Commit Graph

1395 Commits

Author SHA1 Message Date
Adam Jackson
b290884719 Bug #3284: Make the DRI lock reference count per-screen. 2005-09-13 00:44:52 +00:00
Daniel Stone
846f797f37 Add missing fontsproto and fontenc to REQUIRED_MODULES. 2005-09-12 08:24:48 +00:00
Daniel Stone
5b218617fa Add missing resourceproto to REQUIRED_MODULES. 2005-09-12 07:53:16 +00:00
Daniel Stone
ade158d238 Define WITH_VGAHW, since we always build it; add to Xorg headers. 2005-09-12 07:07:59 +00:00
Daniel Stone
34dc481a16 Put DMX AC_CONDITIONALS at the top level, since they must always be called. 2005-09-12 07:02:46 +00:00
Eric Anholt
b0f6fe1f38 Add some optional (disabled) fallback debugging code to the async code, for
better tracking of when we're hitting software.
2005-09-11 21:38:41 +00:00
Eric Anholt
1c003ccf5d Add a pair of hooks, PrepareAccess() and FinishAccess(), which get called
around CPU access to the framebuffer. This allows the hardware to set
    up swappers to deal with endianness, or to tell EXA to move the pixmap
    out to framebuffer if insufficient swappers are available (note: must
    not fail on front buffer!).
Submitted by: benh
2005-09-11 19:08:10 +00:00
Adam Jackson
ca210830bd Simplify life for EXA drivers by reducing some {Con,Dis}joint ops. 2005-09-11 18:43:55 +00:00
Matthieu Herrb
04f81cacb9 OpenBSD also doesn't like defining POSIX_SOURCE: it limits available
functionnality in headers beyond what's used by Xorg.
2005-09-11 18:33:31 +00:00
Daniel Stone
089b4272cf use RGB_DB not RGB_PATH as that's what it's looking for
define HAVE_XKB_CONFIG_H so setting xkb-path works (Jürg Billeter)
2005-09-11 01:16:14 +00:00
Eric Anholt
2f9d01c968 Remove some references to mfb/cfb support that is no longer there. 2005-09-08 01:15:47 +00:00
Daniel Stone
b5d42012f9 Fix typo resulting in failure to swap between RGB/BGR properly. (Stephen P.
Becker)
2005-09-07 01:30:23 +00:00
Daniel Stone
c8de8c23fb Fix tests for maximum number of colours when creating a colourmap, so a
32-bit visual (e.g. ARGB) doesn't overflow an int when attempting to do
1 << 32. (Benjamin Herrenschmidt)
2005-09-05 07:43:51 +00:00
Daniel Stone
5c5c51fa6d Initialise private arrays with calloc, rather than standard malloc.
(Benjamin Herrenschmidt)
2005-09-05 07:40:50 +00:00
Daniel Stone
691669c012 xorg-server.h -> dix-config.h (thinko). 2005-09-03 07:08:58 +00:00
Alan Coopersmith
4b2f5ba1b5 Use macros from xtrans.m4, issue error if not found. 2005-09-03 03:27:55 +00:00
Jesse Barnes
733a6d7a26 Add EXA driver writer documentation for the benefit of future generations
of X hackers.
2005-09-03 02:21:36 +00:00
Daniel Stone
854010d71d Add -include dix-config.h to GL/mesa/X to catch _XSERVER64, et al. 2005-09-02 03:22:01 +00:00
Kristian Høgsberg
287336f3c9 Teach xkb how to optionally run xkbcomp from $bindir instead of
$datadir/xkb.
2005-09-01 19:56:14 +00:00
Daniel Stone
94fbdb5c6d Make RGB_PATH configurable.
Move I2C modules back to $(moduledir)/multimedia.
2005-09-01 14:56:35 +00:00
Kristian Høgsberg
a65c579613 Include xkb-config.h if we have it so we pick up the paths defined there. 2005-08-31 14:58:09 +00:00
Alan Coopersmith
9d3b5e89d6 Save keyboard LED state on startup and restore on exit so text console mode
LEDs match text console mode state. Move push of streams module earlier
    so it's loaded before we start using kbd ioctls provided by the streams
    module.
2005-08-30 22:34:14 +00:00
Kristian Høgsberg
fa6fbd018d Use the $(moduledir) makefile variable instead of @moduledir@ so it can be
overridden at make install time.
Remove driverdir and inputdir from pkg-config file. The directory layout of
    moduledir is fixed and well known by drivers.
2005-08-30 22:31:52 +00:00
Alan Hourihane
b4f4bf028a programs/Xserver/hw/xwin/winmultiwindowshape.c Fix off by one error (Colin
Harrison)
2005-08-30 19:51:59 +00:00
Adam Jackson
0e50af8b7f typo fix 2005-08-30 19:35:06 +00:00
Eric Anholt
0c74799af4 Remove existing broken maxX/maxY code for composite (results in infinite
loops, doesn't deal with failure, doesn't present the interface to
    drivers that I expected) and instead replace it with a simple fallback
    to software when coordinate limits could be violated. Act similarly in
    other acceleration cases as well.
The solution I want to see (and intend to do soon) is to (when necessary)
    create temporary pictures/pixmaps pointing towards the real ones' bits,
    with the offsets adjusted, then render from/to those using adjusted
    coordinates.
2005-08-30 04:41:04 +00:00
Eric Anholt
f20e845b04 More 0 -> NULL for pointers missed in previous commit to this file. 2005-08-30 03:42:07 +00:00
Eric Anholt
7777d325a3 Apply an xserver patch from cworth: Avoid buffer ovverrun when a
trapezoid's right edge is on a pixel boundary.
2005-08-30 03:05:21 +00:00
Eric Anholt
e321f9e7ff Apply the xserver patch from vektor for bug #4208: Use NULL for pointers
instead of 0.
2005-08-30 03:01:38 +00:00
Adam Jackson
20813d3af0 Bug #3974: Fix unaligned memory access on LP64. (Matthieu Herrb) 2005-08-28 19:47:39 +00:00
Adam Jackson
0926cf79c0 Add diagnostic messages for exaDriverInit failure cases. 2005-08-26 20:21:57 +00:00
Adam Jackson
5ffff7cb86 Bug #4160: Fix Altix kernel version check. 2005-08-26 20:08:09 +00:00
Daniel Stone
6b0cdc5dd9 Change use of dix-config.h to xorg-config.h. 2005-08-26 16:46:41 +00:00
Daniel Stone
9d1b349b37 Back out previous change until I figure out something smarter. 2005-08-26 16:34:55 +00:00
Daniel Stone
3075df24e7 Subvert SIGUSR2 to reload all input devices. (Ubuntu #020) 2005-08-26 07:35:55 +00:00
Daniel Stone
89c84575ea Undo rate/period change after function body, so the damage doesn't leak
into other functions. (Debian #050)
2005-08-26 07:24:21 +00:00
Daniel Stone
bb5e934df7 Only open /proc/bus/pci/devices once. (Ubuntu #029) 2005-08-26 07:15:04 +00:00
Daniel Stone
c937faadd0 Better error message on failure to set iopl. (Debian #021) 2005-08-26 06:48:24 +00:00
Daniel Stone
6d34a2ac8a Add Xv symbol from Xext. 2005-08-26 06:35:00 +00:00
Daniel Stone
b8f0d4c3eb Spit out an error when there is no valid FB device, instead of just failing
silently. (Debian #070)
2005-08-26 06:29:15 +00:00
Daniel Stone
b48a24e796 Don't assume that all sun4m CPUs support muldiv; the Cypress CPU, which
implements the 4m MMU but only v7 instructions, does not. (Debian
    #100).
2005-08-26 06:23:41 +00:00
Daniel Stone
fd158d3d52 Get prototype for ntohl from SuSv3-compliant location, fixing it for
systems which define ntohl as a macro only, not both macro and
    function. (Debian #076)
2005-08-26 05:49:44 +00:00
Daniel Stone
53e489c0e3 Work around ATI expansion ROM problem on IA64 caused by prototype HP
McKinley systems. (Bdale Garbee)
2005-08-26 05:47:36 +00:00
Daniel Stone
40374d1149 Accept 'Enabled' and 'Disabled' for Extensions section (Ubuntu #990) 2005-08-26 05:00:07 +00:00
Eric Anholt
ea80b5db25 Fix a use-after-free of cursor data by refcounting for the sprite.current
reference. The particular path seen was XFixes' ReplaceCursor()
    resulting in the sprite.current being freed, but then it getting
    accessed during the ChangeToCursor() that happens as a result of
    WindowHasNewCursor().
2005-08-25 22:11:04 +00:00
Søren Sandmann Pedersen
54cc45b09b Thu Aug 25 17:15:01 2005 Søren Sandmann <sandmann@redhat.com>
Add all the drivers to the module_LTLIBRARIES instead of having separate
    variables for them. Pointed out by Kristian Høgsberg.
2005-08-25 21:22:41 +00:00
Kristian Høgsberg
ff22adc097 Remove sun_inout.s so make dist works (discussed with Alan Coopersmith). 2005-08-25 14:19:43 +00:00
Alan Coopersmith
344a24b622 Use system curses library on Solaris for xorgcfg text mode. 2005-08-25 04:09:45 +00:00
Eric Anholt
0711502f18 Bugzilla #4226: Change the pixmap migration strategy for the CopyNtoN case.
Now, if either source or dest were in framebuffer, try to get both
    there, but prefer system memory for both otherwise. Required making
    exaasync.c go through the try-acceleration path. This significantly
    improves window resizing under composite, because previously the
    pattern of creating a new pixmap and copying default contents from the
    screen caused a fallback every time due to the new destination pixmap
    being in system memory.
2005-08-24 23:48:11 +00:00
Eric Anholt
2261710fe0 Fix a bug where NULL could be dereferenced during the pixmap kick-out
process by referencing the correct offscreen area. Also drive-by the
    comments related to these for clarity.
2005-08-24 23:38:25 +00:00