Even though it's only valid when local, it is possible for a local
client and the server to not match endianness, such as when running
a ppc application under Rosetta.
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 1c8bda798b)
(cherry picked from commit 7be5492bcf)
Even though it's only valid when local, it is possible for a local
client and the server to not match endianness, such as when running
a ppc application under Rosetta.
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
Reviewed-by: Jamey Sharp <jamey@minilop.net>
(cherry picked from commit 14205ade0c)
Conflicts:
hw/xquartz/xpr/appledri.c
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 0ad049706d)
This was a regression.
Introduced by: 08363c5830 and
32db27a7f8
Masked by: 1e69fd4a60
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
Reviewed-by: Jamey Sharp <jamey@minilop.net>
(cherry picked from commit 83fef4235d)
Bug introduced by 9dca441670
xfree86: add a hook to replace the new console handler.
console_handler was not being set, making the server eat up CPU spinning
in WaitForSomething selecting consoleFd over and over again, every time
trying to unregister drain_console without success due to
console_handler being NULL.
Let's just fix the unregistration in xf86SetConsoleHandler() and use that.
But wait, there could be a catch: If some driver replaced the handler using
xf86SetConsoleHandler(), the unregistration in xf86CloseConsole will unregister
that one. I don't understand Xorg well enough to know whether this poses a
problem (could mess up driver deinit somehow or something like that). As it is,
xf86SetConsoleHandler() doesn't offer any way to prevent this (i.e. check which
handler is currently registered).
I had been using it for two days on my machine that previously hit 100% CPU
several times a day. That has now gone away without any new problems appearing.
Signed-off-by: Tomas Trnka <tomastrnka@gmx.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit 323869f329)
(cherry picked from commit 82445286d5)
Caught during review of e095369bf.
Signed-off-by: Peter Harris <pharris@opentext.com>
Reviewed-by-by: Matt Turner <mattst88@gmail.com>
Signed-off-by: Matt Turner <mattst88@gmail.com>
(cherry picked from commit c90903b4f7)
Conflicts:
xkb/xkb.c
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit ec8ee3a9f7)
Use fchmod() to change permissions of the lock file instead
of chmod(), thus avoid the race that can be exploited to set
a symbolic link to any file or directory in the system.
Signed-off-by: Matthieu Herrb <matthieu.herrb@laas.fr>
Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com>
(cherry picked from commit b67581cf82)
use O_NOFOLLOW to open the existing lock file, so symbolic links
aren't followed, thus avoid revealing if it point to an existing
file.
Signed-off-by: Matthieu Herrb <matthieu.herrb@laas.fr>
Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com>
(cherry picked from commit 6ba44b91e3)
Some extension libraries may set this bit before converting the event to
wire protocol and as such range checking the event will cause an invalid
BadValue error to result. As the documentation suggests the the bit
should be "forced on", remove it before doing range checks and continue
to force it on in the server.
Reviewed-by: Jamey Sharp <jamey@minilop.net>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit 2d2dce558d)
(cherry picked from commit e9ae333160)
The corresponding DeviceIntPtr wasn't being gotten properly,
resulting in BadDevice from dixLookupDevice().
Signed-off-by: Carlos Garnacho <carlosg@gnome.org>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit f52d5cd374)
(cherry picked from commit 347f5610ca)
If AddResource fails, it will automatically free the object that was
passed to it by calling the appropriate deleteFunc; and of course
FreeResource also calls the deleteFunc. In both cases it's wrong to call
the destroy hook manually.
Commit by Jamey Sharp and Josh Triplett.
Signed-off-by: Jamey Sharp <jamey@minilop.net>
Signed-off-by: Josh Triplett <josh@joshtriplett.org>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Aaron Plattner <aplattner@nvidia.com>
(cherry picked from commit 0f380a5005)
(cherry picked from commit bd6ea85209)
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>
(cherry picked from commit e32cc0b4c8)
Unconditionally drop the valuators back into the mask when they were there
in the first place. Otherwise, sending identical coordinates from the driver
on a translated device causes the valuator mask to be alternatively
overwritten with the translated value or left as-is. This leads to the
device jumping around between the translated and the original position.
The same could be achieved with a valuator_mask_unset() combination.
Testcase:
xsetwacom set "device name" MapToOutput VGA1
Then press a button on the device, cursor jumps between the two positions.
Introduced in 31737fff08
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Daniel Stone <daniel@fooishbar.org>
Reviewed-by: Chase Douglas <chase.douglas@canonical.com>
Otherwise sys_pitch will be stale when a system memory copy is allocated.
Fixes https://bugs.freedesktop.org/show_bug.cgi?id=38322 and a crash when
unlocking the screen with xscreensaver, reported by Janne Huttunen.
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Tested-by: Janne Huttunen <jahuttun@gmail.com>
Tested-by: Jan Kriho <Erbureth@gmail.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit 4020cab88f)
Fixes assertion failure when calling dixSetPrivate
Debian bug#632549 <http://bugs.debian.org/632549>
Reported-and-tested-by: Mohammed Sameer <msameer@foolab.org>
Signed-off-by: Julien Cristau <jcristau@debian.org>
Reviewed-by: Daniel Stone <daniel@fooishbar.org>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit a1dec7cd6a)
RRPostPendingProperties tries to compare the pending and current
property values to decide whether they're actually changing. However,
it does this using a memcmp that passes in pending_value->size as the
number of bytes. This is actually the number of elements, where each
element is (pending_value->format / 8) bytes long. This causes the
pending value to not be propagated if the first pending_value->size
bytes are the same and only the end of it is changing.
Fix this by computing the total number of bytes to compare in the
memcmp.
Signed-off-by: Aaron Plattner <aplattner@nvidia.com>
Reviewed-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit 08dfff92e8)
For hotplugged devices, xf86AllocateInput does that for us but the xorg.conf
path is different. Since not all drivers reset the fd during PreInit but may
still call close(pInfo->fd) in all cases, this can terminate the logging
early.
Reproducible: add a wacom driver InputDevice section with no Option Device.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Daniel Stone <daniel@fooishbar.org>
(cherry picked from commit 3798dd379c)
xf86ConfigLayout.inputs contains the information from the xorg.conf
file. Passing this into xf86NewInputDevice means the device will get
cleaned up on exit and the pointers in xf86ConfigLayout.inputs are left
dangling. In the second server generation, this results in a server
crash.
Also, rename pDev to pInfo. pDev is pretty much reserved for DeviceIntPtr
types.
Reproducible: AutoAddDevices off and xorg.conf input sections, trigger
server regeneration.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Daniel Stone <daniel@fooishbar.org>
(cherry picked from commit f0d7e9db28)
Devices that succeeded during PreInit and DEVICE_INIT but failed in
DEVICE_ON would be deleted through xf86DeleteInput but not removed from the
list of input devices (and not turned off). The result was a double free on
server shutdown.
Fix this by calling RemoveDevice if EnableDevice fails.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Daniel Stone <daniel@fooishbar.org>
(cherry picked from commit 8ffddbcf72)
Default builds will fail due to BUNDLE_ID_PREFIX not being defined. LAUNCHD_ID_PREFIX
was used in 1.10.x and earlier.
This was fallout from 55ef10975e.
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
This preference allows users to override the related option in Mac OS X's
Mouse/Trackpad preferences. This effectively lets the user determine
which "context" all of X11 fits into for context-based scrolling until
such API exists within X11 itself to pass along to X11 clients.
This is applicable to Mav OS X 10.7+
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit ac43984215)
This occurred to me in hindsight after the last commit. If the
original developer had done this, we would have noticed the
problem sooner.
(cherry picked from commit aa0a57996f)
(cherry picked from commit ec95a9c829)
xp_destroy_surface was called with a surface id of 0, due to some
premature cleanup that set it to 0. This means the surfaces
weren't being destroyed until the window was.
The code that did that was: pDRIDrawablePriv->sid = 0;
In long running applications this leak may or may not have been
harmful. With the old libGL the surfaces weren't destroyed until
the context was destroyed or a new context created. In the new
libGL they are reference counted, and released much sooner, so we
ran into a resource leak more noticeably with some tests.
Make the Apple DRI code dispatch events to the client(s) for
destroyed surfaces, when a resource is destroyed. This seems to
work in my tests, however this clearly wasn't working for a while,
so bugs may result in the future if it enables some new (unexpected)
side effects.
Also add a few helpful comments to aid in understanding the code
in the future.
Tested with the test suite, Pymol, and various Mesa demos.
(cherry picked from commit bede83eb19)
(cherry picked from commit 4fe7df2653)
Also add some comments that weren't merged in from server-1.4-apple's
99babae132
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 0ebe45a717)
This way we'll print an error and still mostly work rather than crashing if
someone installs XQuartz.app incorrectly or tries running the server within
the build system rather than the installed system.
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 88ad050f3c)
The GLXversion member of the __GLXscreen struct
is just cruft since commit ad5c0d9efa,
when we started returning the minimum GLX version supported by all
of the screens on the display, rather than the maximum version supported
by the server.
Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com>
Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk>
(cherry picked from commit 42bc198775)
RawEvents are supposed to be events coming from the driver. When warping the
pointer, this should not generate a raw event.
X.Org Bug 30068 <http://bugs.freedesktop.org/show_bug.cgi?id=30068>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Daniel Stone <daniel@fooishbar.org>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit 23a7832789)
Conflicts:
dix/getevents.c
There are, in general, more fbconfig depths than visual depths.
fbconfigs need not support Window rendering, however any that do must
have an associated visual ID (which we got right), and any that do not
must not claim GLX_WINDOW_BIT in GLX_DRAWABLE_TYPE (which we got
wrong).
Fixes piglit/glx-fbconfig-sanity, assuming you have a sufficiently
recent piglit, as that test formerly wrongly required pixmap-capable
fbconfigs to have a visual.
v2: Additional check for fbconfigs that didn't have GLX_WINDOW_BIT in
the first place, from previous patch by Jon TURNEY; also, also clear
->visualID.
Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk>
Signed-off-by: Adam Jackson <ajax@redhat.com>
(cherry picked from commit 3e8270b869)
Guard the use of backtrace() with HAVE_BACKTRACE, since we already
have the autoconf machinery for setting that.
For the moment, assume dladdr() is available when backtrace() is
Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk>
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 38d55f06cf)
In addition, this change will not call into the X11 activation unless an X11
window was active when we deactivated. We can't rely on the event and current
key windows because the key window will be nil until activated, and the event
will only reference the window if the window was clicked (whereas it will be
nil if we activated via dock or cmd-tab).
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit c8b80a82d9)
XvMCScreenKey is defined as XvMCScreenKeyRec's address, so will never be
NULL. Use the check proposed by Alan Coopersmith and Jeremy Huddleston
instead.
Those warnings go away accordingly:
| CC xvmc.lo
| xvmc.c: In function 'ProcXvMCListSubpictureTypes':
| xvmc.c:490:22: warning: the comparison will always evaluate as 'false' for the address of 'XvMCScreenKeyRec' will never be NULL [-Waddress]
| xvmc.c: In function 'XvMCExtensionInit':
| xvmc.c:671:21: warning: the comparison will always evaluate as 'false' for the address of 'XvMCScreenKeyRec' will never be NULL [-Waddress]
| xvmc.c: In function 'XvMCFindXvImage':
| xvmc.c:749:22: warning: the comparison will always evaluate as 'false' for the address of 'XvMCScreenKeyRec' will never be NULL [-Waddress]
Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Signed-off-by: Cyril Brulebois <kibi@debian.org>
(cherry picked from commit 9da5aa09d7)
Grabbing an SD device temporary floats the device but we must not release
the buttons. Introduced in
commit 9d23459415
Author: Peter Hutterer <peter.hutterer@who-t.net>
Date: Fri Feb 25 11:08:19 2011 +1000
dix: release all buttons and keys before reattaching a device (#34182)
X.Org Bug 36146 <http://bugs.freedesktop.org/show_bug.cgi?id=36146>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit bf2059b07a)
Conflicts:
dix/devices.c
Testcase:
xinput float <keyboard name>
results in the keyboard's enter key being repeated as the device is detached
while the key is still physically down. To avoid this, release all keys and
buttons before reattaching the device.
X.Org Bug 34182 <http://bugs.freedesktop.org/show_bug.cgi?id=34182>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Fernando Carrijo <fcarrijo@freedesktop.org>
(cherry picked from commit 9d23459415)
(cherry picked from commit 81fbb96c54)
