Matthias Hopf
5afc6c1a14
Added linux 2.6 compatible domain aware device scanning code.
...
Additional scanning code uses the /sys interface of 2.6 kernels.
Cleaned up the use of tags and already split domain/bus/dev/fn.
2006-11-03 18:59:51 +01:00
Egbert Eich
6319f7d713
Make int10 fully domain aware.
2006-11-03 18:59:51 +01:00
Matthias Hopf
f4dd2665b0
Added missing domain stripping in already domain aware code.
2006-11-03 18:59:51 +01:00
Matthias Hopf
46901063e8
Build with -D_PC on ix86 only.
2006-11-03 18:59:51 +01:00
Keith Packard
543b397277
Merge branch 'origin' into randr-1.2
2006-11-02 19:00:35 -08:00
Alan Coopersmith
037f23e6f8
Convert Xprt DDX to new motion history api
2006-11-02 18:30:58 -08:00
Alan Coopersmith
7dc54a40e9
Remove references to xf86Info.kbd* from solaris code
2006-11-02 16:38:47 -08:00
Alan Coopersmith
e46f7f78b3
sun_bell.c needs to #include "xf86_OSlib.h"
2006-11-02 16:38:12 -08:00
Daniel Stone
01afa533aa
xfree86/os-support: axe more unused files
2006-11-02 04:32:37 +02:00
Daniel Stone
57c1409151
xfree86/os-support: remove unused file from dist
2006-11-02 04:21:06 +02:00
Daniel Stone
1d65429a9e
xfree86: don't attempt to enable and disable non-DIX devices
...
Don't try to enable and disable devices with no entry in the DIX, such as
the evdev brain.
2006-11-02 04:12:55 +02:00
Daniel Stone
18c246a13b
Merge branch 'input-hotplug'
2006-11-02 03:21:37 +02:00
Daniel Stone
ba9f5138fc
xfree86: allow starting with no input devices
...
Add a server flag (AllowEmptyInput), which will inhibit adding the
standard keyboard and mouse drivers, if there are no input devices in the
config file.
2006-11-02 03:16:10 +02:00
Daniel Stone
be291a6d97
Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug
2006-11-02 03:15:25 +02:00
Daniel Stone
a2d6242106
kdrive:remove Change{Keyboard,Pointer}Device
...
This was removed in the DIX, so just axe it.
2006-11-02 00:46:33 +02:00
Daniel Stone
81728558a0
input: add non-keyboard bell ringing function
...
Add a generic 'ring the bell' function (console bell on Linux and BSD,
/dev/audio on Solaris), and add DDX functions for this. Make this the
core keyboard's bell.
Port Xvfb and Xnest to this.
Port XFree86 to this, with OS-specific hooks for Linux, BSD, and Solaris
taken from foo_io.c in the old layer.
2006-11-01 23:10:26 +02:00
Daniel Stone
3df454719f
kdrive: add KdRingBell prototype
2006-11-01 23:02:57 +02:00
Alan Coopersmith
c03311a1e7
Fix automake error: BUILT_SOURCES was defined multiple times on Solaris
2006-10-31 16:05:48 -08:00
Matthieu Herrb
645d057620
Handle building in a separate objdir
2006-10-29 18:19:56 +01:00
Matthieu Herrb
59584c375f
kill GNU-make'ism.
2006-10-29 17:49:46 +01:00
Thomas Hellstrom
06b6b971d0
Make sure we have 64-bit file-offsets in libdrm.
2006-10-29 15:22:37 +01:00
Daniel Stone
644ee2434a
kdrive: avoid null dereferences
...
Avoid jumping into NULL pointers for OS or card functions which do not
exist.
2006-10-29 03:52:11 +03:00
Daniel Stone
738d2e8817
kdrive: move bell ringing into an OS function
...
Move the bell into an OS function, and use that if it's declared; else,
fall back to using the driver's function.
Remove the Linux keyboard bell function; just move it into the OS layer.
Use named initialisers when converting the old structures, and eliminate
unused functions.
2006-10-29 03:48:02 +03:00
Daniel Stone
96e32805d1
Xi: disallow changing core keyboard and pointer
...
Just short-circuit the change core keyboard/pointer requests.
2006-10-29 03:43:34 +03:00
Thomas Hellstrom
92d04e746b
Import libdrm functions for the drm memory manager.
2006-10-27 18:26:30 +02:00
Myron Stowe
794f2e7291
xfree86: re-enable chipset-specific drivers for Linux/ia64
...
Re-enable chipset-specific support for Linux/ia64, by linking in
lnx_ia64.c.
2006-10-26 20:38:58 +03:00
Daniel Stone
cdc8a4b7b2
Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug
2006-10-26 00:28:30 +03:00
Michel Dänzer
becbda6d51
Fix test for Option "IgnoreABI".
...
This option has plenty of potential for wasting the time of bug triagers
without pretending it's always on.
2006-10-25 11:39:27 +02:00
Keith Packard
59511974db
Merge branch 'master' into randr-1.2
2006-10-24 17:26:20 -07:00
Daniel Stone
65cd5aa4d7
kdrive/input: remove unnecessary #ifdef XINPUTs
...
Xi is now mandatory, so don't bother with the ifdefs.
2006-10-25 01:12:45 +03:00
Daniel Stone
0514d53e10
xfree86: remove motion history handling
...
Remove motion history handling, as we now deal with this in the DIX.
2006-10-25 01:10:44 +03:00
Daniel Stone
b1debebf8f
mi: remove mi motion history
...
This is now unneeded as we do motion history in the DIX.
2006-10-25 01:10:20 +03:00
Daniel Stone
b9e180e632
port all users to the new DIX motion history API
...
Port KDrive, Xvfb, and Xnest, as well as the virtual core devices, to the
new motion history API. Make GetPointerEvents also update the history.
2006-10-25 01:09:19 +03:00
Daniel Stone
2a74b8a91d
xfree86: re-bump input abi to 1.0, yet again
2006-10-25 01:04:53 +03:00
Daniel Stone
0cd6a3d8ef
xfree86/input: add proximity support, free valuators
...
Re-add support for proximity events, and remember to both va_end our
varargs, and free our valuators.
2006-10-23 06:56:07 +03:00
Daniel Stone
a7c2d9a15d
kdrive: numerous warning fixes
2006-10-23 06:08:27 +03:00
Daniel Stone
c8f76fb3a4
xfree86/input: disallow pointer device changes, fix drag event calculation
...
Don't allow users to change the core pointer.
Fix xf86SendDragEvents to check the device button state, not the core
pointer's.
Remove unused xf86CheckButton.
2006-10-23 05:12:15 +03:00
Daniel Stone
c5b5b046e8
Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug
2006-10-23 02:58:30 +03:00
Daniel Stone
fab1d37ecb
xfree86/options: don't blindly dereference NULL pointers
...
Sure, conf_idev should be initialised, but if it's not, try to make the
best of a bad situation.
2006-10-23 02:57:00 +03:00
Daniel Stone
523fe64ba1
kdrive/lots of DDXes: make stubs use KdOsAddInputDrivers()
...
In theory, removing LinuxPciScan should make these drivers OS-independent.
2006-10-23 02:56:28 +03:00
Daniel Stone
57f98e2b72
kdrive/tslib: port to new absolute API, remove debugging
...
Port to new absolute device (not touchscreen) API, and remove some
leftover debugging.
2006-10-23 02:55:41 +03:00
Daniel Stone
a6dc629052
kdrive/os: add KdOsAddInputDrivers
...
Add KdOsAddInputDrivers, which adds all relevant input drivers.
Could possibly be refactored to KdAddInputDrivers, which called through
OsFuncs to a new function, if it existed.
2006-10-23 02:54:59 +03:00
Daniel Stone
ac240b00ff
kdrive/input: use Absolute instead of Touchscreen
...
Adjust to suit Zepheniah's new absolute device API.
2006-10-23 02:54:07 +03:00
Daniel Stone
b1ba4b3e8a
kdrive/neomagic: include vesa.h, use DebugF
2006-10-23 02:53:15 +03:00
Daniel Stone
2f33f4065d
Xnest: port to new input API
...
Port Xnest to Get{Pointer,Keyboard}Events, plus the new mieq API.
2006-10-23 02:52:35 +03:00
Daniel Stone
08928afb05
Xnest: disable XKB, reshuffle code
...
Disable XKB, as we can't yet use it; move Composite disabling to
ddxInitGlobals, along with XKB.
2006-10-23 02:51:52 +03:00
Daniel Stone
cd3b16a57e
Xvfb: port to new mieq API
2006-10-23 02:51:13 +03:00
Daniel Stone
f08b6b2367
kdrive: change DEVICE_TOUCHSCREEN to DEVICE_ABS_{CALIB,AREA}
2006-10-22 12:30:02 +03:00
Drew Parsons
948a97b97e
Minor typo fix to xorg.conf man page.
2006-10-21 23:09:22 +10:00
Zephaniah E. Hull
e26a494f41
Merge branch 'input-hotplug' of git+ssh://git.freedesktop.org/git/xorg/xserver
2006-10-21 04:26:14 -04:00
Zephaniah E. Hull
736b0d5078
DEVICE_TOUCHSCREEN becomes DEVICE_ABS_CALIB.
...
Update the DEVICE_ABS_CALIB stuff to include the new elements.
New DEVICE_ABS_AREA support.
dev->touchscreen becomes dev->absolute, with _CALIB and _AREA stuff in it.
Update xfree86 to compile with this, kdrive needs an update too.
2006-10-21 04:24:49 -04:00
Carl Switzky
edd5f17454
Add ast driver/device info to Xorg server & config utilities
2006-10-19 17:30:54 -07:00
Daniel Stone
aeba855b07
move keymap copy to event processing, from enqueuing
...
Move the keymap copying to event processing time (in
ProcessInputEvents), instead of being at event enqueuing time.
Break SetCore{Pointer,Keyboard} out into separate functions.
Change mieqEnqueue to take a device pointer, that asks for the
_original_ device associated with this event.
2006-10-20 00:43:11 +03:00
Alan Coopersmith
b0780312d8
Pre-release message should tell users to check git, not CVS, for updates
2006-10-20 00:43:10 +03:00
Joshua Baergen
c5dc997baf
Create xorg.conf.example (Gentoo bug #138623 ).
2006-10-20 00:43:10 +03:00
Alan Hourihane
85ac2f16ab
Small modification to blocking signals when switching modes.
2006-10-20 00:42:47 +03:00
Michel Dänzer
a232693c8c
Add per-drawable Xv colour key helper function.
...
This allows overlay Xv adaptors to work slightly better with compositing
managers.
Bump the video driver ABI minor so drivers only need to check for this at build
time.
2006-10-20 00:42:44 +03:00
Alan Coopersmith
5563861ab7
Make sure xorgcfg files are included even when dist made with --disable-xorgcfg
2006-10-20 00:42:07 +03:00
Alan Coopersmith
e1dd1904c6
Pre-release message should tell users to check git, not CVS, for updates
2006-10-19 14:27:15 -07:00
Joshua Baergen
62d2409793
Create xorg.conf.example (Gentoo bug #138623 ).
2006-10-19 11:14:26 -07:00
Zephaniah E. Hull
1b1698af41
Pass SetDeviceValuators down to the driver.
...
NOTE: This changes the LocalDeviceRec struct, which breaks input drivers.
2006-10-18 04:57:22 -04:00
Alan Hourihane
d430e76a16
Small modification to blocking signals when switching modes.
2006-10-16 12:39:05 +01:00
Daniel Stone
4ae7745a0d
xfree86 input: always open devices on NewInputDeviceRequest
2006-10-15 21:48:01 +03:00
Michel Dänzer
f9bfee5098
Add per-drawable Xv colour key helper function.
...
This allows overlay Xv adaptors to work slightly better with compositing
managers.
Bump the video driver ABI minor so drivers only need to check for this at build
time.
2006-10-15 16:55:09 +02:00
Alan Coopersmith
b9108a13fc
Make sure xorgcfg files are included even when dist made with --disable-xorgcfg
2006-10-13 19:05:28 -07:00
Daniel Stone
335b503c5e
Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug
2006-10-13 18:10:45 +03:00
Alan Hourihane
cf948b7b04
Block IO on switchmode just as we do for VT switching
...
to avoid cursor movements signalling io.
2006-10-13 14:48:10 +01:00
Matthias Hopf
1b94c117e0
Fixing probably long-standing bug in domain support.
2006-10-10 19:37:22 +02:00
Daniel Stone
ef7e05e9de
xfree86/xinput: pass first_valuator params, don't clip to first screen bound
...
Pass first_valuator to GetPointerEvents/GetKeyboardValuatorEvents.
Don't clip axis bounds to that of screen 0 when initialising axes.
2006-10-08 21:42:15 +03:00
Daniel Stone
41bb9fce47
mipointer: take device arguments, split miPointerAbsoluteCursor
...
Update mipointer API to take a device argument to (almost) all functions,
and split miPointerAbsoluteCursor into a couple of separate functions.
Remove miPointerAbsoluteCursor call from mieq, as we now deal with it in
GetPointerEvents.
Make miPointerSetPosition (successor of miPointerAbsoluteCursor) take
pointers to x and y, so it can return the clipped values.
Modify callers of miPointer*() functions to generally use the new
functions.
This should fix things with multi-head setups.
2006-10-08 20:34:32 +03:00
Daniel Stone
be8dfafd1d
warning cleanups
...
Fix still more warnings.
2006-10-08 20:32:16 +03:00
Daniel Stone
3ae4d25018
xfree86 Xinput: remove still more excessive debugging
...
There isn't any more debugging left for input events in the XFree86 DDX.
2006-10-08 16:20:42 +03:00
Daniel Stone
58314756ae
GetPointerEvents: add first_valuator parameter
...
Add a first_valuator parameter. Looks correct by inspection, but untested
with first_valuator != 0 as yet.
2006-10-08 16:18:05 +03:00
Daniel Stone
8d8e7f8bae
kdrive/input: remove excessive debugging in NIDR
2006-10-08 15:26:54 +03:00
Daniel Stone
22a836fafd
xfree86/loader: bump input major ABI version
...
Bump input major ABI version to 1.0, since we removed the OS keyboard
layer.
2006-10-07 14:16:51 +03:00
Alan Coopersmith
4c34224630
Merge branch 'XACE-modular'
2006-10-06 18:01:13 -07:00
Daniel Stone
de63a469dc
Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug
2006-10-05 20:29:19 +03:00
Keith Packard
b4659faf9b
Merge branch 'master' into randr-1.2
2006-10-04 14:46:04 -07:00
Alan Coopersmith
8382234a7f
Update pci.ids to Tue 2006-10-03 daily snapshot
2006-10-03 17:49:48 -07:00
Alan Coopersmith
96edf7b853
Don't insert RCS tag into generated pci id header files
2006-10-03 17:48:50 -07:00
Daniel Stone
518db35ca3
Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug
2006-09-29 00:35:21 +03:00
Daniel Stone
ad631afcf3
make core keyboard/pointer map functions act on all core-sending devices
...
Make Set{Keyboard,Modifier,Pointer}Mapping act on all devices which send core
events.
Change LegalModifier to accept a DeviceIntPtr.
2006-09-29 00:34:23 +03:00
Aaron Plattner
4bc5dc2854
Merge branch 'wfb'
2006-09-28 13:27:13 -07:00
Michel Dänzer
2206a92a97
Push information about cliprects of DRI windows to the DRM.
2006-09-27 10:48:49 +02:00
Michel Dänzer
54d371e7a4
Add wrapper for new ioctl to update drawable information in the DRM.
2006-09-27 10:48:41 +02:00
Daniel Stone
ad321fad4b
Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug
2006-09-24 17:56:43 +03:00
Brian
f7c1d94241
Check for visual==NULL in dmxBECreateColormap() before calling XCreateColormap()
...
to prevent potential segfault.
2006-09-23 10:38:10 -06:00
Brian
945b7c6394
The fbcmap.c file used by Xdmx _must_ be compiled with XFree86Server defined.
...
Otherwise, Xdmx generates a slew of protocol errors.
2006-09-23 10:35:25 -06:00
Brian
891e9c3e6c
Replace broken DMXDBG3() with DMXDBG2()
2006-09-23 10:28:24 -06:00
David Nusinow
a10039a100
Allow the xfree86 ddx utils to be optionally built. Patch by Eugene Konev.
2006-09-21 23:58:32 -04:00
Aaron Plattner
c1655f0fd4
Bug 8386: Grow parser buffers to fit an entire line if it's longer than CONFIG_BUF_LEN.
2006-09-21 14:45:17 -07:00
Drew Parsons
8b4ed47c5d
* Install Xprint's Xsession script to $(sysconfdir)/X11/Xsession.d
...
* Removing outdated references to CDE and dt, rename script to
92xprint-xpserverlist.
2006-09-21 22:19:44 +10:00
Donnie Berkholz
d812f486a0
Really fix sparc on 64-bit kernel/32-bit userland.
...
Commit b3a3020fd0
used a sparc64 ifdef instead of
sparc. But for 32-bit userland, __sparc64__ is not defined so the wrong code is
used.
2006-09-20 15:39:39 -07:00
Keith Packard
2be1ac15ae
Remove smashing of CFLAGS from server build.
...
CFLAGS is a user variable, extracted from the environment at configure time
and settable by the user at build time. We must not override this variable.
2006-09-18 12:11:18 -07:00
Adam Jackson
f057de4f73
Don't install librac.a.
...
Thanks, automake.
2006-09-16 03:49:11 -04:00
Eamon Walsh
49a70c8570
Merge branch 'master' into my-XACE-modular
2006-09-15 15:26:57 -04:00
Adam Jackson
46af6d1e95
Always believe the monitor when it reports a reduced-blanking mode.
...
CVT reduced blanking modes are typically only seen on digital connections to
LCDs, but there are some monitors that report them as supported over the
VGA connector too, which is perfectly legitimate, electrically speaking.
2006-09-14 19:28:44 -04:00
Adam Jackson
63acf18b7e
In xf86MatchPciInstances, fail gracefully when there's no PCI device at all.
...
This allows the autoconfig logic to fall through sanely on non-PCI machines,
which importantly includes Xen virtual machines.
2006-09-14 19:26:37 -04:00
Adam Jackson
a8f9936f55
Prefer driver-provided modes when matching name strings to modelines.
...
Well, kinda. Strictly we prefer M_T_BUILTIN strongest since those are modes
where the driver has said it absolutely can't do anything else (VBE). Then
we look for user-defined modes, ie, modelines from the config file. Then
we consider modes reported by the monitor via EDID. Finally if nothing has
matched yet we consider the default mode pool.
Within each of the above-mentioned classes, modes with the M_T_PREFERRED bit
take priority over other modes in the same class.
This logic ensures that the timings sent to the monitor exactly match the
timings it reported as supported, which occasionally don't match the numbers
you might get for that mode from CVT or GTF.
2006-09-14 19:24:41 -04:00
Adam Jackson
81ef1b6d60
Mark EDID modes as driver modes. Infer virtual size from driver modes.
...
This allows the server to guess an appropriate initial virtual size and
resolution. The heuristic is to select the largest driver-reported mode
that matches the monitor's physical aspect ratio. We revalidate this
estimate after mode validation, since we may have filtered away all
modes that would fill that size.
Also, the EDID preferred timing is now marked as M_T_PREFERRED as well.
2006-09-14 19:18:58 -04:00