  <li>tgsi: remove unused function</li>
-  <li>llvmpipe: added some debug assertions, but disabled</li>
-  <li>gallivm: alloca() was called too often for temporary arrays</li>
-  <li>gallivm: add const qualifiers, fix comment string</li>
-  <li>softpipe: disable vertex texturing with draw/llvm</li>
-  <li>mesa: consolidate pixel packing/unpacking code</li>
-  <li>mesa: rename vars in pixel pack/unpack code</li>
-  <li>mesa: implement uint texstore code</li>
-  <li>mesa: remove stray GL_FLOAT case in _mesa_is_legal_format_and_type()</li>
-  <li>mesa: make fixed-pt and byte-valued arrays a runtime feature</li>
-  <li>softpipe: can't no-op depth test stage when occlusion query is enabled</li>
-  <li>mesa: no-op glBufferSubData() on size==0</li>
-  <li>mesa: #include mfeatures.h in enums.h</li>
-  <li>mesa: improve error message</li>
-  <li>mesa: add missing formats in _mesa_format_to_type_and_comps()</li>
-  <li>mesa: handle more pixel types in mipmap generation code</li>
-  <li>mesa: make glIsBuffer() return false for never bound buffers</li>
-  <li>mesa: fix glDeleteBuffers() regression</li>
-  <li>tdfx: s/Format/_BaseFormat/</li>
-  <li>mesa: consolidate assertions in teximage code</li>
-  <li>radeon: set gl_texture_image::TexFormat field in radeonSetTexBuffer2()</li>
-  <li>r600: set gl_texture_image::TexFormat field in r600SetTexBuffer2()</li>
-  <li>r200: set gl_texture_image::TexFormat field in r200SetTexBuffer2()</li>
-  <li>r300: set gl_texture_image::TexFormat field in r300SetTexBuffer2()</li>
-  <li>evergreen: set gl_texture_image::TexFormat field in evergreenSetTexBuffer()</li>
-  <li>st/mesa: fix glDrawPixels(depth/stencil) bugs</li>
-  <li>glsl: fix assorted MSVC warnings</li>
-  <li>mesa: add more work-arounds for acoshf(), asinhf(), atahf()</li>
-  <li>glsl: remove opt_constant_expression.cpp from SConscript</li>
-  <li>mesa: fix error messages and minor reindenting</li>
-  <li>mesa: whitespace cleanups</li>
-  <li>mesa: 80-column wrapping</li>
-  <li>mesa: reorder texture_error_check() params</li>
-  <li>mesa: minor clean-ups in context code</li>
-  <li>mesa: upgrade to glext.h version 66</li>
-  <li>mesa: pass gl_format to _mesa_init_teximage_fields()</li>
-  <li>mesa: fix error msg typo</li>
-  <li>glapi: rename GL3.xml to GL3x.xml as it covers all GL 3.x versions</li>
-  <li>mesa: hook up GL 3.x entrypoints</li>
-  <li>docs: update some GL 3.0 status</li>
-  <li>mesa: fix get_texture_dimensions() for texture array targets</li>
-  <li>swrast: init alpha value to 1.0 in opt_sample_rgb_2d()</li>
-  <li>glsl: fix off by one in register index assertion</li>
-  <li>glsl: use gl_register_file in a few places</li>
-  <li>mesa: rename, make _mesa_register_file_name() non-static</li>
-  <li>mesa: _mesa_valid_register_index() to validate register indexes</li>
-  <li>mesa: replace #defines with new gl_shader_type enum</li>
-  <li>mesa: use gl_shader_type enum</li>
-  <li>glsl: better handling of linker failures</li>
-  <li>glsl: start restoring some geometry shader code</li>
-  <li>mesa: add assertion and update comment in _mesa_format_image_size()</li>
-  <li>mesa: added _mesa_format_image_size64()</li>
-  <li>x11: remove test_proxy_teximage() function</li>
-  <li>st/mesa: fix mapping of zero-sized buffer objects</li>
-  <li>gallivm/llvmpipe: squash merge of the llvm-context branch</li>
-  <li>mesa: raise max texture sizes to 16K</li>
-  <li>softpipe: increase max texture size to 16K</li>
-  <li>mesa: replace large/MAX_WIDTH stack allocations with heap allocations</li>
-  <li>mesa: replace large/MAX_WIDTH stack allocations with heap allocations</li>
-  <li>swrast: avoid large stack allocations in blend code</li>
-  <li>swrast: avoid large stack allocations in tex combine code</li>
-  <li>st/mesa: avoid large stack allocations in readpixels code</li>
-  <li>mesa: replace more MAX_WIDTH stack allocations with heap allocations</li>
-  <li>gallivm/llvmpipe: remove lp_build_context::builder</li>
-  <li>gallivm: fix null builder pointers</li>
-  <li>mesa: fix GL_FRAMEBUFFER_ATTACHMENT_OBJECT_NAME query</li>
-  <li>mesa: return GL_FRAMEBUFFER_DEFAULT as FBO attachment type</li>
-  <li>llvmpipe: fix broken stencil writemask</li>
-  <li>mesa: consolidate some compiler -D flags</li>
-  <li>swrast: allow GL_RG format in glDrawPixels()</li>
-  <li>swrast: fix indentation</li>
-  <li>swrast: accept GL_RG in glReadPixels()</li>
-  <li>swrast: restructure some glReadPixels() code</li>
-  <li>mesa: make glGet*(GL_NONE) generate GL_INVALID_ENUM</li>
-  <li>mesa: remove unneeded cast</li>
-  <li>mesa: update comments, remove dead code</li>
-  <li>st/mesa: new comment about updating state vars</li>
-  <li>mesa: add error margin to clip mask debug/check code</li>
-  <li>gallium/util: minor formatting fixes</li>
-  <li>mesa/llvm: use llvm-config --cppflags</li>
-  <li>st/mesa: fix mipmap generation bug</li>
-  <li>mesa: test for cube map completeness in glGenerateMipmap()</li>
-  <li>mesa: set gl_texture_object::_Complete=FALSE in incomplete()</li>
-  <li>mesa: consolidate glTexImage1/2/3D() code</li>
-  <li>mesa: simplify proxy texture code in texture_error_check()</li>
-  <li>mesa: consolidate the glTexSubImage1/2/3D() functions</li>
-  <li>mesa: consolidate glCopyTexImage1/2D() code</li>
-  <li>mesa: consolidate glCopyTexSubImage1/2/3D() functions</li>
-  <li>mesa: consolidate glCompressedTexImage1/2/3D() functions</li>
-  <li>mesa: make _mesa_test_proxy_teximage() easier to read</li>
-  <li>configure: use llvm-config --cppflags instead of --cflags</li>
-  <li>mesa: revamp error checking for compressed texture images</li>
-  <li>mesa: simplify target checking for TexImage functions</li>
-  <li>draw/llvm: don't flush in vs_llvm_delete()</li>
-  <li>tnl: Initialize gl_program_machine memory in run_vp.</li>
-  <li>tnl: a better way to initialize the gl_program_machine memory</li>
-  <li>mesa, st/mesa: disable GL_ARB_geometry_shader4</li>
-  <li>mesa/meta: fix broken assertion, rename stack depth var</li>
-  <li>glsl: new glsl_strtod() wrapper to fix decimal point interpretation</li>
-  <li>st/mesa: fix renderbuffer pointer check in st_Clear()</li>
-</ul></p>
-
-<p>Brian Rogers (1):
-<ul>
-  <li>mesa: Add missing else in do_row_3D</li>
-</ul></p>
-
-<p>Chad Versace (25):
-<ul>
-  <li>intel_extensions: Add ability to set GLSL version via environment</li>
-  <li>glsl: Add glsl_type::uvecN_type for N=2,3</li>
-  <li>glsl: Add lexer rules for uint and uvecN (N=2..4)</li>
-  <li>glsl: Changes in generated file glsl_lexer.cpp</li>
-  <li>glsl: Add lexer rules for &lt&lt and &gt;&gt; in GLSL 1.30</li>
-  <li>glsl: Change generated file glsl_lexer.cpp</li>
-  <li>glsl: Implement ast-to-hir for binary shifts in GLSL 1.30</li>
-  <li>glsl: Implement constant expr evaluation for bitwise-not</li>
-  <li>glsl: Implement constant expr evaluation for bit-shift ops</li>
-  <li>glsl: Implement constant expr evaluation for bitwise logic ops</li>
-  <li>glsl: Fix ir validation for bit logic ops</li>
-  <li>glsl: Define shift_result_type() in ast_to_hir.cpp</li>
-  <li>glsl: Implement ast-to-hir for bit-shift-assignment</li>
-  <li>glsl: Define bit_logic_result_type() in ast_to_hir.cpp</li>
-  <li>glsl: Implement ast-to-hir for bit-logic ops</li>
-  <li>glsl: Fix lexer rule for ^=</li>
-  <li>glsl: Commit generated file glsl_lexer.cpp</li>
-  <li>glsl: Fix ast-to-hir for ARB_fragment_coord_conventions</li>
-  <li>mesa: Fix C++ includes in sampler.cpp</li>
-  <li>glsl: Fix ir_expression::constant_expression_value()</li>
-  <li>glsl: Fix erroneous cast in ast_jump_statement::hir()</li>
-  <li>glsl: Fix Doxygen tag file in recently renamed files</li>
-  <li>glsl: Improve usage message for glsl_compiler</li>
-  <li>glsl: Fix linker bug in cross_validate_globals()</li>
-  <li>glsl: In ast_to_hir, check sampler array indexing</li>
-</ul></p>
-
-<p>Chia-I Wu (149):
-<ul>
-  <li>glapi: Fix build errors for ES.</li>
-  <li>glapi: Fix ES build errors again.</li>
-  <li>mesa: Update ES APIspec.xml.</li>
-  <li>st/xlib: Notify the context when the front/back buffers are swapped.</li>
-  <li>targets/egl: Use C++ compiler to link GL/ES state trackers.</li>
-  <li>libgl-xlib: Remove unused st_api_create_OpenGL.</li>
-  <li>st/egl: Split modeset code support to modeset.c.</li>
-  <li>st/egl: Rename kms backend to drm.</li>
-  <li>st/egl: s/kms/drm/ on the drm backend.</li>
-  <li>egl: Enable drm platform by default.</li>
-  <li>egl: Check extensions.</li>
-  <li>st/egl: Skip single-buffered configs in EGL.</li>
-  <li>mapi: Fix compiler warnings.</li>
-  <li>st/egl: Drop context argument from egl_g3d_get_egl_image.</li>
-  <li>targets/egl: Fix linking with libdrm.</li>
-  <li>st/vega: Fix version check in context creation.</li>
-  <li>egl: Use attribute names as the _EGLConfig member names.</li>
-  <li>egl: Access config attributes directly.</li>
-  <li>st/egl: Access _EGLConfig directly.</li>
-  <li>st/egl: Do not finish a fence that is NULL.</li>
-  <li>mesa: Remove unused vtxfmt_tmp.h.</li>
-  <li>egl_dri2: Drop the use of _egl[SG]etConfigKey.</li>
-  <li>egl_glx: Drop the use of [SG]ET_CONFIG_ATTRIB.</li>
-  <li>egl_glx: Fix borken driver.</li>
-  <li>egl: Move attributes in _EGLImage to _EGLImageAttribs.</li>
-  <li>egl: Parse image attributes with _eglParseImageAttribList.</li>
-  <li>egl: Move fallback routines to eglfallbacks.c.</li>
-  <li>egl: Drop dpy argument from the link functions.</li>
-  <li>egl: Minor changes to the _EGLConfig interface.</li>
-  <li>egl: Minor changes to the _EGLScreen interface.</li>
-  <li>egl: Fix _eglModeLookup.</li>
-  <li>st/egl: Fix native_mode refresh mode.</li>
-  <li>egl: Add reference count for resources.</li>
-  <li>egl: Use reference counting to replace IsLinked or IsBound.</li>
-  <li>egl: Fix a false negative check in _eglCheckMakeCurrent.</li>
-  <li>st/egl: Use resource reference count for egl_g3d_sync.</li>
-  <li>egl_dri2: Fix a typo that make glFlush be called at wrong time.</li>
-  <li>glapi: Do not use glapidispatch.h.</li>
-  <li>glapi: Move glapidispatch.h to core mesa.</li>
-  <li>glapi: Do not use glapioffsets.h.</li>
-  <li>glapi: Merge glapioffsets.h into glapidispath.h.</li>
-  <li>vbo: Use CALL_* macros.</li>
-  <li>mesa: Remove unnecessary glapitable.h includes.</li>
-  <li>autoconf: Better client API selection.</li>
-  <li>docs: Update egl and openvg docs.</li>
-  <li>autoconf: Update configuration info.</li>
-  <li>Merge branch 'glapi-reorg'</li>
-  <li>targets: Add missing quotes to Makefile.xorg.</li>
-  <li>autoconf: st/vega requires --enable-openvg.</li>
-  <li>st/mesa: Unreference the sampler view in st_bind_surface.</li>
-  <li>autoconf: Tidy configure output for EGL.</li>
-  <li>targets/egl: Fix a warning with --disable-opengl build.</li>
-  <li>egl: Rework _eglGetSearchPath.</li>
-  <li>mesa: Select FEATURE_remap_table when multiple APIs are enabled.</li>
-  <li>mesa: Allow contexts of different APIs to coexist.</li>
-  <li>egl: Set up the pthread key even TLS is used.</li>
-  <li>st/egl: Add native_surface::present callback.</li>
-  <li>st/egl: Use native_surface::present callback.</li>
-  <li>d3d1x: Use native_surface::present.</li>
-  <li>st/egl: Remove flush_frontbuffer and swap_buffers.</li>
-  <li>st/egl: Add support for swap interval and swap behavior.</li>
-  <li>st/egl: Add support for EGL_MATCH_NATIVE_PIXMAP.</li>
-  <li>st/egl: Add extern "C" wrapper to native.h.</li>
-  <li>st/egl: Add native_display_buffer interface.</li>
-  <li>st/egl: Use native_display_buffer for EGL_MESA_drm_image.</li>
-  <li>autoconf: Add --enable-gallium-egl.</li>
-  <li>docs: Update egl docs.</li>
-  <li>st/dri: Add support for surfaceless current contexts.</li>
-  <li>egl_dri2: Fix __DRI_DRI2 version 1 support.</li>
-  <li>st/vega: Do not wait NULL fences.</li>
-  <li>gallium: Add st_api::name.</li>
-  <li>gallium: Add st_context_iface::share to st_api.</li>
-  <li>st/wgl: Use st_context_iface::share for DrvShareLists.</li>
-  <li>st/glx: Replace MESA_VERSION_STRING by xmesa_get_name.</li>
-  <li>mesa: Clean up core.h.</li>
-  <li>scons: Define IN_DRI_DRIVER.</li>
-  <li>tgsi: Add STENCIL to text parser.</li>
-  <li>st/vega: vegaLookupSingle should validate the state.</li>
-  <li>st/vega: Set wrap_r for mask and blend samplers.</li>
-  <li>st/vega: Fix vgReadPixels with a subrectangle.</li>
-  <li>egl_dri2: Fix one context, multiple surfaces.</li>
-  <li>auxiliary: util_blit_pixels_tex should restore the viewport.</li>
-  <li>st/vega: Fix a crash with empty paths.</li>
-  <li>st/vega: Masks and surfaces should share orientation.</li>
-  <li>st/vega: No flipping in vg_prepare_blend_surface.</li>
-  <li>st/vega: Fix a typo in EXTENDED_BLENDER_OVER_FUNC.</li>
-  <li>llvmpipe: Fix build errors on x86.</li>
-  <li>st/vega: Overhaul renderer with renderer states.</li>
-  <li>st/vega: Add DRAWTEX renderer state.</li>
-  <li>st/vega: Add SCISSOR renderer state.</li>
-  <li>st/vega: Add CLEAR renderer state for vgClear.</li>
-  <li>st/vega: Add FILTER renderer state for image filtering.</li>
-  <li>st/vega: Use the renderer for vgMask.</li>
-  <li>st/vega: Add POLYGON_STENCIL and POLYGON_FILL renderer state.</li>
-  <li>st/vega: Delay fb state update to vg_validate_state.</li>
-  <li>st/vega: Use st_framebuffer for fb width/height.</li>
-  <li>st/vega: Move g3d states to renderer.</li>
-  <li>st/vega: Make shader_bind call into the renderer.</li>
-  <li>st/vega: vg_manager should care about only the color buffer.</li>
-  <li>st/vega: Clean up vg_context fields and functions.</li>
-  <li>st/vega: Clean up renderer fields and functions.</li>
-  <li>st/vega: vg_copy_texture and vg_copy_surface should share code.</li>
-  <li>st/vega: Get rid of renderer_copy_texture.</li>
-  <li>st/vega: Update to latest headers.</li>
-  <li>st/vega: Fix image sampler views for alpha-only formats.</li>
-  <li>st/vega: Make path_render and path_stroke take a matrix.</li>
-  <li>st/vega: Make image_draw take a matrix.</li>
-  <li>st/vega: Add primitive text support.</li>
-  <li>st/vega: Revive mask layer support.</li>
-  <li>st/vega: More flexible shader selection.</li>
-  <li>st/vega: Add color transformation support.</li>
-  <li>st/vega: Bump version to 1.1.</li>
-  <li>st/vega: Fix paint coordinates transformations.</li>
-  <li>st/vega: Fix negated logic in image_draw.</li>
-  <li>st/vega: Fix degenerate paints.</li>
-  <li>st/vega: Simplify radial gradient.</li>
-  <li>st/vega: Remove st_inlines.h.</li>
-  <li>st/vega: Delay blend texture creation until needed.</li>
-  <li>st/vega: Create drawing surface mask as needed.</li>
-  <li>st/vega: Initialize pipe states with renderer.</li>
-  <li>st/vega: Avoid unnecessary constant bufer upload.</li>
-  <li>st/vega: Destroy the pipe context with vg_context.</li>
-  <li>st/vega: polygon_array requires a deep free.</li>
-  <li>st/egl: Set pipe_resource::array_size to 1.</li>
-  <li>st/vega: Set pipe_resource::array_size to 1.</li>
-  <li>st/vega: Move vertex transformation to shader.</li>
-  <li>st/vega: Add a missing break.</li>
-  <li>st/vega: Add some comments to pipeline shaders.</li>
-  <li>st/vega: Refactor blend shaders.</li>
-  <li>st/vega: Move masking after blending.</li>
-  <li>st/vega: Add support for per-channel alpha.</li>
-  <li>st/vega: Blending should use premultiplied alpha.</li>
-  <li>st/vega: Fix VG_BLEND_MULTIPLY.</li>
-  <li>st/vega: Add blend shaders for all blend modes.</li>
-  <li>st/vega: Fix pipe blend state for various blend modes.</li>
-  <li>egl: _eglFilterArray should not allocate.</li>
-  <li>mapi: Rewrite mapi_abi.py to get rid of preprocessor magic.</li>
-  <li>vbo: Drop second ATTR macro.</li>
-  <li>vbo: Fix GLES2 glVertexAttrib.</li>
-  <li>mesa: Do not advertise GL_OES_texture_3D.</li>
-  <li>mesa: Fix GL_FIXED arrays.</li>
-  <li>mesa: Fix glTexCoordPointer with type GL_FIXED.</li>
-  <li>st/egl: Plug pbuffer leaks.</li>
-  <li>st/egl: Fix eglCopyBuffers.</li>
-  <li>st/egl: Assorted fixes for dri2_display_get_configs.</li>
-  <li>docs/egl: Update egl.html.</li>
-  <li>st/egl: Fix eglChooseConfig when configs is NULL.</li>
-  <li>docs: Add an example for EGL_DRIVERS_PATH.</li>
-  <li>autoconf: Fix --with-driver=xlib --enable-openvg.</li>
-</ul></p>
-
-<p>Chris Wilson (2):
-<ul>
-  <li>i915g: Fix closure of full batch buffers</li>
-  <li>intel: Check for unsupported texture when finishing using as a render target</li>
-</ul></p>
-
-<p>Christoph Bumiller (80):
-<ul>
-  <li>nv50: import new compiler</li>
-  <li>nouveau: update nouveau_class.h</li>
-  <li>nv50: introduce the big formats table</li>
-  <li>nv50: don't produce MOV immediate to output reg in store opt</li>
-  <li>nv50: change back accidentally swapped UNORM,SNORM vertex type</li>
-  <li>nv50: add/fix some license headers</li>
-  <li>nv50: simple reload elimination and local CSE</li>
-  <li>nv50: fix constant_operand opt mul by 2 case</li>
-  <li>nv50: permit usage of undefined TGSI TEMPs</li>
-  <li>nv50: add missing 2nd source for POW multiplication</li>
-  <li>nv50: add signed RGTC1 to format table, allow 2_10_10_10 for vbufs</li>
-  <li>nv50: fix for empty BBs</li>
-  <li>nv50: insert MOVs also for PHI sources from dominating block</li>
-  <li>nv50: explicitly set src type for SET ops</li>
-  <li>nv50: fixes for nested IFs</li>
-  <li>nv50: don't eliminate loads to dedicated values</li>
-  <li>nv50: fix constbuf validation</li>
-  <li>nv50: build proper phi functions in the first place</li>
-  <li>nv50: fix reg count</li>
-  <li>nv50: begin implementing loops</li>
-  <li>nv50: more constant folding</li>
-  <li>nv50: loops part 2</li>
-  <li>nv50: flatten simple IF/ELSE/ENDIF constructs</li>
-  <li>nv50: fix thinko in store to output reg possible check</li>
-  <li>nv50: generate JOINs for outermost IF clauses</li>
-  <li>nv50: more TGSI opcodes (SIN, SCS, ARL, RET, KILP)</li>
-  <li>nv50: fix PSIZ and PRIMID mapping</li>
-  <li>nv50: check dst compatibility in CSE</li>
-  <li>nv50: initialize edgeflag input index</li>
-  <li>nv50: emit predicate for interp</li>
-  <li>Merge remote branch 'origin/master' into nv50-compiler</li>
-  <li>nv50: DP2, fix ARL</li>
-  <li>nv50: yet another case we need a nop.exit</li>
-  <li>nv50: fix check for sprite/point coord enable</li>
-  <li>nv50: handle TEXTURE_SWIZZLE and GEOMETRY_SHADER4 caps</li>
-  <li>nv50: set the FragDepth output index</li>
-  <li>nv50: turn off verbose debug output by default</li>
-  <li>nv50: attempt at making more complicated loops work</li>
-  <li>nv50: SSG</li>
-  <li>nv50: make FrontFacing -1 or +1</li>
-  <li>nv50: re-add proper TEXBIAS sequence</li>
-  <li>nv50: make use of TGSI immediate type</li>
-  <li>nv50: must join SELECT inputs before MOV inputs</li>
-  <li>nv50: fix XPD, was negated</li>
-  <li>nv50: fix find_dom_frontier</li>
-  <li>nv50: fix build-predicate function</li>
-  <li>Merge remote branch 'origin/master' into nv50-compiler</li>
-  <li>nv50: load address register before using it, not after</li>
-  <li>nv50: save tgsi instructions</li>
-  <li>nv50: prepare for having multiple functions</li>
-  <li>nv50: don't parse again in tgsi_2_nc</li>
-  <li>nv50: use actual loads/stores if TEMPs are accessed indirectly</li>
-  <li>nv50: create value references with the right type</li>
-  <li>nv50: duplicate interps in load_proj_tex_coords</li>
-  <li>nv50: address regs are 16 bit</li>
-  <li>nv50: fix can_load check for 3rd source</li>
-  <li>nv50: reduce bb_reachable_by runtime from pot to linear</li>
-  <li>nv50: minor compiler fixes and cleanups</li>
-  <li>nv50: cannot move from local mem to output reg directly</li>
-  <li>nv50: newlines in shader bincode printing</li>
-  <li>nv50: match TEMP limit with nv50 ir builder</li>
-  <li>nv50: handle TGSI EXP and LOG again</li>
-  <li>nv50: check for immediates when turning MUL ADD into MAD</li>
-  <li>nv50: interp cannot write flags reg</li>
-  <li>nv50: MOV TEMP[0], -CONST[0] must be float32 negation</li>
-  <li>nv50: fix indirect CONST access with large or negative offsets</li>
-  <li>nv50: fix TXP depth comparison value</li>
-  <li>nv50: consider address register in reload elimination</li>
-  <li>nv50: improve and fix modifier folding optimization</li>
-  <li>nv50: put low limit on REG_ALLOC_TEMP and FP_RESULT_COUNT</li>
-  <li>Merge remote branch 'origin/nv50-compiler'</li>
-  <li>nv50: don't segfault on shaders with 0 instructions</li>
-  <li>nv50: get shader fixups/relocations into working state</li>
-  <li>nv50: add relocs for stack and local mem buffers</li>
-  <li>nv50: emit constbuf relocs before uploading constants</li>
-  <li>nv50: fix typo in fifo packet length limit</li>
-  <li>nv50: use formats table in nv50_surface.c</li>
-  <li>nv50: use CLEAR_BUFFERS for surface fills</li>
-  <li>nv50: fix/handle a few more PIPE_CAPs</li>
-  <li>nv50: fix GP state bind and validate</li>
-</ul></p>
-
-<p>Corbin Simpson (8):
-<ul>
-  <li>r600g: Use align() instead of handrolled code.</li>
-  <li>r600g: Trivially deobfuscate r600_hw_states.</li>
-  <li>r600g: Deobfuscate and comment a few more functions in r600_hw_states.</li>
-  <li>r600g: Clean up some indentation and |= vs. | usage.</li>
-  <li>r600g: Fix false and true.</li>
-  <li>r600g: "tmp" is such a bad name for a texture.</li>
-  <li>r600g: Clean up PS setup.</li>
-  <li>r600g: Cleanup viewport floats.</li>
-</ul></p>
-
-<p>Daniel Lichtenberger (1):
-<ul>
-  <li>radeon: fix potential segfault in renderbuffer update</li>
-</ul></p>
-
-<p>Daniel Vetter (21):
-<ul>
-  <li>r200: revalidate after radeon_update_renderbuffers</li>
-  <li>i915g: rip out -&gt;sw_tiled</li>
-  <li>i915g: s/hw_tiled/tiling</li>
-  <li>i915g: add pineview pci ids</li>
-  <li>i915g: kill RGBA/X formats</li>
-  <li>i915g: kill buf-&gt;map_gtt</li>
-  <li>i915g: kill idws-&gt;pool</li>
-  <li>i915g: drop alignment parameter from iws-&gt;buffer_create</li>
-  <li>i915g: add winsys function to create tiled buffers</li>
-  <li>i915g: switch to tiled allocations, kill set_fence</li>
-  <li>i915g: prepare winsys/batchbuffer for execbuf2</li>
-  <li>i915g: return tiling in iws-&gt;buffer_from_handle</li>
-  <li>i915g: implement unfenced color&amp;depth buffer using tiling bits</li>
-  <li>i915g: implement unfenced relocs for textures using tiling bits</li>
-  <li>i915g: postpone mipmap/face offset calculation</li>
-  <li>i915g: don't pot-align stride for tiled buffers</li>
-  <li>i915g: enable X-tiling for textures</li>
-  <li>i915g: switch rendering to mipmapped textures to (x,y) offsets</li>
-  <li>i915g: enable x-tiling for render targets</li>
-  <li>i915g: assert(depth_surface-&gt;offset == 0)</li>
-  <li>i915g: track TODO items</li>
-</ul></p>
-
-<p>Dave Airlie (182):
-<ul>
-  <li>r300g: fix buffer reuse issue caused by previous commit</li>
-  <li>r600g: pull r600_draw struct out into header</li>
-  <li>r600g: use index min/max + index buffer offset.</li>
-  <li>r600g: add vgt dma src defines</li>
-  <li>r600g: fixup texture state on evergreen.</li>
-  <li>r600g: fix texture bos and avoid doing depth blit on evergreen</li>
-  <li>r600g: hide radeon_ctx inside winsys.</li>
-  <li>r600g: attempt to abstract kernel bos from pipe driver.</li>
-  <li>r600g: move constant buffer creation behind winsys abstraction.</li>
-  <li>r600g: use malloc bufmgr for constant buffers</li>
-  <li>r600g: add support for kernel bo</li>
-  <li>r600g: add winsys bo caching.</li>
-  <li>r600g: add upload manager support.</li>
-  <li>r600g: fixup map flushing.</li>
-  <li>r600g: use calloc for ctx bo allocations</li>
-  <li>r600g: oops got the use_mem_constant the wrong way around.</li>
-  <li>r600g; add uses waterfall to asm cf for r6xx.</li>
-  <li>r600g: only emit uses waterfall on r6xx hw.</li>
-  <li>util/r300g: split the r300 index buffer modifier functions out to util</li>
-  <li>r600g: modify index buffers for sizes the hw can't deal with.</li>
-  <li>r600g: send correct surface base update for multi-cbufs</li>
-  <li>r600g: fix fbo-drawbuffers-maxtargets</li>
-  <li>r600g: clean up valgrind issues on maxtargets test.</li>
-  <li>r600g: drop debugging that snuck in</li>
-  <li>r600g: fix tiling support for ddx supplied buffers</li>
-  <li>r600g: add z16 to color setup</li>
-  <li>r600g: add color/texture support for more depth formats.</li>
-  <li>r600g: fix r700 cube map sizing.</li>
-  <li>r600g: fixup r700 CB_SHADER_CONTROL register.</li>
-  <li>r600g: add missing BC_INST wrapper for evergreen</li>
-  <li>r600g: only flush for the correct colorbuffer, not all of them.</li>
-  <li>r600g: deal with overflow of VTX/TEX CF clauses.</li>
-  <li>r600g: set back to correct codepaths.</li>
-  <li>r600g: fixup evergreen miptree setup.</li>
-  <li>r600g: fix eg texture borders.</li>
-  <li>r600g: fix typo in struct member name</li>
-  <li>r600g: cleanup some of the DB blit code</li>
-  <li>r600g: make stencil readback work</li>
-  <li>r600g: disable dirty handling on texture from depth code.</li>
-  <li>r600g: use floats instead of hex for blit vbo</li>
-  <li>r600g: fix depth readback on rv610 and other quirky variants.</li>
-  <li>r600g: fix typo in evergreen register list</li>
-  <li>u_blitter: add a custom blitter call passing a dsa cso</li>
-  <li>r600g: use blitter to do db-&gt;cb flushing.</li>
-  <li>r600g: fix warnings since last commit.</li>
-  <li>egl: fix build since 17eace581d25a626a7d75d9d1205d012cbb14a6e</li>
-  <li>r300g: fix point sprite coord.</li>
-  <li>r600g: add vert support for 16/16 and 16/16/16 floats.</li>
-  <li>r600g: add some more vertex format support.</li>
-  <li>r600g: some more vertex formats</li>
-  <li>r600g: fix draw-elements and draw-elements-base-vertex</li>
-  <li>r600g: drop index_offset parameter to index buffer translate.</li>
-  <li>r600g: fixup tex wrapping.</li>
-  <li>r600g: fixup VP-&gt;FP output-&gt;input routing.</li>
-  <li>r600g: fix typo in r700 alu emit</li>
-  <li>r600g: fixup sprite coord enable.</li>
-  <li>r600g: fix polygon mode</li>
-  <li>mesa/mipmap: fix warning since 1acadebd6270d3604b026842b8a21360968618a0</li>
-  <li>r600g: add eg poly mode code.</li>
-  <li>r600g: make index bias fix for evergreen</li>
-  <li>r600g: add eg db count control register.</li>
-  <li>r300g: fix glsl-fs-pointcoord</li>
-  <li>r600g: add evergreen texture resource properly.</li>
-  <li>r600g: fix db flush breaking config state</li>
-  <li>r600g: on evergreen the centroid isn't set in this register.</li>
-  <li>r600g: add back evergreen name.</li>
-  <li>r600g: add evergreen texture border support to new path</li>
-  <li>r600g: move radeon.h members around to add back map flushing.</li>
-  <li>r600g: add initial vertex translate support.</li>
-  <li>r600g: remove old assert from new codepath</li>
-  <li>Revert "r600g: add initial vertex translate support."</li>
-  <li>r600g: port r300g fix for X* formats in texformat code</li>
-  <li>r600g: add L8A8 unorm.</li>
-  <li>r600g: clean up some code from move to new paths.</li>
-  <li>r600g: return string for chip family</li>
-  <li>r600g: use Elements macro instead of manual sizeofs</li>
-  <li>r600g: fix evergreen depth flushing.</li>
-  <li>r600g: add winsys support for CTL constants.</li>
-  <li>r600g: drop depth quirk on evergreen</li>
-  <li>r600g: add reloc for evergreen color attrib</li>
-  <li>r600g: realign evergreen code with r600 code.</li>
-  <li>r600g: add assembler support for other vtx fetch fields.</li>
-  <li>r600g: fixup vertex format picking.</li>
-  <li>r600g: sync vertex/texture cache on resources on evergreen</li>
-  <li>r600g: add cb flushing for extra buffers + depth buffer on r600/evergreen</li>
-  <li>r600g: fix evergreen draw-buffers</li>
-  <li>r600g: flush SH cache on constant change on evergreen</li>
-  <li>r600g: only set the Z export if shader exports it.</li>
-  <li>r600g: setup basic loop consts on r600 + evergreen.</li>
-  <li>mesa/st: initial attempt at RG support for gallium drivers</li>
-  <li>r600g: break out of search for reloc bo after finding it.</li>
-  <li>r600g: the code to check whether a new vertex shader is needed was wrong</li>
-  <li>r600g: fix wwarning in bo_map function</li>
-  <li>r600g: TODO domain management</li>
-  <li>r600g: add bo fenced list.</li>
-  <li>pb: don't keep checking buffers after first busy</li>
-  <li>r600g: add bo busy backoff.</li>
-  <li>r600g: drop mman allocator</li>
-  <li>r600g: drop use_mem_constant.</li>
-  <li>r600g: avoid unneeded bo wait</li>
-  <li>pb: fix numDelayed accounting</li>
-  <li>r600g: add evergreen stencil support.</li>
-  <li>r600g: use format from the sampler view not from the texture.</li>
-  <li>r600g: fix Z export enable bits.</li>
-  <li>r600g: add some RG texture format support.</li>
-  <li>r600g: drop width/height per level storage.</li>
-  <li>r600g: fix input/output Z export mixup for evergreen.</li>
-  <li>r600g: evergreen has no request size bit in texture word4</li>
-  <li>r600g: enable vertex samplers.</li>
-  <li>r600g: add TXL opcode support.</li>
-  <li>r600g: don't run with scissors.</li>
-  <li>r600g: fix typo in vertex sampling on r600</li>
-  <li>gallium/tgsi: add support for stencil writes.</li>
-  <li>gallium/format: add support for X24S8 and S8X24 formats.</li>
-  <li>gallium/format: add X32_S8X24_USCALED format.</li>
-  <li>gallium/util: add S8 tile sampling support.</li>
-  <li>mesa: add support for FRAG_RESULT_STENCIL.</li>
-  <li>mesa: improve texstore for 8/24 formats and add texstore for S8.</li>
-  <li>softpipe: add support for shader stencil export capability</li>
-  <li>st/mesa: add option to choose a texture format that we won't render to.</li>
-  <li>st/mesa: use shader stencil export to accelerate shader drawpixels.</li>
-  <li>r600g: add support for S8, X24S8 and S8X24 sampler formats.</li>
-  <li>r600g: add shader stencil export support.</li>
-  <li>glsl: add support for shader stencil export</li>
-  <li>st/mesa: enable stencil shader export extension if supported</li>
-  <li>r600g: fix depth0 setting</li>
-  <li>r600g: fix scissor/cliprect confusion</li>
-  <li>r600g: store samplers/views across blit when we need to modify them</li>
-  <li>r600g: reduce size of context structure.</li>
-  <li>r600g: the vs/ps const arrays weren't actually being used.</li>
-  <li>r600g: add copy into tiled texture</li>
-  <li>r600g: split out miptree setup like r300g</li>
-  <li>r600g: use common texture object create function</li>
-  <li>r600g: rename pitch in texture to pitch_in_bytes</li>
-  <li>r600g: remove bpt and start using pitch_in_bytes/pixels.</li>
-  <li>r600g: fix transfer stride.</li>
-  <li>r600g: drop all use of unsigned long</li>
-  <li>r600g: use blitter for hw copy region</li>
-  <li>r600g: evergreen add stencil export bit</li>
-  <li>r600g: add missing eg reg definition</li>
-  <li>r600g: fix stencil export for evergreen harder</li>
-  <li>r600g: drop unused context members</li>
-  <li>r600g: only pick centroid coordinate when asked.</li>
-  <li>r600g: fixup pos/face ena/address properly</li>
-  <li>r600g: fixup typo in macro name</li>
-  <li>r600g: select linear interpolate if tgsi input requests it</li>
-  <li>r300g: clean up warning due to unknown cap.</li>
-  <li>tgsi: add scanner support for centroid inputs</li>
-  <li>r600g: evergreen interpolation support.</li>
-  <li>r600g: add evergreen ARL support.</li>
-  <li>r600g: switch to a common formats.h file since they are in different regs</li>
-  <li>r600g: add defines for tiling</li>
-  <li>r600g: get tiling info from kernel</li>
-  <li>r600g: set tiling bits in hw state</li>
-  <li>r600g: do proper tracking of views/samplers.</li>
-  <li>r600g: fix typo in tiling setup cb code.</li>
-  <li>r600g: depth needs to bound to ds</li>
-  <li>r600g: attempt to cleanup depth blit</li>
-  <li>r600g: fix transfer function for tiling.</li>
-  <li>r600g: retrieve tiling info from kernel for shared buffers.</li>
-  <li>r600g: all non-0 mipmap levels need to be w/h aligned to POT.</li>
-  <li>r600g: move to per-miplevel array mode.</li>
-  <li>r600g: start adding hooks for aligning width/height for tiles.</li>
-  <li>r600g: add r600 surface to store the aligned height.</li>
-  <li>r600g: introduce a per-driver resource flag for transfers.</li>
-  <li>r600g: add texture tiling alignment support.</li>
-  <li>r600g: add texture tiling enable under a debug option.</li>
-  <li>r600g: initial translate state support</li>
-  <li>r600g: start splitting out common code from eg/r600.</li>
-  <li>r600g: not fatal if we can't get tiling info from kernel</li>
-  <li>r600g: merge more of the common r600/evergreen state handling</li>
-  <li>r600g: drop more common state handling code</li>
-  <li>r600g: fix magic 0x1 -&gt;flat shade ena</li>
-  <li>r600g: add assembler support for all the kcache fields.</li>
-  <li>gallium/noop: report GL 2.1</li>
-  <li>r600g: pick correct color swap for A8 fbos.</li>
-  <li>r300g/r600g: bump cache manager timeouts to 1s</li>
-  <li>r600g: it looks like r600 can handle dword offsets in the indices.</li>
-  <li>r300g: try and use all of vertex constant space</li>
-  <li>r300g: fixup rs690 tiling stride alignment calculations.</li>
-  <li>r600g: fix evergreen segfaults.</li>
-  <li>r600g: hack around property unknown issues.</li>
-</ul></p>
-
-<p>Eric Anholt (300):
-<ul>
-  <li>glsl: Add definition of gl_TextureMatrix inverse/transpose builtins.</li>
-  <li>i965: Share the KIL_NV implementation between glsl and non-glsl.</li>
-  <li>i965: Also enable CC statistics when doing OQs.</li>
-  <li>i965: Track the windowizer's dispatch for kill pixel, promoted, and OQ</li>
-  <li>glsl: Rework assignments with write_masks to have LHS chan count match RHS.</li>
-  <li>glsl: Fix copy'n'wasted ir_noop_swizzle conditions.</li>
-  <li>ir_to_mesa: Only compare vector_elements present for any_nequal/all_equal</li>
-  <li>i965: Fix the vector/expression splitting for the write_mask change.</li>
-  <li>i965: When splitting vector variable assignment, ignore unset channels.</li>
-  <li>i965: Update expression splitting for the vector-result change to compares.</li>
-  <li>i965: Warning fix for vector result any_nequal/all_equal change.</li>
-  <li>mesa: Remove the non-required ARB_imaging extension.</li>
-  <li>mesa: Remove EXT_histogram.</li>
-  <li>mesa: Remove SGI_color_table.</li>
-  <li>mesa: Remove SGI_color_matrix.</li>
-  <li>mesa: Remove EXT_convolution.</li>
-  <li>intel: Remove disabled stencil drawpixels acceleration.</li>
-  <li>intel: Remove unnecessary minimum pitch alignment to 32 bytes.</li>
-  <li>intel: Replace my intel_texture_bitmap code with _mesa_meta_Bitmap.</li>
-  <li>radeon: Remove copied minimum pitch alignment code.</li>
-  <li>unichrome: Mostly revert my convolution removal changes.</li>
-  <li>intel: Remove dead intelIsTextureResident().</li>
-  <li>i915: Remove a dead if (0) block.</li>
-  <li>intel: Dead comment removal.</li>
-  <li>intel: Corresponding FinishRenderTexture debug to BeginRenderTexture.</li>
-  <li>i965: Add support for rendering to SARGB8 FBOs.</li>
-  <li>intel: Fix segfault on INTEL_DEBUG=fbo with unsupported framebuffers.</li>
-  <li>intel: Add fallback debug to glGenerateMipmap.</li>
-  <li>intel: More reverting of the sw fallback for depth texture border color.</li>
-  <li>intel: Improve some of the miptree debugging.</li>
-  <li>mesa: Fix type typo in glGenerateMipmap handling of GL_UNSIGNED_INT data.</li>
-  <li>glsl: Fix broadcast_index of lower_variable_index_to_cond_assign.</li>
-  <li>glsl: Add validation that a swizzle only references valid channels.</li>
-  <li>i965: Fix up writemasked assignments in the new FS.</li>
-  <li>i965: Remove swizzling of assignment to vector-splitting single-channel LHS.</li>
-  <li>i965: Handle all_equal/any_nequal in the new FS.</li>
-  <li>i965: Fix vector splitting RHS channel selection with sparse writemasks.</li>
-  <li>i965: Add support for dFdx()/dFdy() to the FS backend.</li>
-  <li>i965: Add support for attribute interpolation on Sandybridge.</li>
-  <li>i965: Set up inputs to the fragment shader according to FP InputsRead.</li>
-  <li>i965: Add support for POW in gen6 FS.</li>
-  <li>i965: Fix negation in the new FS backend.</li>
-  <li>i965: Actually track the "if" depth in loop in the new FS backend.</li>
-  <li>i965: Apply the same set of lowering passes to new FS as to Mesa IR.</li>
-  <li>i965: Fix valgrind complaint about base_ir for new FS debugging.</li>
-  <li>i965: Fix up the FS backend for the variable array indexing pass.</li>
-  <li>i965: Set the variable type when dereferencing an array.</li>
-  <li>i965: Add support for dereferencing structs to the new FS backend.</li>
-  <li>i965: Add support for struct, array, and matrix uniforms to FS backend.</li>
-  <li>i965: Fix all non-snb regression in the snb attribute interpolation commit.</li>
-  <li>i965: Fix up part of my Sandybridge attributes support patch.</li>
-  <li>i965: Add support for gl_FrontFacing to the new FS backend.</li>
-  <li>i965: Subtract instead of adding when computing y delta in new FS backend.</li>
-  <li>mesa: Pull ir_to_mesa's sampler number fetcher out to shared code.</li>
-  <li>i965: Set up sampler numbers in the FS backend.</li>
-  <li>i965: Add support for non-color render target write data to new FS backend.</li>
-  <li>i965: Add support for MRT to the new FS backend.</li>
-  <li>i965: Add support for ir_loop counters to the new FS backend.</li>
-  <li>i965: Add support for ARB_fragment_coord_conventions to the new FS backend.</li>
-  <li>glsl: Also update implicit sizes of varyings at link time.</li>
-  <li>i965: Do interpolation for varying matrices and arrays in the FS backend.</li>
-  <li>i965: Don't try to emit interpolation for unused varying slots.</li>
-  <li>i965: Fix array indexing of arrays of matrices.</li>
-  <li>i965: Clean up obsolete FINISHME comment.</li>
-  <li>mesa: Move the list of builtin uniform info from ir_to_mesa to shared code.</li>
-  <li>i965: Add support for builtin uniforms to the new FS backend.</li>
-  <li>i965: Fix use of undefined mem_ctx in vector splitting.</li>
-  <li>i956: Make new FS discard do its work in a temp, not the null reg!</li>
-  <li>i965: Clean up the virtual GRF handling.</li>
-  <li>ra: First cut at a graph-coloring register allocator for mesa.</li>
-  <li>i965: First cut at register allocation using graph coloring.</li>
-  <li>i965: Add live interval analysis and hook it up to the register allocator.</li>
-  <li>i965: Remove my "safety counter" code from loops.</li>
-  <li>i965: Fix whole-structure/array assignment in new FS.</li>
-  <li>mesa: Don't reference a W component in setting up a vec3 uniform component.</li>
-  <li>i965: Fix new FS handling of builtin uniforms with packed scalars in structs.</li>
-  <li>glsl: Add a lowering pass for texture projection.</li>
-  <li>i965: Use the lowering pass for texture projection.</li>
-  <li>i965: Split the gen4 and gen5 sampler handling apart.</li>
-  <li>i965: Add gen6 attribute interpolation to new FS backend.</li>
-  <li>i965: Fix the gen6 jump size for BREAK/CONT in new FS.</li>
-  <li>i965: Also increment attribute location when skipping unused slots.</li>
-  <li>i965: Pre-gen6, map VS outputs (not FS inputs) to URB setup in the new FS.</li>
-  <li>i965: Add real support for pre-gen5 texture sampling to the new FS.</li>
-  <li>i965: Fix up copy'n'pasteo from moving coordinate setup around for gen4.</li>
-  <li>i965: Restore the forcing of aligned pairs for delta_xy on chips with PLN.</li>
-  <li>i965: When producing a single channel swizzle, don't make a temporary.</li>
-  <li>i965: Add a sanity check for register allocation sizes.</li>
-  <li>i965: Fix off-by-ones in handling the last members of register classes.</li>
-  <li>i965: Don't try to emit code if we failed register allocation.</li>
-  <li>i965: Add support for EXT_texture_swizzle to the new FS backend.</li>
-  <li>i965: Set up swizzling of shadow compare results for GL_DEPTH_TEXTURE_MODE.</li>
-  <li>i965: Fix glean/texSwizzle regression in previous commit.</li>
-  <li>i965: Be more conservative on live interval calculation.</li>
-  <li>i965: Add trivial dead code elimination in the new FS backend.</li>
-  <li>i965: Add initial folding of constants into operand immediate slots.</li>
-  <li>i965: In disasm, gen6 fb writes don't put msg reg # in destreg_conditionalmod.</li>
-  <li>i965: Add support for gen6 FB writes to the new FS.</li>
-  <li>i965: Enable the constant propagation code.</li>
-  <li>i965: Also do constant propagation for the second operand of CMP.</li>
-  <li>i965: Add back gen6 headerless FB writes to the new FS backend.</li>
-  <li>i965: Gen6 no longer has the IFF instruction; always use IF.</li>
-  <li>i965: Fix up IF/ELSE/ENDIF for gen6.</li>
-  <li>i965: Fix botch in the header_present case in the new FS.</li>
-  <li>i965: Add some clarification of the WECtrl field.</li>
-  <li>i965: Don't do 1/w multiplication in new FS for gen6</li>
-  <li>i965: Gen6's sampler messages are the same as Ironlake.</li>
-  <li>i965: Refactor gl_FrontFacing setup out of general variable setup.</li>
-  <li>i965: Add support for gl_FrontFacing on gen6.</li>
-  <li>i965: Don't assume that WPOS is always provided on gen6 in the new FS.</li>
-  <li>i965: Fix gen6 pointsize handling to match pre-gen6.</li>
-  <li>i965: Disable emitting if () statements on gen6 until we really fix them.</li>
-  <li>i965: Normalize cubemap coordinates like is done in the Mesa IR path.</li>
-  <li>mesa: Simplify a bit of _mesa_add_state_reference using memcmp.</li>
-  <li>i965: Drop the check for duplicate _mesa_add_state_reference.</li>
-  <li>i965: Drop the check for YUV constants in the param list.</li>
-  <li>i965: Handle swizzles in the addition of YUV texture constants.</li>
-  <li>i965: Fix gen6 WM push constants updates.</li>
-  <li>i965: Fix new FS gen6 interpolation for sparsely-populated arrays.</li>
-  <li>i965: Enable attribute swizzling (repositioning) in the gen6 SF.</li>
-  <li>i965: Add register coalescing to the new FS backend.</li>
-  <li>i965: Split FS_OPCODE_DISCARD into two steps.</li>
-  <li>i965: Reduce register interference checks for changed FS_OPCODE_DISCARD.</li>
-  <li>i965: Move FS backend structures to a header.</li>
-  <li>i965: Give the math opcodes information on base mrf/mrf len.</li>
-  <li>i965: Give the FB write and texture opcodes the info on base MRF, like math.</li>
-  <li>i965: Compute to MRF in the new FS backend.</li>
-  <li>i965: Don't consider gen6 math instructions to write to MRFs.</li>
-  <li>i965: Add a couple of checks for gen6 math instruction limits.</li>
-  <li>i965: Don't compute-to-MRF in gen6 math instructions.</li>
-  <li>i965: Expand uniform args to gen6 math to full registers to get hstride == 1.</li>
-  <li>i965: Don't compute-to-MRF in gen6 VS math.</li>
-  <li>i965: Fix gen6 pixel_[xy] setup to avoid mixing int and float src operands.</li>
-  <li>i965: Always use the new FS backend on gen6.</li>
-  <li>i965: Fix missing "break;" in i2b/f2b, and missing AND of CMP result.</li>
-  <li>intel: Allow CopyTexSubImage to InternalFormat 3/4 textures, like RGB/RGBA.</li>
-  <li>i965: Don't rebase the index buffer to min 0 if any arrays are in VBOs.</li>
-  <li>i965: Add support for rescaling GL_TEXTURE_RECTANGLE coords to new FS.</li>
-  <li>i965: Set class_sizes[] for the aligned reg pair class.</li>
-  <li>i965: Update the live interval when coalescing regs.</li>
-  <li>i965: Add a pass to the FS to split virtual GRFs to float channels.</li>
-  <li>i965: Add a function for handling the move of boolean values to flag regs.</li>
-  <li>i965: Add peepholing of conditional mod generation from expressions.</li>
-  <li>i965: Enable the new FS backend on pre-gen6 as well.</li>
-  <li>i965: Fix texturing on pre-gen5.</li>
-  <li>i965: Set the type of the null register to fix gen6 FS comparisons.</li>
-  <li>i965: Disable the debug printf I added for FS disasm.</li>
-  <li>i965: Fix a weirdness in NOT handling.</li>
-  <li>i965: Fix assertion failure on gen6 BufferSubData to busy BO.</li>
-  <li>i965: Assert out on gen6 VS constant buffer reads that hang the GPU for now.</li>
-  <li>i965: Fix scissor-offscreen on gen6 like we did pre-gen6.</li>
-  <li>i965: Avoid blits in BufferCopySubdata on gen6.</li>
-  <li>i965: Tell the shader compiler when we expect depth writes for gen6.</li>
-  <li>i965: Remove the gen6 emit_mi_flushes I sprinkled around the driver.</li>
-  <li>i965: Disable thread dispatch when the FS doesn't do any work.</li>
-  <li>i965: Add EU emit support for gen6's new IF instruction with comparison.</li>
-  <li>i965: Set the source operand types for gen6 if/else/endif to integer.</li>
-  <li>i965: Use the new style of IF statement with embedded comparison on gen6.</li>
-  <li>i965: Split register allocation out of the ever-growing brw_fs.cpp.</li>
-  <li>i965: Fix gl_FrontFacing emit on pre-gen6.</li>
-  <li>i965: Add support for register spilling.</li>
-  <li>i965: Don't emit register spill offsets directly into g0.</li>
-  <li>i965: Correct scratch space allocation.</li>
-  <li>i965: Be more aggressive in tracking live/dead intervals within loops.</li>
-  <li>i965: Move the FS disasm/annotation printout to codegen time.</li>
-  <li>i965: Add support for pull constants to the new FS backend.</li>
-  <li>i965: Add EU code for dword scattered reads (constant buffer array indexing).</li>
-  <li>i965: Clarify an XXX comment in FB writes with real info.</li>
-  <li>i965: Use SENDC on the first render target write on gen6.</li>
-  <li>i965: Clear some undefined fields of g0 when using them for gen6 FB writes.</li>
-  <li>i965: Add disasm for the flag register.</li>
-  <li>i965: Add support for discard instructions on gen6.</li>
-  <li>i965: Handle new ir_unop_round_even in channel expression splitting.</li>
-  <li>i965: Fix typo in comment about state flags.</li>
-  <li>i965: Set up the constant buffer on gen6 when it's needed.</li>
-  <li>i965: Add support for constant buffer loads on gen6.</li>
-  <li>i965: Drop the eot argument to read messages, which can never be set.</li>
-  <li>i965: Fix VS URB entry sizing.</li>
-  <li>i965: Disable register spilling on gen6 until it's fixed.</li>
-  <li>i965: Make FS uniforms be the actual type of the uniform at upload time.</li>
-  <li>i965: Add user clip planes support to gen6.</li>
-  <li>i965: Update gen6 SF state when point state (sprite or attenuation) changes.</li>
-  <li>i965: Upload required gen6 VS push constants even when using pull constants.</li>
-  <li>i965: Update the gen6 stencil ref state when stencil state changes.</li>
-  <li>mesa: Make metaops use program refcounts instead of names.</li>
-  <li>mesa: Don't compute an unused texture completeness debug string.</li>
-  <li>intel: For batch, use GTT mapping instead of writing to a malloc and copying.</li>
-  <li>intel: Annotate debug printout checks with unlikely().</li>
-  <li>intel: Remove the magic unaligned memcpy code.</li>
-  <li>i965: Remove dead intel_structs.h file.</li>
-  <li>intel: Avoid taking logbase2 of several things that we max.</li>
-  <li>intel: Remove duplicated teximage miptree to object miptree promotion.</li>
-  <li>intel: Remove leftover dri1 locking fields in the context.</li>
-  <li>mesa: Fix delayed state flagging for EXT_sso-related program changes.</li>
-  <li>intel: Fix the client-side swapbuffers throttling.</li>
-  <li>Revert "intel: Fix the client-side swapbuffers throttling."</li>
-  <li>i965: Allow OPCODE_SWZ to put immediates in the first arg.</li>
-  <li>i965: Add support for math on constants in gen6 brw_wm_glsl.c path.</li>
-  <li>i965: Work around strangeness in swizzling/masking of gen6 math.</li>
-  <li>i965: re-enable gen6 IF statements in the fragment shader.</li>
-  <li>glsl: Free the loop state context when we free the loop state.</li>
-  <li>i965: Fix gl_FragCoord inversion when drawing to an FBO.</li>
-  <li>i965: Shut up spurious gcc warning about GLSL_TYPE enums.</li>
-  <li>mesa: Don't spam the console in a debug build unless some spam is requested.</li>
-  <li>i965: Add state dumping for sampler state.</li>
-  <li>i965: Add dumping of the sampler default color.</li>
-  <li>i965: Fail on loops on gen6 for now until we write the EU emit code for it.</li>
-  <li>i965: Eliminate dead code more aggressively.</li>
-  <li>mesa: Include C++ files in the makedepend of DRI drivers.</li>
-  <li>i965: Fix compute_to_mrf to not move a MRF write up into another live range.</li>
-  <li>i965: Just use memset() to clear most members in FS constructors.</li>
-  <li>i965: Remove extra n at the end of every instruction in INTEL_DEBUG=wm.</li>
-  <li>i965: Fold constants into the second arg of BRW_SEL as well.</li>
-  <li>glsl: Add a helper function for determining if an rvalue could be a saturate.</li>
-  <li>i965: Recognize saturates and turn them into a saturated mov.</li>
-  <li>ir_to_mesa: Detect and emit MOV_SATs for saturate constructs.</li>
-  <li>i965: Improve compute-to-mrf.</li>
-  <li>i965: Remove duplicate MRF writes in the FS backend.</li>
-  <li>i965: Move gen4 blend constant color to the gen4 blending file.</li>
-  <li>i965: Don't upload polygon stipple unless required.</li>
-  <li>i965: Don't upload line stipple pattern unless we're stippling.</li>
-  <li>i965: Don't upload line smooth params unless we're line smoothing.</li>
-  <li>i965: Use the new embedded compare in SEL on gen6 for VS MIN and MAX opcodes.</li>
-  <li>i965: Fix type of gl_FragData[] dereference for FB write.</li>
-  <li>glsl: Make the symbol table's add_function just use the function's name.</li>
-  <li>glsl: Make the symbol table's add_variable just use the variable's name.</li>
-  <li>glsl: Add a helper constructor for expressions that works out result type.</li>
-  <li>glsl: Fix structure and array comparisions.</li>
-  <li>glsl: Quiet unreachable no-return-from-function warning.</li>
-  <li>i965: Dump the WHILE jump distance on gen6.</li>
-  <li>i965: Add support for gen6 DO/WHILE ISA emit.</li>
-  <li>i965: Add support for gen6 BREAK ISA emit.</li>
-  <li>i965: Add support for gen6 CONTINUE instruction emit.</li>
-  <li>i965: Enable IF statements in the VS.</li>
-  <li>i965: Add support for loops in the VS.</li>
-  <li>glsl: Mark the array access for whole-array comparisons.</li>
-  <li>glsl: Fix flipped return of has_value() for array constants.</li>
-  <li>mesa: Add getters for the rest of the supported draw buffers.</li>
-  <li>mesa: Add getters for ARB_copy_buffer's attachment points.</li>
-  <li>intel: Add an env var override to execute for a different GPU revision.</li>
-  <li>i965: Update gen6 WM state on compiled program change, not just FP change.</li>
-  <li>i965: Update gen6 SF state on fragment program change too.</li>
-  <li>i965: Fix compile warning about missing opcodes.</li>
-  <li>i965: Move payload reg setup to compile, not lookup time.</li>
-  <li>i965: Provide delta_xy reg to gen6 non-GLSL path PINTERP.</li>
-  <li>i965: Fix up 16-wide gen6 FB writes after various refactoring.</li>
-  <li>i965: Don't smash a group of coordinates doing gen6 16-wide sampler headers.</li>
-  <li>i965: Fix gen6 interpolation setup for 16-wide.</li>
-  <li>i965: Fix up gen6 samplers for their usage by brw_wm_emit.c</li>
-  <li>i965: Make the sampler's implied move on gen6 be a raw move.</li>
-  <li>i965: Align gen6 push constant size to dispatch width.</li>
-  <li>i965: Add support for the instruction compression bits on gen6.</li>
-  <li>i965: Nuke brw_wm_glsl.c.</li>
-  <li>i965: Remove INTEL_DEBUG=glsl_force now that there's no brw_wm_glsl.c</li>
-  <li>i965: Fix comment about gen6_wm_constants.</li>
-  <li>i965: Handle saturates on gen6 math instructions.</li>
-  <li>i965: Always hand the absolute value to RSQ.</li>
-  <li>i965: Add disabled debug code for dumping out the WM constant payload.</li>
-  <li>i965: Work around gen6 ignoring source modifiers on math instructions.</li>
-  <li>i965: Fix flipped value of the not-embedded-in-if on gen6.</li>
-  <li>i965: Don't try to store gen6 (float) blend constant color in bytes.</li>
-  <li>i965: Set up the color masking for the first drawbuffer on gen6.</li>
-  <li>i965: Set up the per-render-target blend state on gen6.</li>
-  <li>i965: Set the render target index in gen6 fixed-function/ARB_fp path.</li>
-  <li>i965: Use the new pixel mask location for gen6 ARB_fp KIL instructions.</li>
-  <li>i965: Drop KIL_NV from the ff/ARB_fp path since it was only used for GLSL.</li>
-  <li>i965: Drop push-mode reladdr constant loading and always use constant_map.</li>
-  <li>i965: Fix VS constants regression pre-gen6.</li>
-  <li>i965: Clean up VS constant buffer location setup.</li>
-  <li>i965: Set up the correct texture border color state struct for Ironlake.</li>
-  <li>i965: Set render_cache_read_write surface state bit on gen6 constant surfs.</li>
-  <li>i965: remove unused variable since brw_wm_glsl.c removal.</li>
-  <li>intel: Use plain R8 and RG8 for COMPRESSED_RED and COMPRESSED_RG.</li>
-  <li>intel: Set the swizzling for depth textures using the GL_RED depth mode.</li>
-  <li>glsl: Correct the marking of InputsRead/OutputsWritten on in/out matrices.</li>
-  <li>i965: Correct the dp_read message descriptor setup on g4x.</li>
-  <li>intel: Include stdbool so we can stop using GLboolean when we want to.</li>
-  <li>i965: Fix ARL to work on gen6.</li>
-  <li>i956: Fix the old FP path fragment position setup on gen6.</li>
-  <li>i965: Fix gl_FragCoord.z setup on gen6.</li>
-  <li>i965: Add support for using the BLT ring on gen6.</li>
-  <li>intel: Update renderbuffers before looking up CopyTexImage's read buffer.</li>
-  <li>intel: Drop commented intel_flush from copy_teximage.</li>
-  <li>intel: Try to sanely check that formats match for CopyTexImage.</li>
-  <li>intel: Support glCopyTexImage() from XRGB8888 to ARGB8888.</li>
-  <li>i965: Avoid using float type for raw moves, to work around SNB issue.</li>
-  <li>i965: Set the alternative floating point mode on gen6 VS and WM.</li>
-  <li>i965: Add support for gen6 constant-index constant loading.</li>
-  <li>i965: Add support for gen6 reladdr VS constant loading.</li>
-  <li>i965: Improve the hacks for ARB_fp scalar^scalar POW on gen6.</li>
-  <li>i965: Factor out the ir comparision to BRW_CONDITIONAL_* code.</li>
-  <li>i965: Fix regression in FS comparisons on original gen4 due to gen6 changes.</li>
-  <li>i965: Do lowering of array indexing of a vector in the FS.</li>
-  <li>intel: Only do frame throttling at glFlush time when using frontbuffer.</li>
-  <li>intel: Handle forced swrast clears before other clear bits.</li>
-  <li>intel: Use tri clears when we don't know how to blit clear the format.</li>
-  <li>intel: Add spans code for the ARB_texture_rg support.</li>
-  <li>intel: Add a couple of helper functions to reduce rb code duplication.</li>
-  <li>intel: Fix segfaults from trying to use _ColorDrawBuffers in FBO validation.</li>
-  <li>intel: When validating an FBO's combined depth/stencil, use the given FBO.</li>
-</ul></p>
-
-<p>Fabian Bieler (2):
-<ul>
-  <li>r600g: set address of pop instructions to next instruction</li>
-  <li>glsl: fix lowering conditional returns in subroutines</li>
-</ul></p>
-
-<p>Francisco Jerez (51):
-<ul>
-  <li>dri/nv04: Fix PGRAPH_ERRORs when running OA.</li>
-  <li>dri/nv04: Mipmapping fixes.</li>
-  <li>dri/nv04: Align SIFM transfer dimensions.</li>
-  <li>dri/nv04: Fix up color mask.</li>
-  <li>dri/nv04: Fix maximum texture size.</li>
-  <li>dri/nv04: Fix provoking vertex.</li>
-  <li>dri/nouveau: Update nouveau_class.h.</li>
-  <li>dri/nouveau: Add some more extensions.</li>
-  <li>dri/nouveau: Fix glRenderbufferStorage with DEPTH_COMPONENT as internal format.</li>
-  <li>dri/nouveau: Don't request a fake front unnecessarily.</li>
-  <li>dri/nouveau: Don't reemit the BO state in nouveau_state_emit().</li>
-  <li>dri/nouveau: Cleanup references to the old FBOs on glMakeCurrent().</li>
-  <li>meta: Don't bind the created texture object in init_temp_texture().</li>
-  <li>dri/nv10: Fix the CLAMP texture wrap mode.</li>
-  <li>dri/nv04: Use nvgl_wrap_mode().</li>
-  <li>dri/nouveau: Remove unnecessary assertion.</li>
-  <li>dri/nouveau: Cleanup more references to old FBOs and VBOs.</li>
-  <li>dri/nv10-nv20: Fix texturing in some cases after a base level change.</li>
-  <li>dri/nouveau: Fix software mipmap generation on 1x1 textures.</li>
-  <li>dri/nouveau: Have a smaller amount of larger scratch buffers.</li>
-  <li>dri/nouveau: Remove unnecessary flush.</li>
-  <li>dri/nv10: Use fast Z clears.</li>
-  <li>dri/nouveau: Minor cleanup.</li>
-  <li>dri/nv10: Fake fast Z clears for pre-nv17 cards.</li>
-  <li>dri/nouveau: Initialize tile_flags when allocating a render target.</li>
-  <li>nouveau: Get larger push buffers.</li>
-  <li>dri/nouveau: Force a "slow" Z clear if we're getting a new depth buffer.</li>
-  <li>dri/nv20: Clear with the 3D engine.</li>
-  <li>dri/nouveau: Don't assert(0) on compressed internal formats.</li>
-  <li>dri/nv25: Bind a hierarchical depth buffer.</li>
-  <li>dri/nouveau: Call _mesa_update_state() after framebuffer invalidation.</li>
-  <li>dri/nouveau: Honor the access flags in nouveau_bufferobj_map_range.</li>
-  <li>dri/nouveau: Tell the vbo module we want real hardware BOs.</li>
-  <li>dri/nouveau: Split out the scratch helpers to a separate file.</li>
-  <li>dri/nouveau: Avoid recursion in nouveau_bo_context_reset().</li>
-  <li>dri/nouveau: Use a macro to iterate over the bound vertex attributes.</li>
-  <li>dri/nouveau: Split out array handling to its own file.</li>
-  <li>dri/nouveau: Optimize VBO binding re-emission.</li>
-  <li>dri/nouveau: Keep small DYNAMIC_DRAW vertex buffers in system ram.</li>
-  <li>dri/nouveau: Pipeline glTexSubImage texture transfers.</li>
-  <li>dri/nouveau: Fix type promotion issue on 32bit platforms.</li>
-  <li>dri/nouveau: Validate the framebuffer state on read buffer changes.</li>
-  <li>dri/nouveau: Re-emit the BO state when coming back from a software fallback.</li>
-  <li>meta: Don't leak alpha function/reference value changes.</li>
-  <li>meta: Fix incorrect rendering of the bitmap alpha component.</li>
-  <li>vbo: Avoid unnecessary copy to/from current in vertex format upgrade.</li>
-  <li>meta: Don't try to disable cube maps if the driver doesn't expose the extension.</li>
-  <li>meta: Handle bitmaps with alpha test enabled.</li>
-  <li>dri/nouveau: Split hardware/software TNL instantiation more cleanly.</li>
-  <li>dri/nouveau: Fix typo.</li>
-  <li>dri/nouveau: Kill a bunch of ternary operators.</li>
-</ul></p>
-
-<p>Fredrik Höglund (2):
-<ul>
-  <li>r600g: Fix texture sampling with swizzled coords</li>
-  <li>r600g: fix pow(0, 0) evaluating to NaN</li>
-</ul></p>
-
-<p>Guillermo S. Romero (1):
-<ul>
-  <li>r300g: Do not use buf param before checking for NULL.</li>
-</ul></p>
-
-<p>Henri Verbeet (19):
-<ul>
-  <li>r600g: Flush upload buffers before draws instead of before flushes.</li>
-  <li>r600g: Check for other references before checking for existing mappings in radeon_bo_pb_map_internal().</li>
-  <li>r600g: Remove a redundant flush in r600_texture_transfer_map().</li>
-  <li>r600g: Buffer object maps imply a wait.</li>
-  <li>r600g: Respect PB_USAGE_UNSYNCHRONIZED in radeon_bo_pb_map_internal().</li>
-  <li>Revert "r600g: Flush upload buffers before draws instead of before flushes."</li>
-  <li>r600g: fix exports_ps to export a number not a mask.</li>
-  <li>r600g: Mention AMD in the renderer string.</li>
-  <li>r600g: Cleanup the fenced_bo list in r600_context_fini().</li>
-  <li>r600g: Evergreen has two extra frac_bits for the sampler LOD state.</li>
-  <li>r600: Evergreen has two extra frac_bits for the sampler LOD state.</li>
-  <li>r600g: Add PIPE_FORMAT_L8A8_UNORM for Evergreen as well.</li>
-  <li>r600g: Swizzle vertex data only once.</li>
-  <li>r600g: Synchronize supported color formats between Evergreen and r600/r700.</li>
-  <li>r600g: Fix the PIPE_FORMAT_L8A8_UNORM color swaps.</li>
-  <li>r600g: Fix the PIPE_FORMAT_A8_UNORM color swap for Evergreen as well.</li>
-  <li>r600g: Cleanup block bo references in r600_context_fini().</li>
-  <li>r600g: Cleanup fetch shader resources in r600_pipe_shader_destroy().</li>
-  <li>st/mesa: Handle wrapped depth buffers in st_copy_texsubimage().</li>
-</ul></p>
-
-<p>Hui Qi Tay (10):
-<ul>
-  <li>llvmpipe: minor changes in llvm coefficient calcs</li>
-  <li>draw: cliptest and viewport done in a single loop in vertex shader</li>
-  <li>draw: added viewport and cliptest flags</li>
-  <li>draw: sanitize llvm variant key</li>
-  <li>draw: corrections for w coordinate</li>
-  <li>draw: corrections to allow for different cliptest cases</li>
-  <li>llvmpipe: Moved draw pipeline twoside function to llvm setup code</li>
-  <li>llvmpipe: added llvm offset setup code</li>
-  <li>llvmpipe: clean up polygon offset function in lp setup code</li>
-  <li>llvmpipe: fix such that offset/twoside function only does in-place modification</li>
-</ul></p>
-
-<p>Ian Romanick (102):
-<ul>
-  <li>glsl2: Refactor testing for whether a deref is of a matrix or array</li>
-  <li>glsl2: Add flags to enable variable index lowering</li>
-  <li>glsl: Add doxygen comments</li>
-  <li>EGL DRI2: Silence piles of 'unused variable' warnings</li>
-  <li>EGL DRI2: Silence 'missing initializer' warnings</li>
-  <li>egl_glx: Silence piles of 'unused variable' warnings</li>
-  <li>egl: Fix several 'comparison between signed and unsigned integer' warnings</li>
-  <li>dri: Ensure that DRI driver cpp files are in tarballs</li>
-  <li>mesa: Force GL_ARB_copy_buffer to always be enabled</li>
-  <li>mesa: Force GL_SGIS_generate_mipmap to always be enabled</li>
-  <li>Remove GL_MESA_packed_depth_stencil</li>
-  <li>Remove GL_EXT_cull_vertex</li>
-  <li>Regenerate files changed by previous commit</li>
-  <li>Remove unnescessary initializations of UpdateTexturePalette</li>
-  <li>ARB_texture_rg: Add GLX protocol support</li>
-  <li>ARB_texture_rg: Correct some errors in RED / RG internal format handling</li>
-  <li>ARB_texture_rg: Add GL_TEXTURE_{RED,GREEN}_SIZE query support</li>
-  <li>ARB_texture_rg: Add GL_RED as a valid GL_DEPTH_TEXTURE_MODE</li>
-  <li>ARB_texture_rg: Handle RED and RG the same as RGB for tex env</li>
-  <li>ARB_texture_rg: Add R8, R16, RG88, and RG1616 internal formats</li>
-  <li>ARB_texture_rg: Allow RED and RG textures as FBO color buffer attachments</li>
-  <li>mesa: Enable GL_ARB_texture_rg in software paths</li>
-  <li>i965: Enable GL_ARB_texture_rg</li>
-  <li>mesa: Add ARB_texture_compression_rgtc as an alias for EXT_texture_compression_rgtc</li>
-  <li>ARB_texture_rg: Add GL_COMPRESSED_{RED,RG} cases in _mesa_is_color_format</li>
-  <li>mesa: Fix misplaced #endif</li>
-  <li>mesa: Trivial correction to comment</li>
-  <li>rgtc: Detect RGTC formats as color formats and as compressed formats</li>
-  <li>docs: Add list of bugs fixed in 7.9</li>
-  <li>docs: Import 7.9 release notes from 7.9 branch.</li>
-  <li>docs: Import 7.8.x release notes from 7.8 branch.</li>
-  <li>docs: download.html does not need to be updated for each release</li>
-  <li>docs: Update mailing lines from sf.net to freedesktop.org</li>
-  <li>docs: Import news updates from 7.9 branch</li>
-  <li>docs: added news item for 7.9 release</li>
-  <li>glsl: Fail linking if assign_attribute_locations fails</li>
-  <li>glsl: Refactor 'layout' grammar to match GLSL 1.60 spec grammar</li>
-  <li>glsl: Slight refactor of error / warning checking for ARB_fcc layout</li>
-  <li>glsl: Clear type_qualifier using memset</li>
-  <li>glsl: Wrap ast_type_qualifier contents in a struct in a union</li>
-  <li>glsl: Regenerate files modified by previous commits</li>
-  <li>glcpp: Add the define for ARB_explicit_attrib_location when present</li>
-  <li>glcpp: Regenerate files changes by previous commit</li>
-  <li>glsl: Add parser support for GL_ARB_explicit_attrib_location layouts</li>
-  <li>glsl: Regenerate files changes by previous commit</li>
-  <li>glsl: Track explicit location in AST to IR translation</li>
-  <li>glsl: Add linker support for explicit attribute locations</li>
-  <li>main: Enable GL_ARB_explicit_attrib_location for swrast</li>
-  <li>intel: Enable GL_ARB_explicit_attrib_location</li>
-  <li>glsl: Remove const decoration from inlined function parameters</li>
-  <li>docs: skeleton for 7.10 release notes</li>
-  <li>docs: Update status of GL 3.x related extensions</li>
-  <li>mesa: Validate assembly shaders when GLSL shaders are used</li>
-  <li>glsl: Fix incorrect assertion</li>
-  <li>linker: Reject shaders that have unresolved function calls</li>
-  <li>mesa: Silence unused variable warning</li>
-  <li>mesa: Refactor validation of shader targets</li>
-  <li>mesa: Clean up two 'comparison between signed and unsigned' warnings</li>
-  <li>mesa: Clean up various 'unused parameter' warnings in shaderapi</li>
-  <li>glsl: Slightly change the semantic of _LinkedShaders</li>
-  <li>linker: Trivial indention fix</li>
-  <li>i965: Fix indentation after commit 3322fbaf</li>
-  <li>linker: Improve handling of unread/unwritten shader inputs/outputs</li>
-  <li>glapi: Add GL_EXT_separate_shader_objects</li>
-  <li>glapi: Commit files changed by previous commit</li>
-  <li>mesa: Add infrastructure to track GL_EXT_separate_shader_objects</li>
-  <li>mesa: Skeletal support for GL_EXT_separate_shader_objects</li>
-  <li>mesa: Add display list support for GL_EXT_separate_shader_objects functions</li>
-  <li>mesa: Track an ActiveProgram distinct from CurrentProgram</li>
-  <li>Track separate programs for each stage</li>
-  <li>swrast: Enable GL_EXT_separate_shader_objects in software paths</li>
-  <li>intel: Enable GL_EXT_separate_shader_objects in Intel drivers</li>
-  <li>docs: add GL_EXT_separate_shader_objects to release notes</li>
-  <li>glsl: Fix incorrect gl_type of sampler2DArray and sampler1DArrayShadow</li>
-  <li>ir_to_mesa: Refactor code for emitting DP instructions</li>
-  <li>mesa: Allow query of MAX_SAMPLES with EXT_framebuffer_multisample</li>
-  <li>glsl: Refactor is_vec_{zero,one} to be methods of ir_constant</li>
-  <li>glsl: Simplify generation of swizzle for vector constructors</li>
-  <li>glsl: Make is_zero and is_one virtual methods of ir_rvalue</li>
-  <li>ir_to_mesa: Generate smarter code for some conditional moves</li>
-  <li>glsl: Add ir_unop_sin_reduced and ir_unop_cos_reduced</li>
-  <li>glsl: Eliminate assumptions about size of ir_expression::operands</li>
-  <li>glsl: Add ir_rvalue::is_negative_one predicate</li>
-  <li>glsl: Add unary ir_expression constructor</li>
-  <li>glsl: Add ir_quadop_vector expression</li>
-  <li>glsl: Fix matrix constructors with vector parameters</li>
-  <li>i915: Disallow alpha, red, RG, and sRGB as render targets</li>
-  <li>glsl: Use M_LOG2E constant instead of calling log2</li>
-  <li>glsl: Lower ir_binop_pow to a sequence of EXP2 and LOG2</li>
-  <li>i915: Request that POW instructions be lowered</li>
-  <li>i915: Correctly generate unconditional KIL instructions</li>
-  <li>glsl: Ensure that equality comparisons don't return a NULL IR tree</li>
-  <li>i965: Correctly emit constants for aggregate types (array, matrix, struct)</li>
-  <li>glsl: Inherrit type of declared variable from initializer</li>
-  <li>linker: Ensure that unsized arrays have a size after linking</li>
-  <li>linker: Fix regressions caused by previous commit</li>
-  <li>glsl: Inherrit type of declared variable from initializer after processing assignment</li>
-  <li>linker: Allow built-in arrays to have different sizes between shader stages</li>
-  <li>ir_to_mesa: Don't generate swizzles for record derefs of non-scalar/vectors</li>
-  <li>Refresh autogenerated file builtin_function.cpp.</li>
-  <li>glsl: Allow less restrictive uses of sampler array indexing in GLSL &lt= 1.20</li>
-  <li>docs: Import 7.9.1 release notes from 7.9 branch</li>
-</ul></p>
-
-<p>Jakob Bornecrantz (27):
-<ul>
-  <li>rbug: Cast opcode to corrent int size</li>
-  <li>rbug: Add function to get opcode name string</li>
-  <li>scons: Link against talloc in the Gallium DRI drivers</li>
-  <li>i915g: Link with wrapper sw winsys with scons</li>
-  <li>tgsi: Actually care what check_soa_dependencies says</li>
-  <li>tgsi: Fix missing test before check</li>
-  <li>llvmpipe: Move makefile include to before targets</li>
-  <li>wrapper: Fix spelling</li>
-  <li>wrapper: Add a way to dewrap a pipe screen without destroying it</li>
-  <li>egl: Remove unnecessary headers</li>
-  <li>target-helpers: Remove per target software wrapper check</li>
-  <li>graw: Tidy graw xlib scons file a bit</li>
-  <li>scons: Remove old pipebuffer SConscript</li>
-  <li>scons: Detabify</li>
-  <li>scons: Check for pkg-config before trying to use it</li>
-  <li>scons: Check for libdrm_[intel|radeon] as well</li>
-  <li>scons: Move dependancy checks to the main gallium scons file</li>
-  <li>scons: Unify state tracker SConscripts</li>
-  <li>galahad: Correct the name of the scons library</li>
-  <li>graw: Use inline sw helper instead of roll your own loader</li>
-  <li>libgl-xlib: Use sw helper instead of roll your own</li>
-  <li>libgl-xlib: Use inline debug helper instead of non-inline version</li>
-  <li>graw: Use inline debug helper instead of non-inline version</li>
-  <li>gallium: Remove redundant sw and debug target helpers</li>
-  <li>i915g: Improve debug printing for textures</li>
-  <li>i915g: Make sure that new vbo gets updated</li>
-  <li>st/mesa: Unbind all constant buffers</li>
-</ul></p>
-
-<p>Jerome Glisse (75):
-<ul>
-  <li>r600g: alternative command stream building from context</li>
-  <li>r600g: move chip class to radeon common structure</li>
-  <li>r600g: use pipe context for flushing inside map</li>
-  <li>r600g: add back reference check when mapping buffer</li>
-  <li>r600g: directly allocate bo for user buffer</li>
-  <li>r600g: fix multi buffer rendering</li>
-  <li>r600g: occlusion query for new design</li>
-  <li>r600g: flush color buffer after draw command</li>
-  <li>r600g: disable shader rebuild optimization &amp; account cb flush packet</li>
-  <li>r600g: fix multiple occlusion query on same id</li>
-  <li>r600g: initial evergreen support in new path</li>
-  <li>r600g: fix typo in evergreen define (resource are in x range)</li>
-  <li>r600g: move use_mem_constants flags for new designs structure alignment</li>
-  <li>r600g: evergreen fix for new design</li>
-  <li>r600g: fix compilation after change to evergreend.h</li>
-  <li>r600g: fixup some evergreen register definitions</li>
-  <li>r600g: fix evergreen new path</li>
-  <li>r600g: fix reg definition</li>
-  <li>r600g: fix evergreen new path</li>
-  <li>r600g: bring over fix from old path to new path</li>
-  <li>r600g: fix vertex resource &amp; polygon offset</li>
-  <li>r600g: disable early cull optimization when occlusion query running</li>
-  <li>r600g: move around variables to share depth uncompression code</li>
-  <li>r600g: use depth decompression in new path</li>
-  <li>r600g: fix index buffer drawing</li>
-  <li>r600g: build packet header once</li>
-  <li>r600g: fix pointsprite &amp; resource unbinding</li>
-  <li>r600g: fix routing btw vertex &amp; pixel shader</li>
-  <li>r600g: fix occlusion query after change to block structure</li>
-  <li>r600g: use ptr for blit depth uncompress function</li>
-  <li>r600g: fix remaining piglit issue in new design</li>
-  <li>r600g: switch to new design</li>
-  <li>r600g: suspend/resume occlusion query around clear/copy</li>
-  <li>r600g: avoid rebuilding the vertex shader if no change to input format</li>
-  <li>r600g: use a hash table instead of group</li>
-  <li>r600g: delete old path</li>
-  <li>r600g: cleanup</li>
-  <li>r600g: more cleanup</li>
-  <li>r600g: use constant buffer instead of register for constant</li>
-  <li>r600g: fix constant &amp; literal src splitting, also fix mplayer gl2 shader</li>
-  <li>evergreeng: avoid overlapping border color btw VS &amp; PS</li>
-  <li>r600g: indentation fixes</li>
-  <li>r600g: rename radeon_ws_bo to r600_bo</li>
-  <li>r600g: allow r600_bo to be a sub allocation of a big bo</li>
-  <li>r600g: use r600_bo for relocation argument, simplify code</li>
-  <li>r600g: rename radeon_ws_bo to r600_bo</li>
-  <li>r600g: remove dead label &amp; fix indentation</li>
-  <li>r600g: store reloc information in bo structure</li>
-  <li>r600g: improve bo flushing</li>
-  <li>r600g: simplify block relocation</li>
-  <li>r600g: userspace fence to avoid kernel call for testing bo busy status</li>
-  <li>r600g: avoid segfault due to unintialized list pointer</li>
-  <li>r600g: fix dirty state handling</li>
-  <li>r600g: allow driver to work without submitting cmd to GPU</li>
-  <li>gallium/noop: no operation gallium driver</li>
-  <li>r600g: code cleanup (indent, trailing space, empty line ...)</li>
-  <li>r600g: fix occlusion query on evergreen (avoid lockup)</li>
-  <li>r600g: add fetch shader capabilities</li>
-  <li>r600g: dump raw shader output for debugging</li>
-  <li>r600g: update polygon offset only when rasterizer or zbuffer change</li>
-  <li>r600g: indentation fix</li>
-  <li>r600g: more indentation fix + warning silencing + dead code removal</li>
-  <li>r600g: build fetch shader from vertex elements</li>
-  <li>r600g: avoid useless shader rebuild at draw call</li>
-  <li>r600g: remove useless flush map</li>
-  <li>r600g: remove dead code</li>
-  <li>r600g: fix userspace fence against lastest kernel</li>
-  <li>r600g: avoid using pb* helper we are loosing previous cpu cycle with it</li>
-  <li>r600g: specialized upload manager</li>
-  <li>r600g: indentation cleanup</li>
-  <li>r600g: fix bo size when creating bo from handle</li>
-  <li>r600g: fix segfault when translating vertex buffer</li>
-  <li>r600g: need to reference upload buffer as the might still live accross flush</li>
-  <li>r600g: properly unset vertex buffer</li>
-  <li>r600g: avoid segfault</li>
-</ul></p>
-
-<p>Joakim Sindholt (3):
-<ul>
-  <li>util/u_blitter: fix leak</li>
-  <li>radeong: fix leaks</li>
-  <li>r300g: silence guard band cap errors</li>
-</ul></p>
-
-<p>Johann Rudloff (3):
-<ul>
-  <li>radeon: Implement EGL_MESA_no_surface_extension</li>
-  <li>radeon: Implement __DRI_IMAGE and EGL_MESA_image_drm</li>
-  <li>radeon: Implement GL_OES_EGL_image</li>
-</ul></p>
-
-<p>John Doe (3):
-<ul>
-  <li>r600g: misc cleanup</li>
-  <li>r600g: don't double count dirty block</li>
-  <li>r600g: keep a mapping around for each bo</li>
-</ul></p>
-
-<p>Jon TURNEY (1):
-<ul>
-  <li>Ensure -L$(TOP)/$(LIB_DIR) appears in link line before any -L in $LDFLAGS</li>
-</ul></p>
-
-<p>José Fonseca (128):
-<ul>
-  <li>gallivm: Fix address register swizzle.</li>
-  <li>gallivm: Start collecting bitwise arithmetic helpers in a new module.</li>
-  <li>gallivm: Clamp indirect register indices to file_max.</li>
-  <li>util: linearized sRGB values don't fit into 8bits</li>
-  <li>llvmpipe: Default to no threading on single processor systems.</li>
-  <li>tgsi: Don't ignore indirect registers in tgsi_check_soa_dependencies</li>
-  <li>llvmpipe: Describe how to profile llvmpipe.</li>
-  <li>llvmpipe: When failing free fs shader too.</li>
-  <li>util: Flush stdout on util_format.</li>
-  <li>gallivm: Add unorm support to lp_build_lerp()</li>
-  <li>llvmpipe: Special case complementary and identify blend factors in SoA.</li>
-  <li>llvmpipe: Make rgb/alpha bland func/factors match, when there is no alpha.</li>
-  <li>draw: Prevent clipped vertices overflow.</li>
-  <li>draw: Fullfil the new min_lod/max_lod/lod_bias/border_color dynamic state</li>
-  <li>gallivm: Fetch the lod from the dynamic state when min_lod == max_lod.</li>
-  <li>gallivm: Remove dead experimental code.</li>
-  <li>llvmpipe: Decouple sampler view and sampler state updates.</li>
-  <li>scons: New build= option, with support for checked builds.</li>
-  <li>scons: New build= option, with support for checked builds.</li>
-  <li>trace: Fix set_index_buffer and draw_vbo tracing.</li>
-  <li>python/retrace: Handle set_index_buffer and draw_vbo.</li>
-  <li>gallivm: Use SSE4.1's ROUNDSS/ROUNDSD for scalar rounding.</li>
-  <li>gallivm: More comprehensive border usage logic.</li>
-  <li>retrace: Handle clear_render_target and clear_depth_stencil.</li>
-  <li>llvmpipe: Dump a few missing shader key flags.</li>
-  <li>llvmpipe: Fix perspective interpolation for point sprites.</li>
-  <li>llvmpipe: Fix sprite coord perspective interpolation of Q.</li>
-  <li>gallivm: Take the type signedness in consideration in round/ceil/floor.</li>
-  <li>gallivm: Use a faster (and less accurate) log2 in lod computation.</li>
-  <li>gallivm: Fast implementation of iround(log2(x))</li>
-  <li>gallivm: Combined ifloor &amp; fract helper.</li>
-  <li>gallivm: Only apply min/max_lod when necessary.</li>
-  <li>gallivm: Compute lod as integer whenever possible.</li>
-  <li>util: Cleanup util_pack_z_stencil and friends.</li>
-  <li>llvmpipe: Cleanup depth-stencil clears.</li>
-  <li>gallivm: Vectorize the rho computation.</li>
-  <li>gallivm: Do not do mipfiltering when magnifying.</li>
-  <li>gallivm: Simplify lp_build_mipmap_level_sizes' interface.</li>
-  <li>gallivm: Don't compute the second mipmap level when frac(lod) == 0</li>
-  <li>gallivm: Use lp_build_ifloor_fract for lod computation.</li>
-  <li>gallivm: Clamp mipmap level and zero mip weight simultaneously.</li>
-  <li>gallivm: Fix copy'n'paste typo in previous commit.</li>
-  <li>gallivm: Implement brilinear filtering.</li>
-  <li>gallivm: Use the wrappers for SSE pack intrinsics.</li>
-  <li>gallivm: Avoid control flow for two-sided stencil test.</li>
-  <li>gallivm: Warn when doing inefficient integer comparisons.</li>
-  <li>gallivm: Move into the as much of the second level code as possible.</li>
-  <li>llvmpipe: First minify the texture size, then broadcast.</li>
-  <li>gallivm: Help for combined extraction and broadcasting.</li>
-  <li>gallivm: Do size computations simultanously for all dimensions (AoS).</li>
-  <li>llvmpipe: Prevent z &gt; 1.0</li>
-  <li>llvmpipe: Fix MSVC build. Enable the new SSE2 code on non SSE3 systems.</li>
-  <li>gallivm: Handle code have ret correctly.</li>
-  <li>util: Defined M_SQRT2 when not available.</li>
-  <li>gallivm: Less code duplication in log computation.</li>
-  <li>gallivm: Special bri-linear computation path for unmodified rho.</li>
-  <li>gallivm: Don't generate Phis for execution mask.</li>
-  <li>gallivm: Use varilables instead of Phis for cubemap selection.</li>
-  <li>gallivm: Remove support for Phi generation.</li>
-  <li>gallivm: Factor out the SI-&gt;FP texture size conversion for SoA path too</li>
-  <li>gallivm: Simplify if/then/else implementation.</li>
-  <li>gallivm: Cleanup the rest of the flow module.</li>
-  <li>gallivm: Fix a long standing bug with nested if-then-else emission.</li>
-  <li>gallivm: Allow to disable bri-linear filtering with GALLIVM_DEBUG=no_brilinear runtime option</li>
-  <li>gallivm: Use variables instead of Phis in loops.</li>
-  <li>gallivm: Pass texture coords derivates as scalars.</li>
-  <li>llvmpipe: Remove outdated comment about stencil testing.</li>
-  <li>gallivm: Eliminate unsigned integer arithmetic from texture coordinates.</li>
-  <li>gallium: Define C99 restrict keyword where absent.</li>
-  <li>tgsi: Export some names for some tgsi enums.</li>
-  <li>gallivm: More detailed analysis of tgsi shaders.</li>
-  <li>llvmpipe: Use lp_tgsi_info.</li>
-  <li>llvmpipe: Do not dispose the execution engine.</li>
-  <li>llvmpipe: Fix MSVC build.</li>
-  <li>llmvpipe: improve mm_mullo_epi32</li>
-  <li>gallivm: Name anonymous union.</li>
-  <li>llvmpipe: Unbreak Z32_FLOAT.</li>
-  <li>gallivm: More accurate float -&gt; 24bit &amp; 32bit unorm conversion.</li>
-  <li>llvmpipe: Generalize the x8z24 fast path to all depth formats.</li>
-  <li>llvmpipe: Fix depth-stencil regression.</li>
-  <li>llvmpipe: Ensure z_shift and z_width is initialized.</li>
-  <li>gallivm: Fix SoA cubemap derivative computation.</li>
-  <li>llvmpipe: Fix bad refactoring.</li>
-  <li>llvmpipe: Initialize bld ctx via lp_build_context_init instead of ad-hoc and broken code.</li>
-  <li>gallivm: Comment lp_build_insert_new_block().</li>
-  <li>gallivm: Add a note about SSE4.1's nearest mode rounding.</li>
-  <li>llvmpipe: Don't test rounding of x.5 numbers.</li>
-  <li>gallium: Avoid using __doc__ in python scripts.</li>
-  <li>gallivm: always enable LLVMAddInstructionCombiningPass()</li>
-  <li>gallivm: Remove the EMMS opcodes.</li>
-  <li>mesa: Fix windows build (uint -&gt; GLuint).</li>
-  <li>scons: Revamp how to specify targets to build.</li>
-  <li>scons: Fix MinGW cross-compilation.</li>
-  <li>scons: Some pipe drivers are not portable for MSVC</li>
-  <li>scons: Restore x11 tool behavior for backwards compatability.</li>
-  <li>scons: Disable python state tracker when swig is not present.</li>
-  <li>r600g: List recently added files in SConscript.</li>
-  <li>scons: Add aliases for several pipe drivers.</li>
-  <li>scons: i915 can't build on MSVC either.</li>
-  <li>scons: Propagate installation targets.</li>
-  <li>xorg/vmwgfx: Add missing source file to SConscript.</li>
-  <li>st/xorg: Add missing n to error message.</li>
-  <li>st/xorg: Detect libkms with scons too.</li>
-  <li>xorg/vmwgfx: Link libkms when available.</li>
-  <li>r600g: Swap the util_blitter_destroy call order.</li>
-  <li>gallivm: Allocate TEMP/OUT arrays only once.</li>
-  <li>libgl-gdi: Allow to pick softpipe/llvmpipe on runtime.</li>
-  <li>scons: Use inline wrap helpers more consistently.</li>
-  <li>svga: Use consistent hexadecimal representation on debug output.</li>
-  <li>scons: Alias for svga</li>
-  <li>wgl: Stub WGL_ARB_pbuffer support.</li>
-  <li>wgl: More complete WGL_ARB_pbuffer support.</li>
-  <li>svga: Silence debug printf.</li>
-  <li>scons: Move MSVS_VERSION option to common module.</li>
-  <li>vega: Remove extraneous ;</li>
-  <li>retrace: Some fixes.</li>
-  <li>util: C++ safe.</li>
-  <li>wgl: Fix double free. Remove dead code.</li>
-  <li>util: Plug leaks in util_destroy_gen_mipmap.</li>
-  <li>util: __builtin_frame_address() doesn't work on mingw.</li>
-  <li>util: Don't try to use imagehlp on mingw.</li>
-  <li>wgl: Unreference the current framebuffer after the make_current call.</li>
-  <li>WIN32_THREADS -&gt; WIN32</li>
-  <li>mapi: Hack to avoid vgCreateFont being generated as vgCreateFontA.</li>
-  <li>wgl: Fix visual's buffer_mask configuration.</li>
-  <li>mesa: Temporary hack to prevent stack overflow on windows</li>
-  <li>mesa: Bump the number of bits in the register index.</li>
-  <li>llvmpipe: Plug fence leaks.</li>
-</ul></p>
-
-<p>Julien Cristau (1):
-<ul>
-  <li>Makefile: don't include the same files twice in the tarball</li>
-</ul></p>
-
-<p>Keith Whitwell (89):
-<ul>
-  <li>llvmpipe: brackets around macro arg</li>
-  <li>llvmpipe: remove duplicate code</li>
-  <li>llvmpipe: return zero from floor_pot(zero)</li>
-  <li>gallivm: make lp_build_sample_nop public</li>
-  <li>llvmpipe: add LP_PERF flag to disable various aspects of rasterization</li>
-  <li>llvmpipe: add DEBUG_FS to dump variant information</li>
-  <li>llvmpipe: use llvm for attribute interpolant calculation</li>
-  <li>graw: add frag-face shader</li>
-  <li>llvmpipe: fix flatshading in new line code</li>
-  <li>draw: don't apply flatshading to clipped tris with &lt3 verts</li>
-  <li>llvmpipe: handle FACING interpolants in line and point setup</li>
-  <li>llvmpipe: handle up to 8 planes in triangle binner</li>
-  <li>llvmpipe: make debug_fs_variant respect variant-&gt;nr_samplers</li>
-  <li>gallivm: don't apply zero lod_bias</li>
-  <li>llvmpipe: fail gracefully on oom in scene creation</li>
-  <li>llvmpipe: avoid overflow in triangle culling</li>
-  <li>gallivm: special case conversion 4x4f to 1x16ub</li>
-  <li>gallivm: round rather than truncate in new 4x4f-&gt;1x16ub conversion path</li>
-  <li>llvmpipe: clean up setup_tri a little</li>
-  <li>llvmpipe: add rast_tri_4_16 for small lines and points</li>
-  <li>llvmpipe: fix off-by-one in tri_16</li>
-  <li>llvmpipe: defer attribute interpolation until after mask and ztest</li>
-  <li>llvmpipe: use alloca for fs color outputs</li>
-  <li>llvmpipe: store zero into all alloca'd values</li>
-  <li>llvmpipe: dump fragment shader ir and asm when LP_DEBUG=fs</li>
-  <li>gallivm: specialized x8z24 depthtest path</li>
-  <li>gallivm: prefer blendvb for integer arguments</li>
-  <li>gallivm: simpler uint8-&gt;float conversions</li>
-  <li>llvmpipe: try to be sensible about whether to branch after mask updates</li>
-  <li>llvmpipe: clean up shader pre/postamble, try to catch more early-z</li>
-  <li>llvmpipe: simplified SSE2 swz/unswz routines</li>
-  <li>llvmpipe: try to do more of rast_tri_3_16 with intrinsics</li>
-  <li>llvmpipe: add debug helpers for epi32 etc</li>
-  <li>llvmpipe: try to keep plane c values small</li>
-  <li>llvmpipe: fix typo in last commit</li>
-  <li>gallium: move sse intrinsics debug helpers to u_sse.h</li>
-  <li>r600g: add missing file to sconscript</li>
-  <li>gallivm: don't branch on KILLs near end of shader</li>
-  <li>Revert "llvmpipe: try to keep plane c values small"</li>
-  <li>llvmpipe: make sure intrinsics code is guarded with PIPE_ARCH_SSE</li>
-  <li>llvmpipe: don't try to emit non-existent color outputs</li>
-  <li>r600/drm: fix segfaults in winsys create failure path</li>
-  <li>r600g: emit hardware linewidth</li>
-  <li>r600g: handle absolute modifier in shader translator</li>
-  <li>llvmpipe: reintroduce SET_STATE binner command</li>
-  <li>llvmpipe: don't pass frontfacing as a float</li>
-  <li>llvmpipe: slightly shrink the size of a binned triangle</li>
-  <li>llvmpipe: don't store plane.ei value in binned data</li>
-  <li>gallium: move some intrinsics helpers to u_sse.h</li>
-  <li>llvmpipe: do plane calculations with intrinsics</li>
-  <li>llvmpipe: use aligned loads/stores for plane values</li>
-  <li>llvmpipe: fix non-sse build after recent changes</li>
-  <li>llvmpipe: check shader outputs are non-null before using</li>
-  <li>llvmpipe: validate color outputs against key-&gt;nr_cbufs</li>
-  <li>llvmpipe: clean up fields in draw_llvm_variant_key</li>
-  <li>llvmpipe: remove setup fallback path</li>
-  <li>llvmpipe: fail cleanly on malloc failure in lp_setup_alloc_triangle</li>
-  <li>Merge remote branch 'origin/master' into lp-setup-llvm</li>
-  <li>llvmpipe: remove unused file</li>
-  <li>llvmpipe: remove unused arg from jit_setup_tri function</li>
-  <li>Merge branch 'llvm-cliptest-viewport'</li>
-  <li>draw: make sure viewport gets updated in draw llvm shader</li>
-  <li>llvmpipe: turn off draw offset/twoside when we can handle it</li>
-  <li>llvmpipe: avoid generating tri_16 for tris which extend past tile bounds</li>
-  <li>llvmpipe: guard against NULL task-&gt;query pointer</li>
-  <li>st/mesa: unbind constant buffer when not in use</li>
-  <li>r600g: propagate usage flags in texture transfers</li>
-  <li>r600g: propogate resource usage flags to winsys, use to choose bo domains</li>
-  <li>r600g: use a buffer in GTT as intermediate on texture up and downloads</li>
-  <li>r600g: remove unused flink, domain fields from r600_resource</li>
-  <li>r600g: set hardware pixel centers according to gl_rasterization_rules</li>
-  <li>evergreeng: protect against null constant buffers</li>
-  <li>r600g: don't call debug_get_bool_option for tiling more than once</li>
-  <li>evergreeng: respect linewidth state, use integer widths only</li>
-  <li>evergreeng: set hardware pixelcenters according to gl_rasterization_rules</li>
-  <li>r600g: avoid recursion with staged uploads</li>
-  <li>r600g: attempt to turn on DXTn formats</li>
-  <li>r600g: translate ARR instruction</li>
-  <li>r600: fix my pessimism about PIPE_TRANSFER_x flags</li>
-  <li>ws/r600: match bo_busy shared/fence logic in bo_wait</li>
-  <li>r600g: guard experimental s3tc code with R600_ENABLE_S3TC</li>
-  <li>r600g: do not try to use staging resource for depth textures</li>
-  <li>r600g: enforce minimum stride on render target texture images</li>
-  <li>llvmpipe: fix up twoside after recent changes</li>
-  <li>llvmpipe: twoside for specular color also</li>
-  <li>Merge branch 'lp-offset-twoside'</li>
-  <li>llvmpipe: raise dirty flag on transfers to bound constbuf</li>
-  <li>llvmpipe: remove misleading debug string</li>
-  <li>llvmpipe: shortcircuit some calls to set_scene_state</li>
-</ul></p>
-
-<p>Kenneth Graunke (94):
-<ul>
-  <li>glsl: Change from has_builtin_signature to has_user_signature.</li>
-  <li>glsl: Don't print blank (function ...) headers for built-ins.</li>
-  <li>glsl: Properly handle nested structure types.</li>
-  <li>glsl/builtins: Fix equal and notEqual builtins.</li>
-  <li>glsl/builtins: Switch comparison functions to just return an expression.</li>
-  <li>glsl: Add comments to clarify the types of comparison binops.</li>
-  <li>glsl: Fix broken handling of ir_binop_equal and ir_binop_nequal.</li>
-  <li>glsl: "Copyright", not "Constantright"</li>
-  <li>i965: Fix incorrect batchbuffer size in gen6 clip state command.</li>
-  <li>i965: Use logical-not when emitting ir_unop_ceil.</li>
-  <li>glsl: Add front-end support for the "trunc" built-in.</li>
-  <li>glsl: Refresh autogenerated file builtin_function.cpp.</li>
-  <li>i965: Use RNDZ for ir_unop_trunc in the new FS.</li>
-  <li>i965: Correctly emit the RNDZ instruction.</li>
-  <li>i965: Clean up a warning in the old fragment backend.</li>
-  <li>glsl: Add a new ir_unop_round_even opcode for GLSL 1.30's roundEven.</li>
-  <li>glsl: Add front-end support for GLSL 1.30's roundEven built-in.</li>
-  <li>i965: Add support for ir_unop_round_even via the RNDE instruction.</li>
-  <li>glsl: Add support for the 1.30 round() built-in.</li>
-  <li>glsl: Refresh autogenerated file builtin_function.cpp.</li>
-  <li>glsl: Don't return NULL IR for erroneous bit-shift operators.</li>
-  <li>i965: Add missing "break" statement.</li>
-  <li>glsl: Fix copy and paste error in ast_bit_and node creation.</li>
-  <li>glsl: Regenerate parser files.</li>
-  <li>i965: Remove unused variable.</li>
-  <li>glsl: Remove useless ir_shader enumeration value.</li>
-  <li>mesa: Remove FEATURE_ARB_shading_language_120 macro.</li>
-  <li>glcpp: Return NEWLINE token for newlines inside multi-line comments.</li>
-  <li>glcpp: Refresh autogenerated lexer file.</li>
-  <li>glsl: Add support for GLSL 1.30's modf built-in.</li>
-  <li>glsl: Refresh autogenerated file builtin_function.cpp.</li>
-  <li>generate_builtins.py: Output large strings as arrays of characters.</li>
-  <li>Refresh autogenerated file builtin_function.cpp.</li>
-  <li>glsl: Fix constant component count in vector constructor emitting.</li>
-  <li>Fix build on systems where "python" is python 3.</li>
-  <li>i965: Add bit operation support to the fragment shader backend.</li>
-  <li>glsl: Remove unused ARRAY_SIZE macro.</li>
-  <li>glsl/builtins: Rename 'x' to 'y_over_x' in atan(float) implementation.</li>
-  <li>glsl/builtins: Clean up some ugly autogenerated code in atan.</li>
-  <li>Refresh autogenerated file builtin_function.cpp.</li>
-  <li>glsl: Don't print a useless space at the end of an S-Expression list.</li>
-  <li>ir_reader: Return a specific ir_dereference variant.</li>
-  <li>ir_reader: Remove useless error check.</li>
-  <li>ir_reader: Fix some potential NULL pointer dereferences.</li>
-  <li>ir_dead_functions: Actually free dead functions and signatures.</li>
-  <li>glsl: Remove unnecessary "unused variable" warning suppression.</li>
-  <li>glsl: Remove GLSL_TYPE_FUNCTION define.</li>
-  <li>glsl: Convert glsl_type::base_type from #define'd constants to an enum.</li>
-  <li>glsl: Rework reserved word/keyword handling in the lexer.</li>
-  <li>glsl: Add new keywords and reserved words for GLSL 1.30.</li>
-  <li>glsl: Add support for the 'u' and 'U' unsigned integer suffixes.</li>
-  <li>glsl: Refresh autogenerated lexer and parser files.</li>
-  <li>generate_builtins.py: Fix inconsistent use of tabs and spaces warning.</li>
-  <li>glsl: Implement the asinh, acosh, and atanh built-in functions.</li>
-  <li>glsl: Refresh autogenerated file builtin_function.cpp.</li>
-  <li>glsl: Add constant expression handling for asinh, acosh, and atanh.</li>
-  <li>glsl: Remove unused and out of date Makefile.am.</li>
-  <li>glsl: Rename various ir_* files to lower_* and opt_*.</li>
-  <li>glcpp: Define GL_FRAGMENT_PRECISION_HIGH if GLSL version &gt;= 1.30.</li>
-  <li>Refresh autogenerated glcpp parser.</li>
-  <li>glsl: Fix constant expression handling for &lt, &gt;, &lt=, &gt;= on vectors.</li>
-  <li>glsl: Unconditionally define GL_FRAGMENT_PRECISION_HIGH in ES2 shaders.</li>
-  <li>Regenerate glcpp parser.</li>
-  <li>glsl: Reimplement the "cross" built-in without ir_binop_cross.</li>
-  <li>Refresh autogenerated file builtin_function.cpp.</li>
-  <li>glsl: Remove the ir_binop_cross opcode.</li>
-  <li>glsl: Refactor get_num_operands.</li>
-  <li>glsl: Simplify a type check by using type-&gt;is_integer().</li>
-  <li>glsl: Combine many instruction lowering passes into one.</li>
-  <li>mesa: Fix glGet of ES2's GL_MAX_*_VECTORS properties.</li>
-  <li>glsl: Don't inline function prototypes.</li>
-  <li>glsl: Use do_common_optimization in the standalone compiler.</li>
-  <li>glsl: Add a virtual as_discard() method.</li>
-  <li>glsl: Refactor out cloning of function prototypes.</li>
-  <li>glsl: Lazily import built-in function prototypes.</li>
-  <li>glsl: Remove anti-built-in hacks from the print visitor.</li>
-  <li>glsl/linker: Free any IR discarded by optimization passes.</li>
-  <li>glsl: Add an optimization pass to simplify discards.</li>
-  <li>glsl: Add a lowering pass to move discards out of if-statements.</li>
-  <li>glsl: Remove "discard" support from lower_jumps.</li>
-  <li>glsl: Add comments to lower_jumps (from the commit message).</li>
-  <li>ir_print_visitor: Print out constant structure values.</li>
-  <li>glsl: Factor out code which emits a new function into the IR stream.</li>
-  <li>symbol_table: Add support for adding a symbol at top-level/global scope.</li>
-  <li>glsl: Properly add functions during lazy built-in prototype importing.</li>
-  <li>glcpp: Don't emit SPACE tokens in conditional_tokens production.</li>
-  <li>Refresh autogenerated glcpp parser.</li>
-  <li>glsl: Clean up code by adding a new is_break() function.</li>
-  <li>glsl: Consider the "else" branch when looking for loop breaks.</li>
-  <li>Remove OES_compressed_paletted_texture from the ES2 extension list.</li>
-  <li>glsl/builtins: Compute the correct value for smoothstep(vec, vec, vec).</li>
-  <li>glsl: Support if-flattening beyond a given maximum nesting depth.</li>
-  <li>i965: Flatten if-statements beyond depth 16 on pre-gen6.</li>
-  <li>i965: Internally enable GL_NV_blend_square on ES2.</li>
-</ul></p>
-
-<p>Kristian Høgsberg (16):
-<ul>
-  <li>glx: Hold on to drawables if we're just switching to another context</li>
-  <li>intel: Fix GL_ARB_shading_language_120 commit</li>
-  <li>dri2: Make createImageFromName() take a __DRIscreen instead of __DRIcontext</li>
-  <li>glx: Invalidate buffers after binding a drawable</li>
-  <li>dri: Pass the __DRIscreen and the __DRIscreen private back to image lookup</li>
-  <li>glx: Only remove drawables from the hash when we actually delete them</li>
-  <li>gles2: Add GL_EXT_texture_format_BGRA8888 support</li>
-  <li>Get rid of GL/internal/glcore.h</li>
-  <li>gl: Remove unused GLcontextModes fields</li>
-  <li>Rename GLvisual and __GLcontextModes to struct gl_config</li>
-  <li>Drop GLframebuffer typedef and just use struct gl_framebuffer</li>
-  <li>Drop GLcontext typedef and use struct gl_context instead</li>
-  <li>Drop the "neutral" tnl module</li>
-  <li>Only install vtxfmt tables for OpenGL</li>
-  <li>i965: Don't write mrf assignment for pointsize output</li>
-  <li>docs: Fix MESA_drm_image typo</li>
-</ul></p>
-
-<p>Krzysztof Smiechowicz (1):
-<ul>
-  <li>nvfx: Pair os_malloc_aligned() with os_free_aligned().</li>
-</ul></p>
-
-<p>Luca Barbieri (84):
-<ul>
-  <li>auxiliary: fix unintended fallthrough</li>
-  <li>glsl: add pass to lower variable array indexing to conditional assignments</li>
-  <li>auxiliary: fix depth-only and stencil-only clears</li>
-  <li>gallium: avoid the C++ keyword "template" in sw_winsys.h</li>
-  <li>softpipe: make z/s test always pass if no zsbuf, instead of crashing</li>
-  <li>tgsi: add switch/case opcodes to tgsi_opcode_tmp.h</li>
-  <li>softpipe: fix whitespace</li>
-  <li>d3d1x: add new Direct3D 10/11 COM state tracker for Gallium</li>
-  <li>d3d1x: add blob and signature extraction APIs</li>
-  <li>d3d1x: fix compilation with recent Wine versions installed</li>
-  <li>d3d1x: add missing file</li>
-  <li>d3d1x: actually enable and fix blob apis</li>
-  <li>d3d1x: fix build with compilers other than GCC 4.5</li>
-  <li>d3d1x: add template parameters to base class ctor calls for GCC 4.4</li>
-  <li>d3d1x: fix GCC 4.1/4.2 build</li>
-  <li>d3d1x: ignore errors while building docs</li>
-  <li>d3d1x: attempt to fix/workaround bug #30322</li>
-  <li>nvfx: remove gl_PointCoord hack</li>
-  <li>glx: decouple dri2.c and GLX, fixing Gallium EGL and d3d1x build</li>
-  <li>winsys: automatically build sw winsys needed by EGL and d3d1x</li>
-  <li>d3d1x: don't build progs automatically</li>
-  <li>d3d1x: add missing memory barrier</li>
-  <li>d3d1x: link with CXXFLAGS</li>
-  <li>d3d1x: fix cf analysis</li>
-  <li>d3d1x: fix warning</li>
-  <li>d3d1x: fix segfault when hashing</li>
-  <li>d3d1x: destroy native_display on adapter destruction</li>
-  <li>d3d1x: fix GUID declarations</li>
-  <li>d3d1x: redesign the HWND resolver interface</li>
-  <li>d3d1x: fix API name</li>
-  <li>d3d1x: define GUIDs in the normal way</li>
-  <li>d3d1x: add Wine dlls (tri, tex working, but no other testing)</li>
-  <li>d3d1x: properly reference count the backend</li>
-  <li>d3d1x: fix deadlocks on non-recursive mutex</li>
-  <li>d3d1x: bind NULL CSOs before destroying default CSOs on context dtor</li>
-  <li>d3d1x: initialize the mutex</li>
-  <li>d3d1x: autogenerate shader enums and text from def files</li>
-  <li>d3d1x: s/tpf/sm4/g</li>
-  <li>d3d1x: normalize whitespace</li>
-  <li>d3d1x: remove specstrings</li>
-  <li>d3d1x: minifix</li>
-  <li>d3d1x: rename context params</li>
-  <li>d3d11: rename screen params</li>
-  <li>d3d1x: rename params in misc and objects</li>
-  <li>d3d1x: rename parameters in dxgi</li>
-  <li>d3d11: obliterate IDL parameter names</li>
-  <li>d3d1x: remove specstrings.h include</li>
-  <li>d3d1x: flush the pipe context when presenting</li>
-  <li>d3d1x: remove another include specstrings.h</li>
-  <li>d3d1x: flush properly</li>
-  <li>d3d1x: add missing guid.cpp</li>
-  <li>d3d1x: fix build without system EGL/egl.h</li>
-  <li>d3d1x: add autogenerated files as prerequisites, so make builds them</li>
-  <li>d3d1x: obliterate IDL parameter names from d3d10.idl from Wine too</li>
-  <li>d3d1x: add shader dumping</li>
-  <li>d3d1x: add untested support for geometry shader translation</li>
-  <li>d3d1x: don't assert on unsupported resource types</li>
-  <li>d3d1x: fix CheckMultisampleQualityLevels</li>
-  <li>d3d1x: draw to the correct buffer</li>
-  <li>d3d1x: fix linking of dxbc2tgsi</li>
-  <li>nvfx: allow setting NULL constant buffers</li>
-  <li>nvfx: add RGB framebuffer format support in addition to BGR</li>
-  <li>d3d1x: don't crash on drivers not supporting vertex or geometry sampling</li>
-  <li>d3d1x: assert if X visual is not among enumerated visuals</li>
-  <li>d3d1x: stop using GLX in demos, just use the default visual</li>
-  <li>d3d1x: CRLF -&gt; LF in progs</li>
-  <li>mesa: make makedepend an hard requirement</li>
-  <li>gallium: add $(PROGS_DEPS) as dependencies for $(PROGS)</li>
-  <li>d3d1x: fix parallel build</li>
-  <li>d3d1x: add private gitignore file</li>
-  <li>d3d1x: fix progs linking if not all EGL platforms are enabled</li>
-  <li>d3d1x: link progs with CXXFLAGS</li>
-  <li>d3d11: advertise IDXGIDevice1, not just IDXGIDevice</li>
-  <li>d3d11: ignore StructureByteStride</li>
-  <li>d3d1x: link to libdrm for X11 platform too</li>
-  <li>ureg: support centroid interpolation</li>
-  <li>d3d1x: support centroid interpolation</li>
-  <li>d3d1x: properly support specifying MipLevels as 0</li>
-  <li>d3d1x: put proper calling convention in headers, fixes 64-bit builds</li>
-  <li>d3d1x: rework DXGI for occlusion testing and default width/height</li>
-  <li>d3d1x: fix Map</li>
-  <li>d3d11: fix reference counting so devices get freed</li>
-  <li>d3d1x: work around crash in widl</li>
-  <li>glsl: Unroll loops with conditional breaks anywhere (not just the end)</li>
-</ul></p>
-
-<p>Lucas Stach (1):
-<ul>
-  <li>nvfx: fill PIPE_CAP_PRIMITIVE_RESTART and PIPE_CAP_SHADER_STENCIL_EXPORT</li>
-</ul></p>
-
-<p>Marek Olšák (100):
-<ul>
-  <li>r300g: prevent creating multiple winsys BOs for the same handle</li>
-  <li>r300g/swtcl: fix CS overrun</li>
-  <li>st/mesa: fix assertion failure in GetTexImage for cubemaps</li>
-  <li>util: make calling remove_from_list multiple times in a row safe</li>
-  <li>r300g: fixup long-lived BO maps being incorrectly unmapped when flushing</li>
-  <li>r300g: make accessing map_list and buffer_handles thread-safe</li>
-  <li>r300g: fix a copy-paste typo for logging</li>
-  <li>r300g: fix the border color for every format other than PIPE_FORMAT_B8G8R8A8</li>
-  <li>Build r300g by default</li>
-  <li>util: fix util_pack_color for B4G4R4A4</li>
-  <li>r300g: fix macrotiling on R350</li>
-  <li>r300g: code cleanups</li>
-  <li>r300/compiler: fix projective mapping of 2D NPOT textures</li>
-  <li>r300/compiler: do not use copy propagation if SaturateMode is used</li>
-  <li>r300/compiler: fix shadow sampling with swizzled coords</li>
-  <li>r300g: add support for 3D NPOT textures without mipmapping</li>
-  <li>r300g: fix swizzling of texture border color</li>
-  <li>configure.ac: look for libdrm_radeon before building gallium/r300,r600</li>
-  <li>configure.ac: do not build xorg-r300g by default</li>
-  <li>Makefile: ensure Gallium's Makefile.xorg and SConscript.dri are in the tarball</li>
-  <li>r300g: add support for formats beginning with X, like X8R8G8B8</li>
-  <li>r300g: fix conditional rendering in non-wait path</li>
-  <li>r300g: add support for R8G8 colorbuffers</li>
-  <li>r300g: add support for L8A8 colorbuffers</li>
-  <li>update release notes for Gallium</li>
-  <li>r300g: fix microtiling for 16-bits-per-channel formats</li>
-  <li>r300g: do not print get_param errors in non-debug build</li>
-  <li>r300g: say no to PIPE_CAP_STREAM_OUTPUT and PIPE_CAP_PRIMITIVE_RESTART</li>
-  <li>mesa: allow FBO attachments of formats LUMINANCE, LUMINANCE_ALPHA, and INTENSITY</li>
-  <li>r300g: fix texture border for 16-bits-per-channel formats</li>
-  <li>st/mesa: support RGBA16 and use it for RGBA12 as well</li>
-  <li>r300g: add a default channel ordering of texture border for unhandled formats</li>
-  <li>r300g: mention ATI in the renderer string</li>
-  <li>r300g: rename has_hyperz -&gt; can_hyperz</li>
-  <li>r300g: turn magic numbers into names in the hyperz code</li>
-  <li>gallium: add CAPs for indirect addressing and lower it in st/mesa when needed</li>
-  <li>tgsi: fill out CAPs for indirect addressing</li>
-  <li>i915g: fill out CAPs for indirect addressing</li>
-  <li>i965g: fill out CAPs for indirect addressing</li>
-  <li>nv50: fill out CAPs for indirect addressing</li>
-  <li>nvfx: fill out CAPs for indirect addressing</li>
-  <li>r300g: fill out CAPs for indirect addressing</li>
-  <li>r600g: fill out CAPs for indirect addressing</li>
-  <li>svga: fill out CAPs for indirect addressing</li>
-  <li>r300g: fix texture border color for all texture formats</li>
-  <li>r300g: clean up redundancy in draw functions</li>
-  <li>r300g: return shader caps from Draw for SWTCL vertex shaders</li>
-  <li>r300g: remove the hack with OPCODE_RET</li>
-  <li>r300g: print FS inputs uninitialized due to hardware limits to stderr</li>
-  <li>r300g: fix rendering with no vertex elements</li>
-  <li>st/mesa: enable ARB_explicit_attrib_location and EXT_separate_shader_objects</li>
-  <li>docs: add GL 4.1 status</li>
-  <li>gallium: add PIPE_SHADER_CAP_SUBROUTINES</li>
-  <li>st/mesa: set MaxUniformComponents</li>
-  <li>u_blitter: use PIPE_TRANSFER_DISCARD to prevent cpu/gpu stall</li>
-  <li>r300/compiler: fix rc_rewrite_depth_out for it to work with any instruction</li>
-  <li>r300/compiler: remove duplicate function rc_mask_to_swz</li>
-  <li>r300/compiler: add a function for swizzling a mask</li>
-  <li>r300/compiler: move util functions to radeon_compiler_util</li>
-  <li>u_blitter: interpolate clear color using a GENERIC varying instead of COLOR</li>
-  <li>st/mesa: fix texture border color for RED and RG base formats</li>
-  <li>util: rename u_mempool -&gt; u_slab</li>
-  <li>r300g: fix texture border color once again</li>
-  <li>r300/compiler: implement and lower OPCODE_CLAMP</li>
-  <li>ir_to_mesa: Add support for conditional discards.</li>
-  <li>r300g: fix texture swizzling with compressed textures on r400-r500</li>
-  <li>r300g: disable ARB_texture_swizzle if S3TC is enabled on r3xx-only</li>
-  <li>r300g: fix up cubemap texture offset computation</li>
-  <li>r300/compiler: disable the swizzle lowering pass in vertex shaders</li>
-  <li>r300g: fix build</li>
-  <li>r300g: use internal BO handle for add_buffer and write_reloc</li>
-  <li>r300g: implement simple transfer_inline_write for buffers</li>
-  <li>mesa, st/mesa: fix gl_FragCoord with FBOs in Gallium</li>
-  <li>r300g: fix pointer arithmetic with void* in transfer_inline_write</li>
-  <li>r300g: do not remove unused constants if we are not near the limit</li>
-  <li>r300g: add capability bit index_bias_supported</li>
-  <li>r300g: one more r500_index_bias_supported leftover</li>
-  <li>r300g: do not use the index parameter in set_constant_buffer</li>
-  <li>r300g: cleanup winsys</li>
-  <li>r300g: optimize looping over atoms</li>
-  <li>st/mesa: initialize key in st_vp_varient</li>
-  <li>u_blitter: use util_is_format_compatible in the assert</li>
-  <li>r300g: cache packet dwords of 3D_LOAD_VBPNTR in a command buffer if possible</li>
-  <li>r300g: validate buffers only if any of bound buffers is changed</li>
-  <li>r300g: also revalidate the SWTCL vertex buffer after its reallocation</li>
-  <li>r300/compiler: don't terminate regalloc if we surpass max temps limit</li>
-  <li>r300/compiler: add a function to query program stats (alu, tex, temps..)</li>
-  <li>r300/compiler: cleanup rc_run_compiler</li>
-  <li>r300/compiler: do not print pair/tex/presub program stats for vertex shaders</li>
-  <li>r300/compiler: handle DPH and XPD in rc_compute_sources_for_writemask</li>
-  <li>r300/compiler: make lowering passes possibly use up to two less temps</li>
-  <li>r300/compiler: remove at least unused immediates if externals cannot be removed</li>
-  <li>r300/compiler: fix LIT in VS</li>
-  <li>r300/compiler: fix swizzle lowering with a presubtract source operand</li>
-  <li>r300g: fix rendering with a vertex attrib having a zero stride</li>
-  <li>r300g: finally fix the texture corruption on r3xx-r4xx</li>
-  <li>r300g/swtcl: re-enable LLVM</li>
-  <li>r300g: mark vertex arrays as dirty after a buffer_offset change</li>
-  <li>mesa: fix texel store functions for some float formats</li>
-  <li>r300/compiler: disable the rename_regs pass for loops</li>
-</ul></p>
-
-<p>Mario Kleiner (1):
-<ul>
-  <li>mesa/r300classic: Fix dri2Invalidate/radeon_prepare_render for page flipping.</li>
-</ul></p>
-
-<p>Mathias Fröhlich (3):
-<ul>
-  <li>r300g: Avoid returning values in a static array, fixing a potential race</li>
-  <li>r600g: Only compare active vertex elements</li>
-  <li>st/mesa: Set PIPE_TRANSFER_DISCARD for GL_MAP_INVALIDATE_RANGE/BUFFFER_BIT</li>
-</ul></p>
-
-<p>Michal Krol (10):
-<ul>
-  <li>svga: Fix relative addressing translation for pixel shaders.</li>
-  <li>svga: Integer constant register file has a separate namespace.</li>
-  <li>tgsi/exec: Cleanup the remaining arithmetic instructions.</li>
-  <li>tgsi/exec: Get rid of obsolete condition codes.</li>
-  <li>tgsi/build: Reduce interface clutter.</li>
-  <li>graw/gdi: Initial commit.</li>
-  <li>scons: Hook-up graw-gdi target.</li>
-  <li>graw/gdi: Fix window dimensions.</li>
-  <li>os: Open file streams in binary mode.</li>
-  <li>graw: Export graw_save_surface_to_file().</li>
-</ul></p>
-
-<p>Nicolas Kaiser (26):
-<ul>
-  <li>swrast: remove duplicated include</li>
-  <li>egl: remove duplicated include</li>
-  <li>gallium/rtasm: remove duplicated include</li>
-  <li>gallium/util: remove duplicated include</li>
-  <li>gallium/i915: remove duplicated include</li>
-  <li>gallium/llvmpipe: remove duplicated include</li>
-  <li>gallium/softpipe: remove duplicated include</li>
-  <li>gallium/st: remove duplicated includes</li>
-  <li>gallium/winsys: remove duplicated include</li>
-  <li>glx: remove duplicated include</li>
-  <li>dri/common: remove duplicated include</li>
-  <li>dri/i810: remove duplicated include</li>
-  <li>dri/i915: remove duplicated include</li>
-  <li>dri/i965: remove duplicated include</li>
-  <li>dri/intel: remove duplicated include</li>
-  <li>dri/mga: remove duplicated include</li>
-  <li>dri/r128: remove duplicated include</li>
-  <li>dri/r300: remove duplicated include</li>
-  <li>dri/r600: remove duplicated include</li>
-  <li>dri/radeon: remove duplicated includes</li>
-  <li>dri/savage: remove duplicated include</li>
-  <li>main: remove duplicated includes</li>
-  <li>math: remove duplicated includes</li>
-  <li>st: remove duplicated include</li>
-  <li>i965g: use Elements macro instead of manual sizeofs</li>
-  <li>nv50: fix always true conditional in shader optimization</li>
-</ul></p>
-
-<p>Orion Poplawski (1):
-<ul>
-  <li>osmesa: link against libtalloc</li>
-</ul></p>
-
-<p>Owen W. Taylor (1):
-<ul>
-  <li>r600g: Fix location for clip plane registers</li>
-</ul></p>
-
-<p>Peter Clifton (3):
-<ul>
-  <li>intel: Fix emit_linear_blit to use DWORD aligned width blits</li>
-  <li>intel: Add assert check for blitting alignment.</li>
-  <li>meta: Mask Stencil.Clear against stencilMax in _mesa_meta_Clear</li>
-</ul></p>
-
-<p>Robert Hooker (2):
-<ul>
-  <li>intel: Add a new B43 pci id.</li>
-  <li>egl_dri2: Add missing intel chip ids.</li>
-</ul></p>
-
-<p>Roland Scheidegger (16):
-<ul>
-  <li>gallivm: fix copy&amp;paste bug</li>
-  <li>gallivm: don't use URem/UDiv when calculating offsets for blocks</li>
-  <li>gallivm: optimize yuv decoding</li>
-  <li>gallivm: fix trunc/itrunc comment</li>
-  <li>gallivm: faster iround implementation for sse2</li>
-  <li>gallivm: replace sub/floor/ifloor combo with ifloor_fract</li>
-  <li>gallivm: optimize some tex wrap mode calculations a bit</li>
-  <li>gallivm: more linear tex wrap mode calculation simplification</li>
-  <li>gallivm: avoid unnecessary URem in linear wrap repeat case</li>
-  <li>gallivm: optimize soa linear clamp to edge wrap mode a bit</li>
-  <li>gallivm: make use of new iround code in lp_bld_conv.</li>
-  <li>gallivm: fix different handling of [non]normalized coords in linear soa path</li>
-  <li>gallivm: only use lp_build_conv 4x4f -&gt; 1x16 ub fastpath with sse2</li>
-  <li>r200: fix r200 large points</li>
-  <li>mesa: remove unneeded DD_POINT_SIZE and DD_LINE_WIDTH tricaps</li>
-  <li>gallium: support for array textures and related changes</li>
-</ul></p>
-
-<p>Shuang He (1):
-<ul>
-  <li>mesa: allow GLfixed arrays for OpenGL ES 2.0</li>
-</ul></p>
-
-<p>Stephan Schmid (1):
-<ul>
-  <li>r600g: fix relative addressing when splitting constant accesses</li>
-</ul></p>
-
-<p>Thomas Hellstrom (21):
-<ul>
-  <li>st/xorg: Don't try to use option values before processing options</li>
-  <li>xorg/vmwgfx: Make vmwarectrl work also on 64-bit servers</li>
-  <li>st/xorg: Add a customizer option to get rid of annoying cursor update flicker</li>
-  <li>xorg/vmwgfx: Don't hide HW cursors when updating them</li>
-  <li>st/xorg: Don't try to remove invalid fbs</li>
-  <li>st/xorg: Fix typo</li>
-  <li>st/xorg, xorg/vmwgfx: Be a bit more frendly towards cross-compiling environments</li>
-  <li>st/xorg: Fix compilation errors for Xservers compiled without Composite</li>
-  <li>st/xorg: Don't use deprecated x*alloc / xfree functions</li>
-  <li>xorg/vmwgfx: Don't use deprecated x*alloc / xfree functions</li>
-  <li>st/xorg: Fix compilation for Xservers &gt;= 1.10</li>
-  <li>mesa: Make sure we have the talloc cflags when using the talloc headers</li>
-  <li>egl: Add an include for size_t</li>
-  <li>mesa: Add talloc includes for gles</li>
-  <li>st/egl: Fix build for include files in nonstandard places</li>
-  <li>svga/drm: Optionally resolve calls to powf during link-time</li>
-  <li>gallium/targets: Trivial crosscompiling fix</li>
-  <li>st/xorg: Add a function to flush pending rendering and damage</li>
-  <li>gallium/targets/xorg-vmwgfx: Xv fixes</li>
-  <li>xorg/vmwgfx: Flush even if we don't autopaint the color key</li>
-  <li>xorg/vmwgfx: Don't clip video to viewport</li>
-</ul></p>
-
-<p>Tilman Sauerbeck (35):
-<ul>
-  <li>r600g: Fixed a bo leak in r600_blit_state_ps_shader().</li>
-  <li>r600g: Use clamped math for RCP and RSQ.</li>
-  <li>r600g: Formatting fixes.</li>
-  <li>r600g: Added DB_SHADER_CONTROL defines.</li>
-  <li>r600g: Only set PA_SC_EDGERULE on rv770 and greater.</li>
-  <li>r600g: Enable PIPE_SHADER_CAP_TGSI_CONT_SUPPORTED.</li>
-  <li>r600g: Fixed the shift in S_02880C_KILL_ENABLE.</li>
-  <li>glsl2: Empty functions can be inlined.</li>
-  <li>glsl2: Fixed cloning of ir_call error instructions.</li>
-  <li>r600g: Added support for TGSI_SEMANTIC_FACE.</li>
-  <li>gallium/docs: Fixed a typo in the SCS opcode description.</li>
-  <li>r600g: Honour destination operand's writemask in the SCS implementation.</li>
-  <li>r600g: Implemented the Z and W component write for the SCS opcode.</li>
-  <li>python/tests: Fixed tri.py for API and TGSI syntax changes.</li>
-  <li>r600g: Removed debug code.</li>
-  <li>gallium/docs: The RET opcode may appear anywhere in a subroutine.</li>
-  <li>r600g: Destroy the blitter.</li>
-  <li>r600g: Fixed two texture surface leaks in r600_blit_uncompress_depth().</li>
-  <li>r600g: Cleaned up index buffer reference handling in the draw module.</li>
-  <li>r600g: Fixed r600_vertex_element leak.</li>
-  <li>r600g: Added r600_pipe_shader_destroy().</li>
-  <li>r600g: Also clear bc data when we're destroying a shader.</li>
-  <li>r600g: In radeon_bo(), call LIST_INITHEAD early.</li>
-  <li>r600g: Destroy the blitter.</li>
-  <li>r600g: Removed unused 'ptr' argument from radeon_bo().</li>
-  <li>r600g: Made radeon_bo_pb_map_internal() actually call radeon_bo_map().</li>
-  <li>r600g: Fixed unmap condition in radeon_bo_pb_destroy().</li>
-  <li>r600g: Made radeon_bo::map_count signed.</li>
-  <li>r600g: We don't support PIPE_CAP_PRIMITIVE_RESTART.</li>
-  <li>r600g: Delete custom_dsa_flush on shutdown.</li>
-  <li>r600g: Fixed two memory leaks in winsys.</li>
-  <li>r600g: Destroy the winsys in r600_destroy_screen().</li>
-  <li>st/mesa: Reset the index buffer before destroying the pipe context.</li>
-  <li>st/mesa: Reset the constant buffers before destroying the pipe context.</li>
-  <li>r600g: Removed duplicated call to tgsi_split_literal_constant().</li>
-</ul></p>
-
-<p>Timo Wiren (1):
-<ul>
-  <li>Fix typos in comments and debug output strings.</li>
-</ul></p>
-
-<p>Tom Fogal (3):
-<ul>
-  <li>Implement x86_64 atomics for compilers w/o intrinsics.</li>
-  <li>Prefer intrinsics to handrolled atomic ops.</li>
-  <li>Revert "Prefer intrinsics to handrolled atomic ops."</li>
-</ul></p>
-
-<p>Tom Stellard (32):
-<ul>
-  <li>r300/compiler: Refactor the pair instruction data structures</li>
-  <li>r300g: Always try to build libr300compiler.a</li>
-  <li>r300/compiler: Fix two mistakes in the presubtract optimization pass.</li>
-  <li>r300/compiler: Add more helper functions for iterating through sources</li>
-  <li>r300/compiler: Print immediate values after "dead constants" pass</li>
-  <li>r300/compiler: radeon_remove_constants.c: fix indentation</li>
-  <li>r300/compiler: Use rc_for_all_reads_src() in "dead constants" pass</li>
-  <li>r300/compiler: Fix segfault in error path</li>
-  <li>r300/compiler: Don't use rc_error() unless the error is unrecoverable</li>
-  <li>r300/compiler: Don't merge instructions that write output regs and ALU result</li>
-  <li>r300/compiler: Create a helper function for merging presubtract sources</li>
-  <li>r300/compiler: Fix incorrect assumption</li>
-  <li>r300/compiler: Clear empty registers after constant folding</li>
-  <li>r300/compiler: Add a new function for more efficient dataflow analysis</li>
-  <li>r300g: Add new debug option for logging vertex/fragment program stats</li>
-  <li>r300/compiler: Use rc_get_readers_normal() for presubtract optimizations</li>
-  <li>r300/compiler: Don't clobber presubtract sources during optimizations</li>
-  <li>r300/compiler: Don't track readers into an IF block.</li>
-  <li>r300/compiler: Make sure presubtract sources use supported swizzles</li>
-  <li>r300/compiler: Fix register allocator's handling of loops</li>
-  <li>r300/compiler: Fix instruction scheduling within IF blocks</li>
-  <li>r300/compiler: Use zero as the register index for unused sources</li>
-  <li>r300/compiler: Ignore alpha dest register when replicating the result</li>
-  <li>r300/compiler: Add rc_get_readers()</li>
-  <li>r300/compiler: Handle BREAK and CONTINUE in rc_get_readers()</li>
-  <li>r300/compiler: Track readers through branches in rc_get_readers()</li>
-  <li>r300/compiler: Convert RGB to alpha in the scheduler</li>
-  <li>r300/compiler: Use presubtract operations as much as possible</li>
-  <li>r300/compiler: Enable rename_reg pass for r500 cards</li>
-  <li>r300/compiler: Add a more efficient version of rc_find_free_temporary()</li>
-  <li>r300/compiler: Don't allow presubtract sources to be remapped twice</li>
-  <li>r300/compiler: Fix black terrain in Civ4</li>
-</ul></p>
-
-<p>Victor Tseng (1):
-<ul>
-  <li>egl/i965: include inline_wrapper_sw_helper.h</li>
-</ul></p>
-
-<p>Viktor Novotný (6):
-<ul>
-  <li>dri/nouveau: Import headers from rules-ng-ng</li>
-  <li>dri/nouveau: nv04: Use rules-ng-ng headers</li>
-  <li>dri/nouveau: nv10: Use rules-ng-ng headers</li>
-  <li>dri/nouveau nv20: Use rules-ng-ng headers</li>
-  <li>dri/nouveau: Remove nouveau_class.h, finishing switch to rules-ng-ng headers</li>
-  <li>dri/nouveau: Clean up magic numbers in get_rt_format</li>
-</ul></p>
-
-<p>Vinson Lee (214):
-<ul>
-  <li>llvmpipe: Remove unnecessary header.</li>
-  <li>r600g: Remove unnecessary headers.</li>
-  <li>mesa: Include missing header in program.h.</li>
-  <li>glsl: Fix 'format not a string literal and no format arguments' warning.</li>
-  <li>r600g: Silence uninitialized variable warning.</li>
-  <li>r600g: Silence uninitialized variable warning.</li>
-  <li>nvfx: Silence uninitialized variable warnings.</li>
-  <li>r600g: Silence uninitialized variable warning.</li>
-  <li>r600g: Silence uninitialized variable warning.</li>
-  <li>r600g: Silence uninitialized variable warning.</li>
-  <li>r600g: Silence unused variable warning.</li>
-  <li>nv50: Update files in SConscript to match Makefile.</li>
-  <li>nv50: Remove unnecessary headers.</li>
-  <li>nv50: Silence uninitialized variable warning.</li>
-  <li>nv50: Silence uninitialized variable warning.</li>
-  <li>nv50: Silence uninitialized variable warning.</li>
-  <li>gallivm: Remove unnecessary headers.</li>
-  <li>draw: Remove unnecessary header.</li>
-  <li>nv50: Silence uninitialized variable warnings.</li>
-  <li>nv50: Fix 'control reaches end of non-void function' warning.</li>
-  <li>mesa/st: Silence uninitialized variable warning.</li>
-  <li>gallivm: Remove unnecessary header.</li>
-  <li>r600g: Remove unnecessary header.</li>
-  <li>r600g: Remove unnecessary headers.</li>
-  <li>r600g: Fix implicit declaration warning.</li>
-  <li>r600g: Fix memory leak on error path.</li>
-  <li>r600g: Silence uninitialized variable warning.</li>
-  <li>r600g: Silence unused variable warnings.</li>
-  <li>mesa: bump version to 7.10</li>
-  <li>ir_to_mesa: Remove unused member array_indexed from struct statevar_element.</li>
-  <li>mesa: Silence "'valid_texture_object' defined but not used" warning.</li>
-  <li>x86: Silence unused variable warning on Mac OS X.</li>
-  <li>glsl: Fix 'control reaches end of non-void function' warning.</li>
-  <li>nvfx: Remove const qualifer from nvfx_vertprog_translate.</li>
-  <li>nvfx: Silence uninitialized variable warnings.</li>
-  <li>r600g: Remove unused variable.</li>
-  <li>nv50: Silence missing initializer warning.</li>
-  <li>nv50: Remove dead initialization.</li>
-  <li>nv50: Remove dead initialization.</li>
-  <li>tgsi: Remove duplicate case value.</li>
-  <li>glut: Define markWindowHidden for non-Windows only.</li>
-  <li>glut: Define eventParser for non-Windows only.</li>
-  <li>r300g: Silence uninitialized variable warning.</li>
-  <li>intel: Fix implicit declaration of function '_mesa_meta_Bitmap' warning.</li>
-  <li>mesa: Remove unnecessary headers.</li>
-  <li>r600g: Remove unnecessary header.</li>
-  <li>unichrome: Remove unnecessary header.</li>
-  <li>intel: Remove unnecessary headers.</li>
-  <li>r600g: Remove unused variable.</li>
-  <li>r600g: Disable unused variables.</li>
-  <li>r600g: Remove unused variable.</li>
-  <li>r600g: Silence 'control reaches end of non-void function' warning.</li>
-  <li>r600g: Remove unused variable.</li>
-  <li>r600g: Remove unused variable.</li>
-  <li>r600g: Disable unused variables.</li>
-  <li>intel: Remove unnecessary header.</li>
-  <li>st/dri: Remove unnecessary header.</li>
-  <li>r600g: Remove unused variable.</li>
-  <li>r300g: Remove unused variable.</li>
-  <li>r600g: Don't return a value in function returning void.</li>
-  <li>r600g: Remove unused variables.</li>
-  <li>r600g: Include p_compiler.h instead of malloc.h.</li>
-  <li>r600g: Silence uninitialized variable warnings.</li>
-  <li>scons: Add MinGW-w64 prefixes for MinGW build.</li>
-  <li>dri: Add GET_PROGRAM_NAME definition for Mac OS X.</li>
-  <li>scons: Add program/sampler.cpp to SCons build.</li>
-  <li>mesa: Fix printf format warning.</li>
-  <li>mesa: Fix printf format warning.</li>
-  <li>mesa: Fix printf format warning.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>r600g: Update SConscript.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>r600g: Update SConscript.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>r600g: Update SConscript.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>r600g: Fix SCons build.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>r300/compiler: Remove declaration before code.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>glsl: Remove unnecessary header.</li>
-  <li>savage: Remove unnecessary header.</li>
-  <li>r600g: Remove unused variable.</li>
-  <li>r600g: Remove unnecessary headers.</li>
-  <li>r600g: Fix SCons build.</li>
-  <li>r600g: Remove unnecessary header.</li>
-  <li>gallivm: Remove unnecessary header.</li>
-  <li>r600g: Silence uninitialized variable warning.</li>
-  <li>r600g: Silence uninitialized variable warning.</li>
-  <li>r600g: Silence uninitialized variable warning.</li>
-  <li>i915: Silence unused variable warning in non-debug builds.</li>
-  <li>i915: Silence unused variable warning in non-debug builds.</li>
-  <li>i965: Silence unused variable warning on non-debug builds.</li>
-  <li>i965: Silence unused variable warning on non-debug builds.</li>
-  <li>i965: Initialize member variables.</li>
-  <li>r300: Silence uninitialized variable warning.</li>
-  <li>tdfx: Silence unused variable warning on non-debug builds.</li>
-  <li>gallivm: Remove unnecessary header.</li>
-  <li>glsl: Initialize variable in ir_derefence_array::constant_expression_value</li>
-  <li>mesa: Add missing header to shaderobj.h.</li>
-  <li>llvmpipe: Return non-zero exit code for lp_test_round failures.</li>
-  <li>r300/compiler: Remove unused variable.</li>
-  <li>st/xorg: Fix memory leak on error path.</li>
-  <li>llvmpipe: Initialize state variable in debug_bin function.</li>
-  <li>llvmpipe: Initialize variable.</li>
-  <li>draw: Move loop variable declaration outside for loop.</li>
-  <li>r600g: Ensure r600_src is initialized in tgsi_exp function.</li>
-  <li>glsl: Add assert for unhandled ir_shader case.</li>
-  <li>swrast: Print out format on unexpected failure in _swrast_DrawPixels.</li>
-  <li>llvmpipe: Remove unnecessary header.</li>
-  <li>draw: Remove unnecessary header.</li>
-  <li>gallivm: Silence uninitialized variable warnings.</li>
-  <li>gallivm: Silence uninitialized variable warnings.</li>
-  <li>gallivm: Silence uninitialized variable warning.</li>
-  <li>r300g: Silence uninitialized variable warning.</li>
-  <li>mesa: Remove unnecessary headers.</li>
-  <li>r600g: Silence uninitialized variable warnings.</li>
-  <li>st/mesa: Remove unnecessary header.</li>
-  <li>mesa: Remove unnecessary header.</li>
-  <li>egl: Remove unnecessary headers.</li>
-  <li>swrast: Print out format on unexpected failure in _swrast_ReadPixels.</li>
-  <li>st/mesa: Silence uninitialized variable warning.</li>
-  <li>savage: Remove unnecessary header.</li>
-  <li>st/vega: Remove unnecessary headers.</li>
-  <li>dri/nouveau: Silence uninitialized variable warning.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>i965: Silence uninitialized variable warning.</li>
-  <li>i965: Silence uninitialized variable warning.</li>
-  <li>mesa: Clean up header file inclusion in accum.h.</li>
-  <li>mesa: Clean up header file inclusion in version.h.</li>
-  <li>mesa: Clean up header file inclusion in api_loopback.h.</li>
-  <li>mesa: Clean up header file inclusion in api_validate.h.</li>
-  <li>mesa: Include mfeatures.h in api_loopback for FEATURE_beginend.</li>
-  <li>mesa: Include mfeatures.h in api_validate.c for FEATURE_* symbols.</li>
-  <li>mesa: Clean up header file inclusion in arrayobj.h.</li>
-  <li>mesa: Clean up header file inclusion in atifragshader.h.</li>
-  <li>mesa: Clean up header file inclusion in attrib.h.</li>
-  <li>mesa: Clean up header file inclusion in blend.h.</li>
-  <li>mesa: Clean up header file inclusion in buffers.h.</li>
-  <li>mesa: Clean up header file inclusion in colortab.h.</li>
-  <li>mesa: Clean up header file inclusion in convolve.h.</li>
-  <li>mesa: Clean up header file inclusion in debug.h.</li>
-  <li>mesa: Clean up header file inclusion in depth.h.</li>
-  <li>mesa: Clean up header file inclusion in depthstencil.h.</li>
-  <li>mesa: Clean up header file inclusion in drawpix.h.</li>
-  <li>mesa: Clean up header file inclusion in drawtex.h.</li>
-  <li>mesa: Clean up header file inclusion in enable.h.</li>
-  <li>mesa: Clean up header file inclusion in extensions.h.</li>
-  <li>graw: Add struct pipe_surface forward declaration.</li>
-  <li>mesa: Clean up header file inclusion in fbobject.h.</li>
-  <li>mesa: Clean up header file inclusion in ffvertex_prog.h.</li>
-  <li>mesa: Clean up header file inclusion in fog.h.</li>
-  <li>mesa: Clean up header file inclusion in framebuffer.h.</li>
-  <li>mesa: Clean up header file inclusion in hint.h.</li>
-  <li>mesa: Clean up header file inclusion in histogram.h.</li>
-  <li>mesa: Clean up header file inclusion in image.h.</li>
-  <li>mesa: Add missing header and forward declarations in dd.h.</li>
-  <li>mesa: Clean up header file inclusion in light.h.</li>
-  <li>mesa: Clean up header file inclusion in lines.h.</li>
-  <li>mesa: Clean up header file inclusion in matrix.h.</li>
-  <li>mesa: Clean up header file inclusion in multisample.h.</li>
-  <li>mesa: Clean up header file inclusion in nvprogram.h.</li>
-  <li>winsys/xlib: Add cygwin to SConscript.</li>
-  <li>mesa: Clean up header file inclusion in pixel.h.</li>
-  <li>mesa: Clean up header file inclusion in pixelstore.h.</li>
-  <li>mesa: Fix printf format warnings.</li>
-  <li>mesa: Clean up header file inclusion in points.h.</li>
-  <li>i965: Silence uninitialized variable warning.</li>
-  <li>glsl: Add ir_constant_expression.cpp to SConscript.</li>
-  <li>mesa: Add definitions for inverse hyperbolic function on MSVC.</li>
-  <li>glsl: Fix 'control reaches end of non-void function' warning.</li>
-  <li>glsl: Add lower_vector.cpp to SConscript.</li>
-  <li>glsl: Fix type of label 'default' in switch statement.</li>
-  <li>st/mesa: Remove unnecessary headers.</li>
-  <li>swrast: Remove unnecessary header.</li>
-  <li>r600: Remove unnecesary header.</li>
-  <li>intel: Remove unnecessary header.</li>
-  <li>mesa: Clean up header file inclusion in polygon.h.</li>
-  <li>mesa: Clean up header file inclusion in rastpos.h.</li>
-  <li>mesa: Clean up header file inclusion in readpix.h.</li>
-  <li>mesa: Clean up header file inclusion in renderbuffer.h.</li>
-  <li>mesa: Clean up header file inclusion in scissor.h.</li>
-  <li>mesa: Clean up header file inclusion in shaderapi.h.</li>
-  <li>mesa: Clean up header file inclusion in shared.h.</li>
-  <li>mesa: Clean up header file inclusion in stencil.h.</li>
-  <li>r600: Remove unnecessary header.</li>
-  <li>llvmpipe: Remove unnecessary headers.</li>
-  <li>mesa: Clean up header file inclusion in syncobj.h.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>r300/compiler: Move declaration before code.</li>
-  <li>mesa: Clean up header file inclusion in texcompress.h.</li>
-  <li>st/vega: Silence uninitialized variable warning.</li>
-  <li>mesa: Clean up header file inclusion in texcompress_s3tc.h.</li>
-  <li>mesa: Clean up header file inclusion in texenvprogram.h.</li>
-  <li>mesa: Clean up header file inclusion in texformat.h.</li>
-  <li>mesa: Clean up header file inclusion in texgetimage.h.</li>
-  <li>mesa: Clean up header file inclusion in texobj.h.</li>
-  <li>gallium/noop: Add prototype for noop_init_state_functions.</li>
-  <li>mesa: Clean up header file inclusion in texrender.h.</li>
-  <li>mesa: Clean up header file inclusion in transformfeedback.h.</li>
-  <li>mesa: Clean up header file inclusion in varray.h.</li>
-  <li>mesa: Clean up header file inclusion in viewport.h.</li>
-  <li>r200: Silence uninitialized variable warning.</li>
-  <li>r600g: Fix SCons build.</li>
-  <li>i965: Silence uninitialized variable warning.</li>
-</ul></p>
-
-<p>Xavier Chantry (8):
-<ul>
-  <li>nv50: fix size of outputs_written array</li>
-  <li>nv50: apply layout_mask to tile_flags</li>
-  <li>nvfx: only expose one rt on nv30</li>
-  <li>nvfx: fb-&gt;nr_cbufs &lt= 1 on nv30</li>
-  <li>nvfx: reset nvfx-&gt;hw_zeta</li>
-  <li>nvfx: fixes after array textures merge</li>
-  <li>init ps-&gt;context with util_surfaces_get and do_get</li>
-  <li>gallium/trace: check bind_vertex_sampler_states and set_vertex_sampler_views</li>
-</ul></p>
-
-<p>Xiang, Haihao (10):
-<ul>
-  <li>mesa: fix regression from b4bb6680200b5a898583392f4c831c02f41e63f7</li>
-  <li>i965: add support for polygon mode on Sandybridge.</li>
-  <li>i965: fix for flat shading on Sandybridge</li>
-  <li>i965: set minimum/maximum Point Width on Sandybridge</li>
-  <li>meta: allow nested meta operations</li>
-  <li>i965: support for two-sided lighting on Sandybridge</li>
-  <li>i965: fix register region description</li>
-  <li>i965: use align1 access mode for instructions with execSize=1 in VS</li>
-  <li>i965: don't spawn GS thread for LINELOOP on Sandybridge</li>
-  <li>i965: use BLT to clear buffer if possible on Sandybridge</li>
-</ul></p>
-
-<p>Zack Rusin (8):
-<ul>
-  <li>rbug: fix rbug when contexts are being destroyed</li>
-  <li>llvmpipe: fix rasterization of vertical lines on pixel boundaries</li>
-  <li>scons: build the xorg state trackers only when env includes drm</li>
-  <li>gallivm: implement indirect addressing of the output registers</li>
-  <li>gallivm: implement indirect addressing over inputs</li>
-  <li>gallivm: fix storing of the addr register</li>
-  <li>scons: add alias for identity</li>
-  <li>gallium/util: add states relevant to geometry shaders</li>
-</ul></p>
-
-<p>Zhenyu Wang (40):
-<ul>
-  <li>i965: disasm quarter and write enable instruction control on sandybridge</li>
-  <li>i965: new state dump for sandybridge</li>
-  <li>i965: enable accumulator update in PS kernel too on sandybridge</li>
-  <li>i965: Fix color interpolation on sandybridge</li>
-  <li>i965: force zero in clipper to ignore RTAIndex on sandybridge</li>
-  <li>i965: fix point size setting in header on sandybridge</li>
-  <li>i965: ff sync message change for sandybridge</li>
-  <li>i965: ignore quads for GS kernel on sandybridge</li>
-  <li>i965: add sandybridge viewport state bo into validation list</li>
-  <li>i965: VS use SPF mode on sandybridge for now</li>
-  <li>i965: fix jump count on sandybridge</li>
-  <li>i965: Fix sampler on sandybridge</li>
-  <li>i965: fix const register count for sandybridge</li>
-  <li>i965: Add all device ids for sandybridge</li>
-  <li>i965: sandybridge pipe control workaround before write cache flush</li>
-  <li>i965: only allow SIMD8 kernel on sandybridge now</li>
-  <li>i965: don't do calculation for delta_xy on sandybridge</li>
-  <li>i965: fix pixel w interpolation on sandybridge</li>
-  <li>i965: enable polygon offset on sandybridge</li>
-  <li>i965: fix scissor state on sandybridge</li>
-  <li>i965: fix point sprite on sandybridge</li>
-  <li>i965: fix occlusion query on sandybridge</li>
-  <li>i965: fallback bitmap operation on sandybridge</li>
-  <li>i965: Always set tiling for depth buffer on sandybridge</li>
-  <li>i965: fallback lineloop on sandybridge for now</li>
-  <li>Revert "i965: Always set tiling for depth buffer on sandybridge"</li>
-  <li>i965: always set tiling for fbo depth buffer on sandybridge</li>
-  <li>i965: Fix GS hang on Sandybridge</li>
-  <li>Revert "i965: fallback lineloop on sandybridge for now"</li>
-  <li>i965: refresh wm push constant also for BRW_NEW_FRAMENT_PROGRAM on gen6</li>
-  <li>i965: fix dest type of 'endif' on sandybridge</li>
-  <li>Revert "i965: VS use SPF mode on sandybridge for now"</li>
-  <li>i965: also using align1 mode for math2 on sandybridge</li>
-  <li>i965: Fix GS state uploading on Sandybridge</li>
-  <li>i965: upload WM state for _NEW_POLYGON on sandybridge</li>
-  <li>i965: Use MI_FLUSH_DW for blt ring flush on sandybridge</li>
-  <li>i965: explicit tell header present for fb write on sandybridge</li>
-  <li>i965: Fix occlusion query on sandybridge</li>
-  <li>i965: Use last vertex convention for quad provoking vertex on sandybridge</li>
-  <li>i965: Fix provoking vertex select in clip state for sandybridge</li>
-</ul></p>
-
-<p>Zou Nan hai (1):
-<ul>
-  <li>i965: skip too small size mipmap</li>
-</ul></p>
-
-<p>delphi (2):
-<ul>
-  <li>draw: added userclip planes and updated variant_key</li>
-  <li>draw: some changes to allow for runtime changes to userclip planes</li>
-</ul></p>
-
-<p>nobled (3):
-<ul>
-  <li>r300g: Abort if atom allocations fail</li>
-  <li>r300g: Abort if draw_create() fails</li>
-  <li>r300g: Drop unnecessary cast</li>
-</ul></p>
-
-<p>pontus lidman (1):
-<ul>
-  <li>mesa: check for posix_memalign() errors</li>
-</ul></p>
-
-<p>richard (2):
-<ul>
-  <li>evergreen : fix z format setting, enable stencil.</li>
-  <li>r600c : inline vertex format is not updated in an app, switch to use vfetch constants. For the 7.9 and 7.10 branches as well.</li>
-</ul></p>
-
-</body>
-</html>
+<HTML>
+
+<head>
+<TITLE>Mesa Release Notes</TITLE>
+<link rel="stylesheet" type="text/css" href="mesa.css">
+<meta http-equiv="content-type" content="text/html; charset=utf-8" />
+</head>
+
+<BODY>
+
+<body bgcolor="#eeeeee">
+
+<H1>Mesa 7.10 Release Notes / January 7, 2011</H1>
+
+<p>
+Mesa 7.10 is a new development release.
+People who are concerned with stability and reliability should stick
+with a previous release or wait for Mesa 7.10.1.
+</p>
+<p>
+Mesa 7.10 implements the OpenGL 2.1 API, but the version reported by
+glGetString(GL_VERSION) depends on the particular driver being used.
+Some drivers don't support all the features required in OpenGL 2.1.
+</p>
+<p>
+See the <a href="install.html">Compiling/Installing page</a> for prerequisites
+for DRI hardware acceleration.
+</p>
+
+
+<h2>MD5 checksums</h2>
+<pre>
+0a70c15c135561824bdcae92bf232e43  MesaLib-7.10.tar.gz
+33fb94eccc02cbb4d8d1365615e38e46  MesaLib-7.10.tar.bz2
+5cafdc0eda0f9bf370b95c98df3338fa  MesaLib-7.10.zip
+bc644be551ed585fc4f66c16b64a91c9  MesaGLUT-7.10.tar.gz
+5c2677a155672352d62b177e4f0f92e8  MesaGLUT-7.10.tar.bz2
+2ce5001f74496d1ba719ef74d910a5cf  MesaGLUT-7.10.zip
+</pre>
+
+
+<h2>New features</h2>
+<ul>
+<li>GL_ARB_explicit_attrib_location extension (Intel and software drivers).
+<li>GL_ARB_texture_rg (Intel, software drivers, gallium drivers).
+<li>GL_EXT_separate_shader_objects extension (Intel and software drivers).
+<li>GL_NV_primitive_restart extension (Gallium softpipe, llvmpipe).
+<li>New fragment shader back-end for i965-class hardware.
+<li>Support for Sandybridge chipset in i965 DRI driver.
+</ul>
+
+
+<h2>Bug fixes</h2>
+<p>This list is likely incomplete.</p>
+<ul>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28800">Bug 28800</a> - [r300c, r300g] Texture corruption with World of Warcraft</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29420">Bug 29420</a> - Amnesia / HPL2 RendererFeatTest - not rendering correctly</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29946">Bug 29946</a> - [swrast] piglit valgrind glsl-array-bounds-04 fails</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30261">Bug 30261</a> - [GLSL 1.20] allowing inconsistent invariant declaration between two vertex shaders</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30632">Bug 30632</a> - [softpipe] state_tracker/st_manager.c:489: st_context_notify_invalid_framebuffer: Assertion `stfb &amp;&amp; stfb-&gt;iface == stfbi' failed.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30694">Bug 30694</a> - wincopy will crash on Gallium drivers when going to front buffer</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30771">Bug 30771</a> -  [r600g] vert-tex glsl demo</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30787">Bug 30787</a> - Invalid asm shader does not generate draw-time error when used with GLSL shader</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30974">Bug 30974</a> - [llvmpipe] SIGABRT src/gallium/drivers/llvmpipe/lp_state_fs.c:779</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30993">Bug 30993</a> - getFramebufferAttachmentParameteriv wrongly generates error</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31101">Bug 31101</a> -  [glsl2] abort() in ir_validate::visit_enter(ir_assignment *ir)</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31193">Bug 31193</a> -  [regression] aa43176e break water reflections</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31194">Bug 31194</a> - The mesa meta save/restore code doesn't ref the current GLSL program</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31371">Bug 31371</a> - glslparsertest: ir.cpp:358: ir_constant::ir_constant(const glsl_type*, const ir_constant_data*): Assertion `(type->base_type &gt;= 0) &amp;&amp; (type->base_type &lt;= 3)' failed.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31439">Bug 31439</a> - Crash in glBufferSubData() with size == 0</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31495">Bug 31495</a> - [i965 gles2c bisected] OpenGL ES 2.0 conformance GL2Tests_GetBIFD_input.run regressed</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31514">Bug 31514</a> - isBuffer returns true for unbound buffers</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31560">Bug 31560</a> - [tdfx] tdfx_tex.c:702: error: ‘const struct gl_color_table’ has no member named ‘Format’</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31617">Bug 31617</a> - Radeon/Compiz: 'failed to attach dri2 front buffer', error case not handled</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31648">Bug 31648</a> -  [GLSL] array-struct-array gets assertion: `(size &gt;= 1) && (size &lt;= 4)' failed.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31650">Bug 31650</a> - [GLSL] varying gl_TexCoord fails to be re-declared to different size in the second shader</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31673">Bug 31673</a> - GL_FRAGMENT_PRECISION_HIGH preprocessor macro undefined in GLSL ES</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31690">Bug 31690</a> -  i915 shader compiler fails to flatten if in Aquarium webgl demo.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31832">Bug 31832</a> - [i915] Bad renderbuffer format: 21</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31841">Bug 31841</a> - [drm:radeon_cs_ioctl] *ERROR* Invalid command stream !</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31894">Bug 31894</a> - Writing to gl_PointSize with GLES2 corrupts other varyings</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31909">Bug 31909</a> - [i965] brw_fs.cpp:1461: void fs_visitor::emit_bool_to_cond_code(ir_rvalue*): Assertion `expr-&gt;operands[i]-&gt;type-&gt;is_scalar()' failed.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31934">Bug 31934</a> - [gallium] Mapping empty buffer object causes SIGSEGV</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31983">Bug 31983</a> -  [i915 gles2] "if (expression with builtin/varying variables) discard" breaks linkage</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31985">Bug 31985</a> - [GLSL 1.20] initialized uniform array considered as "unsized"</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31987">Bug 31987</a> - [gles2] if input a wrong pname(GL_NONE) to glGetBoolean, it will not case GL_INVALID_ENUM</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32035">Bug 32035</a> - [GLSL bisected] comparing unsized array gets segfault</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32070">Bug 32070</a> - llvmpipe renders stencil demo incorrectly</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32273">Bug 32273</a> - assertion fails when starting vdrift 2010 release with shaders enabled</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32287">Bug 32287</a> - [bisected GLSL] float-int failure</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32311">Bug 32311</a> - [965 bisected] Array look-ups broken on GM45</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32520">Bug 32520</a> -  [gles2] glBlendFunc(GL_ZERO, GL_DST_COLOR) will result in GL_INVALID_ENUM</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32825">Bug 32825</a> - egl_glx driver completely broken in 7.9 branch [fix in master]</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=">Bug </a> - </li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=">Bug </a> - </li>
+
+</ul>
+
+
+<h2>Changes</h2>
+<p>Adam Jackson (2):
+<ul>
+  <li>i965: Update renderer strings for sandybridge</li>
+  <li>drivers/x11: unifdef XFree86Server</li>
+</ul></p>
+
+<p>Alex Deucher (30):
+<ul>
+  <li>r600c: fix mipmap stride on evergreen</li>
+  <li>r600c: add reloc for CB_COLOR0_ATTRIB</li>
+  <li>r600c: pull over 6xx/7xx vertex fixes for evergreen</li>
+  <li>r600c: fix segfault in evergreen stencil code</li>
+  <li>r100: revalidate after radeon_update_renderbuffers</li>
+  <li>r600c: add missing radeon_prepare_render() call on evergreen</li>
+  <li>r600c: properly align mipmaps to group size</li>
+  <li>egl_dri2: Add radeon chip ids</li>
+  <li>r600c/evergreen: texture align is group_bytes just like 6xx/7xx</li>
+  <li>r600g: fix buffer alignment</li>
+  <li>r600g: All EVENT_WRITE packets need the EVENT_INDEX field</li>
+  <li>r600g: translate ARR instruction for evergreen</li>
+  <li>r600g: use meaningful defines for chiprev</li>
+  <li>r600g: use full range of VS resources for vertex samplers</li>
+  <li>r600g: fix additional EVENT_WRITE packet</li>
+  <li>r600g: fix some winsys functions to deal properly with evergreen</li>
+  <li>r600c: add Ontario Fusion APU support</li>
+  <li>r600g: add support for ontario APUs</li>
+  <li>r600c: fix VC flush on cedar and palm</li>
+  <li>gallium/egl: fix r300 vs r600 loading</li>
+  <li>r600c: fix some opcodes on evergreen</li>
+  <li>r600c: bump texture limits to hw limits</li>
+  <li>r600g: bump texture/cb limits appropriately for evergreen</li>
+  <li>radeon: bump mip tree levels to 15</li>
+  <li>r600g: fix rendering with a vertex attrib having a zero stride</li>
+  <li>r600g: remove useless switch statements</li>
+  <li>r600g: add support for NI (northern islands) asics</li>
+  <li>r600c: add support for NI asics</li>
+  <li>r600g: support up to 64 shader constants</li>
+  <li>r600c: fix up SQ setup in blit code for Ontario/NI</li>
+</ul></p>
+
+<p>Andre Maasikas (3):
+<ul>
+  <li>r600c: fix buffer height setting in dri2 case</li>
+  <li>r600g: break alu clause earlier</li>
+  <li>r600g: fix evergreen interpolation setup</li>
+</ul></p>
+
+<p>Andrew Randrianasulu (2):
+<ul>
+  <li>dri/nv04: Don't expose ARB_texture_env_combine/dot3.</li>
+  <li>dri/nv04: Enable eng3dm for A8/L8 textures.</li>
+</ul></p>
+
+<p>Aras Pranckevicius (2):
+<ul>
+  <li>glsl: fix crash in loop analysis when some controls can't be determined</li>
+  <li>glsl: fix matrix type check in ir_algebraic</li>
+</ul></p>
+
+<p>Bas Nieuwenhuizen (3):
+<ul>
+  <li>r600g: set ENABLE_KILL in the shader state in the new design</li>
+  <li>r600g: set ENABLE_KILL on evergreen too</li>
+  <li>r600g: use dirty list to track dirty blocks</li>
+</ul></p>
+
+<p>Ben Skeggs (3):
+<ul>
+  <li>nv50: DST</li>
+  <li>nv50: DPH</li>
+  <li>nv50: silence some unknown get_param warnings</li>
+</ul></p>
+
+<p>Benjamin Franzke (2):
+<ul>
+  <li>st/egl image: multiply drm buf-stride with blocksize</li>
+  <li>r600g: implement texture_get_handle (needed for eglExportDRMImageMESA)</li>
+</ul></p>
+
+<p>Brian Paul (296):
+<ul>
+  <li>glx: add const qualifiers to __indirect_glMultiDrawArraysEXT()</li>
+  <li>glsl2: fix signed/unsigned comparison warning</li>
+  <li>llvmpipe: cast to silence warning</li>
+  <li>llvmpipe: s/boolean/unsigned/ in bitfield to silence warning</li>
+  <li>nv50: use unsigned int for bitfields to silence warnings</li>
+  <li>tgsi: fix incorrect usage_mask for shadow tex instructions</li>
+  <li>gallivm: expand AoS sampling to cover all filtering modes</li>
+  <li>gallivm: fix incorrect vector shuffle datatype</li>
+  <li>gallivm: move i32_vec_type inside the #ifdef</li>
+  <li>mesa: include mfeatures.h in formats.c</li>
+  <li>gallivm: fix wrong return value in bitwise functions</li>
+  <li>tgsi/sse: fix aos_to_soa() loop to handle num_inputs==0</li>
+  <li>gallivm: added missing case for PIPE_TEXTURE_RECT</li>
+  <li>gallium: better docs for pipe_rasterizer_state::sprite_coord_enable</li>
+  <li>gallium: rework handling of sprite_coord_enable state</li>
+  <li>gallium/docs: added new pipeline.txt diagram</li>
+  <li>mesa: don't call valid_texture_object() in non-debug builds</li>
+  <li>glsl2: silence compiler warnings in printf() calls</li>
+  <li>docs: remove old broken link</li>
+  <li>docs: mark as obsolete, remove dead links</li>
+  <li>llvmpipe: fix query bug when no there's no scene</li>
+  <li>gallivm: remove debug code</li>
+  <li>llvmpipe: maintain fragment shader state for draw module</li>
+  <li>llvmpipe: indentation fix</li>
+  <li>llvmpipe: reformatting, remove trailing whitespace, etc</li>
+  <li>llvmpipe: clean-up, comments in setup_point_coefficient()</li>
+  <li>llvmpipe: rename sprite field, add sprite_coord_origin</li>
+  <li>llvmpipe: implement sprite coord origin modes</li>
+  <li>draw: fix test for using the wide-point stage</li>
+  <li>llvmpipe: check bitshift against PIPE_MAX_SHADER_OUTPUTS</li>
+  <li>draw: check bitshift against PIPE_MAX_SHADER_OUTPUS</li>
+  <li>Merge branch 'sprite-coord'</li>
+  <li>draw: new draw_fs.[ch] files</li>
+  <li>glsl2: fix typo in error msg</li>
+  <li>gallivm: fix lp_build_sample_compare()</li>
+  <li>softpipe: add missing calls to set draw vertex samplers/views</li>
+  <li>mesa: don't advertise bogus GL_ARB_shading_language_120 extension</li>
+  <li>configs: remove egl-swrast target from linux-dri config</li>
+  <li>llvmpipe: fix sprite texcoord setup for non-projective texturing</li>
+  <li>mesa: fix assertions to handle srgb formats</li>
+  <li>st/mesa: add missing MESA_FORMAT_S8 case in st_mesa_format_to_pipe_format()</li>
+  <li>st/mesa: use the wrapped renderbuffer in CopyPixels()</li>
+  <li>llvmpipe: make min/max lod and lod bias dynamic state</li>
+  <li>llvmpipe: make texture border_color dynamic state</li>
+  <li>softpipe: fix repeat() function for NPOT textures</li>
+  <li>gallivm: fix repeat() function for NPOT textures</li>
+  <li>swrast: update comments for REMAINDER() macro</li>
+  <li>softpipe: rename sp_state_fs.c -&gt; sp_state_shader.c</li>
+  <li>softpipe: make shader-related functions static</li>
+  <li>softpipe: make blend/stencil/depth functions static</li>
+  <li>softpipe: make sampler state functions static</li>
+  <li>softpipe: make vertex state functions static</li>
+  <li>softpipe: make rasterizer state functions static</li>
+  <li>softpipe: make stream out state functions static</li>
+  <li>softpipe: make clip state functions static</li>
+  <li>softpipe: minor asst. clean-ups</li>
+  <li>softpipe: allocate tile data on demand</li>
+  <li>llvmpipe: fix swizzling of texture border color</li>
+  <li>softpipe: fix swizzling of texture border color</li>
+  <li>draw: pass sampler state down to llvm jit state</li>
+  <li>gallivm: check for level=0 case in lp_build_minify()</li>
+  <li>gallivm: added some comments</li>
+  <li>draw: check for null sampler pointers</li>
+  <li>swrast: fix choose_depth_texture_level() to respect mipmap filtering state</li>
+  <li>st/mesa: replace assertion w/ conditional in framebuffer invalidation</li>
+  <li>glsl2: fix signed/unsigned comparison warning</li>
+  <li>st/xlib: add some comments</li>
+  <li>ir_to_mesa: assorted clean-ups, const qualifiers, new comments</li>
+  <li>mesa: remove assertion w/ undeclared variable texelBytes</li>
+  <li>gallivm: remove newlines</li>
+  <li>draw/llvmpipe: replace DRAW_MAX_TEXTURE_LEVELS with PIPE_MAX_TEXTURE_LEVELS</li>
+  <li>mesa: reformatting, comments, code movement</li>
+  <li>x11: fix breakage from gl_config::visualType removal</li>
+  <li>gallivm: work-around trilinear mipmap filtering regression with LLVM 2.8</li>
+  <li>mesa: remove post-convolution width/height vars</li>
+  <li>gallivm: add compile-time option to emit inst addrs and/or line numbers</li>
+  <li>llvmpipe: code to dump bytecode to file (disabled)</li>
+  <li>gallivm: added lp_build_print_vec4()</li>
+  <li>gallivm: added lp_build_load_volatile()</li>
+  <li>glsl: add ir_unop_round_even case to silence unhandled enum warning</li>
+  <li>st/mesa: fix regressions in glDrawPixels(GL_STENCIL_INDEX)</li>
+  <li>st/mesa: reformatting in st_cb_drawpixels.c</li>
+  <li>st/mesa: use GLuint to avoid problem w/ uint not defined on mingw32</li>
+  <li>st/mesa: update function name, comments</li>
+  <li>gallivm: use util_snprintf()</li>
+  <li>llvmpipe: remove lp_setup_coef*.c files from Makefile</li>
+  <li>mesa: fix mesa version string construction</li>
+  <li>gallivm: fix incorrect type for zero vector in emit_kilp()</li>
+  <li>llvmpipe/draw: always enable LLVMAddInstructionCombiningPass()</li>
+  <li>draw: use float version of LLVM Mul/Add instructions</li>
+  <li>draw: fix typo in comment</li>
+  <li>mesa: add GL_RG case to _mesa_source_buffer_exists()</li>
+  <li>mesa: add missing cases for packing red/green images</li>
+  <li>st/mesa: added cases for GL_COMPRESSED_RED/RG in st_choose_format()</li>
+  <li>docs: update texture red/green support in GL3.txt</li>
+  <li>docs: add GL_ARB_texture_rg to release notes</li>
+  <li>mesa: driver hook for primitive restart</li>
+  <li>mesa: set/get primitive restart state</li>
+  <li>mesa: API spec for primitive restart</li>
+  <li>mesa: regenerated files with primitive restart</li>
+  <li>mesa: plug in primitive restart function</li>
+  <li>vbo: support for primitive restart</li>
+  <li>gallium: new CAP, state for primitive restart</li>
+  <li>st/mesa: support for primitive restart</li>
+  <li>draw: implement primitive splitting for primitive restart</li>
+  <li>softpipe: enable primitive restart</li>
+  <li>llvmpipe: enable primitive restart</li>
+  <li>docs: added GL_NV_primitive_restart extension</li>
+  <li>Merge branch 'primitive-restart-cleanup'</li>
+  <li>winsys/xlib: formatting fixes</li>
+  <li>winsys/xlib: use Bool type for shm field</li>
+  <li>winsys/xlib: fix up allocation/dealloction of XImage</li>
+  <li>winsys/xlib: rename xm-&gt;xlib</li>
+  <li>galahad: silence warnings</li>
+  <li>mesa: move declaration before code</li>
+  <li>docs: updated GL3 status for primitive restart</li>
+  <li>mesa: 80-column wrapping</li>
+  <li>mesa: simplify fbo format checking code</li>
+  <li>mesa: split up the image.c file</li>
+  <li>mesa: add pixel packing for unscaled integer types</li>
+  <li>mesa: _mesa_ClearColorIuiEXT() and _mesa_ClearColorIiEXT()</li>
+  <li>mesa: _mesa_is_format_integer() function</li>
+  <li>mesa: minor reformatting, clean-ups</li>
+  <li>mesa: added _mesa_is_fragment_shader_active() helper</li>
+  <li>mesa: new glDrawPixels error check for integer formats</li>
+  <li>softpipe: added some texture sample debug code (disabled)</li>
+  <li>mesa: added new gl_extensions::EXT_gpu_shader4 field</li>
+  <li>mesa: added new gl_framebuffer::_IntegerColor field</li>
+  <li>mesa: added glGet query for GL_RGBA_INTEGER_MODE_EXT</li>
+  <li>mesa: compute _IntegerColor field in _mesa_test_framebuffer_completeness()</li>
+  <li>mesa: added cases for GL_EXT_texture_integer formats</li>
+  <li>mesa: added cases for GL_EXT_texture_integer</li>
+  <li>st/mesa: add format selection for signed/unsigned integer formats</li>
+  <li>mesa: simplify target_can_be_compressed() function</li>
+  <li>glapi: GL_EXT_texture_integer API</li>
+  <li>glapi: include/build EXT_texture_integer.xml</li>
+  <li>mesa: regenerated API files for GL_EXT_texture_integer</li>
+  <li>mesa: plug in GL_EXT_texture_integer functions</li>
+  <li>mesa: display list support for GL_EXT_texture_integer</li>
+  <li>st/mesa: be smarter choosing texture format for glDrawPixels()</li>
+  <li>softpipe: remove &gt;32bpp color restriction</li>
+  <li>mesa: silence enum comparison warning</li>
+  <li>mesa: fix uninitialized var warning</li>
+  <li>xlib: silence unused var warning</li>
+  <li>util: use pointer_to_func() to silence warning</li>
+  <li>rtasm: use pointer_to_func() to silence warning</li>
+  <li>translate: use function typedefs, casts to silence warnings</li>
+  <li>translate: remove unused prototypes</li>
+  <li>mesa: additional glReadPixels error checks for GL_EXT_texture_integer</li>
+  <li>mesa: additional switch cases for GL_EXT_texture_integer</li>
+  <li>mesa: additional teximage error checks for GL_EXT_texture_integer</li>
+  <li>mesa: do integer FB / shader validation check in _mesa_valid_to_render()</li>
+  <li>mesa: call _mesa_valid_to_render() in glDrawPixels, glCopyPixels, glBitmap</li>
+  <li>mesa: remove the unused _mesa_is_fragment_shader_active() function</li>
+  <li>mesa: fix bug in _mesa_is_format_integer()</li>
+  <li>mesa: rename function to _mesa_is_format_integer_color()</li>
+  <li>mesa: remove 'normalized' parameter from _mesa_VertexAttribIPointer()</li>
+  <li>vbo: re-indent file</li>
+  <li>glapi: xml spec file for GL_EXT_gpu_shader4</li>
+  <li>glapi: include EXT_gpu_shader4.xml</li>
+  <li>glapi: regenerated API files</li>
+  <li>mesa: plug in stubs for glBindFragDataLocation(), glGetFragDataLocation()</li>
+  <li>mesa: add glGetUniformuiv(), plug in uint glUniform funcs</li>
+  <li>mesa: plug in more GL_EXT_gpu_shader4 functions</li>
+  <li>mesa: add new GLvertexformat entries for integer-valued attributes</li>
+  <li>mesa: implement integer-valued vertex attribute functions</li>
+  <li>mesa: add gl_client_array::Integer field and related vertex array state code</li>
+  <li>mesa: consolidate glVertex/Color/etcPointer() code</li>
+  <li>mesa: state/queries for GL_MIN/MAX_PROGRAM_TEXEL_OFFSET_EXT</li>
+  <li>mesa: glArrayElement support for integer-valued arrays</li>
+  <li>mesa: clean-up array element code</li>
+  <li>mesa: add extension table entry for GL_EXT_gpu_shader4</li>
+  <li>mesa: remove obsolete comment</li>
+  <li>mesa: fix incorrect type in _mesa_texstore_rgba_int16()</li>
+  <li>mesa: fix integer cases in _mesa_is_legal_format_and_type()</li>
+  <li>mesa: add const qualifier to _mesa_is_legal_format_and_type()</li>
+  <li>mesa: additional integer formats in _mesa_bytes_per_pixel()</li>
+  <li>mesa: pixel transfer ops do not apply to integer-valued textures</li>
+  <li>mesa: remove dead code</li>
+  <li>osmesa: fix renderbuffer memleak in OSMesaMakeCurrent()</li>
+  <li>mesa: use GLubyte for edge flag arrays</li>
+  <li>mesa: move the gl_config struct declaration</li>
+  <li>dri/util: add a bunch of comments</li>
+  <li>mesa: remove always-false conditional in check_compatible()</li>
+  <li>mesa: fix aux/accum comment and error message mixups</li>
+  <li>llvmpipe: assign context's frag shader pointer before using it</li>
+  <li>llvmpipe: add a cast</li>
+  <li>mesa: silence new warnings in texobj.c</li>
+  <li>egl/gdi: fix typo: xsurf-&gt;gsurf</li>
+  <li>mesa: code to unpack RGBA as uints</li>
+  <li>gallivm: implement scatter stores into temp register file</li>
+  <li>gallivm: add some LLVM var labels</li>
+  <li>gallivm: added debug code to dump temp registers</li>
+  <li>gallivm: add pixel offsets in scatter stores</li>
+  <li>gallivm: added lp_elem_type()</li>
+  <li>gallivm: implement execution mask for scatter stores</li>
+  <li>tgsi: remove unused function</li>
+  <li>llvmpipe: added some debug assertions, but disabled</li>
+  <li>gallivm: alloca() was called too often for temporary arrays</li>
+  <li>gallivm: add const qualifiers, fix comment string</li>
+  <li>softpipe: disable vertex texturing with draw/llvm</li>
+  <li>mesa: consolidate pixel packing/unpacking code</li>
+  <li>mesa: rename vars in pixel pack/unpack code</li>
+  <li>mesa: implement uint texstore code</li>
+  <li>mesa: remove stray GL_FLOAT case in _mesa_is_legal_format_and_type()</li>
+  <li>mesa: make fixed-pt and byte-valued arrays a runtime feature</li>
+  <li>softpipe: can't no-op depth test stage when occlusion query is enabled</li>
+  <li>mesa: no-op glBufferSubData() on size==0</li>
+  <li>mesa: #include mfeatures.h in enums.h</li>
+  <li>mesa: improve error message</li>
+  <li>mesa: add missing formats in _mesa_format_to_type_and_comps()</li>
+  <li>mesa: handle more pixel types in mipmap generation code</li>
+  <li>mesa: make glIsBuffer() return false for never bound buffers</li>
+  <li>mesa: fix glDeleteBuffers() regression</li>
+  <li>tdfx: s/Format/_BaseFormat/</li>
+  <li>mesa: consolidate assertions in teximage code</li>
+  <li>radeon: set gl_texture_image::TexFormat field in radeonSetTexBuffer2()</li>
+  <li>r600: set gl_texture_image::TexFormat field in r600SetTexBuffer2()</li>
+  <li>r200: set gl_texture_image::TexFormat field in r200SetTexBuffer2()</li>
+  <li>r300: set gl_texture_image::TexFormat field in r300SetTexBuffer2()</li>
+  <li>evergreen: set gl_texture_image::TexFormat field in evergreenSetTexBuffer()</li>
+  <li>st/mesa: fix glDrawPixels(depth/stencil) bugs</li>
+  <li>glsl: fix assorted MSVC warnings</li>
+  <li>mesa: add more work-arounds for acoshf(), asinhf(), atahf()</li>
+  <li>glsl: remove opt_constant_expression.cpp from SConscript</li>
+  <li>mesa: fix error messages and minor reindenting</li>
+  <li>mesa: whitespace cleanups</li>
+  <li>mesa: 80-column wrapping</li>
+  <li>mesa: reorder texture_error_check() params</li>
+  <li>mesa: minor clean-ups in context code</li>
+  <li>mesa: upgrade to glext.h version 66</li>
+  <li>mesa: pass gl_format to _mesa_init_teximage_fields()</li>
+  <li>mesa: fix error msg typo</li>
+  <li>glapi: rename GL3.xml to GL3x.xml as it covers all GL 3.x versions</li>
+  <li>mesa: hook up GL 3.x entrypoints</li>
+  <li>docs: update some GL 3.0 status</li>
+  <li>mesa: fix get_texture_dimensions() for texture array targets</li>
+  <li>swrast: init alpha value to 1.0 in opt_sample_rgb_2d()</li>
+  <li>glsl: fix off by one in register index assertion</li>
+  <li>glsl: use gl_register_file in a few places</li>
+  <li>mesa: rename, make _mesa_register_file_name() non-static</li>
+  <li>mesa: _mesa_valid_register_index() to validate register indexes</li>
+  <li>mesa: replace #defines with new gl_shader_type enum</li>
+  <li>mesa: use gl_shader_type enum</li>
+  <li>glsl: better handling of linker failures</li>
+  <li>glsl: start restoring some geometry shader code</li>
+  <li>mesa: add assertion and update comment in _mesa_format_image_size()</li>
+  <li>mesa: added _mesa_format_image_size64()</li>
+  <li>x11: remove test_proxy_teximage() function</li>
+  <li>st/mesa: fix mapping of zero-sized buffer objects</li>
+  <li>gallivm/llvmpipe: squash merge of the llvm-context branch</li>
+  <li>mesa: raise max texture sizes to 16K</li>
+  <li>softpipe: increase max texture size to 16K</li>
+  <li>mesa: replace large/MAX_WIDTH stack allocations with heap allocations</li>
+  <li>mesa: replace large/MAX_WIDTH stack allocations with heap allocations</li>
+  <li>swrast: avoid large stack allocations in blend code</li>
+  <li>swrast: avoid large stack allocations in tex combine code</li>
+  <li>st/mesa: avoid large stack allocations in readpixels code</li>
+  <li>mesa: replace more MAX_WIDTH stack allocations with heap allocations</li>
+  <li>gallivm/llvmpipe: remove lp_build_context::builder</li>
+  <li>gallivm: fix null builder pointers</li>
+  <li>mesa: fix GL_FRAMEBUFFER_ATTACHMENT_OBJECT_NAME query</li>
+  <li>mesa: return GL_FRAMEBUFFER_DEFAULT as FBO attachment type</li>
+  <li>llvmpipe: fix broken stencil writemask</li>
+  <li>mesa: consolidate some compiler -D flags</li>
+  <li>swrast: allow GL_RG format in glDrawPixels()</li>
+  <li>swrast: fix indentation</li>
+  <li>swrast: accept GL_RG in glReadPixels()</li>
+  <li>swrast: restructure some glReadPixels() code</li>
+  <li>mesa: make glGet*(GL_NONE) generate GL_INVALID_ENUM</li>
+  <li>mesa: remove unneeded cast</li>
+  <li>mesa: update comments, remove dead code</li>
+  <li>st/mesa: new comment about updating state vars</li>
+  <li>mesa: add error margin to clip mask debug/check code</li>
+  <li>gallium/util: minor formatting fixes</li>
+  <li>mesa/llvm: use llvm-config --cppflags</li>
+  <li>st/mesa: fix mipmap generation bug</li>
+  <li>mesa: test for cube map completeness in glGenerateMipmap()</li>
+  <li>mesa: set gl_texture_object::_Complete=FALSE in incomplete()</li>
+  <li>mesa: consolidate glTexImage1/2/3D() code</li>
+  <li>mesa: simplify proxy texture code in texture_error_check()</li>
+  <li>mesa: consolidate the glTexSubImage1/2/3D() functions</li>
+  <li>mesa: consolidate glCopyTexImage1/2D() code</li>
+  <li>mesa: consolidate glCopyTexSubImage1/2/3D() functions</li>
+  <li>mesa: consolidate glCompressedTexImage1/2/3D() functions</li>
+  <li>mesa: make _mesa_test_proxy_teximage() easier to read</li>
+  <li>configure: use llvm-config --cppflags instead of --cflags</li>
+  <li>mesa: revamp error checking for compressed texture images</li>
+  <li>mesa: simplify target checking for TexImage functions</li>
+  <li>draw/llvm: don't flush in vs_llvm_delete()</li>
+  <li>tnl: Initialize gl_program_machine memory in run_vp.</li>
+  <li>tnl: a better way to initialize the gl_program_machine memory</li>
+  <li>mesa, st/mesa: disable GL_ARB_geometry_shader4</li>
+  <li>mesa/meta: fix broken assertion, rename stack depth var</li>
+  <li>glsl: new glsl_strtod() wrapper to fix decimal point interpretation</li>
+  <li>st/mesa: fix renderbuffer pointer check in st_Clear()</li>
+</ul></p>
+
+<p>Brian Rogers (1):
+<ul>
+  <li>mesa: Add missing else in do_row_3D</li>
+</ul></p>
+
+<p>Chad Versace (25):
+<ul>
+  <li>intel_extensions: Add ability to set GLSL version via environment</li>
+  <li>glsl: Add glsl_type::uvecN_type for N=2,3</li>
+  <li>glsl: Add lexer rules for uint and uvecN (N=2..4)</li>
+  <li>glsl: Changes in generated file glsl_lexer.cpp</li>
+  <li>glsl: Add lexer rules for &lt&lt and &gt;&gt; in GLSL 1.30</li>
+  <li>glsl: Change generated file glsl_lexer.cpp</li>
+  <li>glsl: Implement ast-to-hir for binary shifts in GLSL 1.30</li>
+  <li>glsl: Implement constant expr evaluation for bitwise-not</li>
+  <li>glsl: Implement constant expr evaluation for bit-shift ops</li>
+  <li>glsl: Implement constant expr evaluation for bitwise logic ops</li>
+  <li>glsl: Fix ir validation for bit logic ops</li>
+  <li>glsl: Define shift_result_type() in ast_to_hir.cpp</li>
+  <li>glsl: Implement ast-to-hir for bit-shift-assignment</li>
+  <li>glsl: Define bit_logic_result_type() in ast_to_hir.cpp</li>
+  <li>glsl: Implement ast-to-hir for bit-logic ops</li>
+  <li>glsl: Fix lexer rule for ^=</li>
+  <li>glsl: Commit generated file glsl_lexer.cpp</li>
+  <li>glsl: Fix ast-to-hir for ARB_fragment_coord_conventions</li>
+  <li>mesa: Fix C++ includes in sampler.cpp</li>
+  <li>glsl: Fix ir_expression::constant_expression_value()</li>
+  <li>glsl: Fix erroneous cast in ast_jump_statement::hir()</li>
+  <li>glsl: Fix Doxygen tag file in recently renamed files</li>
+  <li>glsl: Improve usage message for glsl_compiler</li>
+  <li>glsl: Fix linker bug in cross_validate_globals()</li>
+  <li>glsl: In ast_to_hir, check sampler array indexing</li>
+</ul></p>
+
+<p>Chia-I Wu (149):
+<ul>
+  <li>glapi: Fix build errors for ES.</li>
+  <li>glapi: Fix ES build errors again.</li>
+  <li>mesa: Update ES APIspec.xml.</li>
+  <li>st/xlib: Notify the context when the front/back buffers are swapped.</li>
+  <li>targets/egl: Use C++ compiler to link GL/ES state trackers.</li>
+  <li>libgl-xlib: Remove unused st_api_create_OpenGL.</li>
+  <li>st/egl: Split modeset code support to modeset.c.</li>
+  <li>st/egl: Rename kms backend to drm.</li>
+  <li>st/egl: s/kms/drm/ on the drm backend.</li>
+  <li>egl: Enable drm platform by default.</li>
+  <li>egl: Check extensions.</li>
+  <li>st/egl: Skip single-buffered configs in EGL.</li>
+  <li>mapi: Fix compiler warnings.</li>
+  <li>st/egl: Drop context argument from egl_g3d_get_egl_image.</li>
+  <li>targets/egl: Fix linking with libdrm.</li>
+  <li>st/vega: Fix version check in context creation.</li>
+  <li>egl: Use attribute names as the _EGLConfig member names.</li>
+  <li>egl: Access config attributes directly.</li>
+  <li>st/egl: Access _EGLConfig directly.</li>
+  <li>st/egl: Do not finish a fence that is NULL.</li>
+  <li>mesa: Remove unused vtxfmt_tmp.h.</li>
+  <li>egl_dri2: Drop the use of _egl[SG]etConfigKey.</li>
+  <li>egl_glx: Drop the use of [SG]ET_CONFIG_ATTRIB.</li>
+  <li>egl_glx: Fix borken driver.</li>
+  <li>egl: Move attributes in _EGLImage to _EGLImageAttribs.</li>
+  <li>egl: Parse image attributes with _eglParseImageAttribList.</li>
+  <li>egl: Move fallback routines to eglfallbacks.c.</li>
+  <li>egl: Drop dpy argument from the link functions.</li>
+  <li>egl: Minor changes to the _EGLConfig interface.</li>
+  <li>egl: Minor changes to the _EGLScreen interface.</li>
+  <li>egl: Fix _eglModeLookup.</li>
+  <li>st/egl: Fix native_mode refresh mode.</li>
+  <li>egl: Add reference count for resources.</li>
+  <li>egl: Use reference counting to replace IsLinked or IsBound.</li>
+  <li>egl: Fix a false negative check in _eglCheckMakeCurrent.</li>
+  <li>st/egl: Use resource reference count for egl_g3d_sync.</li>
+  <li>egl_dri2: Fix a typo that make glFlush be called at wrong time.</li>
+  <li>glapi: Do not use glapidispatch.h.</li>
+  <li>glapi: Move glapidispatch.h to core mesa.</li>
+  <li>glapi: Do not use glapioffsets.h.</li>
+  <li>glapi: Merge glapioffsets.h into glapidispath.h.</li>
+  <li>vbo: Use CALL_* macros.</li>
+  <li>mesa: Remove unnecessary glapitable.h includes.</li>
+  <li>autoconf: Better client API selection.</li>
+  <li>docs: Update egl and openvg docs.</li>
+  <li>autoconf: Update configuration info.</li>
+  <li>Merge branch 'glapi-reorg'</li>
+  <li>targets: Add missing quotes to Makefile.xorg.</li>
+  <li>autoconf: st/vega requires --enable-openvg.</li>
+  <li>st/mesa: Unreference the sampler view in st_bind_surface.</li>
+  <li>autoconf: Tidy configure output for EGL.</li>
+  <li>targets/egl: Fix a warning with --disable-opengl build.</li>
+  <li>egl: Rework _eglGetSearchPath.</li>
+  <li>mesa: Select FEATURE_remap_table when multiple APIs are enabled.</li>
+  <li>mesa: Allow contexts of different APIs to coexist.</li>
+  <li>egl: Set up the pthread key even TLS is used.</li>
+  <li>st/egl: Add native_surface::present callback.</li>
+  <li>st/egl: Use native_surface::present callback.</li>
+  <li>d3d1x: Use native_surface::present.</li>
+  <li>st/egl: Remove flush_frontbuffer and swap_buffers.</li>
+  <li>st/egl: Add support for swap interval and swap behavior.</li>
+  <li>st/egl: Add support for EGL_MATCH_NATIVE_PIXMAP.</li>
+  <li>st/egl: Add extern "C" wrapper to native.h.</li>
+  <li>st/egl: Add native_display_buffer interface.</li>
+  <li>st/egl: Use native_display_buffer for EGL_MESA_drm_image.</li>
+  <li>autoconf: Add --enable-gallium-egl.</li>
+  <li>docs: Update egl docs.</li>
+  <li>st/dri: Add support for surfaceless current contexts.</li>
+  <li>egl_dri2: Fix __DRI_DRI2 version 1 support.</li>
+  <li>st/vega: Do not wait NULL fences.</li>
+  <li>gallium: Add st_api::name.</li>
+  <li>gallium: Add st_context_iface::share to st_api.</li>
+  <li>st/wgl: Use st_context_iface::share for DrvShareLists.</li>
+  <li>st/glx: Replace MESA_VERSION_STRING by xmesa_get_name.</li>
+  <li>mesa: Clean up core.h.</li>
+  <li>scons: Define IN_DRI_DRIVER.</li>
+  <li>tgsi: Add STENCIL to text parser.</li>
+  <li>st/vega: vegaLookupSingle should validate the state.</li>
+  <li>st/vega: Set wrap_r for mask and blend samplers.</li>
+  <li>st/vega: Fix vgReadPixels with a subrectangle.</li>
+  <li>egl_dri2: Fix one context, multiple surfaces.</li>
+  <li>auxiliary: util_blit_pixels_tex should restore the viewport.</li>
+  <li>st/vega: Fix a crash with empty paths.</li>
+  <li>st/vega: Masks and surfaces should share orientation.</li>
+  <li>st/vega: No flipping in vg_prepare_blend_surface.</li>
+  <li>st/vega: Fix a typo in EXTENDED_BLENDER_OVER_FUNC.</li>
+  <li>llvmpipe: Fix build errors on x86.</li>
+  <li>st/vega: Overhaul renderer with renderer states.</li>
+  <li>st/vega: Add DRAWTEX renderer state.</li>
+  <li>st/vega: Add SCISSOR renderer state.</li>
+  <li>st/vega: Add CLEAR renderer state for vgClear.</li>
+  <li>st/vega: Add FILTER renderer state for image filtering.</li>
+  <li>st/vega: Use the renderer for vgMask.</li>
+  <li>st/vega: Add POLYGON_STENCIL and POLYGON_FILL renderer state.</li>
+  <li>st/vega: Delay fb state update to vg_validate_state.</li>
+  <li>st/vega: Use st_framebuffer for fb width/height.</li>
+  <li>st/vega: Move g3d states to renderer.</li>
+  <li>st/vega: Make shader_bind call into the renderer.</li>
+  <li>st/vega: vg_manager should care about only the color buffer.</li>
+  <li>st/vega: Clean up vg_context fields and functions.</li>
+  <li>st/vega: Clean up renderer fields and functions.</li>
+  <li>st/vega: vg_copy_texture and vg_copy_surface should share code.</li>
+  <li>st/vega: Get rid of renderer_copy_texture.</li>
+  <li>st/vega: Update to latest headers.</li>
+  <li>st/vega: Fix image sampler views for alpha-only formats.</li>
+  <li>st/vega: Make path_render and path_stroke take a matrix.</li>
+  <li>st/vega: Make image_draw take a matrix.</li>
+  <li>st/vega: Add primitive text support.</li>
+  <li>st/vega: Revive mask layer support.</li>
+  <li>st/vega: More flexible shader selection.</li>
+  <li>st/vega: Add color transformation support.</li>
+  <li>st/vega: Bump version to 1.1.</li>
+  <li>st/vega: Fix paint coordinates transformations.</li>
+  <li>st/vega: Fix negated logic in image_draw.</li>
+  <li>st/vega: Fix degenerate paints.</li>
+  <li>st/vega: Simplify radial gradient.</li>
+  <li>st/vega: Remove st_inlines.h.</li>
+  <li>st/vega: Delay blend texture creation until needed.</li>
+  <li>st/vega: Create drawing surface mask as needed.</li>
+  <li>st/vega: Initialize pipe states with renderer.</li>
+  <li>st/vega: Avoid unnecessary constant bufer upload.</li>
+  <li>st/vega: Destroy the pipe context with vg_context.</li>
+  <li>st/vega: polygon_array requires a deep free.</li>
+  <li>st/egl: Set pipe_resource::array_size to 1.</li>
+  <li>st/vega: Set pipe_resource::array_size to 1.</li>
+  <li>st/vega: Move vertex transformation to shader.</li>
+  <li>st/vega: Add a missing break.</li>
+  <li>st/vega: Add some comments to pipeline shaders.</li>
+  <li>st/vega: Refactor blend shaders.</li>
+  <li>st/vega: Move masking after blending.</li>
+  <li>st/vega: Add support for per-channel alpha.</li>
+  <li>st/vega: Blending should use premultiplied alpha.</li>
+  <li>st/vega: Fix VG_BLEND_MULTIPLY.</li>
+  <li>st/vega: Add blend shaders for all blend modes.</li>
+  <li>st/vega: Fix pipe blend state for various blend modes.</li>
+  <li>egl: _eglFilterArray should not allocate.</li>
+  <li>mapi: Rewrite mapi_abi.py to get rid of preprocessor magic.</li>
+  <li>vbo: Drop second ATTR macro.</li>
+  <li>vbo: Fix GLES2 glVertexAttrib.</li>
+  <li>mesa: Do not advertise GL_OES_texture_3D.</li>
+  <li>mesa: Fix GL_FIXED arrays.</li>
+  <li>mesa: Fix glTexCoordPointer with type GL_FIXED.</li>
+  <li>st/egl: Plug pbuffer leaks.</li>
+  <li>st/egl: Fix eglCopyBuffers.</li>
+  <li>st/egl: Assorted fixes for dri2_display_get_configs.</li>
+  <li>docs/egl: Update egl.html.</li>
+  <li>st/egl: Fix eglChooseConfig when configs is NULL.</li>
+  <li>docs: Add an example for EGL_DRIVERS_PATH.</li>
+  <li>autoconf: Fix --with-driver=xlib --enable-openvg.</li>
+</ul></p>
+
+<p>Chris Wilson (2):
+<ul>
+  <li>i915g: Fix closure of full batch buffers</li>
+  <li>intel: Check for unsupported texture when finishing using as a render target</li>
+</ul></p>
+
+<p>Christoph Bumiller (80):
+<ul>
+  <li>nv50: import new compiler</li>
+  <li>nouveau: update nouveau_class.h</li>
+  <li>nv50: introduce the big formats table</li>
+  <li>nv50: don't produce MOV immediate to output reg in store opt</li>
+  <li>nv50: change back accidentally swapped UNORM,SNORM vertex type</li>
+  <li>nv50: add/fix some license headers</li>
+  <li>nv50: simple reload elimination and local CSE</li>
+  <li>nv50: fix constant_operand opt mul by 2 case</li>
+  <li>nv50: permit usage of undefined TGSI TEMPs</li>
+  <li>nv50: add missing 2nd source for POW multiplication</li>
+  <li>nv50: add signed RGTC1 to format table, allow 2_10_10_10 for vbufs</li>
+  <li>nv50: fix for empty BBs</li>
+  <li>nv50: insert MOVs also for PHI sources from dominating block</li>
+  <li>nv50: explicitly set src type for SET ops</li>
+  <li>nv50: fixes for nested IFs</li>
+  <li>nv50: don't eliminate loads to dedicated values</li>
+  <li>nv50: fix constbuf validation</li>
+  <li>nv50: build proper phi functions in the first place</li>
+  <li>nv50: fix reg count</li>
+  <li>nv50: begin implementing loops</li>
+  <li>nv50: more constant folding</li>
+  <li>nv50: loops part 2</li>
+  <li>nv50: flatten simple IF/ELSE/ENDIF constructs</li>
+  <li>nv50: fix thinko in store to output reg possible check</li>
+  <li>nv50: generate JOINs for outermost IF clauses</li>
+  <li>nv50: more TGSI opcodes (SIN, SCS, ARL, RET, KILP)</li>
+  <li>nv50: fix PSIZ and PRIMID mapping</li>
+  <li>nv50: check dst compatibility in CSE</li>
+  <li>nv50: initialize edgeflag input index</li>
+  <li>nv50: emit predicate for interp</li>
+  <li>Merge remote branch 'origin/master' into nv50-compiler</li>
+  <li>nv50: DP2, fix ARL</li>
+  <li>nv50: yet another case we need a nop.exit</li>
+  <li>nv50: fix check for sprite/point coord enable</li>
+  <li>nv50: handle TEXTURE_SWIZZLE and GEOMETRY_SHADER4 caps</li>
+  <li>nv50: set the FragDepth output index</li>
+  <li>nv50: turn off verbose debug output by default</li>
+  <li>nv50: attempt at making more complicated loops work</li>
+  <li>nv50: SSG</li>
+  <li>nv50: make FrontFacing -1 or +1</li>
+  <li>nv50: re-add proper TEXBIAS sequence</li>
+  <li>nv50: make use of TGSI immediate type</li>
+  <li>nv50: must join SELECT inputs before MOV inputs</li>
+  <li>nv50: fix XPD, was negated</li>
+  <li>nv50: fix find_dom_frontier</li>
+  <li>nv50: fix build-predicate function</li>
+  <li>Merge remote branch 'origin/master' into nv50-compiler</li>
+  <li>nv50: load address register before using it, not after</li>
+  <li>nv50: save tgsi instructions</li>
+  <li>nv50: prepare for having multiple functions</li>
+  <li>nv50: don't parse again in tgsi_2_nc</li>
+  <li>nv50: use actual loads/stores if TEMPs are accessed indirectly</li>
+  <li>nv50: create value references with the right type</li>
+  <li>nv50: duplicate interps in load_proj_tex_coords</li>
+  <li>nv50: address regs are 16 bit</li>
+  <li>nv50: fix can_load check for 3rd source</li>
+  <li>nv50: reduce bb_reachable_by runtime from pot to linear</li>
+  <li>nv50: minor compiler fixes and cleanups</li>
+  <li>nv50: cannot move from local mem to output reg directly</li>
+  <li>nv50: newlines in shader bincode printing</li>
+  <li>nv50: match TEMP limit with nv50 ir builder</li>
+  <li>nv50: handle TGSI EXP and LOG again</li>
+  <li>nv50: check for immediates when turning MUL ADD into MAD</li>
+  <li>nv50: interp cannot write flags reg</li>
+  <li>nv50: MOV TEMP[0], -CONST[0] must be float32 negation</li>
+  <li>nv50: fix indirect CONST access with large or negative offsets</li>
+  <li>nv50: fix TXP depth comparison value</li>
+  <li>nv50: consider address register in reload elimination</li>
+  <li>nv50: improve and fix modifier folding optimization</li>
+  <li>nv50: put low limit on REG_ALLOC_TEMP and FP_RESULT_COUNT</li>
+  <li>Merge remote branch 'origin/nv50-compiler'</li>
+  <li>nv50: don't segfault on shaders with 0 instructions</li>
+  <li>nv50: get shader fixups/relocations into working state</li>
+  <li>nv50: add relocs for stack and local mem buffers</li>
+  <li>nv50: emit constbuf relocs before uploading constants</li>
+  <li>nv50: fix typo in fifo packet length limit</li>
+  <li>nv50: use formats table in nv50_surface.c</li>
+  <li>nv50: use CLEAR_BUFFERS for surface fills</li>
+  <li>nv50: fix/handle a few more PIPE_CAPs</li>
+  <li>nv50: fix GP state bind and validate</li>
+</ul></p>
+
+<p>Corbin Simpson (8):
+<ul>
+  <li>r600g: Use align() instead of handrolled code.</li>
+  <li>r600g: Trivially deobfuscate r600_hw_states.</li>
+  <li>r600g: Deobfuscate and comment a few more functions in r600_hw_states.</li>
+  <li>r600g: Clean up some indentation and |= vs. | usage.</li>
+  <li>r600g: Fix false and true.</li>
+  <li>r600g: "tmp" is such a bad name for a texture.</li>
+  <li>r600g: Clean up PS setup.</li>
+  <li>r600g: Cleanup viewport floats.</li>
+</ul></p>
+
+<p>Daniel Lichtenberger (1):
+<ul>
+  <li>radeon: fix potential segfault in renderbuffer update</li>
+</ul></p>
+
+<p>Daniel Vetter (21):
+<ul>
+  <li>r200: revalidate after radeon_update_renderbuffers</li>
+  <li>i915g: rip out -&gt;sw_tiled</li>
+  <li>i915g: s/hw_tiled/tiling</li>
+  <li>i915g: add pineview pci ids</li>
+  <li>i915g: kill RGBA/X formats</li>
+  <li>i915g: kill buf-&gt;map_gtt</li>
+  <li>i915g: kill idws-&gt;pool</li>
+  <li>i915g: drop alignment parameter from iws-&gt;buffer_create</li>
+  <li>i915g: add winsys function to create tiled buffers</li>
+  <li>i915g: switch to tiled allocations, kill set_fence</li>
+  <li>i915g: prepare winsys/batchbuffer for execbuf2</li>
+  <li>i915g: return tiling in iws-&gt;buffer_from_handle</li>
+  <li>i915g: implement unfenced color&amp;depth buffer using tiling bits</li>
+  <li>i915g: implement unfenced relocs for textures using tiling bits</li>
+  <li>i915g: postpone mipmap/face offset calculation</li>
+  <li>i915g: don't pot-align stride for tiled buffers</li>
+  <li>i915g: enable X-tiling for textures</li>
+  <li>i915g: switch rendering to mipmapped textures to (x,y) offsets</li>
+  <li>i915g: enable x-tiling for render targets</li>
+  <li>i915g: assert(depth_surface-&gt;offset == 0)</li>
+  <li>i915g: track TODO items</li>
+</ul></p>
+
+<p>Dave Airlie (182):
+<ul>
+  <li>r300g: fix buffer reuse issue caused by previous commit</li>
+  <li>r600g: pull r600_draw struct out into header</li>
+  <li>r600g: use index min/max + index buffer offset.</li>
+  <li>r600g: add vgt dma src defines</li>
+  <li>r600g: fixup texture state on evergreen.</li>
+  <li>r600g: fix texture bos and avoid doing depth blit on evergreen</li>
+  <li>r600g: hide radeon_ctx inside winsys.</li>
+  <li>r600g: attempt to abstract kernel bos from pipe driver.</li>
+  <li>r600g: move constant buffer creation behind winsys abstraction.</li>
+  <li>r600g: use malloc bufmgr for constant buffers</li>
+  <li>r600g: add support for kernel bo</li>
+  <li>r600g: add winsys bo caching.</li>
+  <li>r600g: add upload manager support.</li>
+  <li>r600g: fixup map flushing.</li>
+  <li>r600g: use calloc for ctx bo allocations</li>
+  <li>r600g: oops got the use_mem_constant the wrong way around.</li>
+  <li>r600g; add uses waterfall to asm cf for r6xx.</li>
+  <li>r600g: only emit uses waterfall on r6xx hw.</li>
+  <li>util/r300g: split the r300 index buffer modifier functions out to util</li>
+  <li>r600g: modify index buffers for sizes the hw can't deal with.</li>
+  <li>r600g: send correct surface base update for multi-cbufs</li>
+  <li>r600g: fix fbo-drawbuffers-maxtargets</li>
+  <li>r600g: clean up valgrind issues on maxtargets test.</li>
+  <li>r600g: drop debugging that snuck in</li>
+  <li>r600g: fix tiling support for ddx supplied buffers</li>
+  <li>r600g: add z16 to color setup</li>
+  <li>r600g: add color/texture support for more depth formats.</li>
+  <li>r600g: fix r700 cube map sizing.</li>
+  <li>r600g: fixup r700 CB_SHADER_CONTROL register.</li>
+  <li>r600g: add missing BC_INST wrapper for evergreen</li>
+  <li>r600g: only flush for the correct colorbuffer, not all of them.</li>
+  <li>r600g: deal with overflow of VTX/TEX CF clauses.</li>
+  <li>r600g: set back to correct codepaths.</li>
+  <li>r600g: fixup evergreen miptree setup.</li>
+  <li>r600g: fix eg texture borders.</li>
+  <li>r600g: fix typo in struct member name</li>
+  <li>r600g: cleanup some of the DB blit code</li>
+  <li>r600g: make stencil readback work</li>
+  <li>r600g: disable dirty handling on texture from depth code.</li>
+  <li>r600g: use floats instead of hex for blit vbo</li>
+  <li>r600g: fix depth readback on rv610 and other quirky variants.</li>
+  <li>r600g: fix typo in evergreen register list</li>
+  <li>u_blitter: add a custom blitter call passing a dsa cso</li>
+  <li>r600g: use blitter to do db-&gt;cb flushing.</li>
+  <li>r600g: fix warnings since last commit.</li>
+  <li>egl: fix build since 17eace581d25a626a7d75d9d1205d012cbb14a6e</li>
+  <li>r300g: fix point sprite coord.</li>
+  <li>r600g: add vert support for 16/16 and 16/16/16 floats.</li>
+  <li>r600g: add some more vertex format support.</li>
+  <li>r600g: some more vertex formats</li>
+  <li>r600g: fix draw-elements and draw-elements-base-vertex</li>
+  <li>r600g: drop index_offset parameter to index buffer translate.</li>
+  <li>r600g: fixup tex wrapping.</li>
+  <li>r600g: fixup VP-&gt;FP output-&gt;input routing.</li>
+  <li>r600g: fix typo in r700 alu emit</li>
+  <li>r600g: fixup sprite coord enable.</li>
+  <li>r600g: fix polygon mode</li>
+  <li>mesa/mipmap: fix warning since 1acadebd6270d3604b026842b8a21360968618a0</li>
+  <li>r600g: add eg poly mode code.</li>
+  <li>r600g: make index bias fix for evergreen</li>
+  <li>r600g: add eg db count control register.</li>
+  <li>r300g: fix glsl-fs-pointcoord</li>
+  <li>r600g: add evergreen texture resource properly.</li>
+  <li>r600g: fix db flush breaking config state</li>
+  <li>r600g: on evergreen the centroid isn't set in this register.</li>
+  <li>r600g: add back evergreen name.</li>
+  <li>r600g: add evergreen texture border support to new path</li>
+  <li>r600g: move radeon.h members around to add back map flushing.</li>
+  <li>r600g: add initial vertex translate support.</li>
+  <li>r600g: remove old assert from new codepath</li>
+  <li>Revert "r600g: add initial vertex translate support."</li>
+  <li>r600g: port r300g fix for X* formats in texformat code</li>
+  <li>r600g: add L8A8 unorm.</li>
+  <li>r600g: clean up some code from move to new paths.</li>
+  <li>r600g: return string for chip family</li>
+  <li>r600g: use Elements macro instead of manual sizeofs</li>
+  <li>r600g: fix evergreen depth flushing.</li>
+  <li>r600g: add winsys support for CTL constants.</li>
+  <li>r600g: drop depth quirk on evergreen</li>
+  <li>r600g: add reloc for evergreen color attrib</li>
+  <li>r600g: realign evergreen code with r600 code.</li>
+  <li>r600g: add assembler support for other vtx fetch fields.</li>
+  <li>r600g: fixup vertex format picking.</li>
+  <li>r600g: sync vertex/texture cache on resources on evergreen</li>
+  <li>r600g: add cb flushing for extra buffers + depth buffer on r600/evergreen</li>
+  <li>r600g: fix evergreen draw-buffers</li>
+  <li>r600g: flush SH cache on constant change on evergreen</li>
+  <li>r600g: only set the Z export if shader exports it.</li>
+  <li>r600g: setup basic loop consts on r600 + evergreen.</li>
+  <li>mesa/st: initial attempt at RG support for gallium drivers</li>
+  <li>r600g: break out of search for reloc bo after finding it.</li>
+  <li>r600g: the code to check whether a new vertex shader is needed was wrong</li>
+  <li>r600g: fix wwarning in bo_map function</li>
+  <li>r600g: TODO domain management</li>
+  <li>r600g: add bo fenced list.</li>
+  <li>pb: don't keep checking buffers after first busy</li>
+  <li>r600g: add bo busy backoff.</li>
+  <li>r600g: drop mman allocator</li>
+  <li>r600g: drop use_mem_constant.</li>
+  <li>r600g: avoid unneeded bo wait</li>
+  <li>pb: fix numDelayed accounting</li>
+  <li>r600g: add evergreen stencil support.</li>
+  <li>r600g: use format from the sampler view not from the texture.</li>
+  <li>r600g: fix Z export enable bits.</li>
+  <li>r600g: add some RG texture format support.</li>
+  <li>r600g: drop width/height per level storage.</li>
+  <li>r600g: fix input/output Z export mixup for evergreen.</li>
+  <li>r600g: evergreen has no request size bit in texture word4</li>
+  <li>r600g: enable vertex samplers.</li>
+  <li>r600g: add TXL opcode support.</li>
+  <li>r600g: don't run with scissors.</li>
+  <li>r600g: fix typo in vertex sampling on r600</li>
+  <li>gallium/tgsi: add support for stencil writes.</li>
+  <li>gallium/format: add support for X24S8 and S8X24 formats.</li>
+  <li>gallium/format: add X32_S8X24_USCALED format.</li>
+  <li>gallium/util: add S8 tile sampling support.</li>
+  <li>mesa: add support for FRAG_RESULT_STENCIL.</li>
+  <li>mesa: improve texstore for 8/24 formats and add texstore for S8.</li>
+  <li>softpipe: add support for shader stencil export capability</li>
+  <li>st/mesa: add option to choose a texture format that we won't render to.</li>
+  <li>st/mesa: use shader stencil export to accelerate shader drawpixels.</li>
+  <li>r600g: add support for S8, X24S8 and S8X24 sampler formats.</li>
+  <li>r600g: add shader stencil export support.</li>
+  <li>glsl: add support for shader stencil export</li>
+  <li>st/mesa: enable stencil shader export extension if supported</li>
+  <li>r600g: fix depth0 setting</li>
+  <li>r600g: fix scissor/cliprect confusion</li>
+  <li>r600g: store samplers/views across blit when we need to modify them</li>
+  <li>r600g: reduce size of context structure.</li>
+  <li>r600g: the vs/ps const arrays weren't actually being used.</li>
+  <li>r600g: add copy into tiled texture</li>
+  <li>r600g: split out miptree setup like r300g</li>
+  <li>r600g: use common texture object create function</li>
+  <li>r600g: rename pitch in texture to pitch_in_bytes</li>
+  <li>r600g: remove bpt and start using pitch_in_bytes/pixels.</li>
+  <li>r600g: fix transfer stride.</li>
+  <li>r600g: drop all use of unsigned long</li>
+  <li>r600g: use blitter for hw copy region</li>
+  <li>r600g: evergreen add stencil export bit</li>
+  <li>r600g: add missing eg reg definition</li>
+  <li>r600g: fix stencil export for evergreen harder</li>
+  <li>r600g: drop unused context members</li>
+  <li>r600g: only pick centroid coordinate when asked.</li>
+  <li>r600g: fixup pos/face ena/address properly</li>
+  <li>r600g: fixup typo in macro name</li>
+  <li>r600g: select linear interpolate if tgsi input requests it</li>
+  <li>r300g: clean up warning due to unknown cap.</li>
+  <li>tgsi: add scanner support for centroid inputs</li>
+  <li>r600g: evergreen interpolation support.</li>
+  <li>r600g: add evergreen ARL support.</li>
+  <li>r600g: switch to a common formats.h file since they are in different regs</li>
+  <li>r600g: add defines for tiling</li>
+  <li>r600g: get tiling info from kernel</li>
+  <li>r600g: set tiling bits in hw state</li>
+  <li>r600g: do proper tracking of views/samplers.</li>
+  <li>r600g: fix typo in tiling setup cb code.</li>
+  <li>r600g: depth needs to bound to ds</li>
+  <li>r600g: attempt to cleanup depth blit</li>
+  <li>r600g: fix transfer function for tiling.</li>
+  <li>r600g: retrieve tiling info from kernel for shared buffers.</li>
+  <li>r600g: all non-0 mipmap levels need to be w/h aligned to POT.</li>
+  <li>r600g: move to per-miplevel array mode.</li>
+  <li>r600g: start adding hooks for aligning width/height for tiles.</li>
+  <li>r600g: add r600 surface to store the aligned height.</li>
+  <li>r600g: introduce a per-driver resource flag for transfers.</li>
+  <li>r600g: add texture tiling alignment support.</li>
+  <li>r600g: add texture tiling enable under a debug option.</li>
+  <li>r600g: initial translate state support</li>
+  <li>r600g: start splitting out common code from eg/r600.</li>
+  <li>r600g: not fatal if we can't get tiling info from kernel</li>
+  <li>r600g: merge more of the common r600/evergreen state handling</li>
+  <li>r600g: drop more common state handling code</li>
+  <li>r600g: fix magic 0x1 -&gt;flat shade ena</li>
+  <li>r600g: add assembler support for all the kcache fields.</li>
+  <li>gallium/noop: report GL 2.1</li>
+  <li>r600g: pick correct color swap for A8 fbos.</li>
+  <li>r300g/r600g: bump cache manager timeouts to 1s</li>
+  <li>r600g: it looks like r600 can handle dword offsets in the indices.</li>
+  <li>r300g: try and use all of vertex constant space</li>
+  <li>r300g: fixup rs690 tiling stride alignment calculations.</li>
+  <li>r600g: fix evergreen segfaults.</li>
+  <li>r600g: hack around property unknown issues.</li>
+</ul></p>
+
+<p>Eric Anholt (300):
+<ul>
+  <li>glsl: Add definition of gl_TextureMatrix inverse/transpose builtins.</li>
+  <li>i965: Share the KIL_NV implementation between glsl and non-glsl.</li>
+  <li>i965: Also enable CC statistics when doing OQs.</li>
+  <li>i965: Track the windowizer's dispatch for kill pixel, promoted, and OQ</li>
+  <li>glsl: Rework assignments with write_masks to have LHS chan count match RHS.</li>
+  <li>glsl: Fix copy'n'wasted ir_noop_swizzle conditions.</li>
+  <li>ir_to_mesa: Only compare vector_elements present for any_nequal/all_equal</li>
+  <li>i965: Fix the vector/expression splitting for the write_mask change.</li>
+  <li>i965: When splitting vector variable assignment, ignore unset channels.</li>
+  <li>i965: Update expression splitting for the vector-result change to compares.</li>
+  <li>i965: Warning fix for vector result any_nequal/all_equal change.</li>
+  <li>mesa: Remove the non-required ARB_imaging extension.</li>
+  <li>mesa: Remove EXT_histogram.</li>
+  <li>mesa: Remove SGI_color_table.</li>
+  <li>mesa: Remove SGI_color_matrix.</li>
+  <li>mesa: Remove EXT_convolution.</li>
+  <li>intel: Remove disabled stencil drawpixels acceleration.</li>
+  <li>intel: Remove unnecessary minimum pitch alignment to 32 bytes.</li>
+  <li>intel: Replace my intel_texture_bitmap code with _mesa_meta_Bitmap.</li>
+  <li>radeon: Remove copied minimum pitch alignment code.</li>
+  <li>unichrome: Mostly revert my convolution removal changes.</li>
+  <li>intel: Remove dead intelIsTextureResident().</li>
+  <li>i915: Remove a dead if (0) block.</li>
+  <li>intel: Dead comment removal.</li>
+  <li>intel: Corresponding FinishRenderTexture debug to BeginRenderTexture.</li>
+  <li>i965: Add support for rendering to SARGB8 FBOs.</li>
+  <li>intel: Fix segfault on INTEL_DEBUG=fbo with unsupported framebuffers.</li>
+  <li>intel: Add fallback debug to glGenerateMipmap.</li>
+  <li>intel: More reverting of the sw fallback for depth texture border color.</li>
+  <li>intel: Improve some of the miptree debugging.</li>
+  <li>mesa: Fix type typo in glGenerateMipmap handling of GL_UNSIGNED_INT data.</li>
+  <li>glsl: Fix broadcast_index of lower_variable_index_to_cond_assign.</li>
+  <li>glsl: Add validation that a swizzle only references valid channels.</li>
+  <li>i965: Fix up writemasked assignments in the new FS.</li>
+  <li>i965: Remove swizzling of assignment to vector-splitting single-channel LHS.</li>
+  <li>i965: Handle all_equal/any_nequal in the new FS.</li>
+  <li>i965: Fix vector splitting RHS channel selection with sparse writemasks.</li>
+  <li>i965: Add support for dFdx()/dFdy() to the FS backend.</li>
+  <li>i965: Add support for attribute interpolation on Sandybridge.</li>
+  <li>i965: Set up inputs to the fragment shader according to FP InputsRead.</li>
+  <li>i965: Add support for POW in gen6 FS.</li>
+  <li>i965: Fix negation in the new FS backend.</li>
+  <li>i965: Actually track the "if" depth in loop in the new FS backend.</li>
+  <li>i965: Apply the same set of lowering passes to new FS as to Mesa IR.</li>
+  <li>i965: Fix valgrind complaint about base_ir for new FS debugging.</li>
+  <li>i965: Fix up the FS backend for the variable array indexing pass.</li>
+  <li>i965: Set the variable type when dereferencing an array.</li>
+  <li>i965: Add support for dereferencing structs to the new FS backend.</li>
+  <li>i965: Add support for struct, array, and matrix uniforms to FS backend.</li>
+  <li>i965: Fix all non-snb regression in the snb attribute interpolation commit.</li>
+  <li>i965: Fix up part of my Sandybridge attributes support patch.</li>
+  <li>i965: Add support for gl_FrontFacing to the new FS backend.</li>
+  <li>i965: Subtract instead of adding when computing y delta in new FS backend.</li>
+  <li>mesa: Pull ir_to_mesa's sampler number fetcher out to shared code.</li>
+  <li>i965: Set up sampler numbers in the FS backend.</li>
+  <li>i965: Add support for non-color render target write data to new FS backend.</li>
+  <li>i965: Add support for MRT to the new FS backend.</li>
+  <li>i965: Add support for ir_loop counters to the new FS backend.</li>
+  <li>i965: Add support for ARB_fragment_coord_conventions to the new FS backend.</li>
+  <li>glsl: Also update implicit sizes of varyings at link time.</li>
+  <li>i965: Do interpolation for varying matrices and arrays in the FS backend.</li>
+  <li>i965: Don't try to emit interpolation for unused varying slots.</li>
+  <li>i965: Fix array indexing of arrays of matrices.</li>
+  <li>i965: Clean up obsolete FINISHME comment.</li>
+  <li>mesa: Move the list of builtin uniform info from ir_to_mesa to shared code.</li>
+  <li>i965: Add support for builtin uniforms to the new FS backend.</li>
+  <li>i965: Fix use of undefined mem_ctx in vector splitting.</li>
+  <li>i956: Make new FS discard do its work in a temp, not the null reg!</li>
+  <li>i965: Clean up the virtual GRF handling.</li>
+  <li>ra: First cut at a graph-coloring register allocator for mesa.</li>
+  <li>i965: First cut at register allocation using graph coloring.</li>
+  <li>i965: Add live interval analysis and hook it up to the register allocator.</li>
+  <li>i965: Remove my "safety counter" code from loops.</li>
+  <li>i965: Fix whole-structure/array assignment in new FS.</li>
+  <li>mesa: Don't reference a W component in setting up a vec3 uniform component.</li>
+  <li>i965: Fix new FS handling of builtin uniforms with packed scalars in structs.</li>
+  <li>glsl: Add a lowering pass for texture projection.</li>
+  <li>i965: Use the lowering pass for texture projection.</li>
+  <li>i965: Split the gen4 and gen5 sampler handling apart.</li>
+  <li>i965: Add gen6 attribute interpolation to new FS backend.</li>
+  <li>i965: Fix the gen6 jump size for BREAK/CONT in new FS.</li>
+  <li>i965: Also increment attribute location when skipping unused slots.</li>
+  <li>i965: Pre-gen6, map VS outputs (not FS inputs) to URB setup in the new FS.</li>
+  <li>i965: Add real support for pre-gen5 texture sampling to the new FS.</li>
+  <li>i965: Fix up copy'n'pasteo from moving coordinate setup around for gen4.</li>
+  <li>i965: Restore the forcing of aligned pairs for delta_xy on chips with PLN.</li>
+  <li>i965: When producing a single channel swizzle, don't make a temporary.</li>
+  <li>i965: Add a sanity check for register allocation sizes.</li>
+  <li>i965: Fix off-by-ones in handling the last members of register classes.</li>
+  <li>i965: Don't try to emit code if we failed register allocation.</li>
+  <li>i965: Add support for EXT_texture_swizzle to the new FS backend.</li>
+  <li>i965: Set up swizzling of shadow compare results for GL_DEPTH_TEXTURE_MODE.</li>
+  <li>i965: Fix glean/texSwizzle regression in previous commit.</li>
+  <li>i965: Be more conservative on live interval calculation.</li>
+  <li>i965: Add trivial dead code elimination in the new FS backend.</li>
+  <li>i965: Add initial folding of constants into operand immediate slots.</li>
+  <li>i965: In disasm, gen6 fb writes don't put msg reg # in destreg_conditionalmod.</li>
+  <li>i965: Add support for gen6 FB writes to the new FS.</li>
+  <li>i965: Enable the constant propagation code.</li>
+  <li>i965: Also do constant propagation for the second operand of CMP.</li>
+  <li>i965: Add back gen6 headerless FB writes to the new FS backend.</li>
+  <li>i965: Gen6 no longer has the IFF instruction; always use IF.</li>
+  <li>i965: Fix up IF/ELSE/ENDIF for gen6.</li>
+  <li>i965: Fix botch in the header_present case in the new FS.</li>
+  <li>i965: Add some clarification of the WECtrl field.</li>
+  <li>i965: Don't do 1/w multiplication in new FS for gen6</li>
+  <li>i965: Gen6's sampler messages are the same as Ironlake.</li>
+  <li>i965: Refactor gl_FrontFacing setup out of general variable setup.</li>
+  <li>i965: Add support for gl_FrontFacing on gen6.</li>
+  <li>i965: Don't assume that WPOS is always provided on gen6 in the new FS.</li>
+  <li>i965: Fix gen6 pointsize handling to match pre-gen6.</li>
+  <li>i965: Disable emitting if () statements on gen6 until we really fix them.</li>
+  <li>i965: Normalize cubemap coordinates like is done in the Mesa IR path.</li>
+  <li>mesa: Simplify a bit of _mesa_add_state_reference using memcmp.</li>
+  <li>i965: Drop the check for duplicate _mesa_add_state_reference.</li>
+  <li>i965: Drop the check for YUV constants in the param list.</li>
+  <li>i965: Handle swizzles in the addition of YUV texture constants.</li>
+  <li>i965: Fix gen6 WM push constants updates.</li>
+  <li>i965: Fix new FS gen6 interpolation for sparsely-populated arrays.</li>
+  <li>i965: Enable attribute swizzling (repositioning) in the gen6 SF.</li>
+  <li>i965: Add register coalescing to the new FS backend.</li>
+  <li>i965: Split FS_OPCODE_DISCARD into two steps.</li>
+  <li>i965: Reduce register interference checks for changed FS_OPCODE_DISCARD.</li>
+  <li>i965: Move FS backend structures to a header.</li>
+  <li>i965: Give the math opcodes information on base mrf/mrf len.</li>
+  <li>i965: Give the FB write and texture opcodes the info on base MRF, like math.</li>
+  <li>i965: Compute to MRF in the new FS backend.</li>
+  <li>i965: Don't consider gen6 math instructions to write to MRFs.</li>
+  <li>i965: Add a couple of checks for gen6 math instruction limits.</li>
+  <li>i965: Don't compute-to-MRF in gen6 math instructions.</li>
+  <li>i965: Expand uniform args to gen6 math to full registers to get hstride == 1.</li>
+  <li>i965: Don't compute-to-MRF in gen6 VS math.</li>
+  <li>i965: Fix gen6 pixel_[xy] setup to avoid mixing int and float src operands.</li>
+  <li>i965: Always use the new FS backend on gen6.</li>
+  <li>i965: Fix missing "break;" in i2b/f2b, and missing AND of CMP result.</li>
+  <li>intel: Allow CopyTexSubImage to InternalFormat 3/4 textures, like RGB/RGBA.</li>
+  <li>i965: Don't rebase the index buffer to min 0 if any arrays are in VBOs.</li>
+  <li>i965: Add support for rescaling GL_TEXTURE_RECTANGLE coords to new FS.</li>
+  <li>i965: Set class_sizes[] for the aligned reg pair class.</li>
+  <li>i965: Update the live interval when coalescing regs.</li>
+  <li>i965: Add a pass to the FS to split virtual GRFs to float channels.</li>
+  <li>i965: Add a function for handling the move of boolean values to flag regs.</li>
+  <li>i965: Add peepholing of conditional mod generation from expressions.</li>
+  <li>i965: Enable the new FS backend on pre-gen6 as well.</li>
+  <li>i965: Fix texturing on pre-gen5.</li>
+  <li>i965: Set the type of the null register to fix gen6 FS comparisons.</li>
+  <li>i965: Disable the debug printf I added for FS disasm.</li>
+  <li>i965: Fix a weirdness in NOT handling.</li>
+  <li>i965: Fix assertion failure on gen6 BufferSubData to busy BO.</li>
+  <li>i965: Assert out on gen6 VS constant buffer reads that hang the GPU for now.</li>
+  <li>i965: Fix scissor-offscreen on gen6 like we did pre-gen6.</li>
+  <li>i965: Avoid blits in BufferCopySubdata on gen6.</li>
+  <li>i965: Tell the shader compiler when we expect depth writes for gen6.</li>
+  <li>i965: Remove the gen6 emit_mi_flushes I sprinkled around the driver.</li>
+  <li>i965: Disable thread dispatch when the FS doesn't do any work.</li>
+  <li>i965: Add EU emit support for gen6's new IF instruction with comparison.</li>
+  <li>i965: Set the source operand types for gen6 if/else/endif to integer.</li>
+  <li>i965: Use the new style of IF statement with embedded comparison on gen6.</li>
+  <li>i965: Split register allocation out of the ever-growing brw_fs.cpp.</li>
+  <li>i965: Fix gl_FrontFacing emit on pre-gen6.</li>
+  <li>i965: Add support for register spilling.</li>
+  <li>i965: Don't emit register spill offsets directly into g0.</li>
+  <li>i965: Correct scratch space allocation.</li>
+  <li>i965: Be more aggressive in tracking live/dead intervals within loops.</li>
+  <li>i965: Move the FS disasm/annotation printout to codegen time.</li>
+  <li>i965: Add support for pull constants to the new FS backend.</li>
+  <li>i965: Add EU code for dword scattered reads (constant buffer array indexing).</li>
+  <li>i965: Clarify an XXX comment in FB writes with real info.</li>
+  <li>i965: Use SENDC on the first render target write on gen6.</li>
+  <li>i965: Clear some undefined fields of g0 when using them for gen6 FB writes.</li>
+  <li>i965: Add disasm for the flag register.</li>
+  <li>i965: Add support for discard instructions on gen6.</li>
+  <li>i965: Handle new ir_unop_round_even in channel expression splitting.</li>
+  <li>i965: Fix typo in comment about state flags.</li>
+  <li>i965: Set up the constant buffer on gen6 when it's needed.</li>
+  <li>i965: Add support for constant buffer loads on gen6.</li>
+  <li>i965: Drop the eot argument to read messages, which can never be set.</li>
+  <li>i965: Fix VS URB entry sizing.</li>
+  <li>i965: Disable register spilling on gen6 until it's fixed.</li>
+  <li>i965: Make FS uniforms be the actual type of the uniform at upload time.</li>
+  <li>i965: Add user clip planes support to gen6.</li>
+  <li>i965: Update gen6 SF state when point state (sprite or attenuation) changes.</li>
+  <li>i965: Upload required gen6 VS push constants even when using pull constants.</li>
+  <li>i965: Update the gen6 stencil ref state when stencil state changes.</li>
+  <li>mesa: Make metaops use program refcounts instead of names.</li>
+  <li>mesa: Don't compute an unused texture completeness debug string.</li>
+  <li>intel: For batch, use GTT mapping instead of writing to a malloc and copying.</li>
+  <li>intel: Annotate debug printout checks with unlikely().</li>
+  <li>intel: Remove the magic unaligned memcpy code.</li>
+  <li>i965: Remove dead intel_structs.h file.</li>
+  <li>intel: Avoid taking logbase2 of several things that we max.</li>
+  <li>intel: Remove duplicated teximage miptree to object miptree promotion.</li>
+  <li>intel: Remove leftover dri1 locking fields in the context.</li>
+  <li>mesa: Fix delayed state flagging for EXT_sso-related program changes.</li>
+  <li>intel: Fix the client-side swapbuffers throttling.</li>
+  <li>Revert "intel: Fix the client-side swapbuffers throttling."</li>
+  <li>i965: Allow OPCODE_SWZ to put immediates in the first arg.</li>
+  <li>i965: Add support for math on constants in gen6 brw_wm_glsl.c path.</li>
+  <li>i965: Work around strangeness in swizzling/masking of gen6 math.</li>
+  <li>i965: re-enable gen6 IF statements in the fragment shader.</li>
+  <li>glsl: Free the loop state context when we free the loop state.</li>
+  <li>i965: Fix gl_FragCoord inversion when drawing to an FBO.</li>
+  <li>i965: Shut up spurious gcc warning about GLSL_TYPE enums.</li>
+  <li>mesa: Don't spam the console in a debug build unless some spam is requested.</li>
+  <li>i965: Add state dumping for sampler state.</li>
+  <li>i965: Add dumping of the sampler default color.</li>
+  <li>i965: Fail on loops on gen6 for now until we write the EU emit code for it.</li>
+  <li>i965: Eliminate dead code more aggressively.</li>
+  <li>mesa: Include C++ files in the makedepend of DRI drivers.</li>
+  <li>i965: Fix compute_to_mrf to not move a MRF write up into another live range.</li>
+  <li>i965: Just use memset() to clear most members in FS constructors.</li>
+  <li>i965: Remove extra n at the end of every instruction in INTEL_DEBUG=wm.</li>
+  <li>i965: Fold constants into the second arg of BRW_SEL as well.</li>
+  <li>glsl: Add a helper function for determining if an rvalue could be a saturate.</li>
+  <li>i965: Recognize saturates and turn them into a saturated mov.</li>
+  <li>ir_to_mesa: Detect and emit MOV_SATs for saturate constructs.</li>
+  <li>i965: Improve compute-to-mrf.</li>
+  <li>i965: Remove duplicate MRF writes in the FS backend.</li>
+  <li>i965: Move gen4 blend constant color to the gen4 blending file.</li>
+  <li>i965: Don't upload polygon stipple unless required.</li>
+  <li>i965: Don't upload line stipple pattern unless we're stippling.</li>
+  <li>i965: Don't upload line smooth params unless we're line smoothing.</li>
+  <li>i965: Use the new embedded compare in SEL on gen6 for VS MIN and MAX opcodes.</li>
+  <li>i965: Fix type of gl_FragData[] dereference for FB write.</li>
+  <li>glsl: Make the symbol table's add_function just use the function's name.</li>
+  <li>glsl: Make the symbol table's add_variable just use the variable's name.</li>
+  <li>glsl: Add a helper constructor for expressions that works out result type.</li>
+  <li>glsl: Fix structure and array comparisions.</li>
+  <li>glsl: Quiet unreachable no-return-from-function warning.</li>
+  <li>i965: Dump the WHILE jump distance on gen6.</li>
+  <li>i965: Add support for gen6 DO/WHILE ISA emit.</li>
+  <li>i965: Add support for gen6 BREAK ISA emit.</li>
+  <li>i965: Add support for gen6 CONTINUE instruction emit.</li>
+  <li>i965: Enable IF statements in the VS.</li>
+  <li>i965: Add support for loops in the VS.</li>
+  <li>glsl: Mark the array access for whole-array comparisons.</li>
+  <li>glsl: Fix flipped return of has_value() for array constants.</li>
+  <li>mesa: Add getters for the rest of the supported draw buffers.</li>
+  <li>mesa: Add getters for ARB_copy_buffer's attachment points.</li>
+  <li>intel: Add an env var override to execute for a different GPU revision.</li>
+  <li>i965: Update gen6 WM state on compiled program change, not just FP change.</li>
+  <li>i965: Update gen6 SF state on fragment program change too.</li>
+  <li>i965: Fix compile warning about missing opcodes.</li>
+  <li>i965: Move payload reg setup to compile, not lookup time.</li>
+  <li>i965: Provide delta_xy reg to gen6 non-GLSL path PINTERP.</li>
+  <li>i965: Fix up 16-wide gen6 FB writes after various refactoring.</li>
+  <li>i965: Don't smash a group of coordinates doing gen6 16-wide sampler headers.</li>
+  <li>i965: Fix gen6 interpolation setup for 16-wide.</li>
+  <li>i965: Fix up gen6 samplers for their usage by brw_wm_emit.c</li>
+  <li>i965: Make the sampler's implied move on gen6 be a raw move.</li>
+  <li>i965: Align gen6 push constant size to dispatch width.</li>
+  <li>i965: Add support for the instruction compression bits on gen6.</li>
+  <li>i965: Nuke brw_wm_glsl.c.</li>
+  <li>i965: Remove INTEL_DEBUG=glsl_force now that there's no brw_wm_glsl.c</li>
+  <li>i965: Fix comment about gen6_wm_constants.</li>
+  <li>i965: Handle saturates on gen6 math instructions.</li>
+  <li>i965: Always hand the absolute value to RSQ.</li>
+  <li>i965: Add disabled debug code for dumping out the WM constant payload.</li>
+  <li>i965: Work around gen6 ignoring source modifiers on math instructions.</li>
+  <li>i965: Fix flipped value of the not-embedded-in-if on gen6.</li>
+  <li>i965: Don't try to store gen6 (float) blend constant color in bytes.</li>
+  <li>i965: Set up the color masking for the first drawbuffer on gen6.</li>
+  <li>i965: Set up the per-render-target blend state on gen6.</li>
+  <li>i965: Set the render target index in gen6 fixed-function/ARB_fp path.</li>
+  <li>i965: Use the new pixel mask location for gen6 ARB_fp KIL instructions.</li>
+  <li>i965: Drop KIL_NV from the ff/ARB_fp path since it was only used for GLSL.</li>
+  <li>i965: Drop push-mode reladdr constant loading and always use constant_map.</li>
+  <li>i965: Fix VS constants regression pre-gen6.</li>
+  <li>i965: Clean up VS constant buffer location setup.</li>
+  <li>i965: Set up the correct texture border color state struct for Ironlake.</li>
+  <li>i965: Set render_cache_read_write surface state bit on gen6 constant surfs.</li>
+  <li>i965: remove unused variable since brw_wm_glsl.c removal.</li>
+  <li>intel: Use plain R8 and RG8 for COMPRESSED_RED and COMPRESSED_RG.</li>
+  <li>intel: Set the swizzling for depth textures using the GL_RED depth mode.</li>
+  <li>glsl: Correct the marking of InputsRead/OutputsWritten on in/out matrices.</li>
+  <li>i965: Correct the dp_read message descriptor setup on g4x.</li>
+  <li>intel: Include stdbool so we can stop using GLboolean when we want to.</li>
+  <li>i965: Fix ARL to work on gen6.</li>
+  <li>i956: Fix the old FP path fragment position setup on gen6.</li>
+  <li>i965: Fix gl_FragCoord.z setup on gen6.</li>
+  <li>i965: Add support for using the BLT ring on gen6.</li>
+  <li>intel: Update renderbuffers before looking up CopyTexImage's read buffer.</li>
+  <li>intel: Drop commented intel_flush from copy_teximage.</li>
+  <li>intel: Try to sanely check that formats match for CopyTexImage.</li>
+  <li>intel: Support glCopyTexImage() from XRGB8888 to ARGB8888.</li>
+  <li>i965: Avoid using float type for raw moves, to work around SNB issue.</li>
+  <li>i965: Set the alternative floating point mode on gen6 VS and WM.</li>
+  <li>i965: Add support for gen6 constant-index constant loading.</li>
+  <li>i965: Add support for gen6 reladdr VS constant loading.</li>
+  <li>i965: Improve the hacks for ARB_fp scalar^scalar POW on gen6.</li>
+  <li>i965: Factor out the ir comparision to BRW_CONDITIONAL_* code.</li>
+  <li>i965: Fix regression in FS comparisons on original gen4 due to gen6 changes.</li>
+  <li>i965: Do lowering of array indexing of a vector in the FS.</li>
+  <li>intel: Only do frame throttling at glFlush time when using frontbuffer.</li>
+  <li>intel: Handle forced swrast clears before other clear bits.</li>
+  <li>intel: Use tri clears when we don't know how to blit clear the format.</li>
+  <li>intel: Add spans code for the ARB_texture_rg support.</li>
+  <li>intel: Add a couple of helper functions to reduce rb code duplication.</li>
+  <li>intel: Fix segfaults from trying to use _ColorDrawBuffers in FBO validation.</li>
+  <li>intel: When validating an FBO's combined depth/stencil, use the given FBO.</li>
+</ul></p>
+
+<p>Fabian Bieler (2):
+<ul>
+  <li>r600g: set address of pop instructions to next instruction</li>
+  <li>glsl: fix lowering conditional returns in subroutines</li>
+</ul></p>
+
+<p>Francisco Jerez (51):
+<ul>
+  <li>dri/nv04: Fix PGRAPH_ERRORs when running OA.</li>
+  <li>dri/nv04: Mipmapping fixes.</li>
+  <li>dri/nv04: Align SIFM transfer dimensions.</li>
+  <li>dri/nv04: Fix up color mask.</li>
+  <li>dri/nv04: Fix maximum texture size.</li>
+  <li>dri/nv04: Fix provoking vertex.</li>
+  <li>dri/nouveau: Update nouveau_class.h.</li>
+  <li>dri/nouveau: Add some more extensions.</li>
+  <li>dri/nouveau: Fix glRenderbufferStorage with DEPTH_COMPONENT as internal format.</li>
+  <li>dri/nouveau: Don't request a fake front unnecessarily.</li>
+  <li>dri/nouveau: Don't reemit the BO state in nouveau_state_emit().</li>
+  <li>dri/nouveau: Cleanup references to the old FBOs on glMakeCurrent().</li>
+  <li>meta: Don't bind the created texture object in init_temp_texture().</li>
+  <li>dri/nv10: Fix the CLAMP texture wrap mode.</li>
+  <li>dri/nv04: Use nvgl_wrap_mode().</li>
+  <li>dri/nouveau: Remove unnecessary assertion.</li>
+  <li>dri/nouveau: Cleanup more references to old FBOs and VBOs.</li>
+  <li>dri/nv10-nv20: Fix texturing in some cases after a base level change.</li>
+  <li>dri/nouveau: Fix software mipmap generation on 1x1 textures.</li>
+  <li>dri/nouveau: Have a smaller amount of larger scratch buffers.</li>
+  <li>dri/nouveau: Remove unnecessary flush.</li>
+  <li>dri/nv10: Use fast Z clears.</li>
+  <li>dri/nouveau: Minor cleanup.</li>
+  <li>dri/nv10: Fake fast Z clears for pre-nv17 cards.</li>
+  <li>dri/nouveau: Initialize tile_flags when allocating a render target.</li>
+  <li>nouveau: Get larger push buffers.</li>
+  <li>dri/nouveau: Force a "slow" Z clear if we're getting a new depth buffer.</li>
+  <li>dri/nv20: Clear with the 3D engine.</li>
+  <li>dri/nouveau: Don't assert(0) on compressed internal formats.</li>
+  <li>dri/nv25: Bind a hierarchical depth buffer.</li>
+  <li>dri/nouveau: Call _mesa_update_state() after framebuffer invalidation.</li>
+  <li>dri/nouveau: Honor the access flags in nouveau_bufferobj_map_range.</li>
+  <li>dri/nouveau: Tell the vbo module we want real hardware BOs.</li>
+  <li>dri/nouveau: Split out the scratch helpers to a separate file.</li>
+  <li>dri/nouveau: Avoid recursion in nouveau_bo_context_reset().</li>
+  <li>dri/nouveau: Use a macro to iterate over the bound vertex attributes.</li>
+  <li>dri/nouveau: Split out array handling to its own file.</li>
+  <li>dri/nouveau: Optimize VBO binding re-emission.</li>
+  <li>dri/nouveau: Keep small DYNAMIC_DRAW vertex buffers in system ram.</li>
+  <li>dri/nouveau: Pipeline glTexSubImage texture transfers.</li>
+  <li>dri/nouveau: Fix type promotion issue on 32bit platforms.</li>
+  <li>dri/nouveau: Validate the framebuffer state on read buffer changes.</li>
+  <li>dri/nouveau: Re-emit the BO state when coming back from a software fallback.</li>
+  <li>meta: Don't leak alpha function/reference value changes.</li>
+  <li>meta: Fix incorrect rendering of the bitmap alpha component.</li>
+  <li>vbo: Avoid unnecessary copy to/from current in vertex format upgrade.</li>
+  <li>meta: Don't try to disable cube maps if the driver doesn't expose the extension.</li>
+  <li>meta: Handle bitmaps with alpha test enabled.</li>
+  <li>dri/nouveau: Split hardware/software TNL instantiation more cleanly.</li>
+  <li>dri/nouveau: Fix typo.</li>
+  <li>dri/nouveau: Kill a bunch of ternary operators.</li>
+</ul></p>
+
+<p>Fredrik Höglund (2):
+<ul>
+  <li>r600g: Fix texture sampling with swizzled coords</li>
+  <li>r600g: fix pow(0, 0) evaluating to NaN</li>
+</ul></p>
+
+<p>Guillermo S. Romero (1):
+<ul>
+  <li>r300g: Do not use buf param before checking for NULL.</li>
+</ul></p>
+
+<p>Henri Verbeet (19):
+<ul>
+  <li>r600g: Flush upload buffers before draws instead of before flushes.</li>
+  <li>r600g: Check for other references before checking for existing mappings in radeon_bo_pb_map_internal().</li>
+  <li>r600g: Remove a redundant flush in r600_texture_transfer_map().</li>
+  <li>r600g: Buffer object maps imply a wait.</li>
+  <li>r600g: Respect PB_USAGE_UNSYNCHRONIZED in radeon_bo_pb_map_internal().</li>
+  <li>Revert "r600g: Flush upload buffers before draws instead of before flushes."</li>
+  <li>r600g: fix exports_ps to export a number not a mask.</li>
+  <li>r600g: Mention AMD in the renderer string.</li>
+  <li>r600g: Cleanup the fenced_bo list in r600_context_fini().</li>
+  <li>r600g: Evergreen has two extra frac_bits for the sampler LOD state.</li>
+  <li>r600: Evergreen has two extra frac_bits for the sampler LOD state.</li>
+  <li>r600g: Add PIPE_FORMAT_L8A8_UNORM for Evergreen as well.</li>
+  <li>r600g: Swizzle vertex data only once.</li>
+  <li>r600g: Synchronize supported color formats between Evergreen and r600/r700.</li>
+  <li>r600g: Fix the PIPE_FORMAT_L8A8_UNORM color swaps.</li>
+  <li>r600g: Fix the PIPE_FORMAT_A8_UNORM color swap for Evergreen as well.</li>
+  <li>r600g: Cleanup block bo references in r600_context_fini().</li>
+  <li>r600g: Cleanup fetch shader resources in r600_pipe_shader_destroy().</li>
+  <li>st/mesa: Handle wrapped depth buffers in st_copy_texsubimage().</li>
+</ul></p>
+
+<p>Hui Qi Tay (10):
+<ul>
+  <li>llvmpipe: minor changes in llvm coefficient calcs</li>
+  <li>draw: cliptest and viewport done in a single loop in vertex shader</li>
+  <li>draw: added viewport and cliptest flags</li>
+  <li>draw: sanitize llvm variant key</li>
+  <li>draw: corrections for w coordinate</li>
+  <li>draw: corrections to allow for different cliptest cases</li>
+  <li>llvmpipe: Moved draw pipeline twoside function to llvm setup code</li>
+  <li>llvmpipe: added llvm offset setup code</li>
+  <li>llvmpipe: clean up polygon offset function in lp setup code</li>
+  <li>llvmpipe: fix such that offset/twoside function only does in-place modification</li>
+</ul></p>
+
+<p>Ian Romanick (102):
+<ul>
+  <li>glsl2: Refactor testing for whether a deref is of a matrix or array</li>
+  <li>glsl2: Add flags to enable variable index lowering</li>
+  <li>glsl: Add doxygen comments</li>
+  <li>EGL DRI2: Silence piles of 'unused variable' warnings</li>
+  <li>EGL DRI2: Silence 'missing initializer' warnings</li>
+  <li>egl_glx: Silence piles of 'unused variable' warnings</li>
+  <li>egl: Fix several 'comparison between signed and unsigned integer' warnings</li>
+  <li>dri: Ensure that DRI driver cpp files are in tarballs</li>
+  <li>mesa: Force GL_ARB_copy_buffer to always be enabled</li>
+  <li>mesa: Force GL_SGIS_generate_mipmap to always be enabled</li>
+  <li>Remove GL_MESA_packed_depth_stencil</li>
+  <li>Remove GL_EXT_cull_vertex</li>
+  <li>Regenerate files changed by previous commit</li>
+  <li>Remove unnescessary initializations of UpdateTexturePalette</li>
+  <li>ARB_texture_rg: Add GLX protocol support</li>
+  <li>ARB_texture_rg: Correct some errors in RED / RG internal format handling</li>
+  <li>ARB_texture_rg: Add GL_TEXTURE_{RED,GREEN}_SIZE query support</li>
+  <li>ARB_texture_rg: Add GL_RED as a valid GL_DEPTH_TEXTURE_MODE</li>
+  <li>ARB_texture_rg: Handle RED and RG the same as RGB for tex env</li>
+  <li>ARB_texture_rg: Add R8, R16, RG88, and RG1616 internal formats</li>
+  <li>ARB_texture_rg: Allow RED and RG textures as FBO color buffer attachments</li>
+  <li>mesa: Enable GL_ARB_texture_rg in software paths</li>
+  <li>i965: Enable GL_ARB_texture_rg</li>
+  <li>mesa: Add ARB_texture_compression_rgtc as an alias for EXT_texture_compression_rgtc</li>
+  <li>ARB_texture_rg: Add GL_COMPRESSED_{RED,RG} cases in _mesa_is_color_format</li>
+  <li>mesa: Fix misplaced #endif</li>
+  <li>mesa: Trivial correction to comment</li>
+  <li>rgtc: Detect RGTC formats as color formats and as compressed formats</li>
+  <li>docs: Add list of bugs fixed in 7.9</li>
+  <li>docs: Import 7.9 release notes from 7.9 branch.</li>
+  <li>docs: Import 7.8.x release notes from 7.8 branch.</li>
+  <li>docs: download.html does not need to be updated for each release</li>
+  <li>docs: Update mailing lines from sf.net to freedesktop.org</li>
+  <li>docs: Import news updates from 7.9 branch</li>
+  <li>docs: added news item for 7.9 release</li>
+  <li>glsl: Fail linking if assign_attribute_locations fails</li>
+  <li>glsl: Refactor 'layout' grammar to match GLSL 1.60 spec grammar</li>
+  <li>glsl: Slight refactor of error / warning checking for ARB_fcc layout</li>
+  <li>glsl: Clear type_qualifier using memset</li>
+  <li>glsl: Wrap ast_type_qualifier contents in a struct in a union</li>
+  <li>glsl: Regenerate files modified by previous commits</li>
+  <li>glcpp: Add the define for ARB_explicit_attrib_location when present</li>
+  <li>glcpp: Regenerate files changes by previous commit</li>
+  <li>glsl: Add parser support for GL_ARB_explicit_attrib_location layouts</li>
+  <li>glsl: Regenerate files changes by previous commit</li>
+  <li>glsl: Track explicit location in AST to IR translation</li>
+  <li>glsl: Add linker support for explicit attribute locations</li>
+  <li>main: Enable GL_ARB_explicit_attrib_location for swrast</li>
+  <li>intel: Enable GL_ARB_explicit_attrib_location</li>
+  <li>glsl: Remove const decoration from inlined function parameters</li>
+  <li>docs: skeleton for 7.10 release notes</li>
+  <li>docs: Update status of GL 3.x related extensions</li>
+  <li>mesa: Validate assembly shaders when GLSL shaders are used</li>
+  <li>glsl: Fix incorrect assertion</li>
+  <li>linker: Reject shaders that have unresolved function calls</li>
+  <li>mesa: Silence unused variable warning</li>
+  <li>mesa: Refactor validation of shader targets</li>
+  <li>mesa: Clean up two 'comparison between signed and unsigned' warnings</li>
+  <li>mesa: Clean up various 'unused parameter' warnings in shaderapi</li>
+  <li>glsl: Slightly change the semantic of _LinkedShaders</li>
+  <li>linker: Trivial indention fix</li>
+  <li>i965: Fix indentation after commit 3322fbaf</li>
+  <li>linker: Improve handling of unread/unwritten shader inputs/outputs</li>
+  <li>glapi: Add GL_EXT_separate_shader_objects</li>
+  <li>glapi: Commit files changed by previous commit</li>
+  <li>mesa: Add infrastructure to track GL_EXT_separate_shader_objects</li>
+  <li>mesa: Skeletal support for GL_EXT_separate_shader_objects</li>
+  <li>mesa: Add display list support for GL_EXT_separate_shader_objects functions</li>
+  <li>mesa: Track an ActiveProgram distinct from CurrentProgram</li>
+  <li>Track separate programs for each stage</li>
+  <li>swrast: Enable GL_EXT_separate_shader_objects in software paths</li>
+  <li>intel: Enable GL_EXT_separate_shader_objects in Intel drivers</li>
+  <li>docs: add GL_EXT_separate_shader_objects to release notes</li>
+  <li>glsl: Fix incorrect gl_type of sampler2DArray and sampler1DArrayShadow</li>
+  <li>ir_to_mesa: Refactor code for emitting DP instructions</li>
+  <li>mesa: Allow query of MAX_SAMPLES with EXT_framebuffer_multisample</li>
+  <li>glsl: Refactor is_vec_{zero,one} to be methods of ir_constant</li>
+  <li>glsl: Simplify generation of swizzle for vector constructors</li>
+  <li>glsl: Make is_zero and is_one virtual methods of ir_rvalue</li>
+  <li>ir_to_mesa: Generate smarter code for some conditional moves</li>
+  <li>glsl: Add ir_unop_sin_reduced and ir_unop_cos_reduced</li>
+  <li>glsl: Eliminate assumptions about size of ir_expression::operands</li>
+  <li>glsl: Add ir_rvalue::is_negative_one predicate</li>
+  <li>glsl: Add unary ir_expression constructor</li>
+  <li>glsl: Add ir_quadop_vector expression</li>
+  <li>glsl: Fix matrix constructors with vector parameters</li>
+  <li>i915: Disallow alpha, red, RG, and sRGB as render targets</li>
+  <li>glsl: Use M_LOG2E constant instead of calling log2</li>
+  <li>glsl: Lower ir_binop_pow to a sequence of EXP2 and LOG2</li>
+  <li>i915: Request that POW instructions be lowered</li>
+  <li>i915: Correctly generate unconditional KIL instructions</li>
+  <li>glsl: Ensure that equality comparisons don't return a NULL IR tree</li>
+  <li>i965: Correctly emit constants for aggregate types (array, matrix, struct)</li>
+  <li>glsl: Inherrit type of declared variable from initializer</li>
+  <li>linker: Ensure that unsized arrays have a size after linking</li>
+  <li>linker: Fix regressions caused by previous commit</li>
+  <li>glsl: Inherrit type of declared variable from initializer after processing assignment</li>
+  <li>linker: Allow built-in arrays to have different sizes between shader stages</li>
+  <li>ir_to_mesa: Don't generate swizzles for record derefs of non-scalar/vectors</li>
+  <li>Refresh autogenerated file builtin_function.cpp.</li>
+  <li>glsl: Allow less restrictive uses of sampler array indexing in GLSL &lt= 1.20</li>
+  <li>docs: Import 7.9.1 release notes from 7.9 branch</li>
+</ul></p>
+
+<p>Jakob Bornecrantz (27):
+<ul>
+  <li>rbug: Cast opcode to corrent int size</li>
+  <li>rbug: Add function to get opcode name string</li>
+  <li>scons: Link against talloc in the Gallium DRI drivers</li>
+  <li>i915g: Link with wrapper sw winsys with scons</li>
+  <li>tgsi: Actually care what check_soa_dependencies says</li>
+  <li>tgsi: Fix missing test before check</li>
+  <li>llvmpipe: Move makefile include to before targets</li>
+  <li>wrapper: Fix spelling</li>
+  <li>wrapper: Add a way to dewrap a pipe screen without destroying it</li>
+  <li>egl: Remove unnecessary headers</li>
+  <li>target-helpers: Remove per target software wrapper check</li>
+  <li>graw: Tidy graw xlib scons file a bit</li>
+  <li>scons: Remove old pipebuffer SConscript</li>
+  <li>scons: Detabify</li>
+  <li>scons: Check for pkg-config before trying to use it</li>
+  <li>scons: Check for libdrm_[intel|radeon] as well</li>
+  <li>scons: Move dependancy checks to the main gallium scons file</li>
+  <li>scons: Unify state tracker SConscripts</li>
+  <li>galahad: Correct the name of the scons library</li>
+  <li>graw: Use inline sw helper instead of roll your own loader</li>
+  <li>libgl-xlib: Use sw helper instead of roll your own</li>
+  <li>libgl-xlib: Use inline debug helper instead of non-inline version</li>
+  <li>graw: Use inline debug helper instead of non-inline version</li>
+  <li>gallium: Remove redundant sw and debug target helpers</li>
+  <li>i915g: Improve debug printing for textures</li>
+  <li>i915g: Make sure that new vbo gets updated</li>
+  <li>st/mesa: Unbind all constant buffers</li>
+</ul></p>
+
+<p>Jerome Glisse (75):
+<ul>
+  <li>r600g: alternative command stream building from context</li>
+  <li>r600g: move chip class to radeon common structure</li>
+  <li>r600g: use pipe context for flushing inside map</li>
+  <li>r600g: add back reference check when mapping buffer</li>
+  <li>r600g: directly allocate bo for user buffer</li>
+  <li>r600g: fix multi buffer rendering</li>
+  <li>r600g: occlusion query for new design</li>
+  <li>r600g: flush color buffer after draw command</li>
+  <li>r600g: disable shader rebuild optimization &amp; account cb flush packet</li>
+  <li>r600g: fix multiple occlusion query on same id</li>
+  <li>r600g: initial evergreen support in new path</li>
+  <li>r600g: fix typo in evergreen define (resource are in x range)</li>
+  <li>r600g: move use_mem_constants flags for new designs structure alignment</li>
+  <li>r600g: evergreen fix for new design</li>
+  <li>r600g: fix compilation after change to evergreend.h</li>
+  <li>r600g: fixup some evergreen register definitions</li>
+  <li>r600g: fix evergreen new path</li>
+  <li>r600g: fix reg definition</li>
+  <li>r600g: fix evergreen new path</li>
+  <li>r600g: bring over fix from old path to new path</li>
+  <li>r600g: fix vertex resource &amp; polygon offset</li>
+  <li>r600g: disable early cull optimization when occlusion query running</li>
+  <li>r600g: move around variables to share depth uncompression code</li>
+  <li>r600g: use depth decompression in new path</li>
+  <li>r600g: fix index buffer drawing</li>
+  <li>r600g: build packet header once</li>
+  <li>r600g: fix pointsprite &amp; resource unbinding</li>
+  <li>r600g: fix routing btw vertex &amp; pixel shader</li>
+  <li>r600g: fix occlusion query after change to block structure</li>
+  <li>r600g: use ptr for blit depth uncompress function</li>
+  <li>r600g: fix remaining piglit issue in new design</li>
+  <li>r600g: switch to new design</li>
+  <li>r600g: suspend/resume occlusion query around clear/copy</li>
+  <li>r600g: avoid rebuilding the vertex shader if no change to input format</li>
+  <li>r600g: use a hash table instead of group</li>
+  <li>r600g: delete old path</li>
+  <li>r600g: cleanup</li>
+  <li>r600g: more cleanup</li>
+  <li>r600g: use constant buffer instead of register for constant</li>
+  <li>r600g: fix constant &amp; literal src splitting, also fix mplayer gl2 shader</li>
+  <li>evergreeng: avoid overlapping border color btw VS &amp; PS</li>
+  <li>r600g: indentation fixes</li>
+  <li>r600g: rename radeon_ws_bo to r600_bo</li>
+  <li>r600g: allow r600_bo to be a sub allocation of a big bo</li>
+  <li>r600g: use r600_bo for relocation argument, simplify code</li>
+  <li>r600g: rename radeon_ws_bo to r600_bo</li>
+  <li>r600g: remove dead label &amp; fix indentation</li>
+  <li>r600g: store reloc information in bo structure</li>
+  <li>r600g: improve bo flushing</li>
+  <li>r600g: simplify block relocation</li>
+  <li>r600g: userspace fence to avoid kernel call for testing bo busy status</li>
+  <li>r600g: avoid segfault due to unintialized list pointer</li>
+  <li>r600g: fix dirty state handling</li>
+  <li>r600g: allow driver to work without submitting cmd to GPU</li>
+  <li>gallium/noop: no operation gallium driver</li>
+  <li>r600g: code cleanup (indent, trailing space, empty line ...)</li>
+  <li>r600g: fix occlusion query on evergreen (avoid lockup)</li>
+  <li>r600g: add fetch shader capabilities</li>
+  <li>r600g: dump raw shader output for debugging</li>
+  <li>r600g: update polygon offset only when rasterizer or zbuffer change</li>
+  <li>r600g: indentation fix</li>
+  <li>r600g: more indentation fix + warning silencing + dead code removal</li>
+  <li>r600g: build fetch shader from vertex elements</li>
+  <li>r600g: avoid useless shader rebuild at draw call</li>
+  <li>r600g: remove useless flush map</li>
+  <li>r600g: remove dead code</li>
+  <li>r600g: fix userspace fence against lastest kernel</li>
+  <li>r600g: avoid using pb* helper we are loosing previous cpu cycle with it</li>
+  <li>r600g: specialized upload manager</li>
+  <li>r600g: indentation cleanup</li>
+  <li>r600g: fix bo size when creating bo from handle</li>
+  <li>r600g: fix segfault when translating vertex buffer</li>
+  <li>r600g: need to reference upload buffer as the might still live accross flush</li>
+  <li>r600g: properly unset vertex buffer</li>
+  <li>r600g: avoid segfault</li>
+</ul></p>
+
+<p>Joakim Sindholt (3):
+<ul>
+  <li>util/u_blitter: fix leak</li>
+  <li>radeong: fix leaks</li>
+  <li>r300g: silence guard band cap errors</li>
+</ul></p>
+
+<p>Johann Rudloff (3):
+<ul>
+  <li>radeon: Implement EGL_MESA_no_surface_extension</li>
+  <li>radeon: Implement __DRI_IMAGE and EGL_MESA_image_drm</li>
+  <li>radeon: Implement GL_OES_EGL_image</li>
+</ul></p>
+
+<p>John Doe (3):
+<ul>
+  <li>r600g: misc cleanup</li>
+  <li>r600g: don't double count dirty block</li>
+  <li>r600g: keep a mapping around for each bo</li>
+</ul></p>
+
+<p>Jon TURNEY (1):
+<ul>
+  <li>Ensure -L$(TOP)/$(LIB_DIR) appears in link line before any -L in $LDFLAGS</li>
+</ul></p>
+
+<p>José Fonseca (128):
+<ul>
+  <li>gallivm: Fix address register swizzle.</li>
+  <li>gallivm: Start collecting bitwise arithmetic helpers in a new module.</li>
+  <li>gallivm: Clamp indirect register indices to file_max.</li>
+  <li>util: linearized sRGB values don't fit into 8bits</li>
+  <li>llvmpipe: Default to no threading on single processor systems.</li>
+  <li>tgsi: Don't ignore indirect registers in tgsi_check_soa_dependencies</li>
+  <li>llvmpipe: Describe how to profile llvmpipe.</li>
+  <li>llvmpipe: When failing free fs shader too.</li>
+  <li>util: Flush stdout on util_format.</li>
+  <li>gallivm: Add unorm support to lp_build_lerp()</li>
+  <li>llvmpipe: Special case complementary and identify blend factors in SoA.</li>
+  <li>llvmpipe: Make rgb/alpha bland func/factors match, when there is no alpha.</li>
+  <li>draw: Prevent clipped vertices overflow.</li>
+  <li>draw: Fullfil the new min_lod/max_lod/lod_bias/border_color dynamic state</li>
+  <li>gallivm: Fetch the lod from the dynamic state when min_lod == max_lod.</li>
+  <li>gallivm: Remove dead experimental code.</li>
+  <li>llvmpipe: Decouple sampler view and sampler state updates.</li>
+  <li>scons: New build= option, with support for checked builds.</li>
+  <li>scons: New build= option, with support for checked builds.</li>
+  <li>trace: Fix set_index_buffer and draw_vbo tracing.</li>
+  <li>python/retrace: Handle set_index_buffer and draw_vbo.</li>
+  <li>gallivm: Use SSE4.1's ROUNDSS/ROUNDSD for scalar rounding.</li>
+  <li>gallivm: More comprehensive border usage logic.</li>
+  <li>retrace: Handle clear_render_target and clear_depth_stencil.</li>
+  <li>llvmpipe: Dump a few missing shader key flags.</li>
+  <li>llvmpipe: Fix perspective interpolation for point sprites.</li>
+  <li>llvmpipe: Fix sprite coord perspective interpolation of Q.</li>
+  <li>gallivm: Take the type signedness in consideration in round/ceil/floor.</li>
+  <li>gallivm: Use a faster (and less accurate) log2 in lod computation.</li>
+  <li>gallivm: Fast implementation of iround(log2(x))</li>
+  <li>gallivm: Combined ifloor &amp; fract helper.</li>
+  <li>gallivm: Only apply min/max_lod when necessary.</li>
+  <li>gallivm: Compute lod as integer whenever possible.</li>
+  <li>util: Cleanup util_pack_z_stencil and friends.</li>
+  <li>llvmpipe: Cleanup depth-stencil clears.</li>
+  <li>gallivm: Vectorize the rho computation.</li>
+  <li>gallivm: Do not do mipfiltering when magnifying.</li>
+  <li>gallivm: Simplify lp_build_mipmap_level_sizes' interface.</li>
+  <li>gallivm: Don't compute the second mipmap level when frac(lod) == 0</li>
+  <li>gallivm: Use lp_build_ifloor_fract for lod computation.</li>
+  <li>gallivm: Clamp mipmap level and zero mip weight simultaneously.</li>
+  <li>gallivm: Fix copy'n'paste typo in previous commit.</li>
+  <li>gallivm: Implement brilinear filtering.</li>
+  <li>gallivm: Use the wrappers for SSE pack intrinsics.</li>
+  <li>gallivm: Avoid control flow for two-sided stencil test.</li>
+  <li>gallivm: Warn when doing inefficient integer comparisons.</li>
+  <li>gallivm: Move into the as much of the second level code as possible.</li>
+  <li>llvmpipe: First minify the texture size, then broadcast.</li>
+  <li>gallivm: Help for combined extraction and broadcasting.</li>
+  <li>gallivm: Do size computations simultanously for all dimensions (AoS).</li>
+  <li>llvmpipe: Prevent z &gt; 1.0</li>
+  <li>llvmpipe: Fix MSVC build. Enable the new SSE2 code on non SSE3 systems.</li>
+  <li>gallivm: Handle code have ret correctly.</li>
+  <li>util: Defined M_SQRT2 when not available.</li>
+  <li>gallivm: Less code duplication in log computation.</li>
+  <li>gallivm: Special bri-linear computation path for unmodified rho.</li>
+  <li>gallivm: Don't generate Phis for execution mask.</li>
+  <li>gallivm: Use varilables instead of Phis for cubemap selection.</li>
+  <li>gallivm: Remove support for Phi generation.</li>
+  <li>gallivm: Factor out the SI-&gt;FP texture size conversion for SoA path too</li>
+  <li>gallivm: Simplify if/then/else implementation.</li>
+  <li>gallivm: Cleanup the rest of the flow module.</li>
+  <li>gallivm: Fix a long standing bug with nested if-then-else emission.</li>
+  <li>gallivm: Allow to disable bri-linear filtering with GALLIVM_DEBUG=no_brilinear runtime option</li>
+  <li>gallivm: Use variables instead of Phis in loops.</li>
+  <li>gallivm: Pass texture coords derivates as scalars.</li>
+  <li>llvmpipe: Remove outdated comment about stencil testing.</li>
+  <li>gallivm: Eliminate unsigned integer arithmetic from texture coordinates.</li>
+  <li>gallium: Define C99 restrict keyword where absent.</li>
+  <li>tgsi: Export some names for some tgsi enums.</li>
+  <li>gallivm: More detailed analysis of tgsi shaders.</li>
+  <li>llvmpipe: Use lp_tgsi_info.</li>
+  <li>llvmpipe: Do not dispose the execution engine.</li>
+  <li>llvmpipe: Fix MSVC build.</li>
+  <li>llmvpipe: improve mm_mullo_epi32</li>
+  <li>gallivm: Name anonymous union.</li>
+  <li>llvmpipe: Unbreak Z32_FLOAT.</li>
+  <li>gallivm: More accurate float -&gt; 24bit &amp; 32bit unorm conversion.</li>
+  <li>llvmpipe: Generalize the x8z24 fast path to all depth formats.</li>
+  <li>llvmpipe: Fix depth-stencil regression.</li>
+  <li>llvmpipe: Ensure z_shift and z_width is initialized.</li>
+  <li>gallivm: Fix SoA cubemap derivative computation.</li>
+  <li>llvmpipe: Fix bad refactoring.</li>
+  <li>llvmpipe: Initialize bld ctx via lp_build_context_init instead of ad-hoc and broken code.</li>
+  <li>gallivm: Comment lp_build_insert_new_block().</li>
+  <li>gallivm: Add a note about SSE4.1's nearest mode rounding.</li>
+  <li>llvmpipe: Don't test rounding of x.5 numbers.</li>
+  <li>gallium: Avoid using __doc__ in python scripts.</li>
+  <li>gallivm: always enable LLVMAddInstructionCombiningPass()</li>
+  <li>gallivm: Remove the EMMS opcodes.</li>
+  <li>mesa: Fix windows build (uint -&gt; GLuint).</li>
+  <li>scons: Revamp how to specify targets to build.</li>
+  <li>scons: Fix MinGW cross-compilation.</li>
+  <li>scons: Some pipe drivers are not portable for MSVC</li>
+  <li>scons: Restore x11 tool behavior for backwards compatability.</li>
+  <li>scons: Disable python state tracker when swig is not present.</li>
+  <li>r600g: List recently added files in SConscript.</li>
+  <li>scons: Add aliases for several pipe drivers.</li>
+  <li>scons: i915 can't build on MSVC either.</li>
+  <li>scons: Propagate installation targets.</li>
+  <li>xorg/vmwgfx: Add missing source file to SConscript.</li>
+  <li>st/xorg: Add missing n to error message.</li>
+  <li>st/xorg: Detect libkms with scons too.</li>
+  <li>xorg/vmwgfx: Link libkms when available.</li>
+  <li>r600g: Swap the util_blitter_destroy call order.</li>
+  <li>gallivm: Allocate TEMP/OUT arrays only once.</li>
+  <li>libgl-gdi: Allow to pick softpipe/llvmpipe on runtime.</li>
+  <li>scons: Use inline wrap helpers more consistently.</li>
+  <li>svga: Use consistent hexadecimal representation on debug output.</li>
+  <li>scons: Alias for svga</li>
+  <li>wgl: Stub WGL_ARB_pbuffer support.</li>
+  <li>wgl: More complete WGL_ARB_pbuffer support.</li>
+  <li>svga: Silence debug printf.</li>
+  <li>scons: Move MSVS_VERSION option to common module.</li>
+  <li>vega: Remove extraneous ;</li>
+  <li>retrace: Some fixes.</li>
+  <li>util: C++ safe.</li>
+  <li>wgl: Fix double free. Remove dead code.</li>
+  <li>util: Plug leaks in util_destroy_gen_mipmap.</li>
+  <li>util: __builtin_frame_address() doesn't work on mingw.</li>
+  <li>util: Don't try to use imagehlp on mingw.</li>
+  <li>wgl: Unreference the current framebuffer after the make_current call.</li>
+  <li>WIN32_THREADS -&gt; WIN32</li>
+  <li>mapi: Hack to avoid vgCreateFont being generated as vgCreateFontA.</li>
+  <li>wgl: Fix visual's buffer_mask configuration.</li>
+  <li>mesa: Temporary hack to prevent stack overflow on windows</li>
+  <li>mesa: Bump the number of bits in the register index.</li>
+  <li>llvmpipe: Plug fence leaks.</li>
+</ul></p>
+
+<p>Julien Cristau (1):
+<ul>
+  <li>Makefile: don't include the same files twice in the tarball</li>
+</ul></p>
+
+<p>Keith Whitwell (89):
+<ul>
+  <li>llvmpipe: brackets around macro arg</li>
+  <li>llvmpipe: remove duplicate code</li>
+  <li>llvmpipe: return zero from floor_pot(zero)</li>
+  <li>gallivm: make lp_build_sample_nop public</li>
+  <li>llvmpipe: add LP_PERF flag to disable various aspects of rasterization</li>
+  <li>llvmpipe: add DEBUG_FS to dump variant information</li>
+  <li>llvmpipe: use llvm for attribute interpolant calculation</li>
+  <li>graw: add frag-face shader</li>
+  <li>llvmpipe: fix flatshading in new line code</li>
+  <li>draw: don't apply flatshading to clipped tris with &lt3 verts</li>
+  <li>llvmpipe: handle FACING interpolants in line and point setup</li>
+  <li>llvmpipe: handle up to 8 planes in triangle binner</li>
+  <li>llvmpipe: make debug_fs_variant respect variant-&gt;nr_samplers</li>
+  <li>gallivm: don't apply zero lod_bias</li>
+  <li>llvmpipe: fail gracefully on oom in scene creation</li>
+  <li>llvmpipe: avoid overflow in triangle culling</li>
+  <li>gallivm: special case conversion 4x4f to 1x16ub</li>
+  <li>gallivm: round rather than truncate in new 4x4f-&gt;1x16ub conversion path</li>
+  <li>llvmpipe: clean up setup_tri a little</li>
+  <li>llvmpipe: add rast_tri_4_16 for small lines and points</li>
+  <li>llvmpipe: fix off-by-one in tri_16</li>
+  <li>llvmpipe: defer attribute interpolation until after mask and ztest</li>
+  <li>llvmpipe: use alloca for fs color outputs</li>
+  <li>llvmpipe: store zero into all alloca'd values</li>
+  <li>llvmpipe: dump fragment shader ir and asm when LP_DEBUG=fs</li>
+  <li>gallivm: specialized x8z24 depthtest path</li>
+  <li>gallivm: prefer blendvb for integer arguments</li>
+  <li>gallivm: simpler uint8-&gt;float conversions</li>
+  <li>llvmpipe: try to be sensible about whether to branch after mask updates</li>
+  <li>llvmpipe: clean up shader pre/postamble, try to catch more early-z</li>
+  <li>llvmpipe: simplified SSE2 swz/unswz routines</li>
+  <li>llvmpipe: try to do more of rast_tri_3_16 with intrinsics</li>
+  <li>llvmpipe: add debug helpers for epi32 etc</li>
+  <li>llvmpipe: try to keep plane c values small</li>
+  <li>llvmpipe: fix typo in last commit</li>
+  <li>gallium: move sse intrinsics debug helpers to u_sse.h</li>
+  <li>r600g: add missing file to sconscript</li>
+  <li>gallivm: don't branch on KILLs near end of shader</li>
+  <li>Revert "llvmpipe: try to keep plane c values small"</li>
+  <li>llvmpipe: make sure intrinsics code is guarded with PIPE_ARCH_SSE</li>
+  <li>llvmpipe: don't try to emit non-existent color outputs</li>
+  <li>r600/drm: fix segfaults in winsys create failure path</li>
+  <li>r600g: emit hardware linewidth</li>
+  <li>r600g: handle absolute modifier in shader translator</li>
+  <li>llvmpipe: reintroduce SET_STATE binner command</li>
+  <li>llvmpipe: don't pass frontfacing as a float</li>
+  <li>llvmpipe: slightly shrink the size of a binned triangle</li>
+  <li>llvmpipe: don't store plane.ei value in binned data</li>
+  <li>gallium: move some intrinsics helpers to u_sse.h</li>
+  <li>llvmpipe: do plane calculations with intrinsics</li>
+  <li>llvmpipe: use aligned loads/stores for plane values</li>
+  <li>llvmpipe: fix non-sse build after recent changes</li>
+  <li>llvmpipe: check shader outputs are non-null before using</li>
+  <li>llvmpipe: validate color outputs against key-&gt;nr_cbufs</li>
+  <li>llvmpipe: clean up fields in draw_llvm_variant_key</li>
+  <li>llvmpipe: remove setup fallback path</li>
+  <li>llvmpipe: fail cleanly on malloc failure in lp_setup_alloc_triangle</li>
+  <li>Merge remote branch 'origin/master' into lp-setup-llvm</li>
+  <li>llvmpipe: remove unused file</li>
+  <li>llvmpipe: remove unused arg from jit_setup_tri function</li>
+  <li>Merge branch 'llvm-cliptest-viewport'</li>
+  <li>draw: make sure viewport gets updated in draw llvm shader</li>
+  <li>llvmpipe: turn off draw offset/twoside when we can handle it</li>
+  <li>llvmpipe: avoid generating tri_16 for tris which extend past tile bounds</li>
+  <li>llvmpipe: guard against NULL task-&gt;query pointer</li>
+  <li>st/mesa: unbind constant buffer when not in use</li>
+  <li>r600g: propagate usage flags in texture transfers</li>
+  <li>r600g: propogate resource usage flags to winsys, use to choose bo domains</li>
+  <li>r600g: use a buffer in GTT as intermediate on texture up and downloads</li>
+  <li>r600g: remove unused flink, domain fields from r600_resource</li>
+  <li>r600g: set hardware pixel centers according to gl_rasterization_rules</li>
+  <li>evergreeng: protect against null constant buffers</li>
+  <li>r600g: don't call debug_get_bool_option for tiling more than once</li>
+  <li>evergreeng: respect linewidth state, use integer widths only</li>
+  <li>evergreeng: set hardware pixelcenters according to gl_rasterization_rules</li>
+  <li>r600g: avoid recursion with staged uploads</li>
+  <li>r600g: attempt to turn on DXTn formats</li>
+  <li>r600g: translate ARR instruction</li>
+  <li>r600: fix my pessimism about PIPE_TRANSFER_x flags</li>
+  <li>ws/r600: match bo_busy shared/fence logic in bo_wait</li>
+  <li>r600g: guard experimental s3tc code with R600_ENABLE_S3TC</li>
+  <li>r600g: do not try to use staging resource for depth textures</li>
+  <li>r600g: enforce minimum stride on render target texture images</li>
+  <li>llvmpipe: fix up twoside after recent changes</li>
+  <li>llvmpipe: twoside for specular color also</li>
+  <li>Merge branch 'lp-offset-twoside'</li>
+  <li>llvmpipe: raise dirty flag on transfers to bound constbuf</li>
+  <li>llvmpipe: remove misleading debug string</li>
+  <li>llvmpipe: shortcircuit some calls to set_scene_state</li>
+</ul></p>
+
+<p>Kenneth Graunke (94):
+<ul>
+  <li>glsl: Change from has_builtin_signature to has_user_signature.</li>
+  <li>glsl: Don't print blank (function ...) headers for built-ins.</li>
+  <li>glsl: Properly handle nested structure types.</li>
+  <li>glsl/builtins: Fix equal and notEqual builtins.</li>
+  <li>glsl/builtins: Switch comparison functions to just return an expression.</li>
+  <li>glsl: Add comments to clarify the types of comparison binops.</li>
+  <li>glsl: Fix broken handling of ir_binop_equal and ir_binop_nequal.</li>
+  <li>glsl: "Copyright", not "Constantright"</li>
+  <li>i965: Fix incorrect batchbuffer size in gen6 clip state command.</li>
+  <li>i965: Use logical-not when emitting ir_unop_ceil.</li>
+  <li>glsl: Add front-end support for the "trunc" built-in.</li>
+  <li>glsl: Refresh autogenerated file builtin_function.cpp.</li>
+  <li>i965: Use RNDZ for ir_unop_trunc in the new FS.</li>
+  <li>i965: Correctly emit the RNDZ instruction.</li>
+  <li>i965: Clean up a warning in the old fragment backend.</li>
+  <li>glsl: Add a new ir_unop_round_even opcode for GLSL 1.30's roundEven.</li>
+  <li>glsl: Add front-end support for GLSL 1.30's roundEven built-in.</li>
+  <li>i965: Add support for ir_unop_round_even via the RNDE instruction.</li>
+  <li>glsl: Add support for the 1.30 round() built-in.</li>
+  <li>glsl: Refresh autogenerated file builtin_function.cpp.</li>
+  <li>glsl: Don't return NULL IR for erroneous bit-shift operators.</li>
+  <li>i965: Add missing "break" statement.</li>
+  <li>glsl: Fix copy and paste error in ast_bit_and node creation.</li>
+  <li>glsl: Regenerate parser files.</li>
+  <li>i965: Remove unused variable.</li>
+  <li>glsl: Remove useless ir_shader enumeration value.</li>
+  <li>mesa: Remove FEATURE_ARB_shading_language_120 macro.</li>
+  <li>glcpp: Return NEWLINE token for newlines inside multi-line comments.</li>
+  <li>glcpp: Refresh autogenerated lexer file.</li>
+  <li>glsl: Add support for GLSL 1.30's modf built-in.</li>
+  <li>glsl: Refresh autogenerated file builtin_function.cpp.</li>
+  <li>generate_builtins.py: Output large strings as arrays of characters.</li>
+  <li>Refresh autogenerated file builtin_function.cpp.</li>
+  <li>glsl: Fix constant component count in vector constructor emitting.</li>
+  <li>Fix build on systems where "python" is python 3.</li>
+  <li>i965: Add bit operation support to the fragment shader backend.</li>
+  <li>glsl: Remove unused ARRAY_SIZE macro.</li>
+  <li>glsl/builtins: Rename 'x' to 'y_over_x' in atan(float) implementation.</li>
+  <li>glsl/builtins: Clean up some ugly autogenerated code in atan.</li>
+  <li>Refresh autogenerated file builtin_function.cpp.</li>
+  <li>glsl: Don't print a useless space at the end of an S-Expression list.</li>
+  <li>ir_reader: Return a specific ir_dereference variant.</li>
+  <li>ir_reader: Remove useless error check.</li>
+  <li>ir_reader: Fix some potential NULL pointer dereferences.</li>
+  <li>ir_dead_functions: Actually free dead functions and signatures.</li>
+  <li>glsl: Remove unnecessary "unused variable" warning suppression.</li>
+  <li>glsl: Remove GLSL_TYPE_FUNCTION define.</li>
+  <li>glsl: Convert glsl_type::base_type from #define'd constants to an enum.</li>
+  <li>glsl: Rework reserved word/keyword handling in the lexer.</li>
+  <li>glsl: Add new keywords and reserved words for GLSL 1.30.</li>
+  <li>glsl: Add support for the 'u' and 'U' unsigned integer suffixes.</li>
+  <li>glsl: Refresh autogenerated lexer and parser files.</li>
+  <li>generate_builtins.py: Fix inconsistent use of tabs and spaces warning.</li>
+  <li>glsl: Implement the asinh, acosh, and atanh built-in functions.</li>
+  <li>glsl: Refresh autogenerated file builtin_function.cpp.</li>
+  <li>glsl: Add constant expression handling for asinh, acosh, and atanh.</li>
+  <li>glsl: Remove unused and out of date Makefile.am.</li>
+  <li>glsl: Rename various ir_* files to lower_* and opt_*.</li>
+  <li>glcpp: Define GL_FRAGMENT_PRECISION_HIGH if GLSL version &gt;= 1.30.</li>
+  <li>Refresh autogenerated glcpp parser.</li>
+  <li>glsl: Fix constant expression handling for &lt, &gt;, &lt=, &gt;= on vectors.</li>
+  <li>glsl: Unconditionally define GL_FRAGMENT_PRECISION_HIGH in ES2 shaders.</li>
+  <li>Regenerate glcpp parser.</li>
+  <li>glsl: Reimplement the "cross" built-in without ir_binop_cross.</li>
+  <li>Refresh autogenerated file builtin_function.cpp.</li>
+  <li>glsl: Remove the ir_binop_cross opcode.</li>
+  <li>glsl: Refactor get_num_operands.</li>
+  <li>glsl: Simplify a type check by using type-&gt;is_integer().</li>
+  <li>glsl: Combine many instruction lowering passes into one.</li>
+  <li>mesa: Fix glGet of ES2's GL_MAX_*_VECTORS properties.</li>
+  <li>glsl: Don't inline function prototypes.</li>
+  <li>glsl: Use do_common_optimization in the standalone compiler.</li>
+  <li>glsl: Add a virtual as_discard() method.</li>
+  <li>glsl: Refactor out cloning of function prototypes.</li>
+  <li>glsl: Lazily import built-in function prototypes.</li>
+  <li>glsl: Remove anti-built-in hacks from the print visitor.</li>
+  <li>glsl/linker: Free any IR discarded by optimization passes.</li>
+  <li>glsl: Add an optimization pass to simplify discards.</li>
+  <li>glsl: Add a lowering pass to move discards out of if-statements.</li>
+  <li>glsl: Remove "discard" support from lower_jumps.</li>
+  <li>glsl: Add comments to lower_jumps (from the commit message).</li>
+  <li>ir_print_visitor: Print out constant structure values.</li>
+  <li>glsl: Factor out code which emits a new function into the IR stream.</li>
+  <li>symbol_table: Add support for adding a symbol at top-level/global scope.</li>
+  <li>glsl: Properly add functions during lazy built-in prototype importing.</li>
+  <li>glcpp: Don't emit SPACE tokens in conditional_tokens production.</li>
+  <li>Refresh autogenerated glcpp parser.</li>
+  <li>glsl: Clean up code by adding a new is_break() function.</li>
+  <li>glsl: Consider the "else" branch when looking for loop breaks.</li>
+  <li>Remove OES_compressed_paletted_texture from the ES2 extension list.</li>
+  <li>glsl/builtins: Compute the correct value for smoothstep(vec, vec, vec).</li>
+  <li>glsl: Support if-flattening beyond a given maximum nesting depth.</li>
+  <li>i965: Flatten if-statements beyond depth 16 on pre-gen6.</li>
+  <li>i965: Internally enable GL_NV_blend_square on ES2.</li>
+</ul></p>
+
+<p>Kristian Høgsberg (16):
+<ul>
+  <li>glx: Hold on to drawables if we're just switching to another context</li>
+  <li>intel: Fix GL_ARB_shading_language_120 commit</li>
+  <li>dri2: Make createImageFromName() take a __DRIscreen instead of __DRIcontext</li>
+  <li>glx: Invalidate buffers after binding a drawable</li>
+  <li>dri: Pass the __DRIscreen and the __DRIscreen private back to image lookup</li>
+  <li>glx: Only remove drawables from the hash when we actually delete them</li>
+  <li>gles2: Add GL_EXT_texture_format_BGRA8888 support</li>
+  <li>Get rid of GL/internal/glcore.h</li>
+  <li>gl: Remove unused GLcontextModes fields</li>
+  <li>Rename GLvisual and __GLcontextModes to struct gl_config</li>
+  <li>Drop GLframebuffer typedef and just use struct gl_framebuffer</li>
+  <li>Drop GLcontext typedef and use struct gl_context instead</li>
+  <li>Drop the "neutral" tnl module</li>
+  <li>Only install vtxfmt tables for OpenGL</li>
+  <li>i965: Don't write mrf assignment for pointsize output</li>
+  <li>docs: Fix MESA_drm_image typo</li>
+</ul></p>
+
+<p>Krzysztof Smiechowicz (1):
+<ul>
+  <li>nvfx: Pair os_malloc_aligned() with os_free_aligned().</li>
+</ul></p>
+
+<p>Luca Barbieri (84):
+<ul>
+  <li>auxiliary: fix unintended fallthrough</li>
+  <li>glsl: add pass to lower variable array indexing to conditional assignments</li>
+  <li>auxiliary: fix depth-only and stencil-only clears</li>
+  <li>gallium: avoid the C++ keyword "template" in sw_winsys.h</li>
+  <li>softpipe: make z/s test always pass if no zsbuf, instead of crashing</li>
+  <li>tgsi: add switch/case opcodes to tgsi_opcode_tmp.h</li>
+  <li>softpipe: fix whitespace</li>
+  <li>d3d1x: add new Direct3D 10/11 COM state tracker for Gallium</li>
+  <li>d3d1x: add blob and signature extraction APIs</li>
+  <li>d3d1x: fix compilation with recent Wine versions installed</li>
+  <li>d3d1x: add missing file</li>
+  <li>d3d1x: actually enable and fix blob apis</li>
+  <li>d3d1x: fix build with compilers other than GCC 4.5</li>
+  <li>d3d1x: add template parameters to base class ctor calls for GCC 4.4</li>
+  <li>d3d1x: fix GCC 4.1/4.2 build</li>
+  <li>d3d1x: ignore errors while building docs</li>
+  <li>d3d1x: attempt to fix/workaround bug #30322</li>
+  <li>nvfx: remove gl_PointCoord hack</li>
+  <li>glx: decouple dri2.c and GLX, fixing Gallium EGL and d3d1x build</li>
+  <li>winsys: automatically build sw winsys needed by EGL and d3d1x</li>
+  <li>d3d1x: don't build progs automatically</li>
+  <li>d3d1x: add missing memory barrier</li>
+  <li>d3d1x: link with CXXFLAGS</li>
+  <li>d3d1x: fix cf analysis</li>
+  <li>d3d1x: fix warning</li>
+  <li>d3d1x: fix segfault when hashing</li>
+  <li>d3d1x: destroy native_display on adapter destruction</li>
+  <li>d3d1x: fix GUID declarations</li>
+  <li>d3d1x: redesign the HWND resolver interface</li>
+  <li>d3d1x: fix API name</li>
+  <li>d3d1x: define GUIDs in the normal way</li>
+  <li>d3d1x: add Wine dlls (tri, tex working, but no other testing)</li>
+  <li>d3d1x: properly reference count the backend</li>
+  <li>d3d1x: fix deadlocks on non-recursive mutex</li>
+  <li>d3d1x: bind NULL CSOs before destroying default CSOs on context dtor</li>
+  <li>d3d1x: initialize the mutex</li>
+  <li>d3d1x: autogenerate shader enums and text from def files</li>
+  <li>d3d1x: s/tpf/sm4/g</li>
+  <li>d3d1x: normalize whitespace</li>
+  <li>d3d1x: remove specstrings</li>
+  <li>d3d1x: minifix</li>
+  <li>d3d1x: rename context params</li>
+  <li>d3d11: rename screen params</li>
+  <li>d3d1x: rename params in misc and objects</li>
+  <li>d3d1x: rename parameters in dxgi</li>
+  <li>d3d11: obliterate IDL parameter names</li>
+  <li>d3d1x: remove specstrings.h include</li>
+  <li>d3d1x: flush the pipe context when presenting</li>
+  <li>d3d1x: remove another include specstrings.h</li>
+  <li>d3d1x: flush properly</li>
+  <li>d3d1x: add missing guid.cpp</li>
+  <li>d3d1x: fix build without system EGL/egl.h</li>
+  <li>d3d1x: add autogenerated files as prerequisites, so make builds them</li>
+  <li>d3d1x: obliterate IDL parameter names from d3d10.idl from Wine too</li>
+  <li>d3d1x: add shader dumping</li>
+  <li>d3d1x: add untested support for geometry shader translation</li>
+  <li>d3d1x: don't assert on unsupported resource types</li>
+  <li>d3d1x: fix CheckMultisampleQualityLevels</li>
+  <li>d3d1x: draw to the correct buffer</li>
+  <li>d3d1x: fix linking of dxbc2tgsi</li>
+  <li>nvfx: allow setting NULL constant buffers</li>
+  <li>nvfx: add RGB framebuffer format support in addition to BGR</li>
+  <li>d3d1x: don't crash on drivers not supporting vertex or geometry sampling</li>
+  <li>d3d1x: assert if X visual is not among enumerated visuals</li>
+  <li>d3d1x: stop using GLX in demos, just use the default visual</li>
+  <li>d3d1x: CRLF -&gt; LF in progs</li>
+  <li>mesa: make makedepend an hard requirement</li>
+  <li>gallium: add $(PROGS_DEPS) as dependencies for $(PROGS)</li>
+  <li>d3d1x: fix parallel build</li>
+  <li>d3d1x: add private gitignore file</li>
+  <li>d3d1x: fix progs linking if not all EGL platforms are enabled</li>
+  <li>d3d1x: link progs with CXXFLAGS</li>
+  <li>d3d11: advertise IDXGIDevice1, not just IDXGIDevice</li>
+  <li>d3d11: ignore StructureByteStride</li>
+  <li>d3d1x: link to libdrm for X11 platform too</li>
+  <li>ureg: support centroid interpolation</li>
+  <li>d3d1x: support centroid interpolation</li>
+  <li>d3d1x: properly support specifying MipLevels as 0</li>
+  <li>d3d1x: put proper calling convention in headers, fixes 64-bit builds</li>
+  <li>d3d1x: rework DXGI for occlusion testing and default width/height</li>
+  <li>d3d1x: fix Map</li>
+  <li>d3d11: fix reference counting so devices get freed</li>
+  <li>d3d1x: work around crash in widl</li>
+  <li>glsl: Unroll loops with conditional breaks anywhere (not just the end)</li>
+</ul></p>
+
+<p>Lucas Stach (1):
+<ul>
+  <li>nvfx: fill PIPE_CAP_PRIMITIVE_RESTART and PIPE_CAP_SHADER_STENCIL_EXPORT</li>
+</ul></p>
+
+<p>Marek Olšák (100):
+<ul>
+  <li>r300g: prevent creating multiple winsys BOs for the same handle</li>
+  <li>r300g/swtcl: fix CS overrun</li>
+  <li>st/mesa: fix assertion failure in GetTexImage for cubemaps</li>
+  <li>util: make calling remove_from_list multiple times in a row safe</li>
+  <li>r300g: fixup long-lived BO maps being incorrectly unmapped when flushing</li>
+  <li>r300g: make accessing map_list and buffer_handles thread-safe</li>
+  <li>r300g: fix a copy-paste typo for logging</li>
+  <li>r300g: fix the border color for every format other than PIPE_FORMAT_B8G8R8A8</li>
+  <li>Build r300g by default</li>
+  <li>util: fix util_pack_color for B4G4R4A4</li>
+  <li>r300g: fix macrotiling on R350</li>
+  <li>r300g: code cleanups</li>
+  <li>r300/compiler: fix projective mapping of 2D NPOT textures</li>
+  <li>r300/compiler: do not use copy propagation if SaturateMode is used</li>
+  <li>r300/compiler: fix shadow sampling with swizzled coords</li>
+  <li>r300g: add support for 3D NPOT textures without mipmapping</li>
+  <li>r300g: fix swizzling of texture border color</li>
+  <li>configure.ac: look for libdrm_radeon before building gallium/r300,r600</li>
+  <li>configure.ac: do not build xorg-r300g by default</li>
+  <li>Makefile: ensure Gallium's Makefile.xorg and SConscript.dri are in the tarball</li>
+  <li>r300g: add support for formats beginning with X, like X8R8G8B8</li>
+  <li>r300g: fix conditional rendering in non-wait path</li>
+  <li>r300g: add support for R8G8 colorbuffers</li>
+  <li>r300g: add support for L8A8 colorbuffers</li>
+  <li>update release notes for Gallium</li>
+  <li>r300g: fix microtiling for 16-bits-per-channel formats</li>
+  <li>r300g: do not print get_param errors in non-debug build</li>
+  <li>r300g: say no to PIPE_CAP_STREAM_OUTPUT and PIPE_CAP_PRIMITIVE_RESTART</li>
+  <li>mesa: allow FBO attachments of formats LUMINANCE, LUMINANCE_ALPHA, and INTENSITY</li>
+  <li>r300g: fix texture border for 16-bits-per-channel formats</li>
+  <li>st/mesa: support RGBA16 and use it for RGBA12 as well</li>
+  <li>r300g: add a default channel ordering of texture border for unhandled formats</li>
+  <li>r300g: mention ATI in the renderer string</li>
+  <li>r300g: rename has_hyperz -&gt; can_hyperz</li>
+  <li>r300g: turn magic numbers into names in the hyperz code</li>
+  <li>gallium: add CAPs for indirect addressing and lower it in st/mesa when needed</li>
+  <li>tgsi: fill out CAPs for indirect addressing</li>
+  <li>i915g: fill out CAPs for indirect addressing</li>
+  <li>i965g: fill out CAPs for indirect addressing</li>
+  <li>nv50: fill out CAPs for indirect addressing</li>
+  <li>nvfx: fill out CAPs for indirect addressing</li>
+  <li>r300g: fill out CAPs for indirect addressing</li>
+  <li>r600g: fill out CAPs for indirect addressing</li>
+  <li>svga: fill out CAPs for indirect addressing</li>
+  <li>r300g: fix texture border color for all texture formats</li>
+  <li>r300g: clean up redundancy in draw functions</li>
+  <li>r300g: return shader caps from Draw for SWTCL vertex shaders</li>
+  <li>r300g: remove the hack with OPCODE_RET</li>
+  <li>r300g: print FS inputs uninitialized due to hardware limits to stderr</li>
+  <li>r300g: fix rendering with no vertex elements</li>
+  <li>st/mesa: enable ARB_explicit_attrib_location and EXT_separate_shader_objects</li>
+  <li>docs: add GL 4.1 status</li>
+  <li>gallium: add PIPE_SHADER_CAP_SUBROUTINES</li>
+  <li>st/mesa: set MaxUniformComponents</li>
+  <li>u_blitter: use PIPE_TRANSFER_DISCARD to prevent cpu/gpu stall</li>
+  <li>r300/compiler: fix rc_rewrite_depth_out for it to work with any instruction</li>
+  <li>r300/compiler: remove duplicate function rc_mask_to_swz</li>
+  <li>r300/compiler: add a function for swizzling a mask</li>
+  <li>r300/compiler: move util functions to radeon_compiler_util</li>
+  <li>u_blitter: interpolate clear color using a GENERIC varying instead of COLOR</li>
+  <li>st/mesa: fix texture border color for RED and RG base formats</li>
+  <li>util: rename u_mempool -&gt; u_slab</li>
+  <li>r300g: fix texture border color once again</li>
+  <li>r300/compiler: implement and lower OPCODE_CLAMP</li>
+  <li>ir_to_mesa: Add support for conditional discards.</li>
+  <li>r300g: fix texture swizzling with compressed textures on r400-r500</li>
+  <li>r300g: disable ARB_texture_swizzle if S3TC is enabled on r3xx-only</li>
+  <li>r300g: fix up cubemap texture offset computation</li>
+  <li>r300/compiler: disable the swizzle lowering pass in vertex shaders</li>
+  <li>r300g: fix build</li>
+  <li>r300g: use internal BO handle for add_buffer and write_reloc</li>
+  <li>r300g: implement simple transfer_inline_write for buffers</li>
+  <li>mesa, st/mesa: fix gl_FragCoord with FBOs in Gallium</li>
+  <li>r300g: fix pointer arithmetic with void* in transfer_inline_write</li>
+  <li>r300g: do not remove unused constants if we are not near the limit</li>
+  <li>r300g: add capability bit index_bias_supported</li>
+  <li>r300g: one more r500_index_bias_supported leftover</li>
+  <li>r300g: do not use the index parameter in set_constant_buffer</li>
+  <li>r300g: cleanup winsys</li>
+  <li>r300g: optimize looping over atoms</li>
+  <li>st/mesa: initialize key in st_vp_varient</li>
+  <li>u_blitter: use util_is_format_compatible in the assert</li>
+  <li>r300g: cache packet dwords of 3D_LOAD_VBPNTR in a command buffer if possible</li>
+  <li>r300g: validate buffers only if any of bound buffers is changed</li>
+  <li>r300g: also revalidate the SWTCL vertex buffer after its reallocation</li>
+  <li>r300/compiler: don't terminate regalloc if we surpass max temps limit</li>
+  <li>r300/compiler: add a function to query program stats (alu, tex, temps..)</li>
+  <li>r300/compiler: cleanup rc_run_compiler</li>
+  <li>r300/compiler: do not print pair/tex/presub program stats for vertex shaders</li>
+  <li>r300/compiler: handle DPH and XPD in rc_compute_sources_for_writemask</li>
+  <li>r300/compiler: make lowering passes possibly use up to two less temps</li>
+  <li>r300/compiler: remove at least unused immediates if externals cannot be removed</li>
+  <li>r300/compiler: fix LIT in VS</li>
+  <li>r300/compiler: fix swizzle lowering with a presubtract source operand</li>
+  <li>r300g: fix rendering with a vertex attrib having a zero stride</li>
+  <li>r300g: finally fix the texture corruption on r3xx-r4xx</li>
+  <li>r300g/swtcl: re-enable LLVM</li>
+  <li>r300g: mark vertex arrays as dirty after a buffer_offset change</li>
+  <li>mesa: fix texel store functions for some float formats</li>
+  <li>r300/compiler: disable the rename_regs pass for loops</li>
+</ul></p>
+
+<p>Mario Kleiner (1):
+<ul>
+  <li>mesa/r300classic: Fix dri2Invalidate/radeon_prepare_render for page flipping.</li>
+</ul></p>
+
+<p>Mathias Fröhlich (3):
+<ul>
+  <li>r300g: Avoid returning values in a static array, fixing a potential race</li>
+  <li>r600g: Only compare active vertex elements</li>
+  <li>st/mesa: Set PIPE_TRANSFER_DISCARD for GL_MAP_INVALIDATE_RANGE/BUFFFER_BIT</li>
+</ul></p>
+
+<p>Michal Krol (10):
+<ul>
+  <li>svga: Fix relative addressing translation for pixel shaders.</li>
+  <li>svga: Integer constant register file has a separate namespace.</li>
+  <li>tgsi/exec: Cleanup the remaining arithmetic instructions.</li>
+  <li>tgsi/exec: Get rid of obsolete condition codes.</li>
+  <li>tgsi/build: Reduce interface clutter.</li>
+  <li>graw/gdi: Initial commit.</li>
+  <li>scons: Hook-up graw-gdi target.</li>
+  <li>graw/gdi: Fix window dimensions.</li>
+  <li>os: Open file streams in binary mode.</li>
+  <li>graw: Export graw_save_surface_to_file().</li>
+</ul></p>
+
+<p>Nicolas Kaiser (26):
+<ul>
+  <li>swrast: remove duplicated include</li>
+  <li>egl: remove duplicated include</li>
+  <li>gallium/rtasm: remove duplicated include</li>
+  <li>gallium/util: remove duplicated include</li>
+  <li>gallium/i915: remove duplicated include</li>
+  <li>gallium/llvmpipe: remove duplicated include</li>
+  <li>gallium/softpipe: remove duplicated include</li>
+  <li>gallium/st: remove duplicated includes</li>
+  <li>gallium/winsys: remove duplicated include</li>
+  <li>glx: remove duplicated include</li>
+  <li>dri/common: remove duplicated include</li>
+  <li>dri/i810: remove duplicated include</li>
+  <li>dri/i915: remove duplicated include</li>
+  <li>dri/i965: remove duplicated include</li>
+  <li>dri/intel: remove duplicated include</li>
+  <li>dri/mga: remove duplicated include</li>
+  <li>dri/r128: remove duplicated include</li>
+  <li>dri/r300: remove duplicated include</li>
+  <li>dri/r600: remove duplicated include</li>
+  <li>dri/radeon: remove duplicated includes</li>
+  <li>dri/savage: remove duplicated include</li>
+  <li>main: remove duplicated includes</li>
+  <li>math: remove duplicated includes</li>
+  <li>st: remove duplicated include</li>
+  <li>i965g: use Elements macro instead of manual sizeofs</li>
+  <li>nv50: fix always true conditional in shader optimization</li>
+</ul></p>
+
+<p>Orion Poplawski (1):
+<ul>
+  <li>osmesa: link against libtalloc</li>
+</ul></p>
+
+<p>Owen W. Taylor (1):
+<ul>
+  <li>r600g: Fix location for clip plane registers</li>
+</ul></p>
+
+<p>Peter Clifton (3):
+<ul>
+  <li>intel: Fix emit_linear_blit to use DWORD aligned width blits</li>
+  <li>intel: Add assert check for blitting alignment.</li>
+  <li>meta: Mask Stencil.Clear against stencilMax in _mesa_meta_Clear</li>
+</ul></p>
+
+<p>Robert Hooker (2):
+<ul>
+  <li>intel: Add a new B43 pci id.</li>
+  <li>egl_dri2: Add missing intel chip ids.</li>
+</ul></p>
+
+<p>Roland Scheidegger (16):
+<ul>
+  <li>gallivm: fix copy&amp;paste bug</li>
+  <li>gallivm: don't use URem/UDiv when calculating offsets for blocks</li>
+  <li>gallivm: optimize yuv decoding</li>
+  <li>gallivm: fix trunc/itrunc comment</li>
+  <li>gallivm: faster iround implementation for sse2</li>
+  <li>gallivm: replace sub/floor/ifloor combo with ifloor_fract</li>
+  <li>gallivm: optimize some tex wrap mode calculations a bit</li>
+  <li>gallivm: more linear tex wrap mode calculation simplification</li>
+  <li>gallivm: avoid unnecessary URem in linear wrap repeat case</li>
+  <li>gallivm: optimize soa linear clamp to edge wrap mode a bit</li>
+  <li>gallivm: make use of new iround code in lp_bld_conv.</li>
+  <li>gallivm: fix different handling of [non]normalized coords in linear soa path</li>
+  <li>gallivm: only use lp_build_conv 4x4f -&gt; 1x16 ub fastpath with sse2</li>
+  <li>r200: fix r200 large points</li>
+  <li>mesa: remove unneeded DD_POINT_SIZE and DD_LINE_WIDTH tricaps</li>
+  <li>gallium: support for array textures and related changes</li>
+</ul></p>
+
+<p>Shuang He (1):
+<ul>
+  <li>mesa: allow GLfixed arrays for OpenGL ES 2.0</li>
+</ul></p>
+
+<p>Stephan Schmid (1):
+<ul>
+  <li>r600g: fix relative addressing when splitting constant accesses</li>
+</ul></p>
+
+<p>Thomas Hellstrom (21):
+<ul>
+  <li>st/xorg: Don't try to use option values before processing options</li>
+  <li>xorg/vmwgfx: Make vmwarectrl work also on 64-bit servers</li>
+  <li>st/xorg: Add a customizer option to get rid of annoying cursor update flicker</li>
+  <li>xorg/vmwgfx: Don't hide HW cursors when updating them</li>
+  <li>st/xorg: Don't try to remove invalid fbs</li>
+  <li>st/xorg: Fix typo</li>
+  <li>st/xorg, xorg/vmwgfx: Be a bit more frendly towards cross-compiling environments</li>
+  <li>st/xorg: Fix compilation errors for Xservers compiled without Composite</li>
+  <li>st/xorg: Don't use deprecated x*alloc / xfree functions</li>
+  <li>xorg/vmwgfx: Don't use deprecated x*alloc / xfree functions</li>
+  <li>st/xorg: Fix compilation for Xservers &gt;= 1.10</li>
+  <li>mesa: Make sure we have the talloc cflags when using the talloc headers</li>
+  <li>egl: Add an include for size_t</li>
+  <li>mesa: Add talloc includes for gles</li>
+  <li>st/egl: Fix build for include files in nonstandard places</li>
+  <li>svga/drm: Optionally resolve calls to powf during link-time</li>
+  <li>gallium/targets: Trivial crosscompiling fix</li>
+  <li>st/xorg: Add a function to flush pending rendering and damage</li>
+  <li>gallium/targets/xorg-vmwgfx: Xv fixes</li>
+  <li>xorg/vmwgfx: Flush even if we don't autopaint the color key</li>
+  <li>xorg/vmwgfx: Don't clip video to viewport</li>
+</ul></p>
+
+<p>Tilman Sauerbeck (35):
+<ul>
+  <li>r600g: Fixed a bo leak in r600_blit_state_ps_shader().</li>
+  <li>r600g: Use clamped math for RCP and RSQ.</li>
+  <li>r600g: Formatting fixes.</li>
+  <li>r600g: Added DB_SHADER_CONTROL defines.</li>
+  <li>r600g: Only set PA_SC_EDGERULE on rv770 and greater.</li>
+  <li>r600g: Enable PIPE_SHADER_CAP_TGSI_CONT_SUPPORTED.</li>
+  <li>r600g: Fixed the shift in S_02880C_KILL_ENABLE.</li>
+  <li>glsl2: Empty functions can be inlined.</li>
+  <li>glsl2: Fixed cloning of ir_call error instructions.</li>
+  <li>r600g: Added support for TGSI_SEMANTIC_FACE.</li>
+  <li>gallium/docs: Fixed a typo in the SCS opcode description.</li>
+  <li>r600g: Honour destination operand's writemask in the SCS implementation.</li>
+  <li>r600g: Implemented the Z and W component write for the SCS opcode.</li>
+  <li>python/tests: Fixed tri.py for API and TGSI syntax changes.</li>
+  <li>r600g: Removed debug code.</li>
+  <li>gallium/docs: The RET opcode may appear anywhere in a subroutine.</li>
+  <li>r600g: Destroy the blitter.</li>
+  <li>r600g: Fixed two texture surface leaks in r600_blit_uncompress_depth().</li>
+  <li>r600g: Cleaned up index buffer reference handling in the draw module.</li>
+  <li>r600g: Fixed r600_vertex_element leak.</li>
+  <li>r600g: Added r600_pipe_shader_destroy().</li>
+  <li>r600g: Also clear bc data when we're destroying a shader.</li>
+  <li>r600g: In radeon_bo(), call LIST_INITHEAD early.</li>
+  <li>r600g: Destroy the blitter.</li>
+  <li>r600g: Removed unused 'ptr' argument from radeon_bo().</li>
+  <li>r600g: Made radeon_bo_pb_map_internal() actually call radeon_bo_map().</li>
+  <li>r600g: Fixed unmap condition in radeon_bo_pb_destroy().</li>
+  <li>r600g: Made radeon_bo::map_count signed.</li>
+  <li>r600g: We don't support PIPE_CAP_PRIMITIVE_RESTART.</li>
+  <li>r600g: Delete custom_dsa_flush on shutdown.</li>
+  <li>r600g: Fixed two memory leaks in winsys.</li>
+  <li>r600g: Destroy the winsys in r600_destroy_screen().</li>
+  <li>st/mesa: Reset the index buffer before destroying the pipe context.</li>
+  <li>st/mesa: Reset the constant buffers before destroying the pipe context.</li>
+  <li>r600g: Removed duplicated call to tgsi_split_literal_constant().</li>
+</ul></p>
+
+<p>Timo Wiren (1):
+<ul>
+  <li>Fix typos in comments and debug output strings.</li>
+</ul></p>
+
+<p>Tom Fogal (3):
+<ul>
+  <li>Implement x86_64 atomics for compilers w/o intrinsics.</li>
+  <li>Prefer intrinsics to handrolled atomic ops.</li>
+  <li>Revert "Prefer intrinsics to handrolled atomic ops."</li>
+</ul></p>
+
+<p>Tom Stellard (32):
+<ul>
+  <li>r300/compiler: Refactor the pair instruction data structures</li>
+  <li>r300g: Always try to build libr300compiler.a</li>
+  <li>r300/compiler: Fix two mistakes in the presubtract optimization pass.</li>
+  <li>r300/compiler: Add more helper functions for iterating through sources</li>
+  <li>r300/compiler: Print immediate values after "dead constants" pass</li>
+  <li>r300/compiler: radeon_remove_constants.c: fix indentation</li>
+  <li>r300/compiler: Use rc_for_all_reads_src() in "dead constants" pass</li>
+  <li>r300/compiler: Fix segfault in error path</li>
+  <li>r300/compiler: Don't use rc_error() unless the error is unrecoverable</li>
+  <li>r300/compiler: Don't merge instructions that write output regs and ALU result</li>
+  <li>r300/compiler: Create a helper function for merging presubtract sources</li>
+  <li>r300/compiler: Fix incorrect assumption</li>
+  <li>r300/compiler: Clear empty registers after constant folding</li>
+  <li>r300/compiler: Add a new function for more efficient dataflow analysis</li>
+  <li>r300g: Add new debug option for logging vertex/fragment program stats</li>
+  <li>r300/compiler: Use rc_get_readers_normal() for presubtract optimizations</li>
+  <li>r300/compiler: Don't clobber presubtract sources during optimizations</li>
+  <li>r300/compiler: Don't track readers into an IF block.</li>
+  <li>r300/compiler: Make sure presubtract sources use supported swizzles</li>
+  <li>r300/compiler: Fix register allocator's handling of loops</li>
+  <li>r300/compiler: Fix instruction scheduling within IF blocks</li>
+  <li>r300/compiler: Use zero as the register index for unused sources</li>
+  <li>r300/compiler: Ignore alpha dest register when replicating the result</li>
+  <li>r300/compiler: Add rc_get_readers()</li>
+  <li>r300/compiler: Handle BREAK and CONTINUE in rc_get_readers()</li>
+  <li>r300/compiler: Track readers through branches in rc_get_readers()</li>
+  <li>r300/compiler: Convert RGB to alpha in the scheduler</li>
+  <li>r300/compiler: Use presubtract operations as much as possible</li>
+  <li>r300/compiler: Enable rename_reg pass for r500 cards</li>
+  <li>r300/compiler: Add a more efficient version of rc_find_free_temporary()</li>
+  <li>r300/compiler: Don't allow presubtract sources to be remapped twice</li>
+  <li>r300/compiler: Fix black terrain in Civ4</li>
+</ul></p>
+
+<p>Victor Tseng (1):
+<ul>
+  <li>egl/i965: include inline_wrapper_sw_helper.h</li>
+</ul></p>
+
+<p>Viktor Novotný (6):
+<ul>
+  <li>dri/nouveau: Import headers from rules-ng-ng</li>
+  <li>dri/nouveau: nv04: Use rules-ng-ng headers</li>
+  <li>dri/nouveau: nv10: Use rules-ng-ng headers</li>
+  <li>dri/nouveau nv20: Use rules-ng-ng headers</li>
+  <li>dri/nouveau: Remove nouveau_class.h, finishing switch to rules-ng-ng headers</li>
+  <li>dri/nouveau: Clean up magic numbers in get_rt_format</li>
+</ul></p>
+
+<p>Vinson Lee (214):
+<ul>
+  <li>llvmpipe: Remove unnecessary header.</li>
+  <li>r600g: Remove unnecessary headers.</li>
+  <li>mesa: Include missing header in program.h.</li>
+  <li>glsl: Fix 'format not a string literal and no format arguments' warning.</li>
+  <li>r600g: Silence uninitialized variable warning.</li>
+  <li>r600g: Silence uninitialized variable warning.</li>
+  <li>nvfx: Silence uninitialized variable warnings.</li>
+  <li>r600g: Silence uninitialized variable warning.</li>
+  <li>r600g: Silence uninitialized variable warning.</li>
+  <li>r600g: Silence uninitialized variable warning.</li>
+  <li>r600g: Silence unused variable warning.</li>
+  <li>nv50: Update files in SConscript to match Makefile.</li>
+  <li>nv50: Remove unnecessary headers.</li>
+  <li>nv50: Silence uninitialized variable warning.</li>
+  <li>nv50: Silence uninitialized variable warning.</li>
+  <li>nv50: Silence uninitialized variable warning.</li>
+  <li>gallivm: Remove unnecessary headers.</li>
+  <li>draw: Remove unnecessary header.</li>
+  <li>nv50: Silence uninitialized variable warnings.</li>
+  <li>nv50: Fix 'control reaches end of non-void function' warning.</li>
+  <li>mesa/st: Silence uninitialized variable warning.</li>
+  <li>gallivm: Remove unnecessary header.</li>
+  <li>r600g: Remove unnecessary header.</li>
+  <li>r600g: Remove unnecessary headers.</li>
+  <li>r600g: Fix implicit declaration warning.</li>
+  <li>r600g: Fix memory leak on error path.</li>
+  <li>r600g: Silence uninitialized variable warning.</li>
+  <li>r600g: Silence unused variable warnings.</li>
+  <li>mesa: bump version to 7.10</li>
+  <li>ir_to_mesa: Remove unused member array_indexed from struct statevar_element.</li>
+  <li>mesa: Silence "'valid_texture_object' defined but not used" warning.</li>
+  <li>x86: Silence unused variable warning on Mac OS X.</li>
+  <li>glsl: Fix 'control reaches end of non-void function' warning.</li>
+  <li>nvfx: Remove const qualifer from nvfx_vertprog_translate.</li>
+  <li>nvfx: Silence uninitialized variable warnings.</li>
+  <li>r600g: Remove unused variable.</li>
+  <li>nv50: Silence missing initializer warning.</li>
+  <li>nv50: Remove dead initialization.</li>
+  <li>nv50: Remove dead initialization.</li>
+  <li>tgsi: Remove duplicate case value.</li>
+  <li>glut: Define markWindowHidden for non-Windows only.</li>
+  <li>glut: Define eventParser for non-Windows only.</li>
+  <li>r300g: Silence uninitialized variable warning.</li>
+  <li>intel: Fix implicit declaration of function '_mesa_meta_Bitmap' warning.</li>
+  <li>mesa: Remove unnecessary headers.</li>
+  <li>r600g: Remove unnecessary header.</li>
+  <li>unichrome: Remove unnecessary header.</li>
+  <li>intel: Remove unnecessary headers.</li>
+  <li>r600g: Remove unused variable.</li>
+  <li>r600g: Disable unused variables.</li>
+  <li>r600g: Remove unused variable.</li>
+  <li>r600g: Silence 'control reaches end of non-void function' warning.</li>
+  <li>r600g: Remove unused variable.</li>
+  <li>r600g: Remove unused variable.</li>
+  <li>r600g: Disable unused variables.</li>
+  <li>intel: Remove unnecessary header.</li>
+  <li>st/dri: Remove unnecessary header.</li>
+  <li>r600g: Remove unused variable.</li>
+  <li>r300g: Remove unused variable.</li>
+  <li>r600g: Don't return a value in function returning void.</li>
+  <li>r600g: Remove unused variables.</li>
+  <li>r600g: Include p_compiler.h instead of malloc.h.</li>
+  <li>r600g: Silence uninitialized variable warnings.</li>
+  <li>scons: Add MinGW-w64 prefixes for MinGW build.</li>
+  <li>dri: Add GET_PROGRAM_NAME definition for Mac OS X.</li>
+  <li>scons: Add program/sampler.cpp to SCons build.</li>
+  <li>mesa: Fix printf format warning.</li>
+  <li>mesa: Fix printf format warning.</li>
+  <li>mesa: Fix printf format warning.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>r600g: Update SConscript.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>r600g: Update SConscript.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>r600g: Update SConscript.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>r600g: Fix SCons build.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>r300/compiler: Remove declaration before code.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>glsl: Remove unnecessary header.</li>
+  <li>savage: Remove unnecessary header.</li>
+  <li>r600g: Remove unused variable.</li>
+  <li>r600g: Remove unnecessary headers.</li>
+  <li>r600g: Fix SCons build.</li>
+  <li>r600g: Remove unnecessary header.</li>
+  <li>gallivm: Remove unnecessary header.</li>
+  <li>r600g: Silence uninitialized variable warning.</li>
+  <li>r600g: Silence uninitialized variable warning.</li>
+  <li>r600g: Silence uninitialized variable warning.</li>
+  <li>i915: Silence unused variable warning in non-debug builds.</li>
+  <li>i915: Silence unused variable warning in non-debug builds.</li>
+  <li>i965: Silence unused variable warning on non-debug builds.</li>
+  <li>i965: Silence unused variable warning on non-debug builds.</li>
+  <li>i965: Initialize member variables.</li>
+  <li>r300: Silence uninitialized variable warning.</li>
+  <li>tdfx: Silence unused variable warning on non-debug builds.</li>
+  <li>gallivm: Remove unnecessary header.</li>
+  <li>glsl: Initialize variable in ir_derefence_array::constant_expression_value</li>
+  <li>mesa: Add missing header to shaderobj.h.</li>
+  <li>llvmpipe: Return non-zero exit code for lp_test_round failures.</li>
+  <li>r300/compiler: Remove unused variable.</li>
+  <li>st/xorg: Fix memory leak on error path.</li>
+  <li>llvmpipe: Initialize state variable in debug_bin function.</li>
+  <li>llvmpipe: Initialize variable.</li>
+  <li>draw: Move loop variable declaration outside for loop.</li>
+  <li>r600g: Ensure r600_src is initialized in tgsi_exp function.</li>
+  <li>glsl: Add assert for unhandled ir_shader case.</li>
+  <li>swrast: Print out format on unexpected failure in _swrast_DrawPixels.</li>
+  <li>llvmpipe: Remove unnecessary header.</li>
+  <li>draw: Remove unnecessary header.</li>
+  <li>gallivm: Silence uninitialized variable warnings.</li>
+  <li>gallivm: Silence uninitialized variable warnings.</li>
+  <li>gallivm: Silence uninitialized variable warning.</li>
+  <li>r300g: Silence uninitialized variable warning.</li>
+  <li>mesa: Remove unnecessary headers.</li>
+  <li>r600g: Silence uninitialized variable warnings.</li>
+  <li>st/mesa: Remove unnecessary header.</li>
+  <li>mesa: Remove unnecessary header.</li>
+  <li>egl: Remove unnecessary headers.</li>
+  <li>swrast: Print out format on unexpected failure in _swrast_ReadPixels.</li>
+  <li>st/mesa: Silence uninitialized variable warning.</li>
+  <li>savage: Remove unnecessary header.</li>
+  <li>st/vega: Remove unnecessary headers.</li>
+  <li>dri/nouveau: Silence uninitialized variable warning.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>i965: Silence uninitialized variable warning.</li>
+  <li>i965: Silence uninitialized variable warning.</li>
+  <li>mesa: Clean up header file inclusion in accum.h.</li>
+  <li>mesa: Clean up header file inclusion in version.h.</li>
+  <li>mesa: Clean up header file inclusion in api_loopback.h.</li>
+  <li>mesa: Clean up header file inclusion in api_validate.h.</li>
+  <li>mesa: Include mfeatures.h in api_loopback for FEATURE_beginend.</li>
+  <li>mesa: Include mfeatures.h in api_validate.c for FEATURE_* symbols.</li>
+  <li>mesa: Clean up header file inclusion in arrayobj.h.</li>
+  <li>mesa: Clean up header file inclusion in atifragshader.h.</li>
+  <li>mesa: Clean up header file inclusion in attrib.h.</li>
+  <li>mesa: Clean up header file inclusion in blend.h.</li>
+  <li>mesa: Clean up header file inclusion in buffers.h.</li>
+  <li>mesa: Clean up header file inclusion in colortab.h.</li>
+  <li>mesa: Clean up header file inclusion in convolve.h.</li>
+  <li>mesa: Clean up header file inclusion in debug.h.</li>
+  <li>mesa: Clean up header file inclusion in depth.h.</li>
+  <li>mesa: Clean up header file inclusion in depthstencil.h.</li>
+  <li>mesa: Clean up header file inclusion in drawpix.h.</li>
+  <li>mesa: Clean up header file inclusion in drawtex.h.</li>
+  <li>mesa: Clean up header file inclusion in enable.h.</li>
+  <li>mesa: Clean up header file inclusion in extensions.h.</li>
+  <li>graw: Add struct pipe_surface forward declaration.</li>
+  <li>mesa: Clean up header file inclusion in fbobject.h.</li>
+  <li>mesa: Clean up header file inclusion in ffvertex_prog.h.</li>
+  <li>mesa: Clean up header file inclusion in fog.h.</li>
+  <li>mesa: Clean up header file inclusion in framebuffer.h.</li>
+  <li>mesa: Clean up header file inclusion in hint.h.</li>
+  <li>mesa: Clean up header file inclusion in histogram.h.</li>
+  <li>mesa: Clean up header file inclusion in image.h.</li>
+  <li>mesa: Add missing header and forward declarations in dd.h.</li>
+  <li>mesa: Clean up header file inclusion in light.h.</li>
+  <li>mesa: Clean up header file inclusion in lines.h.</li>
+  <li>mesa: Clean up header file inclusion in matrix.h.</li>
+  <li>mesa: Clean up header file inclusion in multisample.h.</li>
+  <li>mesa: Clean up header file inclusion in nvprogram.h.</li>
+  <li>winsys/xlib: Add cygwin to SConscript.</li>
+  <li>mesa: Clean up header file inclusion in pixel.h.</li>
+  <li>mesa: Clean up header file inclusion in pixelstore.h.</li>
+  <li>mesa: Fix printf format warnings.</li>
+  <li>mesa: Clean up header file inclusion in points.h.</li>
+  <li>i965: Silence uninitialized variable warning.</li>
+  <li>glsl: Add ir_constant_expression.cpp to SConscript.</li>
+  <li>mesa: Add definitions for inverse hyperbolic function on MSVC.</li>
+  <li>glsl: Fix 'control reaches end of non-void function' warning.</li>
+  <li>glsl: Add lower_vector.cpp to SConscript.</li>
+  <li>glsl: Fix type of label 'default' in switch statement.</li>
+  <li>st/mesa: Remove unnecessary headers.</li>
+  <li>swrast: Remove unnecessary header.</li>
+  <li>r600: Remove unnecesary header.</li>
+  <li>intel: Remove unnecessary header.</li>
+  <li>mesa: Clean up header file inclusion in polygon.h.</li>
+  <li>mesa: Clean up header file inclusion in rastpos.h.</li>
+  <li>mesa: Clean up header file inclusion in readpix.h.</li>
+  <li>mesa: Clean up header file inclusion in renderbuffer.h.</li>
+  <li>mesa: Clean up header file inclusion in scissor.h.</li>
+  <li>mesa: Clean up header file inclusion in shaderapi.h.</li>
+  <li>mesa: Clean up header file inclusion in shared.h.</li>
+  <li>mesa: Clean up header file inclusion in stencil.h.</li>
+  <li>r600: Remove unnecessary header.</li>
+  <li>llvmpipe: Remove unnecessary headers.</li>
+  <li>mesa: Clean up header file inclusion in syncobj.h.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>r300/compiler: Move declaration before code.</li>
+  <li>mesa: Clean up header file inclusion in texcompress.h.</li>
+  <li>st/vega: Silence uninitialized variable warning.</li>
+  <li>mesa: Clean up header file inclusion in texcompress_s3tc.h.</li>
+  <li>mesa: Clean up header file inclusion in texenvprogram.h.</li>
+  <li>mesa: Clean up header file inclusion in texformat.h.</li>
+  <li>mesa: Clean up header file inclusion in texgetimage.h.</li>
+  <li>mesa: Clean up header file inclusion in texobj.h.</li>
+  <li>gallium/noop: Add prototype for noop_init_state_functions.</li>
+  <li>mesa: Clean up header file inclusion in texrender.h.</li>
+  <li>mesa: Clean up header file inclusion in transformfeedback.h.</li>
+  <li>mesa: Clean up header file inclusion in varray.h.</li>
+  <li>mesa: Clean up header file inclusion in viewport.h.</li>
+  <li>r200: Silence uninitialized variable warning.</li>
+  <li>r600g: Fix SCons build.</li>
+  <li>i965: Silence uninitialized variable warning.</li>
+</ul></p>
+
+<p>Xavier Chantry (8):
+<ul>
+  <li>nv50: fix size of outputs_written array</li>
+  <li>nv50: apply layout_mask to tile_flags</li>
+  <li>nvfx: only expose one rt on nv30</li>
+  <li>nvfx: fb-&gt;nr_cbufs &lt= 1 on nv30</li>
+  <li>nvfx: reset nvfx-&gt;hw_zeta</li>
+  <li>nvfx: fixes after array textures merge</li>
+  <li>init ps-&gt;context with util_surfaces_get and do_get</li>
+  <li>gallium/trace: check bind_vertex_sampler_states and set_vertex_sampler_views</li>
+</ul></p>
+
+<p>Xiang, Haihao (10):
+<ul>
+  <li>mesa: fix regression from b4bb6680200b5a898583392f4c831c02f41e63f7</li>
+  <li>i965: add support for polygon mode on Sandybridge.</li>
+  <li>i965: fix for flat shading on Sandybridge</li>
+  <li>i965: set minimum/maximum Point Width on Sandybridge</li>
+  <li>meta: allow nested meta operations</li>
+  <li>i965: support for two-sided lighting on Sandybridge</li>
+  <li>i965: fix register region description</li>
+  <li>i965: use align1 access mode for instructions with execSize=1 in VS</li>
+  <li>i965: don't spawn GS thread for LINELOOP on Sandybridge</li>
+  <li>i965: use BLT to clear buffer if possible on Sandybridge</li>
+</ul></p>
+
+<p>Zack Rusin (8):
+<ul>
+  <li>rbug: fix rbug when contexts are being destroyed</li>
+  <li>llvmpipe: fix rasterization of vertical lines on pixel boundaries</li>
+  <li>scons: build the xorg state trackers only when env includes drm</li>
+  <li>gallivm: implement indirect addressing of the output registers</li>
+  <li>gallivm: implement indirect addressing over inputs</li>
+  <li>gallivm: fix storing of the addr register</li>
+  <li>scons: add alias for identity</li>
+  <li>gallium/util: add states relevant to geometry shaders</li>
+</ul></p>
+
+<p>Zhenyu Wang (40):
+<ul>
+  <li>i965: disasm quarter and write enable instruction control on sandybridge</li>
+  <li>i965: new state dump for sandybridge</li>
+  <li>i965: enable accumulator update in PS kernel too on sandybridge</li>
+  <li>i965: Fix color interpolation on sandybridge</li>
+  <li>i965: force zero in clipper to ignore RTAIndex on sandybridge</li>
+  <li>i965: fix point size setting in header on sandybridge</li>
+  <li>i965: ff sync message change for sandybridge</li>
+  <li>i965: ignore quads for GS kernel on sandybridge</li>
+  <li>i965: add sandybridge viewport state bo into validation list</li>
+  <li>i965: VS use SPF mode on sandybridge for now</li>
+  <li>i965: fix jump count on sandybridge</li>
+  <li>i965: Fix sampler on sandybridge</li>
+  <li>i965: fix const register count for sandybridge</li>
+  <li>i965: Add all device ids for sandybridge</li>
+  <li>i965: sandybridge pipe control workaround before write cache flush</li>
+  <li>i965: only allow SIMD8 kernel on sandybridge now</li>
+  <li>i965: don't do calculation for delta_xy on sandybridge</li>
+  <li>i965: fix pixel w interpolation on sandybridge</li>
+  <li>i965: enable polygon offset on sandybridge</li>
+  <li>i965: fix scissor state on sandybridge</li>
+  <li>i965: fix point sprite on sandybridge</li>
+  <li>i965: fix occlusion query on sandybridge</li>
+  <li>i965: fallback bitmap operation on sandybridge</li>
+  <li>i965: Always set tiling for depth buffer on sandybridge</li>
+  <li>i965: fallback lineloop on sandybridge for now</li>
+  <li>Revert "i965: Always set tiling for depth buffer on sandybridge"</li>
+  <li>i965: always set tiling for fbo depth buffer on sandybridge</li>
+  <li>i965: Fix GS hang on Sandybridge</li>
+  <li>Revert "i965: fallback lineloop on sandybridge for now"</li>
+  <li>i965: refresh wm push constant also for BRW_NEW_FRAMENT_PROGRAM on gen6</li>
+  <li>i965: fix dest type of 'endif' on sandybridge</li>
+  <li>Revert "i965: VS use SPF mode on sandybridge for now"</li>
+  <li>i965: also using align1 mode for math2 on sandybridge</li>
+  <li>i965: Fix GS state uploading on Sandybridge</li>
+  <li>i965: upload WM state for _NEW_POLYGON on sandybridge</li>
+  <li>i965: Use MI_FLUSH_DW for blt ring flush on sandybridge</li>
+  <li>i965: explicit tell header present for fb write on sandybridge</li>
+  <li>i965: Fix occlusion query on sandybridge</li>
+  <li>i965: Use last vertex convention for quad provoking vertex on sandybridge</li>
+  <li>i965: Fix provoking vertex select in clip state for sandybridge</li>
+</ul></p>
+
+<p>Zou Nan hai (1):
+<ul>
+  <li>i965: skip too small size mipmap</li>
+</ul></p>
+
+<p>delphi (2):
+<ul>
+  <li>draw: added userclip planes and updated variant_key</li>
+  <li>draw: some changes to allow for runtime changes to userclip planes</li>
+</ul></p>
+
+<p>nobled (3):
+<ul>
+  <li>r300g: Abort if atom allocations fail</li>
+  <li>r300g: Abort if draw_create() fails</li>
+  <li>r300g: Drop unnecessary cast</li>
+</ul></p>
+
+<p>pontus lidman (1):
+<ul>
+  <li>mesa: check for posix_memalign() errors</li>
+</ul></p>
+
+<p>richard (2):
+<ul>
+  <li>evergreen : fix z format setting, enable stencil.</li>
+  <li>r600c : inline vertex format is not updated in an app, switch to use vfetch constants. For the 7.9 and 7.10 branches as well.</li>
+</ul></p>
+
+</body>
+</html>
diff --git a/mesalib/docs/relnotes-7.8.2.html b/mesalib/docs/relnotes-7.8.2.html
index 5ee7e12d2..1393b2ede 100644
--- a/mesalib/docs/relnotes-7.8.2.html
+++ b/mesalib/docs/relnotes-7.8.2.html
@@ -1,146 +1,146 @@
-<HTML>
-
-<TITLE>Mesa Release Notes</TITLE>
-
-<head><link rel="stylesheet" type="text/css" href="mesa.css"></head>
-
-<BODY>
-
-<body bgcolor="#eeeeee">
-
-<H1>Mesa 7.8.2 Release Notes / (date tbd)</H1>
-
-<p>
-Mesa 7.8.2 is a bug fix release which fixes bugs found since the 7.8.1 release.
-</p>
-<p>
-Mesa 7.8.2 implements the OpenGL 2.1 API, but the version reported by
-glGetString(GL_VERSION) depends on the particular driver being used.
-Some drivers don't support all the features required in OpenGL 2.1.
-</p>
-<p>
-See the <a href="install.html">Compiling/Installing page</a> for prerequisites
-for DRI hardware acceleration.
-</p>
-
-
-<h2>MD5 checksums</h2>
-<pre>
-c89b63d253605ed40e8ac370d25a833c  MesaLib-7.8.2.tar.gz
-6be2d343a0089bfd395ce02aaf8adb57  MesaLib-7.8.2.tar.bz2
-a04ad3b06ac5ff3969a003fa7bbf7d5b  MesaLib-7.8.2.zip
-7c213f92efeb471f0331670d5079d4c0  MesaDemos-7.8.2.tar.gz
-757d9e2e06f48b1a52848be9b0307ced  MesaDemos-7.8.2.tar.bz2
-8d0e5cfe68b8ebf90265d350ae2c48b1  MesaDemos-7.8.2.zip
-b74482e3f44f35ed395c4aada4fd8240  MesaGLUT-7.8.2.tar.gz
-a471807b65e49c325808ba4551be93ed  MesaGLUT-7.8.2.tar.bz2
-9f190268c42be582ef66e47365ee61e3  MesaGLUT-7.8.2.zip
-</pre>
-
-
-<h2>New features</h2>
-<p>None.</p>
-
-
-<h2>Changes</h2>
-<ul>
-<li>Upgraded glext.h to version 61, and upgraded glxext.h
-</ul>
-
-
-<h2>Bug fixes</h2>
-<ul>
-<li>Fixed Gallium glDrawPixels(GL_DEPTH_COMPONENT).
-<li>Fixed Gallium Cell driver to buildable, runable state
-<li>Fixed bad error checking for glFramebufferRenderbuffer(attachment=GL_DEPTH_STENCIL_ATTACHMENT).
-<li>Fixed incorrect Z coordinate handling in "meta" glDraw/CopyPixels.
-    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=23670">Bug
-    #23670</a>.</li>
-
-<li>Assorted i965 driver fixes.
-    Including but not limited to:
-  <ul>
-    <li>Fix scissoring when width or height is
-    0.  <a href="https://bugs.freedesktop.org/show_bug.cgi?id=27643">Bug
-    #27643</a>.
-    <li>Fix bit allocation for number of color regions for
-    ARB_draw_buffers.</li>
-    <li>Set the correct provoking vertex for clipped first-mode
-    trifans.  <a href="https://bugs.freedesktop.org/show_bug.cgi?id=24470">Bug
-    #24470</a>.</li>
-    <li>Use <code>R16G16B16A16_FLOAT</code> for 3-component half-float.</li>
-    <li>Fix assertion for surface tile offset usage on Ironlake.</li>
-    <li>Fix cube map layouts on Ironlake.</li>
-    <li>When an RB gets a new region, clear the old from the state
-    cache.  <a href="https://bugs.freedesktop.org/show_bug.cgi?id=24119">Bug
-    #24119</a>.</li>
-    <li>Reject shaders with uninlined function calls instead of hanging.</li>
-  </ul>
-</li>
-
-<li>Assorted i915 driver fixes.  Including but not limited to:
-  <ul>
-    <li>Fixed texture LOD clamping in i915 driver.
-    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=24846">Bug
-    #24846</a>.</li>
-    <li>Fix off-by-one for drawing rectangle.
-    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=27408">Bug
-    #27408</a>.</li>
-  </ul>
-</li>
-
-<li>Fixed hangs in etracer on 830 and 845
-    chipsets.  <a href="https://bugs.freedesktop.org/show_bug.cgi?id=26557">Bug
-    #26557</a>.</li>
-<li>Fixed tiling of small textures on all Intel drivers.</li>
-<li>Fixed crash in Savage driver when using <code>_mesa_CopyTexImage2D</code>.
-    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=27652">Bug
-    #27652</a>.</li>
-
-<li>Assorted GLX fixes.  Including but not limited to:
-  <ul>
-    <li>Fixed <code>__glXInitializeVisualConfigFromTags</code>'s handling of
-        unrecognized fbconfig tags.</li>
-    <li>Fixed regression with <code>GLX_USE_GL</code>.
-    <li>Fixed config chooser logic for 'mask' matching.</li>
-    <li>Report swap events correctly in direct rendered case (DRI2)</li>
-    <li>Fixed build with dri2proto which doesn't define
-        <code>X_DRI2SwapInterval</code>.</li>
-    <li>Get <code>GLX_SCREEN</code> first in <code>__glXQueryContextInfo</code>.
-        <a href="https://bugs.freedesktop.org/show_bug.cgi?id=14245">Bug
-	#14245</a>.</li>
-  </ul>
-</li>
-
-<li>Assorted GLSL fixes.  Including but not limited to:
-  <ul>
-    <li>Change variable declared assertion into conditional in GLSL
-        compiler.  <a href="https://bugs.freedesktop.org/show_bug.cgi?id=27921">Bug
-        #27921</a>.</li>
-    <li>Fix instruction indexing
-        bugs.  <a href="https://bugs.freedesktop.org/show_bug.cgi?id=27566">Bug
-        #27566</a>.</li>
-    <li>Updated uniform location / offset encoding to be more like
-        other implementations.</li>
-    <li>Don't overwrite a driver's shader infolog with generic failure
-        message.</li>
-  </ul>
-</li>
-
-<li>Fixed OSMesa build for 16 and 32-bit color channel depth.
-<li>Fixed OSMesa build with hidden symbol visibility. libOSMesa no longer links to libGL.
-    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=28305">Bug
-    #28305</a>.
-<li>Fixed handling of multiple render targets in fixed-function
-    texture envrionmnent programs.</li>
-<li>Fixed conversion errors in <code>signed_rgba8888[rev]</code> texel
-    fetch.</li>
-<li>Don't set srcLevel on <code>GL_TEXTURE_RECTANGLE_ARB</code> targets.</li>
-<li>Various build fixes for OpenBSD.</li>
-<li>Various build fixes for OS X.</li>
-<li>Various build fixes for GCC 3.3.</li>
-</ul>
-
-
-</body>
-</html>
+<HTML>
+
+<TITLE>Mesa Release Notes</TITLE>
+
+<head><link rel="stylesheet" type="text/css" href="mesa.css"></head>
+
+<BODY>
+
+<body bgcolor="#eeeeee">
+
+<H1>Mesa 7.8.2 Release Notes / (date tbd)</H1>
+
+<p>
+Mesa 7.8.2 is a bug fix release which fixes bugs found since the 7.8.1 release.
+</p>
+<p>
+Mesa 7.8.2 implements the OpenGL 2.1 API, but the version reported by
+glGetString(GL_VERSION) depends on the particular driver being used.
+Some drivers don't support all the features required in OpenGL 2.1.
+</p>
+<p>
+See the <a href="install.html">Compiling/Installing page</a> for prerequisites
+for DRI hardware acceleration.
+</p>
+
+
+<h2>MD5 checksums</h2>
+<pre>
+c89b63d253605ed40e8ac370d25a833c  MesaLib-7.8.2.tar.gz
+6be2d343a0089bfd395ce02aaf8adb57  MesaLib-7.8.2.tar.bz2
+a04ad3b06ac5ff3969a003fa7bbf7d5b  MesaLib-7.8.2.zip
+7c213f92efeb471f0331670d5079d4c0  MesaDemos-7.8.2.tar.gz
+757d9e2e06f48b1a52848be9b0307ced  MesaDemos-7.8.2.tar.bz2
+8d0e5cfe68b8ebf90265d350ae2c48b1  MesaDemos-7.8.2.zip
+b74482e3f44f35ed395c4aada4fd8240  MesaGLUT-7.8.2.tar.gz
+a471807b65e49c325808ba4551be93ed  MesaGLUT-7.8.2.tar.bz2
+9f190268c42be582ef66e47365ee61e3  MesaGLUT-7.8.2.zip
+</pre>
+
+
+<h2>New features</h2>
+<p>None.</p>
+
+
+<h2>Changes</h2>
+<ul>
+<li>Upgraded glext.h to version 61, and upgraded glxext.h
+</ul>
+
+
+<h2>Bug fixes</h2>
+<ul>
+<li>Fixed Gallium glDrawPixels(GL_DEPTH_COMPONENT).
+<li>Fixed Gallium Cell driver to buildable, runable state
+<li>Fixed bad error checking for glFramebufferRenderbuffer(attachment=GL_DEPTH_STENCIL_ATTACHMENT).
+<li>Fixed incorrect Z coordinate handling in "meta" glDraw/CopyPixels.
+    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=23670">Bug
+    #23670</a>.</li>
+
+<li>Assorted i965 driver fixes.
+    Including but not limited to:
+  <ul>
+    <li>Fix scissoring when width or height is
+    0.  <a href="https://bugs.freedesktop.org/show_bug.cgi?id=27643">Bug
+    #27643</a>.
+    <li>Fix bit allocation for number of color regions for
+    ARB_draw_buffers.</li>
+    <li>Set the correct provoking vertex for clipped first-mode
+    trifans.  <a href="https://bugs.freedesktop.org/show_bug.cgi?id=24470">Bug
+    #24470</a>.</li>
+    <li>Use <code>R16G16B16A16_FLOAT</code> for 3-component half-float.</li>
+    <li>Fix assertion for surface tile offset usage on Ironlake.</li>
+    <li>Fix cube map layouts on Ironlake.</li>
+    <li>When an RB gets a new region, clear the old from the state
+    cache.  <a href="https://bugs.freedesktop.org/show_bug.cgi?id=24119">Bug
+    #24119</a>.</li>
+    <li>Reject shaders with uninlined function calls instead of hanging.</li>
+  </ul>
+</li>
+
+<li>Assorted i915 driver fixes.  Including but not limited to:
+  <ul>
+    <li>Fixed texture LOD clamping in i915 driver.
+    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=24846">Bug
+    #24846</a>.</li>
+    <li>Fix off-by-one for drawing rectangle.
+    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=27408">Bug
+    #27408</a>.</li>
+  </ul>
+</li>
+
+<li>Fixed hangs in etracer on 830 and 845
+    chipsets.  <a href="https://bugs.freedesktop.org/show_bug.cgi?id=26557">Bug
+    #26557</a>.</li>
+<li>Fixed tiling of small textures on all Intel drivers.</li>
+<li>Fixed crash in Savage driver when using <code>_mesa_CopyTexImage2D</code>.
+    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=27652">Bug
+    #27652</a>.</li>
+
+<li>Assorted GLX fixes.  Including but not limited to:
+  <ul>
+    <li>Fixed <code>__glXInitializeVisualConfigFromTags</code>'s handling of
+        unrecognized fbconfig tags.</li>
+    <li>Fixed regression with <code>GLX_USE_GL</code>.
+    <li>Fixed config chooser logic for 'mask' matching.</li>
+    <li>Report swap events correctly in direct rendered case (DRI2)</li>
+    <li>Fixed build with dri2proto which doesn't define
+        <code>X_DRI2SwapInterval</code>.</li>
+    <li>Get <code>GLX_SCREEN</code> first in <code>__glXQueryContextInfo</code>.
+        <a href="https://bugs.freedesktop.org/show_bug.cgi?id=14245">Bug
+	#14245</a>.</li>
+  </ul>
+</li>
+
+<li>Assorted GLSL fixes.  Including but not limited to:
+  <ul>
+    <li>Change variable declared assertion into conditional in GLSL
+        compiler.  <a href="https://bugs.freedesktop.org/show_bug.cgi?id=27921">Bug
+        #27921</a>.</li>
+    <li>Fix instruction indexing
+        bugs.  <a href="https://bugs.freedesktop.org/show_bug.cgi?id=27566">Bug
+        #27566</a>.</li>
+    <li>Updated uniform location / offset encoding to be more like
+        other implementations.</li>
+    <li>Don't overwrite a driver's shader infolog with generic failure
+        message.</li>
+  </ul>
+</li>
+
+<li>Fixed OSMesa build for 16 and 32-bit color channel depth.
+<li>Fixed OSMesa build with hidden symbol visibility. libOSMesa no longer links to libGL.
+    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=28305">Bug
+    #28305</a>.
+<li>Fixed handling of multiple render targets in fixed-function
+    texture envrionmnent programs.</li>
+<li>Fixed conversion errors in <code>signed_rgba8888[rev]</code> texel
+    fetch.</li>
+<li>Don't set srcLevel on <code>GL_TEXTURE_RECTANGLE_ARB</code> targets.</li>
+<li>Various build fixes for OpenBSD.</li>
+<li>Various build fixes for OS X.</li>
+<li>Various build fixes for GCC 3.3.</li>
+</ul>
+
+
+</body>
+</html>
diff --git a/mesalib/docs/relnotes-7.8.3.html b/mesalib/docs/relnotes-7.8.3.html
index e729ee820..1e9f43313 100644
--- a/mesalib/docs/relnotes-7.8.3.html
+++ b/mesalib/docs/relnotes-7.8.3.html
@@ -1,89 +1,89 @@
-<HTML>
-
-<TITLE>Mesa Release Notes</TITLE>
-
-<head><link rel="stylesheet" type="text/css" href="mesa.css"></head>
-
-<BODY>
-
-<body bgcolor="#eeeeee">
-
-<H1>Mesa 7.8.3 Release Notes / (date tbd)</H1>
-
-<p>
-Mesa 7.8.3 is a bug fix release which fixes bugs found since the 7.8.2 release.
-</p>
-<p>
-Mesa 7.8.3 implements the OpenGL 2.1 API, but the version reported by
-glGetString(GL_VERSION) depends on the particular driver being used.
-Some drivers don't support all the features required in OpenGL 2.1.
-</p>
-<p>
-See the <a href="install.html">Compiling/Installing page</a> for prerequisites
-for DRI hardware acceleration.
-</p>
-
-
-<h2>MD5 checksums</h2>
-<pre>
-x  MesaLib-7.8.3.tar.gz
-x  MesaLib-7.8.3.tar.bz2
-x  MesaLib-7.8.3.zip
-x  MesaDemos-7.8.3.tar.gz
-x  MesaDemos-7.8.3.tar.bz2
-x  MesaDemos-7.8.3.zip
-x  MesaGLUT-7.8.3.tar.gz
-x  MesaGLUT-7.8.3.tar.bz2
-x  MesaGLUT-7.8.3.zip
-</pre>
-
-
-<h2>New features</h2>
-<p>None.</p>
-
-
-<h2>Changes</h2>
-<ul>
-<li>The radeon driver should use less memory when searching for a valid mip
-image.</li>
-</ul>
-
-
-<h2>Bug fixes</h2>
-<ul>
-<li>Fix unsupported FB with D24S8 (bug 
-    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=23670">29116</a>)</li>
-<li>Fix ReadPixels crash when reading depth/stencil from an FBO</li>
-<li>Fixed a bug rendering to 16-bit buffers using swrast.</li>
-<li>Fixed a state tracker/TGSI bug that caused crashes when using Windows'
-    memory debugging features.</li>
-<li>Fixed an issue rendering to 32-bit channels with swrast (bug
-    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=29487">29487</a>)</li>
-<li>GLSL: fix indirect <TT>gl_TextureMatrix</TT> addressing (bug
-    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=28967">28967</a>)</li>
-<li>GLSL: fix for bug
-    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=27216">27216</a></li>
-<li>GLSL: fix zw fragcoord entries in some cases (bug
-    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=29183">29183</a>)</li>
-<li>Fix texture env generation in some cases (bug
-    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=28169">28169</a>)</li>
-<li>osmesa: a fix for calling <TT>OSMesaMakeCurrent</TT> twice was applied (bug 
-    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=10966">10966</a></li>
-<li>A bug was fixed which could cause Mesa to ignore the
-    <TT>MESA_EXTENSION_OVERRIDE</TT> environment variable.</li>
-<li>A bug related to specular highlights on backfaces was fixed.</li>
-<li>A radeon-specific issue with <TT>glCopyTex(Sub)Image</TT> was
-    corrected.</li>
-<li>radeon/wine: flush command stream in more cases, fixing wine d3d9
-    tests.</li>
-<li>r600: fix sin+cos normalization.</li>
-<li>r600: (properly) ignore <TT>GL_COORD_REPLACE</TT> when point sprites are
-    disabled.</li>
-<li>radeon: avoid flushing when the context is not current.</li>
-<li>r300c: a bug affecting unaligned BOs was fixed.</li>
-<li>r300c: a hardlock caused by ARB_half_float_vertex incorrectly advertised on some chipsets.</li>
-</ul>
-
-
-</body>
-</html>
+<HTML>
+
+<TITLE>Mesa Release Notes</TITLE>
+
+<head><link rel="stylesheet" type="text/css" href="mesa.css"></head>
+
+<BODY>
+
+<body bgcolor="#eeeeee">
+
+<H1>Mesa 7.8.3 Release Notes / (date tbd)</H1>
+
+<p>
+Mesa 7.8.3 is a bug fix release which fixes bugs found since the 7.8.2 release.
+</p>
+<p>
+Mesa 7.8.3 implements the OpenGL 2.1 API, but the version reported by
+glGetString(GL_VERSION) depends on the particular driver being used.
+Some drivers don't support all the features required in OpenGL 2.1.
+</p>
+<p>
+See the <a href="install.html">Compiling/Installing page</a> for prerequisites
+for DRI hardware acceleration.
+</p>
+
+
+<h2>MD5 checksums</h2>
+<pre>
+x  MesaLib-7.8.3.tar.gz
+x  MesaLib-7.8.3.tar.bz2
+x  MesaLib-7.8.3.zip
+x  MesaDemos-7.8.3.tar.gz
+x  MesaDemos-7.8.3.tar.bz2
+x  MesaDemos-7.8.3.zip
+x  MesaGLUT-7.8.3.tar.gz
+x  MesaGLUT-7.8.3.tar.bz2
+x  MesaGLUT-7.8.3.zip
+</pre>
+
+
+<h2>New features</h2>
+<p>None.</p>
+
+
+<h2>Changes</h2>
+<ul>
+<li>The radeon driver should use less memory when searching for a valid mip
+image.</li>
+</ul>
+
+
+<h2>Bug fixes</h2>
+<ul>
+<li>Fix unsupported FB with D24S8 (bug 
+    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=23670">29116</a>)</li>
+<li>Fix ReadPixels crash when reading depth/stencil from an FBO</li>
+<li>Fixed a bug rendering to 16-bit buffers using swrast.</li>
+<li>Fixed a state tracker/TGSI bug that caused crashes when using Windows'
+    memory debugging features.</li>
+<li>Fixed an issue rendering to 32-bit channels with swrast (bug
+    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=29487">29487</a>)</li>
+<li>GLSL: fix indirect <TT>gl_TextureMatrix</TT> addressing (bug
+    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=28967">28967</a>)</li>
+<li>GLSL: fix for bug
+    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=27216">27216</a></li>
+<li>GLSL: fix zw fragcoord entries in some cases (bug
+    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=29183">29183</a>)</li>
+<li>Fix texture env generation in some cases (bug
+    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=28169">28169</a>)</li>
+<li>osmesa: a fix for calling <TT>OSMesaMakeCurrent</TT> twice was applied (bug 
+    <a href="https://bugs.freedesktop.org/show_bug.cgi?id=10966">10966</a></li>
+<li>A bug was fixed which could cause Mesa to ignore the
+    <TT>MESA_EXTENSION_OVERRIDE</TT> environment variable.</li>
+<li>A bug related to specular highlights on backfaces was fixed.</li>
+<li>A radeon-specific issue with <TT>glCopyTex(Sub)Image</TT> was
+    corrected.</li>
+<li>radeon/wine: flush command stream in more cases, fixing wine d3d9
+    tests.</li>
+<li>r600: fix sin+cos normalization.</li>
+<li>r600: (properly) ignore <TT>GL_COORD_REPLACE</TT> when point sprites are
+    disabled.</li>
+<li>radeon: avoid flushing when the context is not current.</li>
+<li>r300c: a bug affecting unaligned BOs was fixed.</li>
+<li>r300c: a hardlock caused by ARB_half_float_vertex incorrectly advertised on some chipsets.</li>
+</ul>
+
+
+</body>
+</html>
diff --git a/mesalib/docs/relnotes-7.9.1.html b/mesalib/docs/relnotes-7.9.1.html
index c12db4a50..bef4ef1dd 100644
--- a/mesalib/docs/relnotes-7.9.1.html
+++ b/mesalib/docs/relnotes-7.9.1.html
@@ -1,406 +1,406 @@
-<HTML>
-
-<head>
-<TITLE>Mesa Release Notes</TITLE>
-<link rel="stylesheet" type="text/css" href="mesa.css">
-<meta http-equiv="content-type" content="text/html; charset=utf-8" />
-</head>
-
-<BODY>
-
-<body bgcolor="#eeeeee">
-
-<H1>Mesa 7.9.1 Release Notes / January 7, 2011</H1>
-
-<p>
-Mesa 7.9.1 is a bug fix release which fixes bugs found since the 7.9 release.
-</p>
-<p>
-Mesa 7.9.1 implements the OpenGL 2.1 API, but the version reported by
-glGetString(GL_VERSION) depends on the particular driver being used.
-Some drivers don't support all the features required in OpenGL 2.1.
-</p>
-<p>
-See the <a href="install.html">Compiling/Installing page</a> for prerequisites
-for DRI hardware acceleration.
-</p>
-
-
-<h2>MD5 checksums</h2>
-<pre>
-78422843ea875ad4eac35b9b8584032b  MesaLib-7.9.1.tar.gz
-07dc6cfb5928840b8b9df5bd1b3ae434  MesaLib-7.9.1.tar.bz2
-c8eaea5b3c3d6dee784bd8c2db91c80f  MesaLib-7.9.1.zip
-ee9ecae4ca56fbb2d14dc15e3a0a7640  MesaGLUT-7.9.1.tar.gz
-41fc477d524e7dc5c84da8ef22422bea  MesaGLUT-7.9.1.tar.bz2
-90b287229afdf19317aa989d19462e7a  MesaGLUT-7.9.1.zip
-</pre>
-
-
-<h2>New features</h2>
-<p>None.</p>
-
-<h2>Bug fixes</h2>
-<p>This list is likely incomplete.</p>
-<ul>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28800">Bug 28800</a> - [r300c, r300g] Texture corruption with World of Warcraft</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29420">Bug 29420</a> - Amnesia / HPL2 RendererFeatTest - not rendering correctly</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29946">Bug 29946</a> - [swrast] piglit valgrind glsl-array-bounds-04 fails</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30261">Bug 30261</a> - [GLSL 1.20] allowing inconsistent invariant declaration between two vertex shaders</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30632">Bug 30632</a> - [softpipe] state_tracker/st_manager.c:489: st_context_notify_invalid_framebuffer: Assertion `stfb &amp;&amp; stfb-&gt;iface == stfbi' failed.</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30694">Bug 30694</a> - wincopy will crash on Gallium drivers when going to front buffer</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30787">Bug 30787</a> - Invalid asm shader does not generate draw-time error when used with GLSL shader</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30993">Bug 30993</a> - getFramebufferAttachmentParameteriv wrongly generates error</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31101">Bug 31101</a> -  [glsl2] abort() in ir_validate::visit_enter(ir_assignment *ir)</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31193">Bug 31193</a> -  [regression] aa43176e break water reflections</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31194">Bug 31194</a> - The mesa meta save/restore code doesn't ref the current GLSL program</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31371">Bug 31371</a> - glslparsertest: ir.cpp:358: ir_constant::ir_constant(const glsl_type*, const ir_constant_data*): Assertion `(type->base_type &gt;= 0) &amp;&amp; (type->base_type &lt;= 3)' failed.</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31439">Bug 31439</a> - Crash in glBufferSubData() with size == 0</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31495">Bug 31495</a> - [i965 gles2c bisected] OpenGL ES 2.0 conformance GL2Tests_GetBIFD_input.run regressed</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31514">Bug 31514</a> - isBuffer returns true for unbound buffers</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31560">Bug 31560</a> - [tdfx] tdfx_tex.c:702: error: ‘const struct gl_color_table’ has no member named ‘Format’</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31617">Bug 31617</a> - Radeon/Compiz: 'failed to attach dri2 front buffer', error case not handled</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31648">Bug 31648</a> -  [GLSL] array-struct-array gets assertion: `(size &gt;= 1) && (size &lt;= 4)' failed.</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31650">Bug 31650</a> - [GLSL] varying gl_TexCoord fails to be re-declared to different size in the second shader</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31673">Bug 31673</a> - GL_FRAGMENT_PRECISION_HIGH preprocessor macro undefined in GLSL ES</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31690">Bug 31690</a> -  i915 shader compiler fails to flatten if in Aquarium webgl demo.</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31832">Bug 31832</a> - [i915] Bad renderbuffer format: 21</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31841">Bug 31841</a> - [drm:radeon_cs_ioctl] *ERROR* Invalid command stream !</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31894">Bug 31894</a> - Writing to gl_PointSize with GLES2 corrupts other varyings</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31909">Bug 31909</a> - [i965] brw_fs.cpp:1461: void fs_visitor::emit_bool_to_cond_code(ir_rvalue*): Assertion `expr-&gt;operands[i]-&gt;type-&gt;is_scalar()' failed.</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31934">Bug 31934</a> - [gallium] Mapping empty buffer object causes SIGSEGV</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31983">Bug 31983</a> -  [i915 gles2] "if (expression with builtin/varying variables) discard" breaks linkage</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31985">Bug 31985</a> - [GLSL 1.20] initialized uniform array considered as "unsized"</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31987">Bug 31987</a> - [gles2] if input a wrong pname(GL_NONE) to glGetBoolean, it will not case GL_INVALID_ENUM</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32035">Bug 32035</a> - [GLSL bisected] comparing unsized array gets segfault</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32070">Bug 32070</a> - llvmpipe renders stencil demo incorrectly</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32273">Bug 32273</a> - assertion fails when starting vdrift 2010 release with shaders enabled</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32287">Bug 32287</a> - [bisected GLSL] float-int failure</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32311">Bug 32311</a> - [965 bisected] Array look-ups broken on GM45</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32520">Bug 32520</a> -  [gles2] glBlendFunc(GL_ZERO, GL_DST_COLOR) will result in GL_INVALID_ENUM</li>
-
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32825">Bug 32825</a> - egl_glx driver completely broken in 7.9 branch [fix in master]</li>
-</ul>
-
-
-<h2>Changes</h2>
-<p>The full set of changes can be viewed by using the following GIT command:</p>
-
-<pre>
-  git log mesa-7.9..mesa-7.9.1
-</pre>
-
-<p>Alex Deucher (5):
-<ul>
-  <li>r100: revalidate after radeon_update_renderbuffers</li>
-  <li>r600c: add missing radeon_prepare_render() call on evergreen</li>
-  <li>r600c: properly align mipmaps to group size</li>
-  <li>gallium/egl: fix r300 vs r600 loading</li>
-  <li>r600c: fix some opcodes on evergreen</li>
-</ul></p>
-
-<p>Aras Pranckevicius (2):
-<ul>
-  <li>glsl: fix crash in loop analysis when some controls can't be determined</li>
-  <li>glsl: fix matrix type check in ir_algebraic</li>
-</ul></p>
-
-<p>Brian Paul (27):
-<ul>
-  <li>swrast: fix choose_depth_texture_level() to respect mipmap filtering state</li>
-  <li>st/mesa: replace assertion w/ conditional in framebuffer invalidation</li>
-  <li>egl/i965: include inline_wrapper_sw_helper.h</li>
-  <li>mesa: Add missing else in do_row_3D</li>
-  <li>mesa: add missing formats in _mesa_format_to_type_and_comps()</li>
-  <li>mesa: handle more pixel types in mipmap generation code</li>
-  <li>mesa: make glIsBuffer() return false for never bound buffers</li>
-  <li>mesa: fix glDeleteBuffers() regression</li>
-  <li>swrast: init alpha value to 1.0 in opt_sample_rgb_2d()</li>
-  <li>meta: Mask Stencil.Clear against stencilMax in _mesa_meta_Clear</li>
-  <li>st/mesa: fix mapping of zero-sized buffer objects</li>
-  <li>mesa: check for posix_memalign() errors</li>
-  <li>llvmpipe: fix broken stencil writemask</li>
-  <li>mesa: fix GL_FRAMEBUFFER_ATTACHMENT_OBJECT_NAME query</li>
-  <li>mesa: return GL_FRAMEBUFFER_DEFAULT as FBO attachment type</li>
-  <li>mesa: make glGet*(GL_NONE) generate GL_INVALID_ENUM</li>
-  <li>mesa: test for cube map completeness in glGenerateMipmap()</li>
-  <li>tnl: Initialize gl_program_machine memory in run_vp.</li>
-  <li>tnl: a better way to initialize the gl_program_machine memory</li>
-  <li>mesa, st/mesa: disable GL_ARB_geometry_shader4</li>
-  <li>glsl: fix off by one in register index assertion</li>
-  <li>st/mesa: fix mipmap generation bug</li>
-  <li>glsl: new glsl_strtod() wrapper to fix decimal point interpretation</li>
-  <li>mesa: no-op glBufferSubData() on size==0</li>
-  <li>tdfx: s/Format/_BaseFormat/</li>
-  <li>st/mesa: fix renderbuffer pointer check in st_Clear()</li>
-  <li>mesa: Bump the number of bits in the register index.</li>
-</ul></p>
-
-<p>Chad Versace (5):
-<ul>
-  <li>glsl: Fix lexer rule for ^=</li>
-  <li>glsl: Fix ast-to-hir for ARB_fragment_coord_conventions</li>
-  <li>glsl: Fix ir_expression::constant_expression_value()</li>
-  <li>glsl: Fix erroneous cast in ast_jump_statement::hir()</li>
-  <li>glsl: Fix linker bug in cross_validate_globals()</li>
-</ul></p>
-
-<p>Chia-I Wu (10):
-<ul>
-  <li>targets/egl: Fix linking with libdrm.</li>
-  <li>st/vega: Fix version check in context creation.</li>
-  <li>st/egl: Do not finish a fence that is NULL.</li>
-  <li>egl: Fix a false negative check in _eglCheckMakeCurrent.</li>
-  <li>st/mesa: Unreference the sampler view in st_bind_surface.</li>
-  <li>egl_dri2: Fix __DRI_DRI2 version 1 support.</li>
-  <li>st/vega: Do not wait NULL fences.</li>
-  <li>mesa: Do not advertise GL_OES_texture_3D.</li>
-  <li>egl_glx: Fix borken driver.</li>
-  <li>egl: Check extensions.</li>
-</ul></p>
-
-<p>Daniel Lichtenberger (1):
-<ul>
-  <li>radeon: fix potential segfault in renderbuffer update</li>
-</ul></p>
-
-<p>Daniel Vetter (1):
-<ul>
-  <li>r200: revalidate after radeon_update_renderbuffers</li>
-</ul></p>
-
-<p>Dave Airlie (1):
-<ul>
-  <li>r300g: fixup rs690 tiling stride alignment calculations.</li>
-</ul></p>
-
-<p>Eric Anholt (13):
-<ul>
-  <li>intel: Allow CopyTexSubImage to InternalFormat 3/4 textures, like RGB/RGBA.</li>
-  <li>glsl: Free the loop state context when we free the loop state.</li>
-  <li>i965: Allow OPCODE_SWZ to put immediates in the first arg.</li>
-  <li>i965: Add support for rendering to SARGB8 FBOs.</li>
-  <li>glsl: Add a helper constructor for expressions that works out result type.</li>
-  <li>glsl: Fix structure and array comparisions.</li>
-  <li>glsl: Quiet unreachable no-return-from-function warning.</li>
-  <li>glsl: Mark the array access for whole-array comparisons.</li>
-  <li>glsl: Fix flipped return of has_value() for array constants.</li>
-  <li>mesa: Add getters for the rest of the supported draw buffers.</li>
-  <li>mesa: Add getters for ARB_copy_buffer's attachment points.</li>
-  <li>i965: Correct the dp_read message descriptor setup on g4x.</li>
-  <li>glsl: Correct the marking of InputsRead/OutputsWritten on in/out matrices.</li>
-</ul></p>
-
-<p>Fabian Bieler (1):
-<ul>
-  <li>glsl: fix lowering conditional returns in subroutines</li>
-</ul></p>
-
-<p>Francisco Jerez (3):
-<ul>
-  <li>meta: Don't leak alpha function/reference value changes.</li>
-  <li>meta: Fix incorrect rendering of the bitmap alpha component.</li>
-  <li>meta: Don't try to disable cube maps if the driver doesn't expose the extension.</li>
-</ul></p>
-
-<p>Henri Verbeet (2):
-<ul>
-  <li>r600: Evergreen has two extra frac_bits for the sampler LOD state.</li>
-  <li>st/mesa: Handle wrapped depth buffers in st_copy_texsubimage().</li>
-</ul></p>
-
-<p>Ian Romanick (33):
-<ul>
-  <li>Add 7.9 md5sums</li>
-  <li>docs: Import 7.8.x release notes from 7.8 branch.</li>
-  <li>docs: download.html does not need to be updated for each release</li>
-  <li>docs: Update mailing lines from sf.net to freedesktop.org</li>
-  <li>docs: added news item for 7.9 release</li>
-  <li>mesa: Validate assembly shaders when GLSL shaders are used</li>
-  <li>linker: Reject shaders that have unresolved function calls</li>
-  <li>mesa: Refactor validation of shader targets</li>
-  <li>glsl: Slightly change the semantic of _LinkedShaders</li>
-  <li>linker: Improve handling of unread/unwritten shader inputs/outputs</li>
-  <li>glsl: Commit lexer files changed by previous cherry picking</li>
-  <li>mesa: Make metaops use program refcounts instead of names.</li>
-  <li>glsl: Fix incorrect gl_type of sampler2DArray and sampler1DArrayShadow</li>
-  <li>mesa: Allow query of MAX_SAMPLES with EXT_framebuffer_multisample</li>
-  <li>glsl: better handling of linker failures</li>
-  <li>mesa: Fix glGet of ES2's GL_MAX_*_VECTORS properties.</li>
-  <li>i915: Disallow alpha, red, RG, and sRGB as render targets</li>
-  <li>glsl/linker: Free any IR discarded by optimization passes.</li>
-  <li>glsl: Add an optimization pass to simplify discards.</li>
-  <li>glsl: Add a lowering pass to move discards out of if-statements.</li>
-  <li>i915: Correctly generate unconditional KIL instructions</li>
-  <li>glsl: Add unary ir_expression constructor</li>
-  <li>glsl: Ensure that equality comparisons don't return a NULL IR tree</li>
-  <li>glcpp: Commit changes in generated files cause by previous commit</li>
-  <li>glsl: Inherrit type of declared variable from initializer</li>
-  <li>glsl: Inherrit type of declared variable from initializer after processing assignment</li>
-  <li>linker: Ensure that unsized arrays have a size after linking</li>
-  <li>linker: Fix regressions caused by previous commit</li>
-  <li>linker: Allow built-in arrays to have different sizes between shader stages</li>
-  <li>ir_to_mesa: Don't generate swizzles for record derefs of non-scalar/vectors</li>
-  <li>Refresh autogenerated file builtin_function.cpp.</li>
-  <li>docs: Initial set of release notes for 7.9.1</li>
-  <li>mesa: set version string to 7.9.1</li>
-</ul></p>
-
-<p>Julien Cristau (1):
-<ul>
-  <li>Makefile: don't include the same files twice in the tarball</li>
-</ul></p>
-
-<p>Kenneth Graunke (19):
-<ul>
-  <li>glcpp: Return NEWLINE token for newlines inside multi-line comments.</li>
-  <li>generate_builtins.py: Output large strings as arrays of characters.</li>
-  <li>glsl: Fix constant component count in vector constructor emitting.</li>
-  <li>ir_dead_functions: Actually free dead functions and signatures.</li>
-  <li>glcpp: Define GL_FRAGMENT_PRECISION_HIGH if GLSL version &gt;= 1.30.</li>
-  <li>glsl: Unconditionally define GL_FRAGMENT_PRECISION_HIGH in ES2 shaders.</li>
-  <li>glsl: Fix constant expression handling for &lt, &gt;, &lt=, &gt;= on vectors.</li>
-  <li>glsl: Use do_common_optimization in the standalone compiler.</li>
-  <li>glsl: Don't inline function prototypes.</li>
-  <li>glsl: Add a virtual as_discard() method.</li>
-  <li>glsl: Remove "discard" support from lower_jumps.</li>
-  <li>glsl: Refactor get_num_operands.</li>
-  <li>glcpp: Don't emit SPACE tokens in conditional_tokens production.</li>
-  <li>glsl: Clean up code by adding a new is_break() function.</li>
-  <li>glsl: Consider the "else" branch when looking for loop breaks.</li>
-  <li>Remove OES_compressed_paletted_texture from the ES2 extension list.</li>
-  <li>glsl/builtins: Compute the correct value for smoothstep(vec, vec, vec).</li>
-  <li>Fix build on systems where "python" is python 3.</li>
-  <li>i965: Internally enable GL_NV_blend_square on ES2.</li>
-</ul></p>
-
-<p>Kristian Høgsberg (1):
-<ul>
-  <li>i965: Don't write mrf assignment for pointsize output</li>
-</ul></p>
-
-<p>Luca Barbieri (1):
-<ul>
-  <li>glsl: Unroll loops with conditional breaks anywhere (not just the end)</li>
-</ul></p>
-
-<p>Marek Olšák (17):
-<ul>
-  <li>r300g: fix microtiling for 16-bits-per-channel formats</li>
-  <li>r300g: fix texture border for 16-bits-per-channel formats</li>
-  <li>r300g: add a default channel ordering of texture border for unhandled formats</li>
-  <li>r300g: fix texture border color for all texture formats</li>
-  <li>r300g: fix rendering with no vertex elements</li>
-  <li>r300/compiler: fix rc_rewrite_depth_out for it to work with any instruction</li>
-  <li>r300g: fix texture border color once again</li>
-  <li>r300g: fix texture swizzling with compressed textures on r400-r500</li>
-  <li>r300g: disable ARB_texture_swizzle if S3TC is enabled on r3xx-only</li>
-  <li>mesa, st/mesa: fix gl_FragCoord with FBOs in Gallium</li>
-  <li>st/mesa: initialize key in st_vp_varient</li>
-  <li>r300/compiler: fix swizzle lowering with a presubtract source operand</li>
-  <li>r300g: fix rendering with a vertex attrib having a zero stride</li>
-  <li>ir_to_mesa: Add support for conditional discards.</li>
-  <li>r300g: finally fix the texture corruption on r3xx-r4xx</li>
-  <li>mesa: fix texel store functions for some float formats</li>
-  <li>r300/compiler: disable the rename_regs pass for loops</li>
-</ul></p>
-
-<p>Mario Kleiner (1):
-<ul>
-  <li>mesa/r300classic: Fix dri2Invalidate/radeon_prepare_render for page flipping.</li>
-</ul></p>
-
-<p>Peter Clifton (1):
-<ul>
-  <li>intel: Fix emit_linear_blit to use DWORD aligned width blits</li>
-</ul></p>
-
-<p>Robert Hooker (2):
-<ul>
-  <li>intel: Add a new B43 pci id.</li>
-  <li>egl_dri2: Add missing intel chip ids.</li>
-</ul></p>
-
-<p>Roland Scheidegger (1):
-<ul>
-  <li>r200: fix r200 large points</li>
-</ul></p>
-
-<p>Thomas Hellstrom (17):
-<ul>
-  <li>st/xorg: Don't try to use option values before processing options</li>
-  <li>xorg/vmwgfx: Make vmwarectrl work also on 64-bit servers</li>
-  <li>st/xorg: Add a customizer option to get rid of annoying cursor update flicker</li>
-  <li>xorg/vmwgfx: Don't hide HW cursors when updating them</li>
-  <li>st/xorg: Don't try to remove invalid fbs</li>
-  <li>st/xorg: Fix typo</li>
-  <li>st/xorg, xorg/vmwgfx: Be a bit more frendly towards cross-compiling environments</li>
-  <li>st/xorg: Fix compilation errors for Xservers compiled without Composite</li>
-  <li>st/xorg: Don't use deprecated x*alloc / xfree functions</li>
-  <li>xorg/vmwgfx: Don't use deprecated x*alloc / xfree functions</li>
-  <li>st/xorg: Fix compilation for Xservers &gt;= 1.10</li>
-  <li>mesa: Make sure we have the talloc cflags when using the talloc headers</li>
-  <li>egl: Add an include for size_t</li>
-  <li>mesa: Add talloc includes for gles</li>
-  <li>st/egl: Fix build for include files in nonstandard places</li>
-  <li>svga/drm: Optionally resolve calls to powf during link-time</li>
-  <li>gallium/targets: Trivial crosscompiling fix</li>
-</ul></p>
-
-<p>Tom Stellard (7):
-<ul>
-  <li>r300/compiler: Make sure presubtract sources use supported swizzles</li>
-  <li>r300/compiler: Fix register allocator's handling of loops</li>
-  <li>r300/compiler: Fix instruction scheduling within IF blocks</li>
-  <li>r300/compiler: Use zero as the register index for unused sources</li>
-  <li>r300/compiler: Ignore alpha dest register when replicating the result</li>
-  <li>r300/compiler: Use correct swizzles for all presubtract sources</li>
-  <li>r300/compiler: Don't allow presubtract sources to be remapped twice</li>
-</ul></p>
-
-<p>Vinson Lee (1):
-<ul>
-  <li>glsl: Fix 'control reaches end of non-void function' warning.</li>
-</ul></p>
-
-<p>richard (1):
-<ul>
-  <li>r600c : inline vertex format is not updated in an app, switch to use vfetch constants. For the 7.9 and 7.10 branches as well.</li>
-</ul></p>
-
-</body>
-</html>
+<HTML>
+
+<head>
+<TITLE>Mesa Release Notes</TITLE>
+<link rel="stylesheet" type="text/css" href="mesa.css">
+<meta http-equiv="content-type" content="text/html; charset=utf-8" />
+</head>
+
+<BODY>
+
+<body bgcolor="#eeeeee">
+
+<H1>Mesa 7.9.1 Release Notes / January 7, 2011</H1>
+
+<p>
+Mesa 7.9.1 is a bug fix release which fixes bugs found since the 7.9 release.
+</p>
+<p>
+Mesa 7.9.1 implements the OpenGL 2.1 API, but the version reported by
+glGetString(GL_VERSION) depends on the particular driver being used.
+Some drivers don't support all the features required in OpenGL 2.1.
+</p>
+<p>
+See the <a href="install.html">Compiling/Installing page</a> for prerequisites
+for DRI hardware acceleration.
+</p>
+
+
+<h2>MD5 checksums</h2>
+<pre>
+78422843ea875ad4eac35b9b8584032b  MesaLib-7.9.1.tar.gz
+07dc6cfb5928840b8b9df5bd1b3ae434  MesaLib-7.9.1.tar.bz2
+c8eaea5b3c3d6dee784bd8c2db91c80f  MesaLib-7.9.1.zip
+ee9ecae4ca56fbb2d14dc15e3a0a7640  MesaGLUT-7.9.1.tar.gz
+41fc477d524e7dc5c84da8ef22422bea  MesaGLUT-7.9.1.tar.bz2
+90b287229afdf19317aa989d19462e7a  MesaGLUT-7.9.1.zip
+</pre>
+
+
+<h2>New features</h2>
+<p>None.</p>
+
+<h2>Bug fixes</h2>
+<p>This list is likely incomplete.</p>
+<ul>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28800">Bug 28800</a> - [r300c, r300g] Texture corruption with World of Warcraft</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29420">Bug 29420</a> - Amnesia / HPL2 RendererFeatTest - not rendering correctly</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29946">Bug 29946</a> - [swrast] piglit valgrind glsl-array-bounds-04 fails</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30261">Bug 30261</a> - [GLSL 1.20] allowing inconsistent invariant declaration between two vertex shaders</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30632">Bug 30632</a> - [softpipe] state_tracker/st_manager.c:489: st_context_notify_invalid_framebuffer: Assertion `stfb &amp;&amp; stfb-&gt;iface == stfbi' failed.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30694">Bug 30694</a> - wincopy will crash on Gallium drivers when going to front buffer</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30787">Bug 30787</a> - Invalid asm shader does not generate draw-time error when used with GLSL shader</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30993">Bug 30993</a> - getFramebufferAttachmentParameteriv wrongly generates error</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31101">Bug 31101</a> -  [glsl2] abort() in ir_validate::visit_enter(ir_assignment *ir)</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31193">Bug 31193</a> -  [regression] aa43176e break water reflections</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31194">Bug 31194</a> - The mesa meta save/restore code doesn't ref the current GLSL program</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31371">Bug 31371</a> - glslparsertest: ir.cpp:358: ir_constant::ir_constant(const glsl_type*, const ir_constant_data*): Assertion `(type->base_type &gt;= 0) &amp;&amp; (type->base_type &lt;= 3)' failed.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31439">Bug 31439</a> - Crash in glBufferSubData() with size == 0</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31495">Bug 31495</a> - [i965 gles2c bisected] OpenGL ES 2.0 conformance GL2Tests_GetBIFD_input.run regressed</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31514">Bug 31514</a> - isBuffer returns true for unbound buffers</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31560">Bug 31560</a> - [tdfx] tdfx_tex.c:702: error: ‘const struct gl_color_table’ has no member named ‘Format’</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31617">Bug 31617</a> - Radeon/Compiz: 'failed to attach dri2 front buffer', error case not handled</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31648">Bug 31648</a> -  [GLSL] array-struct-array gets assertion: `(size &gt;= 1) && (size &lt;= 4)' failed.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31650">Bug 31650</a> - [GLSL] varying gl_TexCoord fails to be re-declared to different size in the second shader</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31673">Bug 31673</a> - GL_FRAGMENT_PRECISION_HIGH preprocessor macro undefined in GLSL ES</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31690">Bug 31690</a> -  i915 shader compiler fails to flatten if in Aquarium webgl demo.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31832">Bug 31832</a> - [i915] Bad renderbuffer format: 21</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31841">Bug 31841</a> - [drm:radeon_cs_ioctl] *ERROR* Invalid command stream !</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31894">Bug 31894</a> - Writing to gl_PointSize with GLES2 corrupts other varyings</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31909">Bug 31909</a> - [i965] brw_fs.cpp:1461: void fs_visitor::emit_bool_to_cond_code(ir_rvalue*): Assertion `expr-&gt;operands[i]-&gt;type-&gt;is_scalar()' failed.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31934">Bug 31934</a> - [gallium] Mapping empty buffer object causes SIGSEGV</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31983">Bug 31983</a> -  [i915 gles2] "if (expression with builtin/varying variables) discard" breaks linkage</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31985">Bug 31985</a> - [GLSL 1.20] initialized uniform array considered as "unsized"</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31987">Bug 31987</a> - [gles2] if input a wrong pname(GL_NONE) to glGetBoolean, it will not case GL_INVALID_ENUM</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32035">Bug 32035</a> - [GLSL bisected] comparing unsized array gets segfault</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32070">Bug 32070</a> - llvmpipe renders stencil demo incorrectly</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32273">Bug 32273</a> - assertion fails when starting vdrift 2010 release with shaders enabled</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32287">Bug 32287</a> - [bisected GLSL] float-int failure</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32311">Bug 32311</a> - [965 bisected] Array look-ups broken on GM45</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32520">Bug 32520</a> -  [gles2] glBlendFunc(GL_ZERO, GL_DST_COLOR) will result in GL_INVALID_ENUM</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32825">Bug 32825</a> - egl_glx driver completely broken in 7.9 branch [fix in master]</li>
+</ul>
+
+
+<h2>Changes</h2>
+<p>The full set of changes can be viewed by using the following GIT command:</p>
+
+<pre>
+  git log mesa-7.9..mesa-7.9.1
+</pre>
+
+<p>Alex Deucher (5):
+<ul>
+  <li>r100: revalidate after radeon_update_renderbuffers</li>
+  <li>r600c: add missing radeon_prepare_render() call on evergreen</li>
+  <li>r600c: properly align mipmaps to group size</li>
+  <li>gallium/egl: fix r300 vs r600 loading</li>
+  <li>r600c: fix some opcodes on evergreen</li>
+</ul></p>
+
+<p>Aras Pranckevicius (2):
+<ul>
+  <li>glsl: fix crash in loop analysis when some controls can't be determined</li>
+  <li>glsl: fix matrix type check in ir_algebraic</li>
+</ul></p>
+
+<p>Brian Paul (27):
+<ul>
+  <li>swrast: fix choose_depth_texture_level() to respect mipmap filtering state</li>
+  <li>st/mesa: replace assertion w/ conditional in framebuffer invalidation</li>
+  <li>egl/i965: include inline_wrapper_sw_helper.h</li>
+  <li>mesa: Add missing else in do_row_3D</li>
+  <li>mesa: add missing formats in _mesa_format_to_type_and_comps()</li>
+  <li>mesa: handle more pixel types in mipmap generation code</li>
+  <li>mesa: make glIsBuffer() return false for never bound buffers</li>
+  <li>mesa: fix glDeleteBuffers() regression</li>
+  <li>swrast: init alpha value to 1.0 in opt_sample_rgb_2d()</li>
+  <li>meta: Mask Stencil.Clear against stencilMax in _mesa_meta_Clear</li>
+  <li>st/mesa: fix mapping of zero-sized buffer objects</li>
+  <li>mesa: check for posix_memalign() errors</li>
+  <li>llvmpipe: fix broken stencil writemask</li>
+  <li>mesa: fix GL_FRAMEBUFFER_ATTACHMENT_OBJECT_NAME query</li>
+  <li>mesa: return GL_FRAMEBUFFER_DEFAULT as FBO attachment type</li>
+  <li>mesa: make glGet*(GL_NONE) generate GL_INVALID_ENUM</li>
+  <li>mesa: test for cube map completeness in glGenerateMipmap()</li>
+  <li>tnl: Initialize gl_program_machine memory in run_vp.</li>
+  <li>tnl: a better way to initialize the gl_program_machine memory</li>
+  <li>mesa, st/mesa: disable GL_ARB_geometry_shader4</li>
+  <li>glsl: fix off by one in register index assertion</li>
+  <li>st/mesa: fix mipmap generation bug</li>
+  <li>glsl: new glsl_strtod() wrapper to fix decimal point interpretation</li>
+  <li>mesa: no-op glBufferSubData() on size==0</li>
+  <li>tdfx: s/Format/_BaseFormat/</li>
+  <li>st/mesa: fix renderbuffer pointer check in st_Clear()</li>
+  <li>mesa: Bump the number of bits in the register index.</li>
+</ul></p>
+
+<p>Chad Versace (5):
+<ul>
+  <li>glsl: Fix lexer rule for ^=</li>
+  <li>glsl: Fix ast-to-hir for ARB_fragment_coord_conventions</li>
+  <li>glsl: Fix ir_expression::constant_expression_value()</li>
+  <li>glsl: Fix erroneous cast in ast_jump_statement::hir()</li>
+  <li>glsl: Fix linker bug in cross_validate_globals()</li>
+</ul></p>
+
+<p>Chia-I Wu (10):
+<ul>
+  <li>targets/egl: Fix linking with libdrm.</li>
+  <li>st/vega: Fix version check in context creation.</li>
+  <li>st/egl: Do not finish a fence that is NULL.</li>
+  <li>egl: Fix a false negative check in _eglCheckMakeCurrent.</li>
+  <li>st/mesa: Unreference the sampler view in st_bind_surface.</li>
+  <li>egl_dri2: Fix __DRI_DRI2 version 1 support.</li>
+  <li>st/vega: Do not wait NULL fences.</li>
+  <li>mesa: Do not advertise GL_OES_texture_3D.</li>
+  <li>egl_glx: Fix borken driver.</li>
+  <li>egl: Check extensions.</li>
+</ul></p>
+
+<p>Daniel Lichtenberger (1):
+<ul>
+  <li>radeon: fix potential segfault in renderbuffer update</li>
+</ul></p>
+
+<p>Daniel Vetter (1):
+<ul>
+  <li>r200: revalidate after radeon_update_renderbuffers</li>
+</ul></p>
+
+<p>Dave Airlie (1):
+<ul>
+  <li>r300g: fixup rs690 tiling stride alignment calculations.</li>
+</ul></p>
+
+<p>Eric Anholt (13):
+<ul>
+  <li>intel: Allow CopyTexSubImage to InternalFormat 3/4 textures, like RGB/RGBA.</li>
+  <li>glsl: Free the loop state context when we free the loop state.</li>
+  <li>i965: Allow OPCODE_SWZ to put immediates in the first arg.</li>
+  <li>i965: Add support for rendering to SARGB8 FBOs.</li>
+  <li>glsl: Add a helper constructor for expressions that works out result type.</li>
+  <li>glsl: Fix structure and array comparisions.</li>
+  <li>glsl: Quiet unreachable no-return-from-function warning.</li>
+  <li>glsl: Mark the array access for whole-array comparisons.</li>
+  <li>glsl: Fix flipped return of has_value() for array constants.</li>
+  <li>mesa: Add getters for the rest of the supported draw buffers.</li>
+  <li>mesa: Add getters for ARB_copy_buffer's attachment points.</li>
+  <li>i965: Correct the dp_read message descriptor setup on g4x.</li>
+  <li>glsl: Correct the marking of InputsRead/OutputsWritten on in/out matrices.</li>
+</ul></p>
+
+<p>Fabian Bieler (1):
+<ul>
+  <li>glsl: fix lowering conditional returns in subroutines</li>
+</ul></p>
+
+<p>Francisco Jerez (3):
+<ul>
+  <li>meta: Don't leak alpha function/reference value changes.</li>
+  <li>meta: Fix incorrect rendering of the bitmap alpha component.</li>
+  <li>meta: Don't try to disable cube maps if the driver doesn't expose the extension.</li>
+</ul></p>
+
+<p>Henri Verbeet (2):
+<ul>
+  <li>r600: Evergreen has two extra frac_bits for the sampler LOD state.</li>
+  <li>st/mesa: Handle wrapped depth buffers in st_copy_texsubimage().</li>
+</ul></p>
+
+<p>Ian Romanick (33):
+<ul>
+  <li>Add 7.9 md5sums</li>
+  <li>docs: Import 7.8.x release notes from 7.8 branch.</li>
+  <li>docs: download.html does not need to be updated for each release</li>
+  <li>docs: Update mailing lines from sf.net to freedesktop.org</li>
+  <li>docs: added news item for 7.9 release</li>
+  <li>mesa: Validate assembly shaders when GLSL shaders are used</li>
+  <li>linker: Reject shaders that have unresolved function calls</li>
+  <li>mesa: Refactor validation of shader targets</li>
+  <li>glsl: Slightly change the semantic of _LinkedShaders</li>
+  <li>linker: Improve handling of unread/unwritten shader inputs/outputs</li>
+  <li>glsl: Commit lexer files changed by previous cherry picking</li>
+  <li>mesa: Make metaops use program refcounts instead of names.</li>
+  <li>glsl: Fix incorrect gl_type of sampler2DArray and sampler1DArrayShadow</li>
+  <li>mesa: Allow query of MAX_SAMPLES with EXT_framebuffer_multisample</li>
+  <li>glsl: better handling of linker failures</li>
+  <li>mesa: Fix glGet of ES2's GL_MAX_*_VECTORS properties.</li>
+  <li>i915: Disallow alpha, red, RG, and sRGB as render targets</li>
+  <li>glsl/linker: Free any IR discarded by optimization passes.</li>
+  <li>glsl: Add an optimization pass to simplify discards.</li>
+  <li>glsl: Add a lowering pass to move discards out of if-statements.</li>
+  <li>i915: Correctly generate unconditional KIL instructions</li>
+  <li>glsl: Add unary ir_expression constructor</li>
+  <li>glsl: Ensure that equality comparisons don't return a NULL IR tree</li>
+  <li>glcpp: Commit changes in generated files cause by previous commit</li>
+  <li>glsl: Inherrit type of declared variable from initializer</li>
+  <li>glsl: Inherrit type of declared variable from initializer after processing assignment</li>
+  <li>linker: Ensure that unsized arrays have a size after linking</li>
+  <li>linker: Fix regressions caused by previous commit</li>
+  <li>linker: Allow built-in arrays to have different sizes between shader stages</li>
+  <li>ir_to_mesa: Don't generate swizzles for record derefs of non-scalar/vectors</li>
+  <li>Refresh autogenerated file builtin_function.cpp.</li>
+  <li>docs: Initial set of release notes for 7.9.1</li>
+  <li>mesa: set version string to 7.9.1</li>
+</ul></p>
+
+<p>Julien Cristau (1):
+<ul>
+  <li>Makefile: don't include the same files twice in the tarball</li>
+</ul></p>
+
+<p>Kenneth Graunke (19):
+<ul>
+  <li>glcpp: Return NEWLINE token for newlines inside multi-line comments.</li>
+  <li>generate_builtins.py: Output large strings as arrays of characters.</li>
+  <li>glsl: Fix constant component count in vector constructor emitting.</li>
+  <li>ir_dead_functions: Actually free dead functions and signatures.</li>
+  <li>glcpp: Define GL_FRAGMENT_PRECISION_HIGH if GLSL version &gt;= 1.30.</li>
+  <li>glsl: Unconditionally define GL_FRAGMENT_PRECISION_HIGH in ES2 shaders.</li>
+  <li>glsl: Fix constant expression handling for &lt, &gt;, &lt=, &gt;= on vectors.</li>
+  <li>glsl: Use do_common_optimization in the standalone compiler.</li>
+  <li>glsl: Don't inline function prototypes.</li>
+  <li>glsl: Add a virtual as_discard() method.</li>
+  <li>glsl: Remove "discard" support from lower_jumps.</li>
+  <li>glsl: Refactor get_num_operands.</li>
+  <li>glcpp: Don't emit SPACE tokens in conditional_tokens production.</li>
+  <li>glsl: Clean up code by adding a new is_break() function.</li>
+  <li>glsl: Consider the "else" branch when looking for loop breaks.</li>
+  <li>Remove OES_compressed_paletted_texture from the ES2 extension list.</li>
+  <li>glsl/builtins: Compute the correct value for smoothstep(vec, vec, vec).</li>
+  <li>Fix build on systems where "python" is python 3.</li>
+  <li>i965: Internally enable GL_NV_blend_square on ES2.</li>
+</ul></p>
+
+<p>Kristian Høgsberg (1):
+<ul>
+  <li>i965: Don't write mrf assignment for pointsize output</li>
+</ul></p>
+
+<p>Luca Barbieri (1):
+<ul>
+  <li>glsl: Unroll loops with conditional breaks anywhere (not just the end)</li>
+</ul></p>
+
+<p>Marek Olšák (17):
+<ul>
+  <li>r300g: fix microtiling for 16-bits-per-channel formats</li>
+  <li>r300g: fix texture border for 16-bits-per-channel formats</li>
+  <li>r300g: add a default channel ordering of texture border for unhandled formats</li>
+  <li>r300g: fix texture border color for all texture formats</li>
+  <li>r300g: fix rendering with no vertex elements</li>
+  <li>r300/compiler: fix rc_rewrite_depth_out for it to work with any instruction</li>
+  <li>r300g: fix texture border color once again</li>
+  <li>r300g: fix texture swizzling with compressed textures on r400-r500</li>
+  <li>r300g: disable ARB_texture_swizzle if S3TC is enabled on r3xx-only</li>
+  <li>mesa, st/mesa: fix gl_FragCoord with FBOs in Gallium</li>
+  <li>st/mesa: initialize key in st_vp_varient</li>
+  <li>r300/compiler: fix swizzle lowering with a presubtract source operand</li>
+  <li>r300g: fix rendering with a vertex attrib having a zero stride</li>
+  <li>ir_to_mesa: Add support for conditional discards.</li>
+  <li>r300g: finally fix the texture corruption on r3xx-r4xx</li>
+  <li>mesa: fix texel store functions for some float formats</li>
+  <li>r300/compiler: disable the rename_regs pass for loops</li>
+</ul></p>
+
+<p>Mario Kleiner (1):
+<ul>
+  <li>mesa/r300classic: Fix dri2Invalidate/radeon_prepare_render for page flipping.</li>
+</ul></p>
+
+<p>Peter Clifton (1):
+<ul>
+  <li>intel: Fix emit_linear_blit to use DWORD aligned width blits</li>
+</ul></p>
+
+<p>Robert Hooker (2):
+<ul>
+  <li>intel: Add a new B43 pci id.</li>
+  <li>egl_dri2: Add missing intel chip ids.</li>
+</ul></p>
+
+<p>Roland Scheidegger (1):
+<ul>
+  <li>r200: fix r200 large points</li>
+</ul></p>
+
+<p>Thomas Hellstrom (17):
+<ul>
+  <li>st/xorg: Don't try to use option values before processing options</li>
+  <li>xorg/vmwgfx: Make vmwarectrl work also on 64-bit servers</li>
+  <li>st/xorg: Add a customizer option to get rid of annoying cursor update flicker</li>
+  <li>xorg/vmwgfx: Don't hide HW cursors when updating them</li>
+  <li>st/xorg: Don't try to remove invalid fbs</li>
+  <li>st/xorg: Fix typo</li>
+  <li>st/xorg, xorg/vmwgfx: Be a bit more frendly towards cross-compiling environments</li>
+  <li>st/xorg: Fix compilation errors for Xservers compiled without Composite</li>
+  <li>st/xorg: Don't use deprecated x*alloc / xfree functions</li>
+  <li>xorg/vmwgfx: Don't use deprecated x*alloc / xfree functions</li>
+  <li>st/xorg: Fix compilation for Xservers &gt;= 1.10</li>
+  <li>mesa: Make sure we have the talloc cflags when using the talloc headers</li>
+  <li>egl: Add an include for size_t</li>
+  <li>mesa: Add talloc includes for gles</li>
+  <li>st/egl: Fix build for include files in nonstandard places</li>
+  <li>svga/drm: Optionally resolve calls to powf during link-time</li>
+  <li>gallium/targets: Trivial crosscompiling fix</li>
+</ul></p>
+
+<p>Tom Stellard (7):
+<ul>
+  <li>r300/compiler: Make sure presubtract sources use supported swizzles</li>
+  <li>r300/compiler: Fix register allocator's handling of loops</li>
+  <li>r300/compiler: Fix instruction scheduling within IF blocks</li>
+  <li>r300/compiler: Use zero as the register index for unused sources</li>
+  <li>r300/compiler: Ignore alpha dest register when replicating the result</li>
+  <li>r300/compiler: Use correct swizzles for all presubtract sources</li>
+  <li>r300/compiler: Don't allow presubtract sources to be remapped twice</li>
+</ul></p>
+
+<p>Vinson Lee (1):
+<ul>
+  <li>glsl: Fix 'control reaches end of non-void function' warning.</li>
+</ul></p>
+
+<p>richard (1):
+<ul>
+  <li>r600c : inline vertex format is not updated in an app, switch to use vfetch constants. For the 7.9 and 7.10 branches as well.</li>
+</ul></p>
+
+</body>
+</html>
diff --git a/mesalib/docs/relnotes-7.9.html b/mesalib/docs/relnotes-7.9.html
index c94ebeef5..fe1fd5e41 100644
--- a/mesalib/docs/relnotes-7.9.html
+++ b/mesalib/docs/relnotes-7.9.html
@@ -1,137 +1,137 @@
-<HTML>
-
-<TITLE>Mesa Release Notes</TITLE>
-
-<head><link rel="stylesheet" type="text/css" href="mesa.css"></head>
-
-<BODY>
-
-<body bgcolor="#eeeeee">
-
-<H1>Mesa 7.9 Release Notes / October 4, 2010</H1>
-
-<p>
-Mesa 7.9 is a new development release.
-People who are concerned with stability and reliability should stick
-with a previous release or wait for Mesa 7.9.1.
-</p>
-<p>
-Mesa 7.9 implements the OpenGL 2.1 API, but the version reported by
-glGetString(GL_VERSION) depends on the particular driver being used.
-Some drivers don't support all the features required in OpenGL 2.1.
-</p>
-<p>
-See the <a href="install.html">Compiling/Installing page</a> for prerequisites
-for DRI hardware acceleration.
-</p>
-
-
-<h2>MD5 checksums</h2>
-<pre>
-ed65ab425b25895c7f473d0a5e6e64f8  MesaLib-7.9.tar.gz
-82c740c49d572baa6da2b1a1eee90bca  MesaLib-7.9.tar.bz2
-cd2b6ecec759b0457475e94bbb38fedb  MesaLib-7.9.zip
-7b54af9fb9b1f6a1a65db2520f50848f  MesaGLUT-7.9.tar.gz
-20d07419d1929f833fdb36bced290ad5  MesaGLUT-7.9.tar.bz2
-62a7edecd7c92675cd6029b05217eb0a  MesaGLUT-7.9.zip
-</pre>
-
-
-<h2>New features</h2>
-<ul>
-<li>New, improved GLSL compiler written by Intel.
-    See the <a href="shading.html"> Shading Language</a> page for
-    more information.
-<li>New, very experimental Gallium driver for R600-R700 Radeons.
-<li>Support for AMD Evergreen-based Radeons (HD 5xxx)
-<li>GL_EXT_timer_query extension (i965 driver and softpipe only)
-<li>GL_EXT_framebuffer_multisample extension (intel drivers, MAX_SAMPLES = 1)
-<li>GL_ARB_texture_swizzle extension (alias of GL_EXT_texture_swizzle)
-<li>GL_ARB_draw_elements_base_vertex, GL_ARB_fragment_program_shadow,
-    GL_ARB_window_pos, GL_EXT_gpu_program_parameters,
-    GL_ATI_texture_env_combine3, GL_MESA_pack_invert, and GL_OES_EGL_image
-    extensions in Gallium drivers
-<li>GL_ARB_depth_clamp and GL_NV_depth_clamp extensions (in nv50 and r600
-    Gallium drivers)
-<li>GL_ARB_half_float_vertex extension (in nvfx, r300, r600, softpipe,
-    and llvmpipe Gallium drivers)
-<li>GL_EXT_draw_buffers2 (in nv50, r600, softpipe, and llvmpipe Gallium
-    drivers)
-<li>GL_EXT_texture_swizzle (in nvfx, r300, r600, softpipe, and llvmpipe
-    Gallium drivers)
-<li>GL_ATI_texture_mirror_once (in nvfx, nv50, r300, r600, softpipe, and
-    llvmpipe Gallium drivers)
-<li>GL_NV_conditional_render (in r300 Gallium driver)
-<li>Initial "signs of life" support for Sandybridge hardware in i965 DRI
-    driver.
-</ul>
-
-
-<h2>Bug fixes</h2>
-<p>This list is likely incomplete.</p>
-<ul>
-<li>Massive improvements to the Gallium driver for R300-R500 Radeons; this
-    driver is now considered stable for use as a DRI (OpenGL) driver.
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=10908">Bug 10908</a> - GLSL: gl_FogParamaters gl_Fog built-in uniform not functioning</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=13753">Bug 13753</a> - Numerous bugs in GLSL uniform handling</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=16854">Bug 16854</a> - GLSL function call at global scope causes SEGV</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=16856">Bug 16856</a> - GLSL indexing of unsized array results in assertion failure</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=18659">Bug 18659</a> - Crash in shader/slang/slang_codegen.c _slang_gen_function_call_name()</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=19089">Bug 19089</a> - [GLSL] glsl1/shadow2D() cases fail</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=22622">Bug 22622</a> - [GM965 GLSL] noise*() cause GPU lockup</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=23743">Bug 23743</a> - For loop from 0 to 0 not optimized out</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=24553">Bug 24553</a> - shader compilation times explode when using more () pairs</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25664">Bug 25664</a> - [GLSL] re-declaring an empty array fails to compile</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25769">Bug 25769</a> - [GLSL] "float" can be implicitly converted to "int"</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25808">Bug 25808</a> - [GLSL] const variable is modified successfully</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25826">Bug 25826</a> - [GLSL] declaring an unsized array then re-declaring with a size fails</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25827">Bug 25827</a> - [GLSL] vector constructor accepts too many arguments successfully</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25829">Bug 25829</a> - [GLSL] allowing non-void function without returning value</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25830">Bug 25830</a> - [GLSL] allowing non-constant-expression as const declaration initializer</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25877">Bug 25877</a> - [GLSL 1.10] implicit conversion from "int" to "float" should not be allowed</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25878">Bug 25878</a> - [GLSL] sampler is converted to int successfully</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25994">Bug 25994</a> - [GM45][GLSL] 'return' statement in vertex shader unsupported</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25999">Bug 25999</a> - [GLSL] embedded structure constructor fails to compile</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=26000">Bug 26000</a> - [GLSL] allowing different parameter qualifier between the function definition and declaration</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=26001">Bug 26001</a> - [GLSL 1.10] constructing matrix from matrix succeeds</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=26224">Bug 26224</a> - [GLSL] Cannot get location of a uniform struct member</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=26990">Bug 26990</a> - [GLSL] variable declaration in "while" fails to compile</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=27004">Bug 27004</a> - [GLSL] allowing macro redefinition</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=27060">Bug 27060</a> - [965] piglit glsl-fs-raytrace failure due to lack of function calls.</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=27216">Bug 27216</a> - Assignment with a function call in an if statement causes an assertion failure</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=27261">Bug 27261</a> - GLSL Compiler fails on the following vertex shader</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=27265">Bug 27265</a> - GLSL Compiler doesnt link the attached vertex shader</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=27388">Bug 27388</a> - [i965] piglit glsl-vs-arrays failure</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=27403">Bug 27403</a> - GLSL struct causing "Invalid src register file ..." error</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=27914">Bug 27914</a> - GLSL compiler uses MUL+ADD where it could use MAD</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28055">Bug 28055</a> - glsl-texcoord-array fails GLSL compilation</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28374">Bug 28374</a> - SIGSEGV shader/slang/slang_typeinfo.c:534</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28748">Bug 28748</a> - [i965] uninlined function calls support</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28833">Bug 28833</a> - piglit/shaders/glsl-texcoord-array fail</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28834">Bug 28834</a> - Add support for system fpclassify to GL_OES_query_matrix function for OpenBSD / NetBSD</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28837">Bug 28837</a> - varying vec4 index support</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28845">Bug 28845</a> - The GLU tesselator code has some warnings</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28889">Bug 28889</a> - [regression] wine game crash</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28894">Bug 28894</a> - slang build fails if absolute path contains spaces</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28913">Bug 28913</a> - [GLSL] allowing two version statements</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28931">Bug 28931</a> - Floating Point Exception in Warzone2100 Trunk version</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28966">Bug 28966</a> - [r300g] Dynamic branching 3 demo does not run</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28967">Bug 28967</a> - slang/slang_emit.c:350: storage_to_src_reg: Assertion `index &gt;= 0' failed.</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29013">Bug 29013</a> - [r300g] translate_rgb_op: unknown opcode ILLEGAL OPCODE</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29020">Bug 29020</a> - [r300g] Wine d3d9 tests hardlock</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29910">Bug 29910</a> - Mesa advertises bogus GL_ARB_shading_language_120</li>
-<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30196">Bug 30196</a> - [GLSL] gl_TextureMatrix{Inverse,Transpose,InverseTranspose} unsupported</li>
-</ul>
-
-
-<h2>Changes</h2>
-<ul>
-<li>The Mesa demo/test programs have been moved into a separate git
-repository.
-<li>GL/glext.h file upgraded to version 64
-<li>GL/glxext.h file upgraded to version 32
-<li>GL/wglext.h file upgraded to version 22
-</ul>
-
-</body>
-</html>
+<HTML>
+
+<TITLE>Mesa Release Notes</TITLE>
+
+<head><link rel="stylesheet" type="text/css" href="mesa.css"></head>
+
+<BODY>
+
+<body bgcolor="#eeeeee">
+
+<H1>Mesa 7.9 Release Notes / October 4, 2010</H1>
+
+<p>
+Mesa 7.9 is a new development release.
+People who are concerned with stability and reliability should stick
+with a previous release or wait for Mesa 7.9.1.
+</p>
+<p>
+Mesa 7.9 implements the OpenGL 2.1 API, but the version reported by
+glGetString(GL_VERSION) depends on the particular driver being used.
+Some drivers don't support all the features required in OpenGL 2.1.
+</p>
+<p>
+See the <a href="install.html">Compiling/Installing page</a> for prerequisites
+for DRI hardware acceleration.
+</p>
+
+
+<h2>MD5 checksums</h2>
+<pre>
+ed65ab425b25895c7f473d0a5e6e64f8  MesaLib-7.9.tar.gz
+82c740c49d572baa6da2b1a1eee90bca  MesaLib-7.9.tar.bz2
+cd2b6ecec759b0457475e94bbb38fedb  MesaLib-7.9.zip
+7b54af9fb9b1f6a1a65db2520f50848f  MesaGLUT-7.9.tar.gz
+20d07419d1929f833fdb36bced290ad5  MesaGLUT-7.9.tar.bz2
+62a7edecd7c92675cd6029b05217eb0a  MesaGLUT-7.9.zip
+</pre>
+
+
+<h2>New features</h2>
+<ul>
+<li>New, improved GLSL compiler written by Intel.
+    See the <a href="shading.html"> Shading Language</a> page for
+    more information.
+<li>New, very experimental Gallium driver for R600-R700 Radeons.
+<li>Support for AMD Evergreen-based Radeons (HD 5xxx)
+<li>GL_EXT_timer_query extension (i965 driver and softpipe only)
+<li>GL_EXT_framebuffer_multisample extension (intel drivers, MAX_SAMPLES = 1)
+<li>GL_ARB_texture_swizzle extension (alias of GL_EXT_texture_swizzle)
+<li>GL_ARB_draw_elements_base_vertex, GL_ARB_fragment_program_shadow,
+    GL_ARB_window_pos, GL_EXT_gpu_program_parameters,
+    GL_ATI_texture_env_combine3, GL_MESA_pack_invert, and GL_OES_EGL_image
+    extensions in Gallium drivers
+<li>GL_ARB_depth_clamp and GL_NV_depth_clamp extensions (in nv50 and r600
+    Gallium drivers)
+<li>GL_ARB_half_float_vertex extension (in nvfx, r300, r600, softpipe,
+    and llvmpipe Gallium drivers)
+<li>GL_EXT_draw_buffers2 (in nv50, r600, softpipe, and llvmpipe Gallium
+    drivers)
+<li>GL_EXT_texture_swizzle (in nvfx, r300, r600, softpipe, and llvmpipe
+    Gallium drivers)
+<li>GL_ATI_texture_mirror_once (in nvfx, nv50, r300, r600, softpipe, and
+    llvmpipe Gallium drivers)
+<li>GL_NV_conditional_render (in r300 Gallium driver)
+<li>Initial "signs of life" support for Sandybridge hardware in i965 DRI
+    driver.
+</ul>
+
+
+<h2>Bug fixes</h2>
+<p>This list is likely incomplete.</p>
+<ul>
+<li>Massive improvements to the Gallium driver for R300-R500 Radeons; this
+    driver is now considered stable for use as a DRI (OpenGL) driver.
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=10908">Bug 10908</a> - GLSL: gl_FogParamaters gl_Fog built-in uniform not functioning</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=13753">Bug 13753</a> - Numerous bugs in GLSL uniform handling</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=16854">Bug 16854</a> - GLSL function call at global scope causes SEGV</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=16856">Bug 16856</a> - GLSL indexing of unsized array results in assertion failure</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=18659">Bug 18659</a> - Crash in shader/slang/slang_codegen.c _slang_gen_function_call_name()</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=19089">Bug 19089</a> - [GLSL] glsl1/shadow2D() cases fail</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=22622">Bug 22622</a> - [GM965 GLSL] noise*() cause GPU lockup</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=23743">Bug 23743</a> - For loop from 0 to 0 not optimized out</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=24553">Bug 24553</a> - shader compilation times explode when using more () pairs</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25664">Bug 25664</a> - [GLSL] re-declaring an empty array fails to compile</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25769">Bug 25769</a> - [GLSL] "float" can be implicitly converted to "int"</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25808">Bug 25808</a> - [GLSL] const variable is modified successfully</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25826">Bug 25826</a> - [GLSL] declaring an unsized array then re-declaring with a size fails</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25827">Bug 25827</a> - [GLSL] vector constructor accepts too many arguments successfully</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25829">Bug 25829</a> - [GLSL] allowing non-void function without returning value</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25830">Bug 25830</a> - [GLSL] allowing non-constant-expression as const declaration initializer</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25877">Bug 25877</a> - [GLSL 1.10] implicit conversion from "int" to "float" should not be allowed</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25878">Bug 25878</a> - [GLSL] sampler is converted to int successfully</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25994">Bug 25994</a> - [GM45][GLSL] 'return' statement in vertex shader unsupported</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=25999">Bug 25999</a> - [GLSL] embedded structure constructor fails to compile</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=26000">Bug 26000</a> - [GLSL] allowing different parameter qualifier between the function definition and declaration</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=26001">Bug 26001</a> - [GLSL 1.10] constructing matrix from matrix succeeds</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=26224">Bug 26224</a> - [GLSL] Cannot get location of a uniform struct member</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=26990">Bug 26990</a> - [GLSL] variable declaration in "while" fails to compile</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=27004">Bug 27004</a> - [GLSL] allowing macro redefinition</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=27060">Bug 27060</a> - [965] piglit glsl-fs-raytrace failure due to lack of function calls.</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=27216">Bug 27216</a> - Assignment with a function call in an if statement causes an assertion failure</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=27261">Bug 27261</a> - GLSL Compiler fails on the following vertex shader</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=27265">Bug 27265</a> - GLSL Compiler doesnt link the attached vertex shader</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=27388">Bug 27388</a> - [i965] piglit glsl-vs-arrays failure</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=27403">Bug 27403</a> - GLSL struct causing "Invalid src register file ..." error</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=27914">Bug 27914</a> - GLSL compiler uses MUL+ADD where it could use MAD</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28055">Bug 28055</a> - glsl-texcoord-array fails GLSL compilation</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28374">Bug 28374</a> - SIGSEGV shader/slang/slang_typeinfo.c:534</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28748">Bug 28748</a> - [i965] uninlined function calls support</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28833">Bug 28833</a> - piglit/shaders/glsl-texcoord-array fail</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28834">Bug 28834</a> - Add support for system fpclassify to GL_OES_query_matrix function for OpenBSD / NetBSD</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28837">Bug 28837</a> - varying vec4 index support</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28845">Bug 28845</a> - The GLU tesselator code has some warnings</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28889">Bug 28889</a> - [regression] wine game crash</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28894">Bug 28894</a> - slang build fails if absolute path contains spaces</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28913">Bug 28913</a> - [GLSL] allowing two version statements</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28931">Bug 28931</a> - Floating Point Exception in Warzone2100 Trunk version</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28966">Bug 28966</a> - [r300g] Dynamic branching 3 demo does not run</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28967">Bug 28967</a> - slang/slang_emit.c:350: storage_to_src_reg: Assertion `index &gt;= 0' failed.</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29013">Bug 29013</a> - [r300g] translate_rgb_op: unknown opcode ILLEGAL OPCODE</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29020">Bug 29020</a> - [r300g] Wine d3d9 tests hardlock</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29910">Bug 29910</a> - Mesa advertises bogus GL_ARB_shading_language_120</li>
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30196">Bug 30196</a> - [GLSL] gl_TextureMatrix{Inverse,Transpose,InverseTranspose} unsupported</li>
+</ul>
+
+
+<h2>Changes</h2>
+<ul>
+<li>The Mesa demo/test programs have been moved into a separate git
+repository.
+<li>GL/glext.h file upgraded to version 64
+<li>GL/glxext.h file upgraded to version 32
+<li>GL/wglext.h file upgraded to version 22
+</ul>
+
+</body>
+</html>
diff --git a/mesalib/docs/shading.html b/mesalib/docs/shading.html
index cc1f9f1ec..aba395974 100644
--- a/mesalib/docs/shading.html
+++ b/mesalib/docs/shading.html
@@ -1,268 +1,268 @@
-<HTML>
-
-<TITLE>Shading Language Support</TITLE>
-
-<link rel="stylesheet" type="text/css" href="mesa.css"></head>
-
-<BODY>
-
-<H1>Shading Language Support</H1>
-
-<p>
-This page describes the features and status of Mesa's support for the
-<a href="http://opengl.org/documentation/glsl/" target="_parent">
-OpenGL Shading Language</a>.
-</p>
-
-<p>
-Contents
-</p>
-<ul>
-<li><a href="#envvars">Environment variables</a>
-<li><a href="#120">GLSL 1.20 support</a>
-<li><a href="#unsup">Unsupported Features</a>
-<li><a href="#notes">Implementation Notes</a>
-<li><a href="#hints">Programming Hints</a>
-<li><a href="#standalone">Stand-alone GLSL Compiler</a>
-<li><a href="#implementation">Compiler Implementation</a>
-<li><a href="#validation">Compiler Validation</a>
-</ul>
-
-
-
-<a name="envvars">
-<h2>Environment Variables</h2>
-
-<p>
-The <b>MESA_GLSL</b> environment variable can be set to a comma-separated
-list of keywords to control some aspects of the GLSL compiler and shader
-execution.  These are generally used for debugging.
-</p>
-<ul>
-<li><b>dump</b> - print GLSL shader code to stdout at link time
-<li><b>log</b> - log all GLSL shaders to files.
-    The filenames will be "shader_X.vert" or "shader_X.frag" where X
-    the shader ID.
-<li><b>nopt</b> - disable compiler optimizations
-<li><b>opt</b> - force compiler optimizations
-<li><b>uniform</b> - print message to stdout when glUniform is called
-<li><b>nopvert</b> - force vertex shaders to be a simple shader that just transforms
-    the vertex position with ftransform() and passes through the color and
-    texcoord[0] attributes.
-<li><b>nopfrag</b> - force fragment shader to be a simple shader that passes
-    through the color attribute.
-<li><b>useprog</b> - log glUseProgram calls to stderr
-</ul>
-<p>
-Example:  export MESA_GLSL=dump,nopt
-</p>
-
-
-<a name="120">
-<h2>GLSL Version</h2>
-
-<p>
-The GLSL compiler currently supports version 1.20 of the shading language.
-</p>
-
-<p>
-Several GLSL extensions are also supported:
-</p>
-<ul>
-<li>GL_ARB_draw_buffers
-<li>GL_ARB_texture_rectangle
-<li>GL_ARB_fragment_coord_conventions
-<li>GL_EXT_texture_array
-</ul>
-
-
-<a name="unsup">
-<h2>Unsupported Features</h2>
-
-<p>XXX update this section</p>
-
-<p>
-The following features of the shading language are not yet fully supported
-in Mesa:
-</p>
-
-<ul>
-<li>Linking of multiple shaders does not always work.  Currently, linking
-    is implemented through shader concatenation and re-compiling.  This
-    doesn't always work because of some #pragma and preprocessor issues.
-<li>gl_ClipVertex
-<li>The gl_Color and gl_SecondaryColor varying vars are interpolated
-    without perspective correction
-</ul>
-
-<p>
-All other major features of the shading language should function.
-</p>
-
-
-<a name="notes">
-<h2>Implementation Notes</h2>
-
-<ul>
-<li>Shading language programs are compiled into low-level programs
-    very similar to those of GL_ARB_vertex/fragment_program.
-<li>All vector types (vec2, vec3, vec4, bvec2, etc) currently occupy full
-    float[4] registers.
-<li>Float constants and variables are packed so that up to four floats
-    can occupy one program parameter/register.
-<li>All function calls are inlined.
-<li>Shaders which use too many registers will not compile.
-<li>The quality of generated code is pretty good, register usage is fair.
-<li>Shader error detection and reporting of errors (InfoLog) is not
-    very good yet.
-<li>The ftransform() function doesn't necessarily match the results of
-    fixed-function transformation.
-</ul>
-
-<p>
-These issues will be addressed/resolved in the future.
-</p>
-
-
-<a name="hints">
-<h2>Programming Hints</h2>
-
-<ul>
-<li>Use the built-in library functions whenever possible.
-    For example, instead of writing this:
-<pre>
-        float x = 1.0 / sqrt(y);
-</pre>
-    Write this:
-<pre>
-        float x = inversesqrt(y);
-</pre>
-</li>
-</ul>
-
-
-<a name="standalone">
-<h2>Stand-alone GLSL Compiler</h2>
-
-<p>
-The stand-alone GLSL compiler program can be used to compile GLSL shaders
-into low-level GPU code.
-</p>
-
-<p>
-This tool is useful for:
-<p>
-<ul>
-<li>Inspecting GPU code to gain insight into compilation
-<li>Generating initial GPU code for subsequent hand-tuning
-<li>Debugging the GLSL compiler itself
-</ul>
-
-<p>
-After building Mesa, the compiler can be found at src/glsl/glsl_compiler
-</p>
-
-<p>
-Here's an example of using the compiler to compile a vertex shader and
-emit GL_ARB_vertex_program-style instructions:
-</p>
-<pre>
-    src/glsl/glsl_compiler --dump-ast myshader.vert
-</pre>
-
-Options include
-<ul>
-<li><b>--dump-ast</b> - dump GPU code
-<li><b>--dump-hir</b> - dump high-level IR code
-<li><b>--dump-lir</b> - dump low-level IR code
-<li><b>--link</b> - ???
-</ul>
-
-
-
-
-<a name="implementation">
-<h2>Compiler Implementation</h2>
-
-<p>
-The source code for Mesa's shading language compiler is in the
-<code>src/glsl/</code> directory.
-</p>
-
-<p>
-XXX provide some info about the compiler....
-</p>
-
-<p>
-The final vertex and fragment programs may be interpreted in software
-(see prog_execute.c) or translated into a specific hardware architecture
-(see drivers/dri/i915/i915_fragprog.c for example).
-</p>
-
-<h3>Code Generation Options</h3>
-
-<p>
-Internally, there are several options that control the compiler's code
-generation and instruction selection.
-These options are seen in the gl_shader_state struct and may be set
-by the device driver to indicate its preferences:
-
-<pre>
-struct gl_shader_state
-{
-   ...
-   /** Driver-selectable options: */
-   GLboolean EmitHighLevelInstructions;
-   GLboolean EmitCondCodes;
-   GLboolean EmitComments;
-};
-</pre>
-
-<ul>
-<li>EmitHighLevelInstructions
-<br>
-This option controls instruction selection for loops and conditionals.
-If the option is set high-level IF/ELSE/ENDIF, LOOP/ENDLOOP, CONT/BRK
-instructions will be emitted.
-Otherwise, those constructs will be implemented with BRA instructions.
-</li>
-
-<li>EmitCondCodes
-<br>
-If set, condition codes (ala GL_NV_fragment_program) will be used for
-branching and looping.
-Otherwise, ordinary registers will be used (the IF instruction will
-examine the first operand's X component and do the if-part if non-zero).
-This option is only relevant if EmitHighLevelInstructions is set.
-</li>
-
-<li>EmitComments
-<br>
-If set, instructions will be annoted with comments to help with debugging.
-Extra NOP instructions will also be inserted.
-</br>
-
-</ul>
-
-
-<a name="validation">
-<h2>Compiler Validation</h2>
-
-<p>
-Developers working on the GLSL compiler should test frequently to avoid
-regressions.
-</p>
-
-<p>
-The <a href="http://people.freedesktop.org/~nh/piglit/">Piglit</a> project
-has many GLSL tests and the
-<a href="http://glean.sf.net" target="_parent">Glean</a> glsl1 test 
-tests GLSL features.
-</p>
-
-<p>
-The Mesa demos repository also has some good GLSL tests.
-</p>
-
-</BODY>
-</HTML>
+<HTML>
+
+<TITLE>Shading Language Support</TITLE>
+
+<link rel="stylesheet" type="text/css" href="mesa.css"></head>
+
+<BODY>
+
+<H1>Shading Language Support</H1>
+
+<p>
+This page describes the features and status of Mesa's support for the
+<a href="http://opengl.org/documentation/glsl/" target="_parent">
+OpenGL Shading Language</a>.
+</p>
+
+<p>
+Contents
+</p>
+<ul>
+<li><a href="#envvars">Environment variables</a>
+<li><a href="#120">GLSL 1.20 support</a>
+<li><a href="#unsup">Unsupported Features</a>
+<li><a href="#notes">Implementation Notes</a>
+<li><a href="#hints">Programming Hints</a>
+<li><a href="#standalone">Stand-alone GLSL Compiler</a>
+<li><a href="#implementation">Compiler Implementation</a>
+<li><a href="#validation">Compiler Validation</a>
+</ul>
+
+
+
+<a name="envvars">
+<h2>Environment Variables</h2>
+
+<p>
+The <b>MESA_GLSL</b> environment variable can be set to a comma-separated
+list of keywords to control some aspects of the GLSL compiler and shader
+execution.  These are generally used for debugging.
+</p>
+<ul>
+<li><b>dump</b> - print GLSL shader code to stdout at link time
+<li><b>log</b> - log all GLSL shaders to files.
+    The filenames will be "shader_X.vert" or "shader_X.frag" where X
+    the shader ID.
+<li><b>nopt</b> - disable compiler optimizations
+<li><b>opt</b> - force compiler optimizations
+<li><b>uniform</b> - print message to stdout when glUniform is called
+<li><b>nopvert</b> - force vertex shaders to be a simple shader that just transforms
+    the vertex position with ftransform() and passes through the color and
+    texcoord[0] attributes.
+<li><b>nopfrag</b> - force fragment shader to be a simple shader that passes
+    through the color attribute.
+<li><b>useprog</b> - log glUseProgram calls to stderr
+</ul>
+<p>
+Example:  export MESA_GLSL=dump,nopt
+</p>
+
+
+<a name="120">
+<h2>GLSL Version</h2>
+
+<p>
+The GLSL compiler currently supports version 1.20 of the shading language.
+</p>
+
+<p>
+Several GLSL extensions are also supported:
+</p>
+<ul>
+<li>GL_ARB_draw_buffers
+<li>GL_ARB_texture_rectangle
+<li>GL_ARB_fragment_coord_conventions
+<li>GL_EXT_texture_array
+</ul>
+
+
+<a name="unsup">
+<h2>Unsupported Features</h2>
+
+<p>XXX update this section</p>
+
+<p>
+The following features of the shading language are not yet fully supported
+in Mesa:
+</p>
+
+<ul>
+<li>Linking of multiple shaders does not always work.  Currently, linking
+    is implemented through shader concatenation and re-compiling.  This
+    doesn't always work because of some #pragma and preprocessor issues.
+<li>gl_ClipVertex
+<li>The gl_Color and gl_SecondaryColor varying vars are interpolated
+    without perspective correction
+</ul>
+
+<p>
+All other major features of the shading language should function.
+</p>
+
+
+<a name="notes">
+<h2>Implementation Notes</h2>
+
+<ul>
+<li>Shading language programs are compiled into low-level programs
+    very similar to those of GL_ARB_vertex/fragment_program.
+<li>All vector types (vec2, vec3, vec4, bvec2, etc) currently occupy full
+    float[4] registers.
+<li>Float constants and variables are packed so that up to four floats
+    can occupy one program parameter/register.
+<li>All function calls are inlined.
+<li>Shaders which use too many registers will not compile.
+<li>The quality of generated code is pretty good, register usage is fair.
+<li>Shader error detection and reporting of errors (InfoLog) is not
+    very good yet.
+<li>The ftransform() function doesn't necessarily match the results of
+    fixed-function transformation.
+</ul>
+
+<p>
+These issues will be addressed/resolved in the future.
+</p>
+
+
+<a name="hints">
+<h2>Programming Hints</h2>
+
+<ul>
+<li>Use the built-in library functions whenever possible.
+    For example, instead of writing this:
+<pre>
+        float x = 1.0 / sqrt(y);
+</pre>
+    Write this:
+<pre>
+        float x = inversesqrt(y);
+</pre>
+</li>
+</ul>
+
+
+<a name="standalone">
+<h2>Stand-alone GLSL Compiler</h2>
+
+<p>
+The stand-alone GLSL compiler program can be used to compile GLSL shaders
+into low-level GPU code.
+</p>
+
+<p>
+This tool is useful for:
+<p>
+<ul>
+<li>Inspecting GPU code to gain insight into compilation
+<li>Generating initial GPU code for subsequent hand-tuning
+<li>Debugging the GLSL compiler itself
+</ul>
+
+<p>
+After building Mesa, the compiler can be found at src/glsl/glsl_compiler
+</p>
+
+<p>
+Here's an example of using the compiler to compile a vertex shader and
+emit GL_ARB_vertex_program-style instructions:
+</p>
+<pre>
+    src/glsl/glsl_compiler --dump-ast myshader.vert
+</pre>
+
+Options include
+<ul>
+<li><b>--dump-ast</b> - dump GPU code
+<li><b>--dump-hir</b> - dump high-level IR code
+<li><b>--dump-lir</b> - dump low-level IR code
+<li><b>--link</b> - ???
+</ul>
+
+
+
+
+<a name="implementation">
+<h2>Compiler Implementation</h2>
+
+<p>
+The source code for Mesa's shading language compiler is in the
+<code>src/glsl/</code> directory.
+</p>
+
+<p>
+XXX provide some info about the compiler....
+</p>
+
+<p>
+The final vertex and fragment programs may be interpreted in software
+(see prog_execute.c) or translated into a specific hardware architecture
+(see drivers/dri/i915/i915_fragprog.c for example).
+</p>
+
+<h3>Code Generation Options</h3>
+
+<p>
+Internally, there are several options that control the compiler's code
+generation and instruction selection.
+These options are seen in the gl_shader_state struct and may be set
+by the device driver to indicate its preferences:
+
+<pre>
+struct gl_shader_state
+{
+   ...
+   /** Driver-selectable options: */
+   GLboolean EmitHighLevelInstructions;
+   GLboolean EmitCondCodes;
+   GLboolean EmitComments;
+};
+</pre>
+
+<ul>
+<li>EmitHighLevelInstructions
+<br>
+This option controls instruction selection for loops and conditionals.
+If the option is set high-level IF/ELSE/ENDIF, LOOP/ENDLOOP, CONT/BRK
+instructions will be emitted.
+Otherwise, those constructs will be implemented with BRA instructions.
+</li>
+
+<li>EmitCondCodes
+<br>
+If set, condition codes (ala GL_NV_fragment_program) will be used for
+branching and looping.
+Otherwise, ordinary registers will be used (the IF instruction will
+examine the first operand's X component and do the if-part if non-zero).
+This option is only relevant if EmitHighLevelInstructions is set.
+</li>
+
+<li>EmitComments
+<br>
+If set, instructions will be annoted with comments to help with debugging.
+Extra NOP instructions will also be inserted.
+</br>
+
+</ul>
+
+
+<a name="validation">
+<h2>Compiler Validation</h2>
+
+<p>
+Developers working on the GLSL compiler should test frequently to avoid
+regressions.
+</p>
+
+<p>
+The <a href="http://people.freedesktop.org/~nh/piglit/">Piglit</a> project
+has many GLSL tests and the
+<a href="http://glean.sf.net" target="_parent">Glean</a> glsl1 test 
+tests GLSL features.
+</p>
+
+<p>
+The Mesa demos repository also has some good GLSL tests.
+</p>
+
+</BODY>
+</HTML>