The macro has been changed to do this already, no need for double
not-not-ing.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit 23e3d1f233)
Simply returning the mask bit breaks checks like
BitIsOn(mask, 0) != BitIsOn(mask, 1);
as used in 048e93593e.
The naming of this macro suggests that it should return boolean values
anyway. This patch also adds a few simple tests for these macros to make
sure they don't accidentally break in the future.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Pat Kane <pekane52@gmail.com>
Reviewed-by: Julien Cristau <jcristau@debian.org>
Reviewed-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit 42dc91e32a)
When RandR 1.2's transformation code is enabled, it rotates the cursor
image so that it appears upright on a rotated screen. This code
completely mangles 2-color cursors on hardware where the the mask and
source images are not interleaved due to two problems:
1. stride is calculated as (width / 4) rather than (width / 8), so the
expression (y * stride) skips two lines instead of one for every
time y is incremented.
2. cursor_bitpos ignores the 'mask' parameter if the hardware doesn't
specify any of the HARDWARE_CURSOR_SOURCE_MASK_INTERLEAVE_* flags.
To fix this, refactor the code to pass the whole xf86CursorInfoPtr
through to cursor_bitpos and compute the correct stride there based on
the flags. If none of the SOURCE_MASK_INTERLEAVE flags are set, use
the total cursor size to move the 'image' variable into the mask part
of the image before computing the desired byte pointer.
Signed-off-by: Aaron Plattner <aplattner@nvidia.com>
Reviewed-by: Robert Morell <rmorell@nvidia.com>
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Tested-by: Cyril Brulebois <kibi@debian.org>
(cherry picked from commit ffcbfa0063)
screenInfo.numScreens is not a valid screen number, they go from 0 to
numScreens - 1.
Signed-off-by: Cyril Brulebois <kibi@debian.org>
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit 279ef1ffd7)
screenInfo.numScreens is not a valid screen number, they go from 0 to
numScreens - 1.
Signed-off-by: Julien Cristau <jcristau@debian.org>
Reviewed-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit 8684543021)
This fixes a gamma issue on vt switch observed with KDM. VT switching away
and back would result in a black screen. Avoid this by storing the current
gamma information on init.
https://bugzilla.redhat.com/show_bug.cgi?id=533217
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit 566d09a5cd)
When entering the VT, re-apply the saved gamma settings for each screen.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit c050aa2f5f)
If any part of the stack calls XIGetKnownProperty during device shutdown
the property is re-initialized before the server generation resets, leaving
the value invalid again.
Move the reset to the extension init which happens before input devices are
initialized before the first property is requested.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Julien Cristau <jcristau@debian.org>
(cherry picked from commit 5cd11d2356)
The version check is hackish, and I highly doubt anyone using this code is on
a version of OS X that predates this support (10.2 has support for it).
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 6eed13ad59)
Simple typo, should have been adjusting the horizontal timings
consistently since we're not trying to mangle vertical at all.
Signed-off-by: Adam Jackson <ajax@redhat.com>
Reviewed-by: Patrick E. Kane <pekane52@gmail.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit b2f9ce201c)
Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=24348
Before (data flag ignored -> broken):
66 DATA:
e944f1 JMP 1ff6
After (fixed):
66 DATA:
e944f1ffff JMP 00001ff8
This subtle difference in the length of decoded instruction meant
that the VBE call jumped to the routine setting AX=0x14F (VBE Failed)
instead of the routine that set AX=0x4F (VBE success).
The ability to run the same code in vm86 significantly aided the
debugging of this issue. Those X.org developers who would like to drop
vm86 better take special care towards _all_ vesa bugs, as those will
expose further issues.
Patch applies easily to even xserver 1.4.2.
Signed-off-by: Luc Verhaegen <libv@skynet.be>
Tested-by: Luc Verhaegen <libv@skynet.be>
Reviewed-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit cc2c73ddcb)
Disable timer/keyboard trapping on GNU/Hurd for now
Trapping disabled for now, as some VBIOSes (mga-g450 notably) use these
ports, and the int10 wrapper is not emulating them.
It's effectively what happens in the Linux variant too, as iopl() is used there,
making the ioperm() meaningless.
Signed-off-by: Olaf Buddenhagen <antrik@users.sf.net>
Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org>
Reviewed-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit a6c64d9645)
DDX driver may implement schedule swap without GetMSC. In that case we
can't call GetMSC in DRI2SwapBuffers.
Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com>
Reviewed-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de>
CC: Keith Packard <keithp@keithp.com>
CC: Kristian Høgsberg <krh@bitplanet.net>
Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit 383dfe23f1)
After we infer the aspect ratio for the screen, we pick the largest
mode matching that aspect ratio from the best mode pool available.
We then clamp virtual size to that mode, and run the resulting mode
list through the driver's ValidMode hook. In doing so we might filter
away our initial guess. If this happens we shrink the default mode
to the next largest mode from _any_ mode pool. This is usually wrong,
and we should instead pick the next aspect-matched mode from the best
available mode pool (as always, user then driver then default).
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Adam Jackson <ajax@redhat.com>
(cherry picked from commit 7bb653bedc)
This fixes a typo introduced in commit
80b5d3a326. The pointer pDst was changed
unintentionally to pWin from a copy/paste error. This resulted in all
QT-based apps and some tcl/tk ones (like fontforge) to crash X 1.9 on
starting up, when Xinerama was enabled.
Bug report: https://bbs.archlinux.org/viewtopic.php?id=106125
Signed-off-by: Elie Bleton <drozofil@gmail.com>
Reviewed-by: Adam Jackson <ajax@redhat.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Tested-by: Linus Arver <linusarver@gmail.com>
(cherry picked from commit 78f94f19aa)
Commit de8be07cc0 adds a requirement to
event to core conversion that at least one of the X or Y valuators are
set in the valuator mask. This commit fixes the event conversion test to
be compliant.
Signed-off-by: Chase Douglas <chase.douglas@canonical.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit 290af0418f)
Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
For whatever reason, some (broken) monitors will crash if you do this.
We're not actually using this information for anything, so let's just
not do it.
Originally reported as http://bugzilla.redhat.com/620333
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Adam Jackson <ajax@redhat.com>
(cherry picked from commit d2064fbb68)
Otherwise, building fails with CFLAGS="-Wall -Werror" this way:
| protocol-common.c: In function ‘init_simple’:
| protocol-common.c:159: error: implicit declaration of function ‘XkbInitPrivates’
| protocol-common.c:159: error: nested extern declaration of ‘XkbInitPrivates’
Signed-off-by: Cyril Brulebois <kibi@debian.org>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit c481dae0dc)
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit 291c39dfef)
Change the damage extension reporter to queue up events after we chain
to the wrapped functions. Damage events are typically sent out after
the rendering happens anyway, since we submit batch buffers from the
flush callback chain and then flush client io buffers. Compositing
managers relie on this order, and there is no way we could reliably
provide damage events to clients before the rendering happens anyway.
By queueing up the damage events before the rendering happens, there's
a risk that the client io buffer may overflow and send the damage
events to the client before the driver has even seen the rendering
request. Reporting damage events after the rendering fixes this
corner case and better corresponds with how we expect this to work.
Signed-off-by: Kristian Høgsberg <krh@bitplanet.net>
Reviewed-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit 8d7b7a0d71)
[anholt: re-applied to revert the revert, now that the cause of the
revert is fixed]
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit 333b6ed26e)
In all these cases, any rendering implied by this damage has already
occurred, and we want to get the damage out to the client. Some of
the DamageRegionAppend calls were explicitly telling damage to flush
the reportAfter damage out, but not all.
Bug #30260. Fixes the compiz wallpaper plugin with client damage
changed to reportAfter.
Signed-off-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit f36153e3ef)
Commit cf88363db0 fixed the handling of
BigReq requests that are way too large and handles the case where the
read() syscall returns a short read. However, it neglected to handle
the case where it returns a long read, which happens when the client
has another request in the queue after the bogus large one.
Handle the long read case by subtracting the smaller of 'needed' and
'gotnow' from oci->ignoreBytes. If needed < gotnow, simply subtract
the two, leaving gotnow equal to the number of extra bytes read.
Since the code immediately following the (oci->ignoreBytes > 0) block
tries to handle the next request, advance oci->bufptr immediately
instead of setting oci->lenLastReq and letting the next call to
ReadRequestFromClient do it.
Fixes the XTS pChangeKeyboardMapping-3 test.
CASES TESTS PASS UNSUP UNTST NOTIU WARN FIP FAIL UNRES UNIN ABORT
-Xproto 122 389 367 2 19 0 0 0 1 0 0 0
+Xproto 122 389 368 2 19 0 0 0 0 0 0 0
Signed-off-by: Aaron Plattner <aplattner@nvidia.com>
Reviewed-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit c80c41767e)
Lines of length greater than 46340 can be drawn with one of the
coordinates being negative. However for dashed lines, miPolyBuildPoly
overflows the int type when setting up edges for a section of the
dashed line. This results in the dashed segments not being drawn at
all.
Signed-off-by: Siddhesh Poyarekar <siddhesh.poyarekar@gmail.com>
Reviewed-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit be7cf14c36)
os/strlc{at,py}.c were trying to include xorg-config.h, which is not
available in dix.
Signed-off-by: Julien Cristau <jcristau@debian.org>
Acked-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit 69e8e1b0b9)
Fix Xserver on GNU/Hurd into using the "mem" device instead of
the deprecated "iopl" device.
Reviewed-by: Olaf Buddenhagen <antrik@users.sf.net>
Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org>
(cherry picked from commit f72aadd38b)
This commit breaks the Compiz "Wallpaper" plugin.
This reverts commit 8d7b7a0d71.
Reviewed-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit 1a0d9324b3)
Change the library order in TSLIB_LIBS variable, placing ${TSLIB_LIBS}
after the local static libraries. This fixes linking with
-Wl,--as-needed.
Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=30600
Signed-off-by: Michał Górny <mgorny@gentoo.org>
Reviewed-by: Daniel Stone <daniel@fooishbar.org>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit 8806a04f7f)
If pkg-config is unable to find tslib but the fallback check does find
it, the compilation continues with tslib support enabled though
TSLIB_LIBS are unset. Thus, the compilation fails with a linking error
on tslib functions.
This patch sets TSLIB_LIBS to '-lts' whenever the tslib fallback check
succeeds.
Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=30599
Signed-off-by: Michał Górny <mgorny@gentoo.org>
Reviewed-by: Daniel Stone <daniel@fooishbar.org>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit c7e436e998)
If ts_open() fails and return NULL, then next call to ts_fd()
segfaults because of NULL dereference. There is no need to
check output of ts_fd() as ts_open() did this internally.
Signed-off-by: Paulius Zaleckas <paulius.zaleckas@gmail.com>
Reviewed-by: Daniel Stone <daniel@fooishbar.org>
Reviewed-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit 8990b31214)
Without this patch, any negative valuator value is wrong when returned
from XQueryDeviceState(). This is a regression from at least xserver
1.4.
Valuator data is set in dix/getevents.c:set_valuators() by copying
signed int values into an unsigned int field
DeviceEvent.valuators.data.
That data is converted into a double with an implicit cast by
assignment to axisVal[i] in Xi/exevents.c:UpdateDeviceState().
That double is converted back to a signed int in
queryst.c:ProcXQueryDeviceState(). If the original value in
set_valuators() is negative, the double value will be > 2^31 and the
conversion back to a signed int is undefined. (Although I
consistently see the value -2^31.)
Fix this by changing the definition of DeviceEvent.valuators.data from
uint32_t to int32_t.
Signed-off-by: Joe Shaw <joeshaw@litl.com>
Reviewed-by: Chase Douglas <chase.douglas@canonical.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit e354ccac36)
Helps debugging greatly, random 8 or 16 bit values can sometimes look like
valid values, causing much excitement on the client front.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Daniel Stone <daniel@fooishbar.org>
(cherry picked from commit ca21a26622)
From the original bug reporter Ezra Reeves:
"I did some more digging on this today, and I found that an HP branded
wireless USB mouse has the same issue. With this mouse (as well as the
logitech wireless mouse), the return from:
xdev = XListInputDevices(GDK_WINDOW_XDISPLAY(rootwin), &ndevices_return);
lists the USB device twice, but both have xdev[num].use == 3
(IsXExtensionKeyboard as defined in X11/XI.h).
[...]
Swapping the order of the test in Xi/listdev.c that determines whether a
device is a pointer or a keyboard properly detects my devices (OEM USB
wireless mouse/kb combo) -- one as a keyboard and one as a pointer."
X.Org Bug 29046 <http://bugs.freedesktop.org/show_bug.cgi?id=29046>
Reported-by: Erik Kilfoil <ekilfoil@gmail.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit fe8d122b13)
EventToCore as of the commit below won't generate core motion events if the
valuator mask for x/y isn't set. For DGA, we work around this check by
forcibly setting the mask in the event we pass down.
commit de8be07cc0
Author: Peter Hutterer <peter.hutterer@who-t.net>
Date: Tue Aug 17 12:08:52 2010 +1000
dix: don't create core motion events for non-x/y valuators.
X.Org Bug 30267 <http://bugs.freedesktop.org/show_bug.cgi?id=30267>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Tested-by: Andrew Randrianasulu <randrik@mail.ru>
Tested-by: Andy Furniss
Reviewed-by: Daniel Stone <daniel@fooishbar.org>
(cherry picked from commit 9872e0f011)
Devices that send motion events with valuators other than x/y get core
motion events with unchanged x/y coordinates. This confuses some
applications.
If the DeviceEvent does not have the x/y valuators set, return BadMatch on
core conversion, thus skipping the event altogether.
Reported-by: Bartosz Brachaczek <b.brachaczek@gmail.com>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Tested-by: Bartosz Brachaczek <b.brachaczek@gmail.com>
(cherry picked from commit de8be07cc0)
Christof Wolf has reported a regression that seems to be caused by
this change, so reverting the change in the 1.9 branch. We'll
investigate a proper fix in master for 1.10.
This reverts commit c89f052104.
event_size is never assigned to in mieqProcessInputEvents(), so realloc()
is always called. This is benign, but I'm guessing not intended.
Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk>
Reviewed-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit da4eebe58e)
Just use memcpy, seriously.
Reviewed-by: Rémi Cardona <remi@gentoo.org>
Signed-off-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
(cherry picked from commit 853d7ebfa3)
