a95bb52b43
devPrivates rework: add missing include of dix/privates.h
2007-11-19 16:34:38 -05:00
ea9c63e93b
DEFAULT_DPI was undefined here.
2007-11-19 15:53:49 +01:00
db9ae86353
Bump DEFAULT_DPI to 96.
...
75 is just nonsense.
2007-11-18 11:57:01 -05:00
ee2d4626dc
Bug #12932 : Use DEFAULT_DPI in randr1.2 instead of hardcoded 96.
2007-11-18 11:56:31 -05:00
a46c30c3be
Bug #12531 : RRModesForScreen can fail to allocate.
2007-11-18 11:45:13 -05:00
fac7e7e4e1
Document the requirement for interleaved code and declarations.
2007-11-18 11:39:26 -05:00
bad96e5a86
dix: getevents.c remove trailing whitespaces.
2007-11-18 15:13:45 +10:30
75e5e9f88c
Xi: add missing XI_DeviceClassesChangedMask to XI filters array.
2007-11-18 15:12:50 +10:30
d15339a92c
Bug #9725 : Don't look in root's $HOME for config files, that's just confusing.
2007-11-17 22:12:10 -05:00
748cfbc820
Disinfect mi/ of mfb.
2007-11-17 21:23:05 -05:00
a969db091c
XKB: Don't ring the bell when we don't have a BellProc (bug #13246 )
2007-11-17 22:51:39 +01:00
c89b543198
exa: set driverPriv to NULL before it might get used later with garbage
2007-11-17 18:20:49 +10:00
514ba4ca72
Bug #1612 : Use a stronger PRNG.
...
Currently just reads from /dev/urandom, and only on Linux.
2007-11-16 19:53:11 -05:00
12dd6e9911
dix: reset MD's devPrivate classes to NULL before device initialisation.
...
XkbInitKeyboardDefviceStruct may call FatalError if it fails. FatalError then
cleans up all the devices, resulting in a segfault if the pointer is
uninitialised.
2007-11-16 17:23:09 +10:30
be3321c2e9
dix: Free both current classes and original classes when closing an MD.
2007-11-16 12:12:41 +10:30
3c39dd19ec
Xi: Deep-copy full list of FeedbackClasses.
...
All feedback classes are linked lists and the whole list has to be duplicated,
not just the first entry.
Xkb stuff still missing.
2007-11-16 12:12:14 +10:30
497862df2f
dix: explicitly float all attached SDs before closing down devices.
...
Some drivers flush on shutdown, if our SD is still attached we'd be trying to
route an event through a non-existing device.
2007-11-16 11:20:22 +10:30
9de1ebe2a8
dix: Fix up class restoring when last SD disconnects.
...
Old code was fundamentally broken, fixes now are:
- free the MDs current device classes
- copy the device classes instead of flipping the pointers
- check for the old MD, not the new one.
2007-11-16 10:45:28 +10:30
20fd478324
Small static cleanups on dix/
2007-11-15 17:01:33 -05:00
70e50fa51f
Allocate RRCrtcRecs with calloc.
2007-11-15 17:00:37 -05:00
8d0cd1cd2c
Fix a really dumb typo.
2007-11-15 12:21:25 -08:00
f797c96845
Save pixmap allocation hints into the PixmapRec.
2007-11-15 12:21:25 -08:00
6bc50de021
Simplify system resource range setup.
...
osRes only existed to get copied into Acc. Waste of effort.
2007-11-15 15:03:58 -05:00
e1ff14a924
Delete some dead code in X -configure.
2007-11-15 14:45:49 -05:00
01cfba7522
Nuke the debugging allocator.
...
This has never been hooked up in the modular build, and can not possibly
have built since before 6.7. Clearly no one's using it.
2007-11-15 14:27:03 -05:00
c67b9c5fc3
Clean up some garbage in driver enumeration.
2007-11-15 14:24:36 -05:00
0706e5e790
Eliminate some redundancy in autoconfiguration.
...
We already synthesize Monitor and Module sections for you, no need to
specify them explicitly in the fake config buffer.
2007-11-15 14:15:09 -05:00
2c01a49bf0
Don't sleep(1) at exit on any platform.
2007-11-15 13:32:59 -05:00
83926cb8be
Xi: Only alloc modifierKeyMap if <slave device>->maxKeysPerModifier > 0
...
Sometimes (e.g. on my debian ppc box) maxKeysPerModifier of the SD is 0. So we
try to malloc(0), bringing the whole server down with a FatalError because it
looks as if the malloc failed. This is bad, so only alloc if we actually have
something to alloc.
2007-11-15 17:41:07 +10:30
070195dbf8
Xi: fix up sloppy class copying causing segfaults.
2007-11-15 16:27:18 +10:30
53539688ca
dix: SetKeySymMap should alloc the map if dst doesn't have one already.
2007-11-15 16:23:48 +10:30
b40646dc10
dix: Add FreeDeviceClass and FreeFeedbackClass for centralised xfree.
...
Ensures that we only have one way of freeing a device class to avoid leaks in
ChangeMasterDeviceClasses and other places.
2007-11-15 15:43:44 +10:30
4c9cc82fc4
For some reason "-nozap" appeared twice. Weird.
2007-11-15 01:46:11 -02:00
18833d648f
Xi: reset xkb-stuff to NULL after copying (DeepCopyDeviceClasses)
...
Having two devices point to the same xkb stuff causes SIGABRTs.
Also, don't init a MD's xkbInfo unless the SD has an xkbInfo.
2007-11-15 12:13:59 +10:30
1635832c16
Revert "xkb: disable xkb key repeats (temporarily)"
...
This reverts commit 2b1d946392
2007-11-15 11:35:07 +10:30
b05246696d
Xi: free XkbSrvLedInfos as well when freeing an MD's device classes.
2007-11-15 11:31:43 +10:30
c758e5a664
dix: Make sure core MDs copy their initial classes before they are used.
...
Anything in dev->key, dev->valuator etc. of a MD must always be a copy of the
original class. The intial classes of an MD (the ones set up before an SD is
attached) as well, as we may have to restore them if no SD is attached
anymore.
2007-11-15 10:47:20 +10:30
a08665d4d3
mi: don't call SwitchCoreKeyboard, we switch during event processing now.
2007-11-15 10:42:29 +10:30
64711a0948
Xi: When switching MD classes, make a deep copy instead of pointer flip.
...
Turns out it's really really hard synchronising device state across multiple
duplicated events if they all share the same struct. So instead of doing so,
when the SD changes deep-copy all it's classes into the MD. The MD then has
the same capabilities, but the state can be set separately. This should fix
xkb, key state, repeat etc. problems.
Updating the device state allows us to remove the SwitchCoreKeyboard from the
event gathering, it's all done during event processing now.
2007-11-15 10:41:34 +10:30
3dde66f96b
Start 1.4.99
2007-11-14 15:10:59 -05:00
c3897ca709
Add -pogo option for init/teardown performance testing.
2007-11-14 14:46:41 -05:00
cecac79445
Don't sleep(1) at server exit.
2007-11-14 14:46:22 -05:00
1603130236
Merge branch 'master' into XACE-SELINUX
...
Conflicts:
Xext/xace.c
Xext/xace.h
2007-11-14 13:35:50 -05:00
f207e69d62
xselinux: adjust receive hook to use new synthetic_event class.
2007-11-14 12:23:29 -05:00
cc4586df60
Xi: toggle the public.devPrivate as well when switching device classes.
...
The master needs to have the same devPrivate as the slave, in case a client
issues a request that goes down to the driver.
Example: if a driver wants to ring the keyboard bell, it'll pick a keyboard.
The KeyClassPtr will direct it to the matching method in the driver, but
because the MD doesn't have the devPrivate set the driver segfaults.
Even if all drivers were updated to not dereference the nullpointer, nothing
would actually ever happen.
To avoid this, we flip the master's public.devPrivate to the last SDs
devPrivate.
2007-11-14 14:27:52 +10:30
b44c1118f3
dix: Return Success from ProcBell if no BellProc is set.
...
We must not return BadDevice, this causes applications to crash. If no
BellProc is set, just quietly do nothing and report a Success.
2007-11-14 14:08:21 +10:30
240b10fa9c
Xi: Move updating the device state from POE into separate function.
...
POE now only deals with processing the event and calling the appropriate
delivery methods. Actually modifying the device state is done in
UpdateDeviceState. This separation should make it easier to avoid setting the
state twice when master events are processed.
2007-11-14 13:38:56 +10:30
0bdfeb55fc
Xi: remove RegisterPairingClient handling. Deprecated with device hierarchy.
2007-11-13 17:31:20 +10:30
e96d926d64
Xi: remove trailing whitespaces.
2007-11-13 17:14:35 +10:30
5031238aad
dix: remove trailing/wrong whitespaces from devices.c and events.c
2007-11-13 16:58:23 +10:30
Eamon Walsh
Michel Dänzer
Adam Jackson
Tormod Volden
Jernej Azarija
Peter Hutterer
Daniel Stone
Ben Skeggs
Adam Jackson
Aaron Plattner
Tiago Vignatti