Alan Hourihane [Wed, 24 Aug 2005 09:08:06 +0000 (09:08 +0000)]
Fix Windows build problems.
GCC's visibility attribute is for ELF systems, so undef INTERNAL
GCC's fastcall attribute interferes with __stdcall so undef FASTCALL
Brian Paul [Tue, 23 Aug 2005 14:50:02 +0000 (14:50 +0000)]
GLw demo doesn't belong here
Brian Paul [Tue, 23 Aug 2005 13:40:30 +0000 (13:40 +0000)]
removed lines that caused some files to be included in tarballs multiple times
Brian Paul [Tue, 23 Aug 2005 03:50:12 +0000 (03:50 +0000)]
simplify, clean-up texel addressing macros
Brian Paul [Tue, 23 Aug 2005 00:45:54 +0000 (00:45 +0000)]
update version to 6.5.0
Brian Paul [Mon, 22 Aug 2005 14:28:19 +0000 (14:28 +0000)]
added GL_X_RENDERABLE to glXChooseFBConfig (bug 4181)
Brian Paul [Mon, 22 Aug 2005 14:27:33 +0000 (14:27 +0000)]
bump version to 6.5
Brian Paul [Mon, 22 Aug 2005 14:27:13 +0000 (14:27 +0000)]
added .S DRI files to tarball (Donnie Berkholz), bumped version to 6.5
Dave Airlie [Sun, 21 Aug 2005 03:17:52 +0000 (03:17 +0000)]
fix up libdrm for egl
Brian Paul [Fri, 19 Aug 2005 23:44:35 +0000 (23:44 +0000)]
fix typo
Brian Paul [Fri, 19 Aug 2005 23:42:29 +0000 (23:42 +0000)]
md5 sums for 6.3.2
Brian Paul [Fri, 19 Aug 2005 23:34:10 +0000 (23:34 +0000)]
update MESA_DEBUG with FP info
Adam Jackson [Fri, 19 Aug 2005 20:20:03 +0000 (20:20 +0000)]
Convert libGL and DRI drivers to require libdrm.
libdrm can be had from:
http://people.freedesktop.org/~ajax/libdrm/
Ian Romanick [Fri, 19 Aug 2005 18:53:26 +0000 (18:53 +0000)]
Fix a realloc problem with indirect vertex arrays. The actual head pointer
wasn't tracked and used for the realloc, so it tended to explode.
Brian Paul [Fri, 19 Aug 2005 16:57:50 +0000 (16:57 +0000)]
updates for 6.3.2 release, plus other assorted clean-ups
Alan Hourihane [Fri, 19 Aug 2005 16:07:33 +0000 (16:07 +0000)]
Commit missing piece from Egbert's 32/64 bit patches
Ian Romanick [Fri, 19 Aug 2005 00:48:25 +0000 (00:48 +0000)]
Enable ARB_texture_compression. The code has been in place for a long time,
but, for whatever reason, the extension wasn't enabled.
Add some comments to i810ChooseTextureFormat. There's some strangeness with
i810 texture formats.
Ian Romanick [Thu, 18 Aug 2005 22:27:37 +0000 (22:27 +0000)]
Add support for NV_blend_square. This was tested with
progs/tests/blendsquare.
Ian Romanick [Thu, 18 Aug 2005 22:16:46 +0000 (22:16 +0000)]
Make sure extension entry point offsets are properly initialized. This is
primarilly needed to support functions that are internally implemented using
extension entry points (e.g., glBlendFunc uses glBlendFuncSeparate).
Ian Romanick [Thu, 18 Aug 2005 20:01:01 +0000 (20:01 +0000)]
Don't try to update hardware texture fields if there is no hardware texture.
This can happen when all the texture enables for a texture unit are disabled.
This fixes bugzilla #3195.
Ian Romanick [Thu, 18 Aug 2005 19:52:44 +0000 (19:52 +0000)]
Add simple regression test for bugzilla #3195.
Ian Romanick [Thu, 18 Aug 2005 19:51:11 +0000 (19:51 +0000)]
Make the i810 driver use driGetRendererString like all the other drivers.
The GL_RENDERER string now accurately contains the chip name. This was only
tested on a "plain" i810 (PCI ID 8086:7121).
Brian Paul [Thu, 18 Aug 2005 14:33:26 +0000 (14:33 +0000)]
Replace mult/div operators with bitwise operators in texel fetch routines
for better performance. (Roland Scheidegger)
Brian Paul [Tue, 16 Aug 2005 23:13:25 +0000 (23:13 +0000)]
Bump version to 6.3.2
Brian Paul [Tue, 16 Aug 2005 23:12:37 +0000 (23:12 +0000)]
updates for 6.3.2
Brian Paul [Mon, 15 Aug 2005 23:25:33 +0000 (23:25 +0000)]
include stdint.h if WIN32 and GNUC (Colin Harrison). Patch forwarded to SGI.
Brian Paul [Mon, 15 Aug 2005 22:57:30 +0000 (22:57 +0000)]
test the standard stencil op modes too (Philipp Klaus Krause)
Roland Scheidegger [Mon, 15 Aug 2005 18:24:12 +0000 (18:24 +0000)]
In _mesa_TexEnvfv, do not accept values for GL_OPERAND2_ALPHA and GL_OPERAND2_RGB which are only valid with ARB_texture_env_combine but not with EXT_texture_env_combine, when only EXT_texture_env_combine is supported.
Ian Romanick [Mon, 15 Aug 2005 15:01:39 +0000 (15:01 +0000)]
Put back '-I/usr/X11R6/include'.
Make 'USING_EGL=0' the default for now. It's causing too much confusion for
too many non-developers (which in turn wastes too much developer time). It
is trivially overridden from the make command line. See comments in
linux-dri.
Alan Hourihane [Mon, 15 Aug 2005 06:59:24 +0000 (06:59 +0000)]
Add Egberts fixes for 64bit architectures
Add additional checks for the *DRIRec info structure passed in from the
device driver. This ensures that things fallback to indirect rendering if
the DDX driver has had modifications (i.e. removal of the drmAddress field).
Alan Hourihane [Sun, 14 Aug 2005 12:16:51 +0000 (12:16 +0000)]
Add a check which compares the passed in size of the DDX driver private DRI
structure to the 3D drivers view of it, and abort if they don't match.
This traps the case with the drmAddress removal in the DDX driver.
Ian Romanick [Sat, 13 Aug 2005 01:43:28 +0000 (01:43 +0000)]
Enable support for EXT_stencil_wrap. The code was already there, but the
extension wasn't enabled. I have verified its correct function with Mesa's
stencil_wrap test. It is enabled on both Savage4 and Savage3D. Since
Savage3D uses a software fallback for *all* stencil operations, this is safe.
Brian Paul [Fri, 12 Aug 2005 18:56:56 +0000 (18:56 +0000)]
added FREXPF() macro (bug 4060)
Ian Romanick [Thu, 11 Aug 2005 20:29:45 +0000 (20:29 +0000)]
Remove the logic that determines at compile time whether or not HAVE_ALIAS
should be defined. It was flawed on some platforms (e.g., Darwin & mingw).
Instead, rely on the build system to define it on the compiler command line.
This also reverts ajax's hand-edit to indirect_size.c. I'll fix that on the
X.org side of things later today.
Ian Romanick [Thu, 11 Aug 2005 19:47:06 +0000 (19:47 +0000)]
Remove support for the R200_NO_VTXFMT environment variable. Instead use
'tcl_mode=1'. This fixes bugzilla #3972.
Ian Romanick [Thu, 11 Aug 2005 19:37:55 +0000 (19:37 +0000)]
Makefile cleanups suggested by George Fufutos (bugzilla #3899).
Ian Romanick [Thu, 11 Aug 2005 19:29:25 +0000 (19:29 +0000)]
Don't link with libGL, libEGL, or libEGLdri. On EGL setups libGL isn't
available, and on libGL setups libEGL and libEGLdri aren't available.
Ian Romanick [Thu, 11 Aug 2005 19:23:23 +0000 (19:23 +0000)]
Since Brian's recent commit, "-I/usr/X11R6/include/X11/extensions" is no
longer needed.
Brian Paul [Thu, 11 Aug 2005 14:18:53 +0000 (14:18 +0000)]
fix X #includes so they're relative to X11/ directory (bug 4047)
Ian Romanick [Thu, 11 Aug 2005 04:25:36 +0000 (04:25 +0000)]
Make ffb driver build on x86-64. The fix is two fold. First, use the
proper DRM_CAS_RESULT to declare the variable used to store the result
of DRM_CAS. Second, only use the "real" versions of LOCK_HARDWARE and
UNLOCK_HARDWARE on SPARC. That's the only platform where the hardware
can really exist.
Ian Romanick [Wed, 10 Aug 2005 23:54:15 +0000 (23:54 +0000)]
Remove _glapi_check_multithread from the interface exported by the loader to
the driver. The loader now takes care of this for the driver.
Remove _glapi_DispatchTSD and give _glapi_Dispatch its semantic (i.e.,
having a NULL value means that the application is multithreaded and
_glapi_get_dispatch must be called).
Gut all of the dispatch override code. This removes _glapi_RealDispatch,
_glapi_tls_RealDispatch, _glapi_begin_dispatch_override,
_glapi_end_dispatch_override, and _glapi_get_override_dispatch.
Remove _glapi_get_proc_address, _glapi_get_proc_name, _glapi_get_version,
and _glapi_check_table from the loader / driver interface.
Reviewed by: Brian Paul
Ian Romanick [Wed, 10 Aug 2005 23:39:42 +0000 (23:39 +0000)]
Mesa uses the glVertexAttrib*NV functions to implement fixed-function
attributes in display lists. If a driver doesn't add entry points for
GL_NV_vertex_program, it will explode. This patch adds GL_NV_vertex_program
to the list of default extensions whose entry points are added by
driInitExtensions.
Thanks go to Roland Scheidegger for tracking the problem down and suggesting
the fix.
Brian Paul [Wed, 10 Aug 2005 17:04:03 +0000 (17:04 +0000)]
the PIXEL_MAP_I_TO_I table should be GLfloat, not GLint
Brian Paul [Wed, 10 Aug 2005 14:42:21 +0000 (14:42 +0000)]
fix static assertion problem for gcc (bug 4022)
Brian Paul [Wed, 10 Aug 2005 14:40:05 +0000 (14:40 +0000)]
disable the pointer size assertions (bug 4021)
Brian Paul [Wed, 10 Aug 2005 14:38:39 +0000 (14:38 +0000)]
fix indentation
Keith Whitwell [Wed, 10 Aug 2005 14:13:45 +0000 (14:13 +0000)]
Check for prim/vertex store non-null before messing with refcounts.
Ian Romanick [Tue, 9 Aug 2005 18:54:12 +0000 (18:54 +0000)]
Add rules and target to build src/mesa/x86-64/glapi_x86-64.S.
Keith Whitwell [Tue, 9 Aug 2005 16:11:28 +0000 (16:11 +0000)]
Decrement the context's refcounts on part-used storage structs on
context destroy. Fixes memory leak.
Brian Paul [Tue, 9 Aug 2005 15:17:19 +0000 (15:17 +0000)]
actually free the renderbuffer in xmesa_delete_renderbuffer(), fixes a memory leak
Brian Paul [Tue, 9 Aug 2005 15:14:51 +0000 (15:14 +0000)]
delete display lists upon exit
Brian Paul [Tue, 9 Aug 2005 15:13:32 +0000 (15:13 +0000)]
free display lists and destroy window upon exit
Brian Paul [Tue, 9 Aug 2005 15:07:55 +0000 (15:07 +0000)]
_mesa_free_framebuffer_data() call frees the renderbuffers, no need to call the Delete methods after
Brian Paul [Tue, 9 Aug 2005 14:23:28 +0000 (14:23 +0000)]
free context structure (bug 1252316) and s/free/_mesa_free/
Ian Romanick [Mon, 8 Aug 2005 23:22:46 +0000 (23:22 +0000)]
Make the linux-dri-x86 builds work on x86-64 again. mklib now
determines the bits (either 32 or 64) for libraries without the lib
prefix. progs/egl/Makefile passes CFLAGS on the link commands so that
things like '-m32' get propagated.
Brian Paul [Mon, 8 Aug 2005 21:42:51 +0000 (21:42 +0000)]
s/Disaptch/Dispatch/ (bug 3918)
Brian Paul [Mon, 8 Aug 2005 21:30:12 +0000 (21:30 +0000)]
added directfbgl.h to tarball
Ian Romanick [Mon, 8 Aug 2005 17:22:44 +0000 (17:22 +0000)]
Since glcontextmodes.[ch] are no longer used in the DRI drivers, removed the
IN_DRI_DRIVER cruft. This was suggested by George Fufutos.
Dave Airlie [Mon, 8 Aug 2005 04:34:35 +0000 (04:34 +0000)]
fix up color tiling in miniglx radeon server code
Dave Airlie [Mon, 8 Aug 2005 04:17:52 +0000 (04:17 +0000)]
fix miniglx setting virtual size of 832 for 800x600
Jon Smirl [Mon, 8 Aug 2005 03:58:26 +0000 (03:58 +0000)]
Remove platform specific bits from radeon_egl.c
Eric Anholt [Mon, 8 Aug 2005 03:27:28 +0000 (03:27 +0000)]
Dike out the linux-specific EGL bits unless it's a linux build.
Eric Anholt [Mon, 8 Aug 2005 03:26:18 +0000 (03:26 +0000)]
Since this isn't a bash script, don't use function and instead inline the usage
into the one place that it's used.
Vladimir Dergachev [Sun, 7 Aug 2005 21:50:33 +0000 (21:50 +0000)]
Allow R300 driver to build on x86-64 and PowerPC.
Jon Smirl [Sun, 7 Aug 2005 02:30:32 +0000 (02:30 +0000)]
Add missing egldri.h header file
Jon Smirl [Sun, 7 Aug 2005 02:15:27 +0000 (02:15 +0000)]
Fix up some of the sample to work with the r200 EGL driver
Jon Smirl [Sun, 7 Aug 2005 02:13:14 +0000 (02:13 +0000)]
Enable EGL support in 'linux-dri' config
Vladimir Dergachev [Sat, 6 Aug 2005 05:19:42 +0000 (05:19 +0000)]
Follow suggestion by Aapo Tahkola to fix giant memory leak from forgetting to free the key when entry was already found (presumably a common case).
Vladimir Dergachev [Sat, 6 Aug 2005 04:39:27 +0000 (04:39 +0000)]
Fix card_extensions by following corresponding r200 code.
Fix suggested by Ian Romanick.
R300 driver now works without segfault.
Ian Romanick [Fri, 5 Aug 2005 23:13:13 +0000 (23:13 +0000)]
Fixes bugzilla #1730.
Make sure that all the values used in the clean-up code are initialized at
some point. It is still unclear to me as to why GCC does not complain. I
suspect the problem may be due to the depth of if-statement nesting.
Ian Romanick [Fri, 5 Aug 2005 22:12:49 +0000 (22:12 +0000)]
Fixes bugzillla #812.
The weak version of _glapi_DispatchTSD was removed from dri_util.c. As
stated in the bug report, this was only needed to support versions of libGL
that did not have this symbol. Versions of libGL without this symbol are no
longer supported at all, so it is totally irrelevant.
Alan Hourihane [Fri, 5 Aug 2005 20:47:59 +0000 (20:47 +0000)]
remove the drmAddress
Ian Romanick [Fri, 5 Aug 2005 19:13:51 +0000 (19:13 +0000)]
In order to create the set of available extensions, we have to know if
direct-rendering is possible on a given screen. It is not possible to
determine this by just looking at driDisplay.private in the display
structure. However, it is possible to determine this by looking at
driScreen.private in the screen structure. This field is only non-NULL if a
DRI driver has been successfully loaded and bootstrapped. Once we get to
that point, we know that direct-rendering is at least theoretically
possible.
This fixes bugzilla #2875.
Ian Romanick [Fri, 5 Aug 2005 18:13:37 +0000 (18:13 +0000)]
Fix recent problems with display lists and other parts of the code.
CALL_by_offset, SET_by_offset, and GET_by_offset all had various problems.
The core issue is that parts of the device-independent code in Mesa assumes
that all functions have slots in the dispatch table. This is especially
true in the display list code. It will merrilly try to set dispatch
pointers for glVertexAttrib1fARB even if GL_ARB_vertex_program is not
supported. When the GET/SET/CALL macros are invoked, they would read a 0
from the remap table. The problem is that 0 is the dispatch offset for
glNewList!
One change is that the remap table is now initialized to be full of -1
values. In addtion, all of the *_by_offset marcos misbehave in an obvious
way if the specified offset is -1. SET_by_offset will do nothing,
GET_by_offset will return NULL, and CALL_by_offset, since it uses
GET_by_offset, will segfault.
I also had to add GL_EXT_blend_func_separate to the list of default
extensions in all_mesa_extensions (src/mesa/drivers/dri/common/utils.c).
Even though many drivers do not export this extension, glBlendFunc is
internally implemented by calling glBlendFuncSeparate. Without this
addition, glBlendFunc stopped working on drivers (such as mga) that do not
export GL_EXT_blend_func_separate.
There are still a few assertions / crashes in GL_ARB_vertex_program tests,
but I don't think that these are related to any of my changes.
Jon Smirl [Fri, 5 Aug 2005 03:48:15 +0000 (03:48 +0000)]
Remove EGL from linux-solo makefile
Jon Smirl [Fri, 5 Aug 2005 03:33:59 +0000 (03:33 +0000)]
More EGL prep. No impact on anything outside of EGL
Jon Smirl [Thu, 4 Aug 2005 22:23:34 +0000 (22:23 +0000)]
New files for radeon egl driver.
Still a work in progress.
Ian Romanick [Wed, 3 Aug 2005 23:05:25 +0000 (23:05 +0000)]
Convert all instances of XTHREADS to USE_XTHREADS. This fixes one of
serveral things that are broken when building on a system with X.org 7.0rc0
installed.
Dave Airlie [Wed, 3 Aug 2005 10:22:31 +0000 (10:22 +0000)]
Add color tiling support to miniglx for radeon
(Stephane Marchesin + Dave Airlie)
Ian Romanick [Mon, 1 Aug 2005 16:30:24 +0000 (16:30 +0000)]
Fix build problems related to finding Xthreads.h.
Ian Romanick [Fri, 29 Jul 2005 17:30:18 +0000 (17:30 +0000)]
Remove unused structures and data fields. These had been left in because
DRI drivers using the "old" interface relied on certain fields being at
certain offsets. Removing the structures would break that. Now that the
old interface is no longer supported, the dead data can finally die.
Ian Romanick [Fri, 29 Jul 2005 17:25:50 +0000 (17:25 +0000)]
Add a new config called linux-indirect. This build the GLX libGL just like
linux-dri, but it disables support for direct-rendering in it. This config
mainly exists so that I can verify that changes to libGL haven't broken
builds on non-DRI platforms.
Jon Smirl [Fri, 29 Jul 2005 16:53:02 +0000 (16:53 +0000)]
Leave the AGP_BASE code in. It is pointless to bump the interface
for X on a miniglx fix. It can be removed later.
Jon Smirl [Fri, 29 Jul 2005 13:58:22 +0000 (13:58 +0000)]
Remove OUTREG(RADEON_AGP_BASE, info->ringHandle); from DRI driver
Code is already in DRM driver to do this.
Jon Smirl [Fri, 29 Jul 2005 00:48:12 +0000 (00:48 +0000)]
Commit Ian's fixes from Bug 3877
I have tested these on miniglx and they seem to be working.
I'm sure Dave will give them a better test.
Jon Smirl [Fri, 29 Jul 2005 00:19:36 +0000 (00:19 +0000)]
Fix a == vs !- typo for glx_enable_extension
Ian Romanick [Thu, 28 Jul 2005 23:33:18 +0000 (23:33 +0000)]
Use the routines supplied in the __DRIinterfaceMethods structure to create a
destroy __GLcontextModes structures. This eliminates the need for DRI
drivers to ever link with glcontextmodes.c.
Ian Romanick [Thu, 28 Jul 2005 21:57:17 +0000 (21:57 +0000)]
glxEnableExtension is a platform-specific function exported by the loader.
Therefore, drivers should not require it to run.
Ian Romanick [Thu, 28 Jul 2005 20:55:26 +0000 (20:55 +0000)]
Remove '#if 0' block that was accidentally left in.
Brian Paul [Thu, 28 Jul 2005 02:46:26 +0000 (02:46 +0000)]
SPARC, DRI interface changes for 6.3.1
Ian Romanick [Thu, 28 Jul 2005 01:15:45 +0000 (01:15 +0000)]
The dri/*/depend files don't exist on a raw check-out, and that makes 'make
tarballs' have issues.
Ian Romanick [Thu, 28 Jul 2005 00:29:51 +0000 (00:29 +0000)]
Major rip-up of internal function insertion interface. The old
_glapi_add_entrypoint has been replaced by a new routine called
_glapi_add_dispatch. This new routine dynamically assignes dispatch offsets
to functions added. This allows IHVs to add support for extension functions
that do not have assigned dispatch offsets.
It also means that a driver has no idea what offset will be assigned to a
function. The vast majority of the changes in this commit account for that.
An additional table, driDispatchRemapTable, is added. Functions not in the
Linux OpenGL ABI (i.e., anything not in GL 1.2 + ARB_multitexture) has a
fixed offset in this new table. The entry in this table specifies the
offset in of the function in the real dispatch table.
The internal interface was also bumped from version
20050725 to
20050727.
This has been tested with various programs in progs/demos on:
radeon (Radeon Mobility M6)
r128 (Rage 128 Pro)
mga (G400)
Ian Romanick [Thu, 28 Jul 2005 00:11:10 +0000 (00:11 +0000)]
Fix SPARC assembly for 64-bit
- The test for whether or not we're building for 64-bit is '#ifdef __arch64__'.
This appears to be correct on both Linux and Solaris.
- The test for Solaris is now '#if defined(SVR4) || defined(__SVR4) ||
defined(__svr4__)'. GCC 3.4.0 has all three defined on Solaris 9.
- Enables assembly language clip routines.
- Fixes to make GLSL code build on Solaris.
- Update gl_SPARC_asm.py.
Philippe Houdoin [Wed, 27 Jul 2005 19:56:33 +0000 (19:56 +0000)]
This precious BeOS include file was lost since some released MesaLibs. Fixed. ;-)
Roland Scheidegger [Wed, 27 Jul 2005 18:21:34 +0000 (18:21 +0000)]
test for the proper drm version, at least 1.17 is required
Alan Hourihane [Tue, 26 Jul 2005 23:09:50 +0000 (23:09 +0000)]
typo
Alan Hourihane [Tue, 26 Jul 2005 23:06:10 +0000 (23:06 +0000)]
support GLX_DIRECT_COLOR in 1.4.0
Ian Romanick [Tue, 26 Jul 2005 22:53:38 +0000 (22:53 +0000)]
Remove the last remnants of GLX_BUILT_IN_XMESA. This allows the removal of
the evil, ugly GLX_PREFIX macro as well.
Alan Hourihane [Tue, 26 Jul 2005 21:53:16 +0000 (21:53 +0000)]
Bump the required ddx to 1.4.0 as drmAddress has been removed so that 32/64bit
combinations work.
Ian Romanick [Tue, 26 Jul 2005 05:57:24 +0000 (05:57 +0000)]
No, really, get rid of all the remaining references to DRI_USE_NEW_INTERFACE.
A couple other remnants of the old interfaces hit the dust too. Thanks
Jon. :)