summaryrefslogtreecommitdiff
path: root/hw
AgeCommit message (Collapse)AuthorFilesLines
2011-05-25xserver: remove AbsoluteClass, breaking the A(P|B)ISimon Thum1-1/+1
This struct was unused and has been effectively removed in commit 633b81e8ba09cc6a1ea8b43f323874fda2cf0bde Refs: xorg-server-1.10.0-133-g633b81e Remove the remainder, with an ABI bump to 13.0. Signed-off-by: Simon Thum <simon.thum@gmx.de> Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
2011-05-25xfree86: Allow "MatchLayout" statements in config filesOleh Nykyforchyn5-0/+54
Usage example (tested on a dual-seat PC): Section "InputClass" Identifier "keyboard-all" MatchIsKeyboard "on" MatchDevicePath "/dev/input/event*" MatchLayout "!GeForce|!Matrox" Driver "evdev" Option "XkbLayout" "us" Option "XkbOptions" "terminate:ctrl_alt_bksp" EndSection It disables auto keyboard configuration for layouts "GeForce" and "Matrox". Note that "" in patterns means "no Layout sections found", e.g. MatchLayout "GeForce|" is "in layout GeForce or without explicit layout at all". Signed-off-by: Oleh Nykyforchyn <oleh.nyk@gmail.com> Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> Acked-by: Dan Nicholson <dbn.lists@gmail.com> Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
2011-05-25Xephyr: fix pointer coordinate translation when screen is rotatedTomas Frydrych1-2/+23
In the Xephyr case the position of the pointer relative toward the Xephyr window is controlled by the host server without taking into account rotation of the Xephyr screen. Consequently the pointer coords must always be translated when the fb is rotated. Signed-off-by: Tomas Frydrych <tomas@sleepfive.com> Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
2011-05-25Xephyr: added dummy ephyrDeviceCursorCleanup() to avoid crashingTomas Frydrych1-1/+6
The DeviceCursorCleanup in miPointerSpriteFuncRec can no longer be NULL it seems. Signed-off-by: Tomas Frydrych <tomas@sleepfive.com> Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
2011-05-24Fix XWin compilation after updates for input API changesColin Harrison2-6/+4
Fix XWin compilation after updates for input API changes in commits e7150db5 8670c46b and 20fb07f4 Also remove a no longer needed InternalEvent* variable Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
2011-05-23xfree86: bump to video ABI 11Peter Hutterer1-1/+1
We've broken the ABI with some commit and drivers built against ABI 10 happily segfault now. (The relevant patch is 51f353d0a0d116af16d7d9590cadef6c56328746 which changed the ATOM typedef from unsigned long to uint32_t, thanks to Cyril Brulebois <kibi@debian.org> for figuring this out) Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> Signed-off-by: Keith Packard <keithp@keithp.com>
2011-05-22XQuartz: Don't crash if CG increases our display resolutionJeremy Huddleston1-2/+3
miPaintWindow would cause fbFill() to overwrite pScreen's pixmap which was sized for the old resolution. Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-22XQuartz: RandR: Don't crash if X11 is launched while there are no attached ↵Jeremy Huddleston3-100/+117
displays If CG reports no displays when launching, we could crash in RandR. Instead, just provide a fake 800x600 display until we are notified about displays being attached. Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-22XQuartz: Mark functions _X_NORETURNJeremy Huddleston3-2/+4
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-22XQuartz: Silence clang warnings about shadow declarationsJeremy Huddleston2-4/+4
X11Application.m:1272:26: warning: declaration shadows a local variable [-Wshadow,Semantic Issue] xp_error e; ^ X11Application.m:1098:36: note: previous declaration is here - (void) sendX11NSEvent:(NSEvent *)e { ^ 1 warning generated. bundle-main.c:648:36: warning: declaration shadows a local variable [-Wshadow,Semantic Issue] int max_files, i; ^ bundle-main.c:594:9: note: previous declaration is here int i; ^ 1 warning generated. Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-22XQuartz: Update DEBUG_LOG to report to ASLJeremy Huddleston5-17/+38
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-14XQuartz: RandR: Avoid over-releasing if we are unable to determine the ↵Jeremy Huddleston1-8/+11
current display mode. Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-14XQuartz: Don't call mieqEnqueue during server shutdownJeremy Huddleston2-11/+11
Found-by: GuardMalloc Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-14XQuartz: Fix an array-index-out-of-bounds crasherJeremy Huddleston1-1/+1
Found-by: GuardMalloc Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-13dmx: fix warning for doxygen explicit linksGaetan Nadon1-2/+2
Explicit links to functions in another file are not supported. Reviewed-by Jeremy Huddleston <jeremyhu@apple.com> Signed-off-by: Gaetan Nadon <memsize@videotron.ca> Signed-off-by: Keith Packard <keithp@keithp.com>
2011-05-13dmx: modernize doxygen generation.Gaetan Nadon6-745/+954
The configuration and stylesheet were very old. The stylesheet is not checked-in, use the generated one. The header is not checked-in, use the generated one. Add datetime and projectname in default footer. Developer documentation is not installed and not included in tarball. Reviewed-by Jeremy Huddleston <jeremyhu@apple.com> Signed-off-by: Gaetan Nadon <memsize@videotron.ca> Signed-off-by: Keith Packard <keithp@keithp.com>
2011-05-13dmx: split DocBook/XML and Doxygen makefile targetsGaetan Nadon9-261/+278
Maintaining either requires full knowledge of both. It's not obvious one has to check the usage of global variables in devbook.am when maintaining doxygen target. Or vice-versa. Being in their respective directory, one less thing to worry about. Reviewed-by Jeremy Huddleston <jeremyhu@apple.com> Signed-off-by: Gaetan Nadon <memsize@videotron.ca> Signed-off-by: Keith Packard <keithp@keithp.com>
2011-05-13dmx/doc: use common makefile for developers documentationGaetan Nadon1-4/+9
The user/specs docs now have external references support. Developers doc are not installed so they do not participate. However, using a similar makefile shared amongst developers document reduces maintenance and is forward looking. Reviewed-by Jeremy Huddleston <jeremyhu@apple.com> Signed-off-by: Gaetan Nadon <memsize@videotron.ca> Signed-off-by: Keith Packard <keithp@keithp.com>
2011-05-13doc: use common makefile for developers documentationGaetan Nadon9-46/+22
The user/specs docs now have external references support. Developers doc are not installed so they do not participate. However, using a similar makefile shared amongst developers document reduces maintenance and is forward looking. Man pages being out of here, reorg developers docs under the same roof. Drop the obsolete sgml subdir. Reviewed-by Jeremy Huddleston <jeremyhu@apple.com> Signed-off-by: Gaetan Nadon <memsize@videotron.ca> Signed-off-by: Keith Packard <keithp@keithp.com>
2011-05-13dri2: Don't send so many needless invalidate eventsVille Syrjälä1-1/+6
Only send invalidate events for drawables if some client has requested some buffers. Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> Reviewed-by: Michel Dänzer <michel@daenzer.net> Signed-off-by: Keith Packard <keithp@keithp.com>
2011-05-13Merge remote-tracking branch 'whot/for-keith'Keith Packard11-176/+56
2011-05-13Merge remote-tracking branch 'jeremyhu/master'Keith Packard13-89/+660
2011-05-13XQuartz: Don't circumvent NDEBUGJeremy Huddleston1-6/+0
If someone wants to turn off asserts with NDEBUG, let them. Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-12XQuartz: Redirect stdout/stderr to aslJeremy Huddleston5-3/+493
In order to improve logging in XQuartz, stdout and stderr should be redirected to asl (syslog). Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-12XQuartz: Add a LOGGING section to our man pageJeremy Huddleston1-1/+24
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-12XQuartz: stub: Log directly to ASL rather than stdout/stderrJeremy Huddleston2-40/+47
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-12Fix a typo: laucnd instead of launchdJeremy Huddleston1-19/+19
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-10XQuartz: Don't call into CoreFoundation after fork() and before exec()Jeremy Huddleston1-5/+24
After fork()ing, we should just limit ourselves to setting up the environment, file descriptors, and exec()ing. Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-11input: remove DDX event list handlingPeter Hutterer11-66/+27
The current approach to event posting required the DDX to request the event list (allocated by the DIX) and then pass that list into QueuePointerEvent and friends. Remove this step and use the DIX event list directly. This means that QueuePointerEvent is not reentrant but it wasn't before anyway. Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-11input: replace EventListPtr with InternalEvent arrayPeter Hutterer9-13/+13
EventListPtr is a relic from pre-1.6, when we had protocol events in the event queue and thus events of varying size. Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-11input: Provide Queue{Button|Keyboard|Proximity}Event helpersPeter Hutterer8-135/+54
Don't require every caller to use GPE + mieqEnqueue, provide matching Queue...Event functions instead. Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-11xfree86: print the device ID to the log when adding a device.Peter Hutterer1-2/+2
Sometimes the name isn't enough, it's handy to see the device ID's from the log file. Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
2011-05-03XQuartz: prefs_copy_url and prefs_get_copy return retained objectsJeremy Huddleston1-1/+2
No functional change. This just annotates the return policy. Found-by: clang static analyzer Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-03XQuartz: Ensure that {CF,NS}_RETURNS{,_NOT}_RETAINED are definedJeremy Huddleston1-1/+38
These will be used in subsequent patches to denote proper retain counts in XQuartz Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-05-01Make xorg.conf.example rule compatible with Solaris makeAlan Coopersmith1-1/+1
Solaris make won't substitute $< in explicit rules, only implicit ones Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> Reviewed-by: Gaetan Nadon <memsize@videotron.ca>
2011-04-29XQuartz: Fix incorrect typedefs with XPLUGIN_VERSION < 4Jeremy Huddleston1-3/+3
Ok, this time for sure... how many brown bags can I fit over my face? Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-04-29XQuartz: Use a rwlock instead of a mutex to protect window_hash in the ↵Jeremy Huddleston1-10/+10
pthread case Concurrent reads are acceptable, so using an rwlock should be better. Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-04-29Merge remote-tracking branch 'jturney/remove-opengl-spec-download'Keith Packard5-36/+11
2011-04-28hw/xwin: wglext.h should be provided by w32api, rather than downloadedJon TURNEY4-9/+5
wglext.h should be provided by the w32api package, rather than downloaded. if it's not, do 'wget -P /usr/include/w32api/GL http://www.opengl.org/registry/api/wglext.h' Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> Reviewed-by: Colin Harrison <colin.harrison@virgin.net> Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
2011-04-28configure: Look for Khronos OpenGL spec files using pkg-configJon TURNEY2-29/+8
Look for Khronos OpenGL spec files using pkg-config, rather than downloading them Also add a --with-khronos-spec-dir=PATH configure option so XWin can be directed where to find these files without using the khronos-spec-files package XWin with AIGLX requires OpenGL spec files in order to generate wrapper code which: (1) thunks from the glapi dispatch table which uses the default cdecl calling convention to native GL functions using the stdcall calling convention. (2) performs function address lookup for OpenGL 1.2+ functions, which are treated as extensions and so not directly linkable. v2: KHRONOS_SPEC_DIR is only valid when XWIN_GLX_WINDOWS is defined. Avoid 'make dist' seeing invalid dependencies by only including rules using KHRONOS_SPEC_DIR if XWIN_GLX_WINDOWS is defined Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> Reviewed-by: Colin Harrison <colin.harrison@virgin.net> Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
2011-04-27XQuartz: BuildFix to build correctly with XPLUGIN_VERSION < 4Jeremy Huddleston1-12/+6
This fixes a regression introduced by d79cc14a51f3e8d4d2f66aad055092672cab1526 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2011-04-27Merge remote-tracking branch 'jturney/master'Keith Packard15-52/+103
2011-04-27Cygwin/X: Handle failure during winScreenInit()Jon TURNEY6-5/+14
Handle failure during winScreenInit() a bit more cleanly, rather than crashing This avoids a crash with 'XWin -fullscreen -screen 0 @2 -screen 1 @1' Also document that fullscreen may only be applied to one screen. Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> Reviewed-by: Colin Harrison <colin.harrison@virgin.net> Tested-by: Colin Harrison <colin.harrison@virgin.net>
2011-04-27Cygwin/X: Make winOverrrideStyle() thread-safeJon TURNEY3-26/+24
Make winOverrrideStyle() thread-safe winOverrideStyle() is called from the internal WM client thread. Accessing server-internal data structures to get window name and class is not safe, as there is no lock to ensure we do not collide with these data structures being updated in the server thread. Rewrite so the internal client thread uses X client calls to obtain this data safely Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> Reviewed-by: Colin Harrison <colin.harrison@virgin.net> Tested-by: Colin Harrison <colin.harrison@virgin.net>
2011-04-27Cygwin/X: Cosmetic fixes to logging of result from X*TextPropertyToTextList()Jon TURNEY1-1/+4
Report XLocaleNotSupported result from X*TextPropertyToTextList() Fix formatting for unknown results reported for X*TextPropertyToTextList() Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> Reviewed-by: Colin Harrison <colin.harrison@virgin.net> Tested-by: Colin Harrison <colin.harrison@virgin.net>
2011-04-27Cygwin/X: Decorate function pointers retrieved via GetProcAddress with WINAPIJon TURNEY2-2/+2
Decorate function pointers retrieved via GetProcAddress which are currently missing it with WINAPI, to ensure stdcall convention is used when calling them. This fixes a crash currently seen when compiled -O2 and the -screen option uses a size and monitor number e.g. -screen 0 1280x1000@2 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> Reviewed-by: Colin Harrison <colin.harrison@virgin.net> Tested-by: Colin Harrison <colin.harrison@virgin.net>
2011-04-27Cygwin/X: Don't make InputOnly windows visibleJon TURNEY1-6/+11
Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> Reviewed-by: Colin Harrison <colin.harrison@virgin.net> Tested-by: Colin Harrison <colin.harrison@virgin.net>
2011-04-27Cygwin/X: Fix a GDI bitmap resource leak of window iconsJon TURNEY3-12/+24
Ensure any icon created specially for a window is destroyed when the window is destroyed Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> Reviewed-by: Colin Harrison <colin.harrison@virgin.net> Tested-by: Colin Harrison <colin.harrison@virgin.net>
2011-04-27Cygwin/X: Internal WM workaround for Java AWT bugJon TURNEY1-0/+24
Java applications using AWT on JRE 1.6.0 break with non-reparenting WMs AWT doesn't explicitly know about (See sun bug #6434227) XDecoratedPeer.handleConfigureNotifyEvent() only processes non-synthetic ConfigureNotify events to update window location if it's identified the WM as a non-reparenting WM it knows about (compiz or lookingglass) Rather than tell all sorts of lies to get XWM to recognize us as one of those, simply send a synthetic ConfigureNotify for every non-synthetic one Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> Reviewed-by: Colin Harrison <colin.harrison@virgin.net> Tested-by: Colin Harrison <colin.harrison@virgin.net>
2011-04-26configure: Let configure --enable/disable-aiglx control building of AIGLX ↵Jon TURNEY1-1/+1
for all DDXs Let configure --enable/disable-aiglx control building of AIGLX for all DDXs. Currently we can't use --enable/disable-aiglx to control if Xwin DDX is built with AIGLX enabled, as at the moment it's forced off if we aren't building the X.Org DDX DRI or DRI2 loader Rearrange things a bit, introducing a new automake conditional, AIGLX_DRI_LOADER to specifically indicate if the X.Org DDX DRI/DRI2 loader convenience library should be built, and replace the previous X.Org DDX-specific uses of the AIGLX conditional with that As before, AIGLX_DRI_LOADER is only enabled if --enable-glx, --enable-aiglx and at least one of --enable-dri or --enable-dri2 are enabled This allows the general conditional AIGLX to control if AIGLX is built for the XWin DDX as well The C #define AIGLX set by AC_DEFINE(AIGLX) seems to be obsolete, I can't find anything which checks it Updated for ajax's "glx: Make --disable-dri not disable AIGLX" patch, which allows DRI2 to be enabled independently of DRI1 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> Reviewed-by: Colin Harrison <colin.harrison@virgin.net> Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
generated by cgit v1.2.3 (git 2.39.1) at 2024-11-22 03:51:11 +0000