| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
| |
we'll need the remote xkb in KeyboardProc in future so let's move it up.
|
| |
|
| |
|
|
|
|
| |
(partially) fixes ArcticaProject/nx-libs#666
|
|
|
|
| |
(partially) fixes ArcticaProject/nx-libs#666
|
|
|
|
| |
(partially) fixes ArcticaProject/nx-libs#666
|
|
|
|
| |
(partially) fixes ArcticaProject/nx-libs#666
|
|
|
|
| |
(partially) fixes ArcticaProject/nx-libs#666
|
|
|
|
| |
(partially) fixes ArcticaProject/nx-libs#666
|
|
|
|
| |
(partially) fixes ArcticaProject/nx-libs#666
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Backport of this commit:
commit a2880699e8f1f576e1a48ebf25e8982463323f84
Author: Keith Packard <keithp@keithp.com>
Date: Tue Mar 25 08:21:16 2014 -0700
fb: fix fast-path blt detection
The width parameter is used to disable the blit fast-path (memcpy) when
source and destination rows overlap in memory. This check was added in [0].
Unfortunately, the calculation to determine if source and destination
lines overlapped was incorrect:
(1) it converts width from pixels to bytes, but width is actually in
bits, not pixels.
(2) it adds this byte offset to dst/srcLine, which implicitly converts
the offset from bytes to sizeof(FbBits).
Fix both of these by converting addresses to byte pointers and width
to bytes and doing comparisons on the resulting byte address.
For example:
A 32-bpp 1366 pixel-wide row will have
width = 1366 * 32 = 43712 bits
bpp = 32
(bpp >> 3) = 4
width * (bpp >> 3) = 174848 FbBits
(FbBits *)width => 699392 bytes
So, "careful" was true if the destination line was within 699392 bytes,
instead of just within its 1366 * 4 = 5464 byte row.
This bug causes us to take the slow path for large non-overlapping rows
that are "close" in memory. As a data point, XGetImage(1366x768) on my
ARM chromebook was taking ~140 ms, but with this fixed, it now takes
about 60 ms.
XGetImage() -> exaGetImage() -> fbGetImage -> fbBlt()
[0] commit e32cc0b4c85c78cd8743a6e1680dcc79054b57ce
Author: Adam Jackson <ajax@redhat.com>
Date: Thu Apr 21 16:37:11 2011 -0400
fb: Fix memcpy abuse
The memcpy fast path implicitly assumes that the copy walks
left-to-right. That's not something memcpy guarantees, and newer glibc
on some processors will indeed break that assumption. Since we walk a
line at a time, check the source and destination against the width of
the blit to determine whether we can be sloppy enough to allow memcpy.
(Having done this, we can remove the check for !reverse as well.)
v3: Convert to byte units
This first checks to make sure the blt is byte aligned, converts all
of the data to byte units and then compares for byte address range
overlap between source and dest.
Signed-off-by: Keith Packard <keithp@keithp.com>
Reviewed-by: Daniel Kurtz <djkurtz@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Fixes ArcticaProject/nx-libs#750
Backport of this commit:
commit e32cc0b4c85c78cd8743a6e1680dcc79054b57ce
Author: Adam Jackson <ajax@redhat.com>
Date: Thu Apr 21 16:37:11 2011 -0400
fb: Fix memcpy abuse
The memcpy fast path implicitly assumes that the copy walks
left-to-right. That's not something memcpy guarantees, and newer glibc
on some processors will indeed break that assumption. Since we walk a
line at a time, check the source and destination against the width of
the blit to determine whether we can be sloppy enough to allow memcpy.
(Having done this, we can remove the check for !reverse as well.)
On an Intel Core i7-2630QM with an NVIDIA GeForce GTX 460M running in
NoAccel, the broken code and various fixes for -copywinwin{10,100,500}
gives (edited to fit in 80 columns):
1: Disable the fastpath entirely
2: Replace memcpy with memmove
3: This fix
4: The code before this fix
1 2 3 4 Operation
------ --------------- --------------- --------------- ------------
258000 269000 ( 1.04) 544000 ( 2.11) 552000 ( 2.14) Copy 10x10
21300 23000 ( 1.08) 43700 ( 2.05) 47100 ( 2.21) Copy 100x100
960 962 ( 1.00) 1990 ( 2.09) 1990 ( 2.07) Copy 500x500
So it's a modest performance hit, but correctness demands it, and it's
probably worth keeping the 2x speedup from having the fast path in the
first place.
Signed-off-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Fixes this valgrind finding
==16977== Warning: invalid file descriptor -1 in syscall close()
==16977== Conditional jump or move depends on uninitialised value(s)
==16977== at 0x544B6B: XkbSendNewKeyboardNotify (xkbEvents.c:62)
==16977== by 0x540481: ProcXkbGetKbdByName (xkb.c:5330)
==16977== by 0x4341C5: Dispatch (NXdispatch.c:482)
==16977== by 0x40EB02: main (main.c:353)
==16977== Uninitialised value was created by a heap allocation
==16977== at 0x4C2DB8F: malloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==16977== by 0x431BD7: NextAvailableClient (dispatch.c:3719)
==16977== by 0x47B297: AllocNewConnection (connection.c:821)
==16977== by 0x47B297: EstablishNewConnections (connection.c:910)
==16977== by 0x463DFE: ProcessWorkQueue (dixutils.c:541)
==16977== by 0x47635E: WaitForSomething (WaitFor.c:213)
==16977== by 0x434089: Dispatch (NXdispatch.c:360)
==16977== by 0x40EB02: main (main.c:353)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Backport of
commit dc9ce695a69ca0787f58f8d160212a7a41acb703
Author: Rami Ylimäki <rami.ylimaki@vincit.fi>
Date: Wed Mar 9 15:45:40 2011 +0200
xkb: Initialize pad bytes sent in replies of geometry requests.
Valgrind complains about uninitialized data being written to clients.
Reviewed-by: Erkki Seppälä <erkki.seppala@vincit.fi>
Signed-off-by: Rami Ylimäki <rami.ylimaki@vincit.fi>
Reviewed-by: Daniel Stone <daniel@fooishbar.org>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Backport of
commit 81b3b0cce088866dc3cda099d7c8d6655849fd43
Author: Tomas Janousek <tomi@nomi.cz>
Date: Wed May 20 15:03:01 2009 +0200
Bug #6428, #16458, #21464: Fix crash due to uninitialized VModMap fields.
In ProcXkbGetKbdByName, mrep.firstVModMapKey, .nVModMapKeys and
.totalVModMapKeys were not initialized, contained random values and caused
accesses to unallocated and later modified memory, causing
XkbSizeVirtualModMap and XkbWriteVirtualModMap to see different number of
nonzero values, resulting in writes past the end of an array in XkbSendMap.
This patch initializes those values sensibly and reverts commits 5c0a2088 and
6dd4fc46, which have been plain non-sense.
Signed-off-by: Tomas Janousek <tomi@nomi.cz>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This commit + some more adaption in ddxLoad.c which where no longer
present when Alan did the patch. However, our code still has them:
Author: Alan Coopersmith <alan.coopersmith@oracle.com>
Date: Sat Nov 27 20:09:04 2010 -0800
Convert existing Xprintf style calls to asprintf style
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com>
Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net>
Backported-to-NX-by: Ulrich Sibiller <uli42@gmx.de>
|
|
|
|
|
|
|
|
|
|
|
| |
Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com>
Date: Thu Jul 22 11:34:54 2010 +0300
xkb: Don't check for NULL before calling free
Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Backported-to-NX-by: Ulrich Sibiller <uli42@gmx.de>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com>
Date: Thu Jul 22 11:33:33 2010 +0300
xkb: Fix memory leak if opening file fails
If fopen fails pointer in buf would be overwriten with a new pointer.
Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Backported-to-NX-by: Ulrich Sibiller <uli42@gmx.de>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Backported from Xorg:
commit f292de2ef13dc994a38029cee9e2642576893332
Author: Daniel Stone <daniel@fooishbar.org>
Date: Wed Mar 21 02:04:12 2007 +0200
XKB: Fix size_syms calculation bug
Apparently it needed to be nSyms*15/10, not *12/10; make it match the
other allocation code.
Backported-to-nx-by: Ulrich Sibiller <uli42@gmx.de>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
commit 7c4c00649cae855e141ec9a3667bfe399e3156d0
Author: Daniel Stone <daniel@fooishbar.org>
Date: Thu Jan 1 05:25:52 2009 +1100
XKB: Remove unused DDX functions
They were complete no-ops anyway.
Signed-off-by: Daniel Stone <daniel@fooishbar.org>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Backported-to-nx-by: Ulrich Sibiller <uli42@gmx.de>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
commit 6aef4e96affcc26b9415c6bc9c9bebb5af05a1a7
Author: Daniel Stone <daniel@fooishbar.org>
Date: Sat Jan 17 13:42:46 2009 +0200
XKB: Fix logic error
Use logical or instead of bitwise or.
Signed-off-by: Daniel Stone <daniel@fooishbar.org>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Backported-to-nx-by: Ulrich Sibiller <uli42@gmx.de>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
commit 8311cd5f89ca6781842bb24671b8122cdf8be148
Author: Jeremy Huddleston <jeremyhu@freedesktop.org>
Date: Tue Mar 23 10:37:36 2010 -0700
XKB: Fix garbage initialization
XkbEnableDisableControls set extra garbage bits on the xkbControlsNotify
changedControls mask because it was uninitialized on the stack.
Found by clang
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Backported-to-nx-by: Ulrich Sibiller <uli42@gmx.de>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We have fixed that (independendly) before, but missed one line.
Backported-to-nx-by: Ulrich Sibiller <uli42@gmx.de>
commit 85f9017393c9bb19553e9afcf554673a44a09993
Author: Ander Conselvan de Oliveira <ander.conselvan-de-oliveira@nokia.com>
Date: Tue Feb 8 11:10:10 2011 +0200
ProcXkbGetXkbByName: fix use of uninitialised bytes valgrind error.
==9999== Syscall param writev(vector[...]) points to uninitialised byte(s)
==9999== at 0x4AB5154: writev (writev.c:51)
==9999== by 0x7C7C3: _XSERVTransWritev (Xtrans.c:912)
==9999== by 0x61C8B: FlushClient (io.c:924)
==9999== by 0x62423: WriteToClient (io.c:846)
==9999== by 0xCE39B: XkbSendMap (xkb.c:1408)
==9999== by 0xD247B: ProcXkbGetKbdByName (xkb.c:5814)
==9999== by 0x4AB53: Dispatch (dispatch.c:432)
==9999== by 0x205BF: main (main.c:291)
==9999== Address 0x557eb68 is 40 bytes inside a block of size 4,096 alloc'd
==9999== at 0x48334A4: calloc (vg_replace_malloc.c:467)
==9999== by 0x62567: WriteToClient (io.c:1065)
==9999== by 0x452EB: ProcEstablishConnection (dispatch.c:3685)
==9999== by 0x4AB53: Dispatch (dispatch.c:432)
==9999== by 0x205BF: main (main.c:291)
==9999== Uninitialised value was created by a stack allocation
==9999== at 0xD1910: ProcXkbGetKbdByName (xkb.c:5559)
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Oliver McFadden <oliver.mcfadden@nokia.com>
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan-de-oliveira@nokia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Backported-to-nx-by: Ulrich Sibiller <uli42@gmx.de>
commit 8a34d7a8532c7ca013e67307f3baf200167abb92
Author: Ander Conselvan de Oliveira <ander.conselvan-de-oliveira@nokia.com>
Date: Tue Feb 8 11:10:11 2011 +0200
XkbSendNames: fix use of uninitialised bytes valgrind error.
==537== Syscall param writev(vector[...]) points to uninitialised byte(s)
==537== at 0x4AB7154: writev (writev.c:51)
==537== by 0x8935B: _XSERVTransWritev (Xtrans.c:912)
==537== by 0x6C55F: FlushClient (io.c:924)
==537== by 0x6CCF3: WriteToClient (io.c:846)
==537== by 0xD51D3: XkbSendNames (xkb.c:3765)
==537== by 0xD8183: ProcXkbGetKbdByName (xkb.c:5825)
==537== by 0x27B7B: Dispatch (dispatch.c:432)
==537== by 0x205B7: main (main.c:291)
==537== Address 0x55899f2 is 154 bytes inside a block of size 1,896 alloc'd
==537== at 0x4834C48: malloc (vg_replace_malloc.c:236)
==537== by 0xD47AF: XkbSendNames (xkb.c:3642)
==537== by 0xD8183: ProcXkbGetKbdByName (xkb.c:5825)
==537== by 0x27B7B: Dispatch (dispatch.c:432)
==537== by 0x205B7: main (main.c:291)
==537== Uninitialised value was created by a heap allocation
==537== at 0x4834C48: malloc (vg_replace_malloc.c:236)
==537== by 0xD47AF: XkbSendNames (xkb.c:3642)
==537== by 0xD8183: ProcXkbGetKbdByName (xkb.c:5825)
==537== by 0x27B7B: Dispatch (dispatch.c:432)
==537== by 0x205B7: main (main.c:291)
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Oliver McFadden <oliver.mcfadden@nokia.com>
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan-de-oliveira@nokia.com>
|
| |
|
|
|
|
| |
an empty nxagentKeyboard variable is no reason to fall back to no-XKB mode
|
|
|
|
| |
most of the code was existing twice
|
|
|
|
| |
manpage does not tell us if NULL is valid
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
use the ones from xkb
|
|
|
|
| |
Just use the existing xkb code
|
|
|
|
| |
Defines regarding XKB-BASE_DIRECTORY are used only xkb.
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|