While VT-switched, FB access is disabled and should remain so. Trying to switch
modes in that state would re-enable it, potentially causing crashes if trying
to access it before the driver has recovered from the mode switch.
Signed-off-by: Pierre-Loup A. Griffais <pgriffais@nvidia.com>
Reviewed-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
numScreens is always being assigned to 0 in dix for any server generation.
Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
This replaces a globally-allocated array that depended on MAXSCREENS.
Signed-off-by: Jamey Sharp <jamey@minilop.net>
Acked-by: Tiago Vignatti <tiago.vignatti@nokia.com>
This patch was created with:
git ls-files '*.[ch]' | while read f; do unifdef -B -DRENDER -o $f $f; done
Signed-off-by: Keith Packard <keithp@keithp.com>
High resolution devices was generating integer overflow.
For instance the wacom Cintiq 21UX has an axis value up to
87000. Thus the term (dSx * (Cx - Rxlow)) is greater than
MAX_INT32.
Using 64bits integer avoids such problem.
Signed-off-by: Philippe Ribet <ribet@cena.fr>
Signed-off-by: Benjamin Tissoires <tissoire@cena.fr>
Reviewed-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Commit 8736d112af changed the priority
ordering of the InputClass option merging to be "last match wins". This
fixes the handling of Option "Ignore" to follow that logic.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Reviewed-by: Keith Packard <keithp@keithp.com>
In addition to the conf files found in /etc/X11 or $sysconfdir/X11 used
for local administration, we also reserve a system directory for vendor
and package usage. The simple search path is:
/usr/share/X11/xorg.conf.d
$datadir/X11/xorg.conf.d
Files from these directories will have the lowest config priority. The
directory $datadir/X11/xorg.conf.d is exported from xorg-server.pc in
the variable "sysconfigdir". Packages should install their .conf files
to the directory specified by:
`pkg-config --variable=sysconfigdir xorg-server`
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
There's no reason to carry all the oddities from xorg.conf like appended
hostname to the search path for xorg.conf.d. This changes it to something
very simple:
/etc/X11/<cmdline>
$sysconfdir/X11/<cmdline>
/etc/X11/xorg.conf.d
$sysconfdir/X11/xorg.conf.d
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
vga arbiter will be locked in one device while AbortDDX will call LeaveVT
routines from the other device. Fail!
Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com>
Reviewed-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
When doing driver autoconfiguration with some parts of the config file
present but no driver set (e.g. only input configuration) fix the case
that we may have multiple drivers to try.
Create a screen section for each driver and let them be tried in a row.
Signed-off-by: Ruediger Oertel <ro@suse.de>
Reviewed-by: Dan Nicholson <dbn.lists@gmail.com>
Tested-by: Timo Aaltonen <timo.aaltonen@aalto.fi>
Signed-off-by: Keith Packard <keithp@keithp.com>
A driver that is assigned by an input class is only present as idev->driver.
The driver itself has no access to this information once PreInit is called.
For devices that rely on chain-hotplugging (wacom), this means that for the
second device the driver information is lost and the second device cannot be
initialized through NewInputDeviceRequest. Although this could be worked
around by hardcoding the driver name in the wacom driver, having the
assigned driver in the options seems like the better solution.
This issue only manifests itself with the udev backend. With HAL, the driver
is assigned by HAL and the option is duplicated in config/hal.c.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Dan Nicholson <dbn.lists@gmail.com>
buffer_alloc: Called allocating function "realloc" which allocated memory dictated by parameter "len + strlen(displaySize_string)"
alloc_strlen: Allocated memory does not have space for the terminating NUL of the string
var_assign: Assigned "ptr->mon_comment" to storage allocated by "realloc(ptr->mon_comment, len + strlen(displaySize_string))"
Signed-off-by: Oliver McFadden <oliver.mcfadden@nokia.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
xf86Config.c: In function 'configInputDevices':
xf86Config.c:1514: error: request for member 'lay_identifier' in something
not a structure or union
make[5]: *** [xf86Config.lo] Error 1
Introduced with e1165632bd.
X.Org Bug 26971 <http://bugs.freedesktop.org/show_bug.cgi?id=26971>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Acked-by: Dan Nicholson <dbn.lists@gmail.com>
In the vast majority of cases there is no xorg.conf that specifies a core
pointer/keyboard. Skip this warning, since we'll get another notification
about how the server relies on the config backend for input devices anyway.
Leave the warning in for the error case (AEI off).
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Fernando Carrijo <fcarrijo@yahoo.com.br>
Reviewed-by: Dan Nicholson <dbn.lists@gmail.com>
When the compat output is missing (I don't think this is actually
possible), or is disabled (and hence has no crtc), we would like to
avoid dereferencing NULL pointers. This patch creates inline functions
to extract the current compat output, crtc or associated RandR crtc
structure, carefully checking for NULL pointers everywhere.
Reviewed-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
The problem fixed by this patch can be reproduced on Linux with the
following steps.
- Access NULL pointer intentionally in ProcessOtherEvent on key press.
- Instead of saving core dump to a file, write it into a pipe.
echo "|/usr/sbin/my-core-dumper" > /proc/sys/kernel/core_pattern
- Dump the core by pressing a key.
While the core is being dumped into the pipe, the smart schedule timer
will cause a pending SIGALRM. Linux kernel stops writing data to the
pipe when there are pending signals. This causes the core dump to be
truncated. On my system I'm expecting a 6 MB dump but the size will be
60 kB instead. The problem is solved if we block the SIGALRM caused by
expired smart schedule timer.
I haven't been able to reproduce this problem in the following cases.
- Save core dump to a file instead of a pipe.
- kill -SEGV `pidof Xorg`
- Press a key to dump core while gdb is attached to Xorg.
- Give option -dumbSched to Xorg.
Also note that the fix works only when NoTrapSignals has the default
value FALSE. The problem can still be reproduced if error signals
aren't trapped. In addition to pending SIGALRM, there is a similar
problem with pending SIGIO from the keyboard driver during core dump.
Signed-off-by: Rami Ylimaki <ext-rami.ylimaki@nokia.com>
Reviewed-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
Currently the config and InputClasses are merged together so that the
options from the config backend have the highest priority. This is bad
since it means options such as a default XKB layout set by the backend
cannot be changed by the user.
This patch changes order of precedence to be:
1. xorg.conf
2. xorg.conf.d (later files have higher priority)
3. config backend
In order to allow this ordering, the config parsing has been changed to
read the xorg.conf.d files before xorg.conf. This has the consequence
that the core device picking which looks for the first InputDevice may
not find it in xorg.conf.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
The only DDX currently using hotplugging is the xfree86 one and it looks
like it'll stay that way for a bit. Move the initialization to the DDX,
since Xephyr, Xnest, and friends don't need HAL or udev notifications.
Add CloseInput (counterpart to InitInput) to be able to clean up the config
initialization from the DDX as well.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Julien Cristau <jcristau@debian.org>
Reviewed-by: Dan Nicholson <dbn.lists@gmail.com>
The qxl driver is for the QXL virtualized graphics device.
Signed-off-by: Søren Sandmann Pedersen <ssp@redhat.com>
Reviewed-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
Tags may be a list of comma-separated strings that match against a MatchTag
InputClass section. If any of the tags specified for a device match against
the MatchTag of the section, this match is evaluated true and passed on to
the next match condition.
Tags are specified as "input.tags" (hal) or "ID_INPUT.tags" (udev), the
value of the tags is case-sensitive and require an exact match (not a
substring match).
i.e. "quirk" will not match "QUIRK", "need_quirk" or "quirk_needed".
Example configuration:
udev:
ENV{ID_INPUT.tags}="foo,bar"
hal:
<merge key="input.tags" type="string">foo,bar</merge>
xorg.conf:
Section "InputClass"
Identifier "foobar quirks"
MatchTag "foo|foobar"
Option "Foobar" "on"
EndSection
Where the xorg.conf section matches against any device with the tag "foo"
or tag "foobar" set.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Tested-by: Dan Nicholson <dbn.lists@gmail.com>
Reviewed-by: Dan Nicholson <dbn.lists@gmail.com>
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
In order to keep the number of InputClass sections manageable, allow
matches to contain multiple arguments. The arguments will be separated
by the '|' character. This allows a policy to apply to multiple types of
devices. For example:
Section "InputClass"
Identifier "Inverted Mice"
MatchProduct "Crazy Mouse|Silly Mouse"
Option "InvertX" "yes"
EndSection
This applies to the MatchProduct, MatchVendor and MatchDevicePath
entries. Currently there is no way to escape characters, so names or
patterns cannot contain '|'.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Sometimes it is desirable to skip adding specific input devices to the
server. The "Ignore" option is used similarly to Monitor sections so
that matched devices will not be added. BadIDChoice is returned to the
config backend so that it will clean up all resources.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Drivers and options specified in InputClass sections work on a "first
match wins" strategy. Let's be consistent when documenting it.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
This is RAC's remnant. Any sane person would use a more wise method of
debugging instead.
X.Org Bug 26074 <http://bugs.freedesktop.org/show_bug.cgi?id=26074>
Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com>
Reviewed-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
Fixes crash when xscreensaver tries to use GammaRamp calls to fade out
http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6915712
Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com>
Reviewed-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
Adds new function xf86Activate to the OS-specific *VTsw*.c files
and calls it from xf86ProcessActionEvent
Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com>
Tested-by: Tiago Vignatti <tiago.vignatti@nokia.com> (GNU/Linux)
Fix for OpenSolaris bug 6876992: "[vconsole] Ctrl+Alt+F12 switchs to blank
console screen with hotkeys property turned-off"
http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6876992
Xorg needs to do sanity test for the VT it is commanded to switch to.
If the VT is not opened by any process, discard the switching request.
The changes also contain the fix for some flaws discovered when
getting the new gdm to run.
Signed-off-by: Aaron Zang <Aaron.Zang@Sun.COM>
Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com>
xf86Xinput.c relied on xkbsrv.h's definition of True/False which seems odd
at first and weird on second glance.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Dan Nicholson <dbn.lists@gmail.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
With InputClass support, it makes more sense to cover all
aspects of acceleration in options. Previously, one could only set the
default on the command line.
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>
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>
ba2d39dd54 introduced warnings:
xf86Mode.c: In function ‘xf86CheckModeForDriver’:
xf86Mode.c:986: warning: passing argument 1 of ‘modeInClockRange’ from incompatible pointer type
xf86Mode.c:253: note: expected ‘ClockRangePtr’ but argument is of type ‘ClockRangesPtr’
xf86Mode.c:1002: warning: passing argument 1 of ‘modeInClockRange’ from incompatible pointer type
xf86Mode.c:253: note: expected ‘ClockRangePtr’ but argument is of type ‘ClockRangesPtr’
Because I foolishly didn't notice that we had types with nearly
identical members named ClockRange and ClockRanges. The latter
contained an extra 'strategy' member at the end, which claimed to be
needed by the vidmode extension. Of course, this was a lie: the only time
we'd use it was in mode validation, for drivers using LOOKUP_CLKDIV2 with
non-programmable clocks. The only driver using LOOKUP_CLKDIV2 is
rendition, which has a programmable clock. The only driver using the
ClockRanges type was smi, which did not use it for its 'strategy' member,
so has been fixed to use ClockRange instead.
Signed-off-by: Adam Jackson <ajax@redhat.com>
Reviewed-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
Add a backend using libudev for input hotplug, and disable the hal and
dbus backends if this one is enabled.
XKB configuration happens using xkb{rules,model,layout,variant,options}
properties (case-insensitive) on the device. We fill in InputAttributes
to allow configuration through InputClass in Xorg.
Requires udev 148 for the input_id helper and ID_INPUT* properties.
Signed-off-by: Julien Cristau <jcristau@debian.org>
Acked-by: Dan Nicholson <dbn.lists@gmail.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
While the identifier is likely set before the input classes are merged, the
driver may not be. Hence don't check for a driver before we've completed
configuration for this device.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Dan Nicholson <dbn.lists@gmail.com>
Signed-off-by: Julien Cristau <jcristau@debian.org>
Currently Xorg uses hal's fdi files to decide what configuration options
are applied to automatically added input devices. This is sub-optimal
since it requires users to use a new and different configuration store
than xorg.conf.
The InputClass section attempts to provide a system similar to hal where
configuration can be applied to all devices with certain attributes. For
now, devices can be matched to:
* A substring of the product name via a MatchProduct entry
* A substring of the vendir name via a MatchVendor entry
* A pathname pattern of the device file via a MatchDevicePath entry
* A device type via boolean entries for MatchIsKeyboard, MatchIsPointer,
MatchIsJoystick, MatchIsTablet, MatchIsTouchpad and MatchIsTouchscreen
See the INPUTCLASS section in xorg.conf(5) for more details.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
In order to give NewInputDeviceRequest more information, a new
InputAttributes type is introduced. Currently, this collects the product
and vendor name, device path, and sets booleans for attributes such as
having keys and/or a pointer. Only the HAL backend fills in the
attributes, though.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Reviewed-by: Peter Hutterer <peter.hutterer at who-t.net>
Refactored code into the parser to allow the freeform boolean types used
in Option entries to be used in other configuration entries. This isn't
as powerful as allowing "No" to precede the option names, but it atleast
gives a common handling of "yes", "no", etc.
A type xf86TriState has been added to support an optional boolean. This
allows the boolean sense of the value to be kept while providing a means
to signal that it is unset.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Reviewed-by: Peter Hutterer <peter.hutterer at who-t.net>
Any input device with this option will be automatically added to whichever
server layout is selected at startup. This removes the need to reference a
device from the ServerLayout section. The two following configuration are
identical:
CONFIG 1:
Section "ServerLayout"
InputDevice "foo"
EndSection
Section "InputDevice"
Identifier "foo"
...
EndSection
CONFIG 2:
Section "InputDevice"
Identifier "foo"
Option "AutoServerLayout" "on"
...
EndSection
The selection of the server layout affects both explicitly specified
layouts and the implicit layout.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Keith Packard <keithp at keithp.com>
Add a new command line parameter, -configdir, to specify the config
directory to be used. Rules are the same as -config for root vs. user
privileges.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Reviewed-by: Peter Hutterer <peter.hutterer at who-t.net>
Currently there is a single file, xorg.conf, for configuring the server.
This works fine most of the time, but it becomes a problem when packages
or system services need to adjust the configuration. Instead, allow
multiple configuration files to live in a directory. Typically this will
be /etc/X11/xorg.conf.d.
Files with a suffix of .conf will be read and added to the server
configuration after xorg.conf. The server won't fall back to using the
auto configuration unless there is no config file and there are no files
in the config directory.
Right now this uses a simpler search template than the config file
search path by not using the command line or environment variable
parameters. The matching code was refactored a bit to make this more
coherent. Any DDX wanting to read the config files will need to call
xf86initConfigFiles before opening/reading them. This is to allow
xf86openConfigFile without xf86openConfigDirFiles and vice-versa.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Reviewed-by: Peter Hutterer <peter.hutterer at who-t.net>
I don't think this one has been in use since 2003.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
Convert all calls of CreateNewResourceType to pass name argument
Breaks DIX ABI.
ABI versions bumped:
Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com>
Reviewed-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
Default remains the same - on for most OS'es on i386 (except Solaris),
off for everyone else. Can be manually toggled via --enable-pc98 or
--disable-pc98.
Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
Most of the Xv Put/Get operations have an off by one error in the
viewport clipping.
Apparently PutImage was fixed at some point but the same code was
already copy-pasted all over the place, and so the other operations
still suffer from the bug.
Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com>
Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
Previously it was trying to set the same value as the default one. Sigh.
Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com>
Acked-by: Rami Ylimaki <ext-rami.ylimaki@nokia.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
Should probably also be applied to stabler xserver branches too.
Luc Verhaegen.
From a22bc20721bad506d8fa9772b1258568cbffe7d2 Mon Sep 17 00:00:00 2001
From: Luc Verhaegen <libv@skynet.be>
Date: Wed, 11 Nov 2009 15:52:39 +0100
Subject: [PATCH] Xv: Fix AdjustFrame when driver implements ReputImage.
Finally fixes fd.o #4653, filed more than 4 years ago.
Patch can be happily applied to all modular Xorg versions.
Signed-off-by: Luc Verhaegen <libv@skynet.be>
Reviewed-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
Xorg creates its log file following the umask of the user running
startx, which may result in a world-writable log. Set umask to 022 to
prevent this.
Debian bug#555308 <http://bugs.debian.org/555308>
See also http://thread.gmane.org/gmane.comp.security.oss.general/2299
Signed-off-by: Julien Cristau <jcristau@debian.org>
Reviewed-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
Let's let glibc do the right thing for dense/sparse selection.
The _alpha_iobase code has been unused since the switch to libpciaccess. It
really should have been killed by fba700f1f6.
Signed-off-by: Matt Turner <mattst88@gmail.com>
Tested-by: Michael Cree <mcree@orcon.net.nz>
Signed-off-by: Keith Packard <keithp@keithp.com>
Using common defaults will reduce errors and maintenance.
Only the very small or inexistent custom section need periodic maintenance
when the structure of the component changes. Do not edit defaults.
Reviewed-By: Jeremy Huddleston <jeremyhu@apple.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
The event fd may be invalidated by the pInfo->read_input call. If it is
invalidated, the subsequent FD_CLR call will segfault. Thus, the FD_CLR
call must precede the pInfo->read_input call.
Signed-off-by: Chase Douglas <chasedouglas@gmail.com>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Keith Packard <keithp@keithp.com>
Existing video drivers will get the console enabled by default.
Signed-off-by: Jamey Sharp <jamey@minilop.net>
Signed-off-by: Keith Packard <keithp@keithp.com>
This was initially fixed by commit 3932a84857
but then (presumably not intentionally) undone by commit
1d54479cb3 .
Signed-off-by: Hans Nieser <hnsr@xs4all.nl>
Signed-off-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Marcin Baczyński <marbacz@gmail.com>
Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
* mattst88/master:
[alpha] assume we have __NR_pciconfig_iobase
[alpha] don't return from void functions
Fix undefined symbols on alpha
Fix breakage on alpha caused by c7680befe5
Revert "alpha: kill xf86SlowBCopyToBus and xf86SlowBCopyFromBus"
When testing if an fd is valid, the required construct is >= 0, not > 0.
[Daniel: Fixed up the Linux MTRR case as well.]
Signed-off-by: Martin Ettl <ettl.martin@gmx.de>
Signed-off-by: Daniel Stone <daniel@fooishbar.org>
The vesa driver still uses slowbcopy_frombus and slowbcopy_tobus.
This reverts commit 5ef53a94ce.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
All input drivers use xf86PostKeyEventP indirectly now and have been since
it exists. I guess that qualifies it as tested - no need to spam the logs.
Reported-by: Felix Wenk
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Daniel Stone <daniel@fooishbar.org>
Removing DGA ended up breaking any drivers calling into the old
xf86DiDGAInit function as it tried to see if DGA was already enabled
and ended up crashing if the VT wasn't completely initialized. Oops.
Also, if the driver initializes DGA itself, have the DiDGA
initialization overwrite that information as the DiDGA code will call
ReInit on mode detect.
Signed-off-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
The DIX event queue is allocated before InitInput is called, so fetch
the pointer there and not randomly at other times. This avoids failing
to fetch the pointer sometimes during server regen and then smashing
memory through the stale pointer from the previous server generation.
Signed-off-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
xf86MatchDevice returned malloc'd storage containing the list of
devices to look at; make sure that gets freed.
Signed-off-by: Keith Packard <keithp@keithp.com>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Boolean option to enable/disable SIGIO handlers is set by the first
of these found:
- UseSIGIO option is set in xorg.conf ServerFlags
- Default set at build time by ./configure --enable-use-sigio-by-default
- Platform default value: Solaris = no, all others = yes
This matches the current settings on all platforms except Solaris.
This reverts Solaris (for now) to the settings used in Xorg 1.6, before
SIGIO support for Solaris was added, due to some system level bugs that
won't be resolved in time for Xorg 1.7 release, but allows us to enable
when those are resolved (or when we need to test if they're resolved).
See http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6879897
Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
So far there are no apparently issues on not closing the fd. But let's do the
right job here.
Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com>
Acked-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
In practice, some of the native drivers for older Geode products
have become deprecated due to lack of e.g. libpciaccess upgrade,
but that's OK, since most distributions don't ship them anymore.
In that case, we'll let X server fall back to good old VESA.
xf86SlowBCopyToBus and xf86SlowBCopyFromBus cause segfaults on my
system.
Also remove associated slowbcopy_tobus/slowbcopy_frombus macros.
Signed-off-by: Matt Turner <mattst88@gmail.com>
All architectures should be able to use the same unaligned access code,
regardless of whether they need special unaligned access instructions.
Let's let gcc do the heavy lifting.
In the case that we're not using a gcc-compatible compiler, use memmove.
The xserver already requires pixman, so include pixman.h for its uint*_t
types.
Signed-off-by: Matt Turner <mattst88@gmail.com>
Checks for __GNUC__ are superfluous since the only other compiler for
the platform is Compaq C, and it doesn't support GCC style inline
assembly.
Signed-off-by: Matt Turner <mattst88@gmail.com>
This adds support for using the libpciaccess interface for
vga arbitration support on top of a kernel which supports it.
Currently patches are queued for kernel 2.6.32 in jbarnes
pci tree, and shipping in Fedora kernel.
Co-authors:
Tiago Vignatti <tiago.vignatti@nokia.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This function was used as the default motion event queue API until
including XINPUT_ABI 2 (server 1.5).
This API was broken with 1883485 in May 2008 (wrong casting of parameters)
and isn't in use by input drivers past ABI 3.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com>
No one is using bus notifications now. We hope that the kernel take care of
this properly.
For other not-so-urgent-notifications (ACPI wakeups, etc) we can just register
a handler on server's scheduler (using xf86AddGeneralHandler). And for
external applications, the "trend" is to use HAL to kick notifications. So
we're already provided of enough notification schemes.
Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com>
Signed-off-by: Aaron Plattner <aplattner@nvidia.com>
commit 48ee555833 (OpenSolaris VT support)
broke the autoconfiguration code in xf86AutoConfig.c that uses the
Solaris-specific VIS_GETIDENTIFIER ioctl on a frame buffer device like
/dev/fb by changing xf86Info.consoleFd from /dev/fb to a /dev/vt/*
device.
This fixes it by reworking the code to split the console device
(/dev/vt/*, the vtXX CLI option) from the frame buffer device
(/dev/fb, -dev option) to allow both VT and autoconfig to work.
It also fixes the console device to use /dev/fb when VT's are not
supported instead of throwing a Fatal Error because it can't open
/dev/vt/0.
Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com>
The xorg.conf generator was not assigning correctly the primary device
("bootable") as screen zero. So just skip this kind of routines for now.
Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com>
If you want to run a pre-1999 kernel, you'll need a pre-2009 X server
[Some pre-Solaris 8 VT support is left by this patch to allow reuse by
the new Solaris VT support that follows in the next patch.]
Signed-off-by: Aaron Zang <Aaron.Zang@Sun.COM>
Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com>
xf86PostKeyboardEvent also makes use of xf86PostKeyEventP to avoid code
duplication, and the valuator verification has been split into the
XI_VERIFY_VALUATORS macro.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
InternalEvents shouldn't be used anywhere outside the X server itself. Split
up into events.h for opaque typedefs for the events needed by various
headers and eventstr.h for the actual struct definitions.
eventstr.h must only be included by code that requires internal events and
is not part of the SDK.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Inside a windowing system, it's not the place to probe for devices. Goodbye
-probe and -probeonly.
Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com>
Such stupid and ugly way to dump PCI information! Oh boy... Anyway, this
doesn't belong to the X server at all. Go away!
Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com>
This changes the ABI, but since the video ABI is at 6 already
it should be fine.
driver changes are in the pipeline after this.
Signed-off-by: Dave Airlie <airlied@redhat.com>
Fixes screensaver fadeout effects.
Also make all RandR 1.2 compatibility code for XF86VidMode operate only on the
CRTC associated with the compatibility output, not all CRTCs at once.
This was a vestige from the days before we'd make the mode list from the
EDID data, and from CRT technology when you could reasonably assume that
higher refresh rates were better. Also it did not function as advertised,
acting as a high-pass filter instead of a band-pass.
xextproto had Xlib client headers moved into libXext.
Protocol header files are named fooproto.h, header files with constants
foo.h or fooconst.h where foo.h was already in use for client-side headers.
These old interfaces are no longer supported by the server, removing them
requires bumping the video driver ABI. Note that this is not guaranteed to
be the last change in ABI version 6.
Signed-off-by: Keith Packard <keithp@keithp.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Otherwise no subsequent driver will be able to claim this pci slot.
Example for this: fbdev tries to claim, but framebuffer device is not
available. Later on VESA cannot claim the device.
GIT change
http://cgit.freedesktop.org/xorg/xserver/commit/?id=45c8bd0fe54273039fdaa1eeeafb81b5774f2c75
changed the default symbol visibility of the Xserver. As a result 2 symbols
that are needed by the RandR 1.2/1.3 implementation in the fglrx driver are no
longer visible:
xf86configptr
xf86CursorScreenKey
We would like to get these two symbols _X_EXPORT'ed before Xserver 1.7 is
released. Otherwise it will be problematic for fglrx to support RandR 1.3 on
Xserver 1.7.
In the future, we may want to sync our RandR implementation to later versions
of the RandR implementation in hw/xfree86/modes. Therefore it would be nice if
all symbols used by the Xserver RandR implementation were _X_EXPORT'ed in the
future.
Previously when compiling on freebsd amd64 we'd end up at xi86
block (line 1315) which would define mem_barrier and write_mem_barrier
to be NOP's. Instead they should be valid, as per the linux amd64 setup.
This stops the hangs experienced by many when using the nv driver
which would hang due to out of order dma requests as noticed in
http://bugs.freedesktop.org/show_bug.cgi?id=3168
Signed-off-by: Benjamin Close <Benjamin.Close@clearchain.com>
isMaster is not enough as long as we differ between master pointers and
keyboard. With flexible device classes, the usual checks for whether a
master device is a pointer (currently check for ->button, ->valuators or
->key) do not work as an SD may post an event through a master and mess this
check up.
Example, a device with valuators but no buttons would remove the button
class from the VCP and thus result in the
IsPointerDevice(inputInfo.pointer) == FALSE.
This will become worse in the future when new device classes are introduced
that aren't provided in the current system (e.g. a switch class).
This patch replaces isMaster with "type", one of SLAVE, MASTER_POINTER and
MASTER_KEYBOARD. All checks for dev->isMaster are replaced with an
IsMaster(dev).
Historically, if no input device was referenced in the ServerLayout,
the server would pick the first "mouse" device found in the xorg.conf.
This patch gives evdev, synaptics, vmmouse and void the same status. If
there is a section in the config file using this driver - use it as the core
pointer.
Device selection is in driver-order, not in config-order. If a "mouse"
device is listed after a "synaptics" device, the "mouse" device gets
preference. This replicates the original behaviour.
This code only takes effect if AllowEmptyInput is off and there is no core
pointer in the server layout.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Historically, if no input device was referenced in the ServerLayout,
the server would pick the first "mouse" device found in the xorg.conf.
This patch gives evdev, synaptics, vmmouse and void the same status. If
there is a section in the config file using this driver - use it as the core
pointer.
Device selection is in driver-order, not in config-order. If a "mouse"
device is listed after a "synaptics" device, the "mouse" device gets
preference. This replicates the original behaviour.
This code only takes effect if AllowEmptyInput is off and there is no core
pointer in the server layout.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
This approach is broken anyway. DIPT only checked for the XInput type
"MOUSE" and the only user of this is xf86ActivateDevice when it sets the
Activate/DeactivateGrab functions.
Since synaptics and wacom set their own types, evdev only sets MOUSE for,
well, mice half the devices didn't have this set correctly anyway.
Instead, ActivatePointerGrab should be merged together with
ActivateKeyboardGrab.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
There's only two reasons for hierarchy events:
- device is added, removed, etc. In this case we want to send the event as
it happens.
- devices are added in a XIChangeDeviceHierarchy request. In this case we
only want one event cumulating all changes.
This way clients querying the gamma value via the VidMode extension at least
get the last value set via the same, rather than always something bogus.
Signed-off-by: Michel Dänzer <daenzer@vmware.com>
We only put internal events into the queue now, so let's check for ET_Motion
rather than the MotionNotify.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
We only put internal events into the queue now, so let's check for ET_Motion
rather than the MotionNotify.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Zapping is triggered by xkb these days, so note in the man page that it's the
Terminate_Server action. Since it's XKB, personal preferences towards or
against zapping should be achieved through xkb rulesets.
If Terminate_Server is not in the xkb actions, then we can't zap anyway and we
don't need a default of DontZap "on".
This patch restores the old meaning of DontZap - disallow zapping altogether,
regardless of XKB's current keymap.
Ideally, this patch should be accompanied by b0f64bdab00db652e in
xkeyboard-config.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
After the call to xf86ActivateDevice, the new device will be added to
inputInfo.devices. However, if the subsequent call to ActivateDevice
fails, the correponding InputInfoRec for the device is deleted but an
entry still remains in inputInfo.devices. This might lead to a server
crash later on (on InitAndStartDevices for instance) when the device
control proc would be called for an invalid device.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Replace multi-stage filtering with simple linear velocity,
tracked several instances backwards. A heuristic ensures
only approximately linear motion is considered, so velocity
remains valid in any case. Numerical stability is much
better, and nothing changes to people who didn't tune the
advanced features of the previous algorithm.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
remove a few lines which redo part of the pointer acceleration
init. Properties is the way to go for them.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Fix this bug report:
,----< from http://bugzilla.freedesktop.org/show_bug.cgi?id=20504 >
| Using the Visual StaticGray (8 bit depth) is missing one gray level.
| The gray level of index zero and index one are the same and all
| other levels are shifted by one. The max level (255) cannot be used.
`----
Signed-off-by: James Cloos <cloos@jhcloos.com>
DeleteInputDeviceRequest function doesn't handle "virtual" devices well.
TightVNC libvnc.so module to X (which makes bare Xorg VNC capable) uses such
kind of devices.
Bare Xvnc (it is something like Xvfb) simply uses AddInputDevice &
RegisterDevice functions. Xvnc uses DeleteInputDeviceRequest from Xi/stubs.c
so everything works fine (now I see that DeleteInputDeviceRequest in
Xi/stubs.c should call RemoveDevice function, shouldn't it? :) )
Situation is quite different when you use libvnc.so module. It uses same
schema as Xvnc, so it simply calls AddInputDevice & RegisterDevice. Thus
device is created correctly. When server is terminated it calls
DeleteInputDeviceRequest (now from hw/xfree86/common/xf86Xinput.c) for each
device. Here is the difference - Xvnc calls DeleteInputDeviceRequest from
Xi/stubs.c as I wrote above. Thus Xorg gets sigsegv because "VNC" devices
don't have real input driver.
X.Org Bug 20087 <http://bugs.freedesktop.org/show_bug.cgi?id=20087>
[This isn't really a fix (libVNC should behave correctly) but not crashing the
server sounds like an improvement.]
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
With the API change, we can now purge the XI conversion from POE.
Note: this commit breaks DGA even more.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Note that this breaks DGA. Life is tough.
EnqueueEvent is a somewhat half-baked solution, we immediately drop back into
XI and store them. But it should in theory work.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Don't let the dcce be random data.
Was only used to provide a list of input devices that XF86-Misc could use,
now that XF86-Misc is gone, was parsed and logged, then completely ignored.
(Depends on previous patch that introduces OBSOLETE_TOKEN in parser to
make obsolete keywords like InputDevices & RgbPath be non-fatal errors.)
Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com>
Acked-by: Adam Jackson <ajax@redhat.com>
This was all a glorified no-op. We rely on pciaccess to create device
maps anyway, so we should have no reason to care about what the host
address is.
Acked-by: Ian Romanick <ian.d.romanick at intel.com>
Signed-off-by: Adam Jackson <ajax@redhat.com>
By making the "Unable to open config file" header a warning, it was
not appearing with the filename when a config file was specified and
not found. Now we make it an error message again, but only issue
the error if a filename was specified - if none was specified, then
we don't even issue a warning, just the "Using autoconfig" info message.
Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com>
Instead of always keeping two copies of the keymap, only generate the
core keymap from the XKB keymap when we really need to, and use the XKB
keymap as the canonical keymap.
Signed-off-by: Daniel Stone <daniel@fooishbar.org>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
We already have state fully stored within XKB, so instead of duplicating it,
just generate the values to send to clients when required.
Signed-off-by: Daniel Stone <daniel@fooishbar.org>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
XkbInitKeyboardDeviceStruct is now the only valid keyboard
initialisation: all the details are hidden behind here. This now makes
it impossible to supply a core keymap at startup.
If dev->key is valid, dev->key->xkbInfo->desc is also valid.
Signed-off-by: Daniel Stone <daniel@fooishbar.org>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
No more #ifdef XKB, because you can't disable the build, and no more
noXkbExtension either.
Signed-off-by: Daniel Stone <daniel@fooishbar.org>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
If we can't enable a device, bail out of NewInputDeviceRequest rather than
blithely continuing. Also, be more verbose when initialization failed. Also,
be more verbose when initialization failed. Also, be more verbose when
initialization failed. Also, be more verbose when initialization failed.
Signed-off-by: Daniel Stone <daniel@fooishbar.org>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
There are several other direct calls to free, check
% egrep '\<free\(' `find . -name \*.c`
but they are free'ing memory from explicit malloc calls.
This one was not intended, and corrected, so that it would
both, follow the conventions everywhere (and work on some
libc that doesn't like free(0)), and make it easier to use
malloc wrappers.
The builtin-fonts configure option was removed, as it at best should
have been a runtime option. Instead, now it always register all "font
path element" backends, and adds built-ins fonts at the end of the
default font path.
This should be a more reasonable solution, to "correct" the most
common Xorg FAQ (could not open default font 'fixed'), and also don't
break by default applications that use only the standard/historical
X Font rendering.
Add a ClipNotify helper that lets the driver know about changes in the
clipping of an Xv backing drawable.
Signed-off-by: Daniel Stone <daniel@fooishbar.org>
Devices are only activated once - right after they've been added to the
server. If a device failes activation, it's dead. There's no reason to
continue. Return the error code from ActivateDevice() without setting up
sprite information or even sending a event to the client.
Then - in the DDX - just remove the device again.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
- The Gamma values from the monitor section are now used during initial config.
- The old colormap system is disabled when gamma set hook is available.
- Gamma values are now persistent for the lifetime of the xserver.
- This requires no driver changes and should be driver ABI compatible.
A grep on xorg/* revealed there's no consumer of this define.
Quote Alan Coopersmith:
"The consumer was in past versions of the headers now located
in proto/x11proto - for instance, in X11R6.0's xc/include/Xproto.h,
all the event definitions were only available if NEED_EVENTS were
defined, and all the reply definitions required NEED_REPLIES.
Looks like Xproto.h dropped them by X11R6.3, which didn't have
the #ifdef's anymore, so these are truly ancient now."
Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com>
Signed-off-by: Adam Jackson <ajax@redhat.com>
The awk script was incorrectly referencing the struct name, and
not the struct variable.
Also added some comments to sdksyms.sh, for the reason it generates
the "symbol table" and add a message to the generated file, telling
is was automatically generated.
The kbd driver may send events during device initialisation, and these events
need the EQ set up already.
X.Org Bug 18890 <http://bugs.freedesktop.org/show_bug.cgi?id=18890>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
As suggested by Julien Cristau
This is an follow-up to
commit 9c5dd7337f
Author: Peter Hutterer <peter.hutterer@redhat.com>
Date: Wed Dec 3 14:24:25 2008 +1000
Let the DDX decide on the XkbRulesDefaults.
Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com>
Acked-by: Julien Cristau <jcristau@debian.org>
Signed-off-by: Daniel Stone <daniel@fooishbar.org>
drv and idev are only set for SDs, but are only dereferenced for SDs too, so
initializing them to NULL is safe.
Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com>
It is declared as
<hash>ifdef HAVE_ACPI
Bool xf86acpiDisableFlag = FALSE;
<hash>endif
in hw/xfree86/common/xf86Globals.c
but not protected by the ifdef in the sdk header xf86Priv.h, what
caused a build failure in the tinderbox, due to the address of the
symbol being taken (to ensure it is available) in sdksyms.c.
All .a libraries were converted to .la, and instead of linking the
Xorg binary with a mix of .a and .la, and adding some libraries more
then once in the command line, etc, now it generates a single libxorg.la
from all the required convenience libraries, and links with a dummy
xorg.c (that should usually be the file with the main function...).
This removes the requirement of some things like libosandcommon and
libinit, that existed to circumvent problems when linking multiple
.a and .la in the final Xorg binary.
The "symbol table" is now generated dynamically, by a shell script,
with an embedded gawk parser that parses cpp output. The new file
sdksyms.sh is generated by hand by analyzing all Makefile.am's and
making it create a sdksyms.c file, that includes all sdk headers that
will add symbols for the Xorg binary. Module headers aren't read, and
a in 2 files it was required to add a "<hash>ifndef XorgLoader" around
declarations shared between the Xorg binary and libextmod. A few
other changes were added to other sdk headers, like preventing
multiple inclusion, or including other headers to satisfy dependencies.
This should be a lot more portable, and better (hopefully properly)
using libtool to generate convenience libraries.
Those tables were once used to decide what symbols are visible to
modules, but they were outdated. The only real usage was that, since
it was taking the address of symbols, linkage should fail if the
symbols were not available.
Now the proper way to make symbols available to modules should
be to use the _X_EXPORT macro, or not compile with hidden symbols,
so that all symbols would be available.
All symbols in the tables were revised to ensure they are exported,
and only symbols that were not exported are ClientSleepUntil() and
DuplicateModule(), that were not in the sdk for quite some time
already, and should not have any users outside of the X Server
(and/or builtin modules).