Keith Whitwell [Fri, 27 Nov 2009 13:59:00 +0000 (13:59 +0000)]
st/xorg: fix composite after texture size changes
José Fonseca [Wed, 11 Nov 2009 00:56:43 +0000 (16:56 -0800)]
wgl: Fix copy'n'paste typo in comment.
José Fonseca [Sun, 15 Nov 2009 20:14:03 +0000 (12:14 -0800)]
svga: Handle comment tokens when dumping.
José Fonseca [Tue, 24 Nov 2009 14:43:30 +0000 (14:43 +0000)]
svga: Use consistent names for public symbol names of shader dumping facilities.
José Fonseca [Tue, 24 Nov 2009 14:37:45 +0000 (14:37 +0000)]
svga: Use consistent file names for dumping facilities.
Keith Whitwell [Fri, 27 Nov 2009 12:24:42 +0000 (12:24 +0000)]
st/xorg: speculatively round textures up to nearest POT
I'm not sure if this is a great change, but helps with caching.
Probably we want to turn this on/off on a driver-by-driver basis.
Keith Whitwell [Fri, 27 Nov 2009 12:22:43 +0000 (12:22 +0000)]
st/xorg: free last fence
Keith Whitwell [Fri, 27 Nov 2009 12:19:28 +0000 (12:19 +0000)]
svga: flush our command buffer after the 8th distinct render target
This helps improve the surface cache behaviour in the face of the
large number of single-use render targets generated by EXA and the xorg
state tracker. Without this we can reference hundreds of individual
render targets from a command buffer, which leaves little scope for
sharing or reuse of those targets.
Flushing early means we can start reusing textures much sooner.
This shouldn't have much effect on normal 3d rendering as it's pretty
rare to have a command buffer with >8 different render targets in that
world.
Keith Whitwell [Fri, 27 Nov 2009 12:18:22 +0000 (12:18 +0000)]
svga: add DEBUG_CACHE option
Vinson Lee [Fri, 27 Nov 2009 04:02:49 +0000 (23:02 -0500)]
svga: Prevent potential null pointer dereference in vmw_surface.c.
Jakob Bornecrantz [Tue, 24 Nov 2009 21:54:00 +0000 (22:54 +0100)]
vmware/xorg: Remove gem include
Michel Dänzer [Thu, 26 Nov 2009 15:46:13 +0000 (16:46 +0100)]
st/xorg: Use pipe clear hook for solid fills of whole pixmaps.
Can give a little boost e.g. for anti-aliased text rendering.
Keith Whitwell [Thu, 26 Nov 2009 15:25:09 +0000 (15:25 +0000)]
svga: hash the whole key, not just the first four bytes
Michel Dänzer [Thu, 26 Nov 2009 15:00:06 +0000 (16:00 +0100)]
st/xorg: Make sure DRI2 blits use GPU copy contents even for software fallback.
Fixes 3D apps not updating with a non-GL compositing manager and Option
"2DAccel" "off".
Also clean up a little pixmap vs. drawable mess.
Keith Whitwell [Thu, 26 Nov 2009 14:23:24 +0000 (14:23 +0000)]
st/xorg: fix refcounting bugs introduced in earlier commit
Keith Whitwell [Thu, 26 Nov 2009 14:23:07 +0000 (14:23 +0000)]
st/xorg: formatting
This directory needs indent run over it.
Keith Whitwell [Thu, 26 Nov 2009 12:52:45 +0000 (12:52 +0000)]
st/xorg: remove debugging flush
Accidentally committed in
7fbdbad5c02e3d5bfbf0e641e2aec224e39fa974
('st/xorg: consolidate some dest surface state setting')
José Fonseca [Thu, 26 Nov 2009 12:02:14 +0000 (12:02 +0000)]
svga: Remove spurious argument to SVGA_DBG.
Keith Whitwell [Thu, 26 Nov 2009 11:55:47 +0000 (11:55 +0000)]
st/xorg: render throttling in block handler
Similar to the classic swapbuffer throttling in GL drivers, put an
upper bound on the number of outstanding chunks of rendering the state
tracker can generate -- where calling the block handler denotes a
chunk.
Currently that number is set at around 4 "chunks", but could be
tweaked up or down. If a better measure for the amount of outstanding
rendering is found, that would be fine too.
As it stands, this improves interactivity by preventing the X server
from queueing up arbitary amounts of rendering.
Keith Whitwell [Thu, 26 Nov 2009 11:17:06 +0000 (11:17 +0000)]
st/xorg: split copy operation into prepare/copy/done phases
Any high-overhead one-off tasks are moved into the prepare hook.
Keith Whitwell [Thu, 26 Nov 2009 10:40:40 +0000 (10:40 +0000)]
st/xorg: don't set up constant buffer for non-xv fragment shaders
These currently don't reference any constants. Can add this back
if newer shaders need them, but in the meantime don't create
a new constant buffer every time we do a blit.
Keith Whitwell [Thu, 26 Nov 2009 10:34:28 +0000 (10:34 +0000)]
st/xorg: split up shared Done call
The two users of composite (Composite and Solid) now call a new
xorg_composite_done() from their Done functions, while CopyDone
is directly implemented on top of xorg_renderer.c.
Keith Whitwell [Thu, 26 Nov 2009 10:15:01 +0000 (10:15 +0000)]
st/xorg: set up rasterizer state in init
Keith Whitwell [Wed, 25 Nov 2009 18:45:20 +0000 (18:45 +0000)]
st/xorg: remove redundant clipping code
Keith Whitwell [Wed, 25 Nov 2009 18:42:54 +0000 (18:42 +0000)]
st/xorg: don't bother with cso save and restore in copy func
Keith Whitwell [Wed, 25 Nov 2009 18:41:11 +0000 (18:41 +0000)]
st/xorg: consolidate some dest surface state setting
Vinson Lee [Thu, 26 Nov 2009 05:35:31 +0000 (00:35 -0500)]
glu/sgi: Fix memory leak in gluBuild2DMipmapLevelsCore.
Jakob Bornecrantz [Tue, 24 Nov 2009 18:20:59 +0000 (19:20 +0100)]
vmware/xorg: Link against libkms
If the system doesn't have libkms installed it wont try to link against it.
Jakob Bornecrantz [Tue, 24 Nov 2009 18:16:37 +0000 (19:16 +0100)]
vmware/xorg: Clean Makefile a bit
Jakob Bornecrantz [Tue, 24 Nov 2009 17:47:15 +0000 (18:47 +0100)]
vmware/xorg: Don't link against libdrm_intel
Jakob Bornecrantz [Tue, 24 Nov 2009 17:44:39 +0000 (18:44 +0100)]
vmware/xorg: Stage driver in lib/gallium
Jakob Bornecrantz [Tue, 24 Nov 2009 13:53:03 +0000 (14:53 +0100)]
st/xorg: Pass mode types from the kernel to X
Jakob Bornecrantz [Tue, 24 Nov 2009 11:47:38 +0000 (12:47 +0100)]
st/xorg: Add libkms integration
Jakob Bornecrantz [Tue, 24 Nov 2009 01:49:57 +0000 (02:49 +0100)]
st/xorg: Touch up xorg_crtc.c
Jakob Bornecrantz [Tue, 24 Nov 2009 03:19:07 +0000 (04:19 +0100)]
st/xorg: Rename driver functions
Jakob Bornecrantz [Tue, 24 Nov 2009 01:54:24 +0000 (02:54 +0100)]
st/xorg: Rename output functions
Jakob Bornecrantz [Tue, 24 Nov 2009 01:28:09 +0000 (02:28 +0100)]
st/xorg: Rename dri2 functions
Jakob Bornecrantz [Tue, 24 Nov 2009 01:21:16 +0000 (02:21 +0100)]
st/xorg: Standardise all function names defined in xorg_tracker.h
Keith Whitwell [Wed, 25 Nov 2009 17:13:04 +0000 (17:13 +0000)]
svga: try harder to make the cachable flag work
It doesn't though.
Zack Rusin [Mon, 23 Nov 2009 06:00:34 +0000 (01:00 -0500)]
st/xorg: accelerate src luminance
Zack Rusin [Sun, 22 Nov 2009 20:21:14 +0000 (15:21 -0500)]
util: also print out memory statistics
Keith Whitwell [Wed, 25 Nov 2009 11:44:41 +0000 (11:44 +0000)]
svga: revert packing of surface key
Over-ambitious packing of values broke my cursor.
Dave Airlie [Wed, 25 Nov 2009 10:23:22 +0000 (20:23 +1000)]
radeon: fix context destroy needing lock for flushing.
Thanks to Intel code which I've just stolen pretty much as usual.
This fixes fdo bug 22851 which is a dri1 regression since rewrite.
Tested by: fpiobaf (Fabio) on #radeon
Signed-off-by: Dave Airlie <airlied@redhat.com>
Vinson Lee [Wed, 25 Nov 2009 05:39:37 +0000 (00:39 -0500)]
glu/sgi: Fix memory leak in gluBuild1DMipmapLevelsCore.
Vinson Lee [Wed, 25 Nov 2009 05:28:56 +0000 (00:28 -0500)]
glu/sgi: Fix memory leak in gluBuild2DMipmapLevelsCore.
Vinson Lee [Wed, 25 Nov 2009 05:18:49 +0000 (00:18 -0500)]
glu/sgi: Fix memory leak in gluBuild3DMipmapLevelsCore.
Keith Whitwell [Tue, 24 Nov 2009 21:13:18 +0000 (21:13 +0000)]
svga: cache textures as well as buffers
Keith Whitwell [Tue, 24 Nov 2009 20:48:12 +0000 (20:48 +0000)]
st/xorg: use surface_copy for blits if available
Even if its not available, we really want to be coalescing blit
operations better.
Jakob Bornecrantz [Tue, 24 Nov 2009 00:14:03 +0000 (01:14 +0100)]
svga: Filter out pendantic and ansi flags
Rather have the driver compile without the flags then
having to disable them.
Vinson Lee [Tue, 24 Nov 2009 06:23:12 +0000 (01:23 -0500)]
glu/sgi: Fix memory leak in gluBuild3DMipmapLevelsCore.
Vinson Lee [Tue, 24 Nov 2009 05:57:55 +0000 (00:57 -0500)]
dri: Fix potential null pointer deference in dri_put_drawable.
Maciej Cencora [Mon, 23 Nov 2009 20:59:08 +0000 (21:59 +0100)]
radeon: fix errors in miptree related function
- typo
- memory leak
- off by one (spotted by airlied)
Maciej Cencora [Mon, 23 Nov 2009 20:28:22 +0000 (21:28 +0100)]
Merge branch 'mesa_7_7_branch' of anongit.freedesktop.org/git/mesa/mesa into mesa_7_7_branch
Jakob Bornecrantz [Mon, 23 Nov 2009 18:59:02 +0000 (19:59 +0100)]
svga: Scrub Makefiles a bit
Remove x86 specific hacks. Not that they will ever be used on
none x86 arches, but they are built by default. And the way the
flags where added was a hack.
Vinson Lee [Mon, 23 Nov 2009 06:52:59 +0000 (01:52 -0500)]
mesa: Initialize variable in _mesa_get_texel_fetch_func.
Vinson Lee [Mon, 23 Nov 2009 06:30:32 +0000 (01:30 -0500)]
glx: Prevent potential null pointer deference in driCreateContext.
Vinson Lee [Mon, 23 Nov 2009 06:09:06 +0000 (01:09 -0500)]
glu/sgi: Fix memory leak in bitmapBuild2DMipmaps.
Vinson Lee [Mon, 23 Nov 2009 05:57:37 +0000 (00:57 -0500)]
glu/sgi: Fix memory leak in gluScaleImage3D.
Maciej Cencora [Sun, 22 Nov 2009 14:08:46 +0000 (15:08 +0100)]
r300: fix VP source conflict resolution on 64-bit machines
On 32bit machines we were lucky because the sizeof(reg) == sizeof(rc_src_register).
On 64bit machines pointers are 8 bytes long, so we were overwriting other data.
Maciej Cencora [Sun, 22 Nov 2009 13:13:18 +0000 (14:13 +0100)]
r300: fix SIN/COS/SCS instructions for R300 fp
Vinson Lee [Sun, 22 Nov 2009 06:57:35 +0000 (01:57 -0500)]
glu/sgi: Fix memory leak in gluScaleImage.
Vinson Lee [Sun, 22 Nov 2009 06:45:53 +0000 (01:45 -0500)]
glu: Fix memory leak in __gl_meshMakeEdge.
Vinson Lee [Sun, 22 Nov 2009 06:26:32 +0000 (01:26 -0500)]
svga: Prevent potential null pointer deference in vmw_surface.c.
Vinson Lee [Sun, 22 Nov 2009 06:20:07 +0000 (01:20 -0500)]
svga: Fix memory leak in vmw_screen_ioctl.c
Vinson Lee [Sun, 22 Nov 2009 06:10:38 +0000 (01:10 -0500)]
st/xorg: Prevent potential null pointer deference in xorg_xv.c.
Vinson Lee [Sun, 22 Nov 2009 05:55:52 +0000 (00:55 -0500)]
st/xorg: Reorder cases in switch statement.
Silences missing break statement warning.
Maciej Cencora [Sat, 21 Nov 2009 14:56:23 +0000 (15:56 +0100)]
radeon: fix compressed mipmapped textures
Tested on r300 only, other cards may require adjusting texture_compressed_row_align.
Maciej Cencora [Sat, 21 Nov 2009 14:56:02 +0000 (15:56 +0100)]
radeon: fix glCompressedTexSubImage
Maciej Cencora [Sat, 21 Nov 2009 14:56:23 +0000 (15:56 +0100)]
radeon: fix compressed mipmapped textures
Tested on r300 only, other cards may require adjusting texture_compressed_row_align.
Maciej Cencora [Sat, 21 Nov 2009 14:56:02 +0000 (15:56 +0100)]
radeon: fix glCompressedTexSubImage
Roland Scheidegger [Sat, 21 Nov 2009 12:39:30 +0000 (04:39 -0800)]
intel: make CopyTex[Sub]Image fallback debug messages more consistent
Roland Scheidegger [Sat, 21 Nov 2009 12:22:50 +0000 (04:22 -0800)]
mesa: handle different RowStride in _mesa_get_compressed_teximage
drivers storing compressed textures with non-native stride but relying on
_mesa_get_compressed_teximage for GetCompressedTexImage otherwise won't
work correctly (for instance i965 compressed mipmaps).
Eric Anholt [Sat, 21 Nov 2009 12:33:21 +0000 (13:33 +0100)]
i965: Fix several memory leaks on exit.
Bug #25194.
Vinson Lee [Sat, 21 Nov 2009 00:33:25 +0000 (16:33 -0800)]
st/egl: Fix memory leak in egl_tracker.c.
Vinson Lee [Sat, 21 Nov 2009 00:03:48 +0000 (16:03 -0800)]
draw: Initialize variable in draw_pt.c.
Vinson Lee [Fri, 20 Nov 2009 23:52:53 +0000 (15:52 -0800)]
progs/xdemos: Fix memory leak in pbinfo.c.
Vinson Lee [Fri, 20 Nov 2009 23:48:29 +0000 (15:48 -0800)]
progs/xdemos: Fix memory leak in pbdemo.c.
Vinson Lee [Fri, 20 Nov 2009 23:18:19 +0000 (15:18 -0800)]
progs/xdemos: Fix memory leak in glxpbdemo.c.
Vinson Lee [Fri, 20 Nov 2009 23:06:56 +0000 (15:06 -0800)]
progs/xdemos: Remove duplicate code in glxinfo.c.
Vinson Lee [Fri, 20 Nov 2009 23:03:16 +0000 (15:03 -0800)]
progs/xdemos: Add assert in corender.c.
Vinson Lee [Fri, 20 Nov 2009 22:31:42 +0000 (14:31 -0800)]
i915g: Add missing break statement in i915_debug.c.
Vinson Lee [Fri, 20 Nov 2009 22:20:15 +0000 (14:20 -0800)]
gallium/util: Initialize variables in u_pack_color.h.
Jakob Bornecrantz [Wed, 18 Nov 2009 16:45:39 +0000 (17:45 +0100)]
st/xorg: Flush any pending operations on upload
Jakob Bornecrantz [Wed, 18 Nov 2009 16:38:49 +0000 (17:38 +0100)]
st/xorg: Unbind any textures in solid
Helps debuging with rbug
Alan Hourihane [Fri, 20 Nov 2009 18:09:10 +0000 (18:09 +0000)]
Fix vega compilation.
Alan Hourihane [Fri, 20 Nov 2009 18:08:29 +0000 (18:08 +0000)]
Fix memory leak.
Vinson Lee [Fri, 20 Nov 2009 21:09:42 +0000 (13:09 -0800)]
progs/xdemos: Silence unused value warnings in sharedtex_mt.c.
Vinson Lee [Fri, 20 Nov 2009 21:01:22 +0000 (13:01 -0800)]
progs/xdemos: Fix memory leak in glxinfo.c.
Vinson Lee [Fri, 20 Nov 2009 20:48:44 +0000 (12:48 -0800)]
progs/tests: Silence unused value warnings in sharedtex.c.
Eric Anholt [Fri, 20 Nov 2009 20:42:06 +0000 (21:42 +0100)]
mesa: Fix NULL deref in optimizer when NumInstructions == 0.
Bug #24984.
Vinson Lee [Fri, 20 Nov 2009 20:34:21 +0000 (12:34 -0800)]
progs/tests: Initialize variable in getprocaddress.c.
Vinson Lee [Fri, 20 Nov 2009 20:11:55 +0000 (12:11 -0800)]
progs/util: Exit RawImageGetData early if malloc fails.
Prevents a null pointer deference later on.
Vinson Lee [Fri, 20 Nov 2009 19:34:21 +0000 (11:34 -0800)]
progs/slang: Check for ftell failure in vstest.c.
Vinson Lee [Fri, 20 Nov 2009 19:22:13 +0000 (11:22 -0800)]
progs/util: Fix memory leak if malloc fails in RawImageOpen.
Vinson Lee [Fri, 20 Nov 2009 08:03:10 +0000 (00:03 -0800)]
st/xorg: Fix type of 'unknown'.
Vinson Lee [Fri, 20 Nov 2009 07:56:07 +0000 (23:56 -0800)]
progs/vpglsl: Fix psiz-mul.glsl compilation error on Mac OS.
Vinson Lee [Fri, 20 Nov 2009 06:59:01 +0000 (22:59 -0800)]
progs/vp: Fix memory leak in vp-tris.c.
Vinson Lee [Fri, 20 Nov 2009 06:52:05 +0000 (22:52 -0800)]
progs/trivial: Redraw upon keypress.
Vinson Lee [Fri, 20 Nov 2009 06:46:19 +0000 (22:46 -0800)]
progs/glsl: Redraw upon keypress.
Dave Airlie [Fri, 20 Nov 2009 01:45:43 +0000 (11:45 +1000)]
r100: fix texture_from_pixmap and compiz.
r100 state emission has separate rect and non-rect states,
if we are doing TFP for a TEXTURE_2D we shouldn't use the rect
states as they won't get emitted properly.
Signed-off-by: Dave Airlie <airlied@redhat.com>