mesa.git
18 years agoBump version to 6.3.2
Brian Paul [Tue, 16 Aug 2005 23:13:25 +0000 (23:13 +0000)]
Bump version to 6.3.2

18 years agoupdates for 6.3.2
Brian Paul [Tue, 16 Aug 2005 23:12:37 +0000 (23:12 +0000)]
updates for 6.3.2

18 years agoinclude stdint.h if WIN32 and GNUC (Colin Harrison). Patch forwarded to SGI.
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.

18 years agotest the standard stencil op modes too (Philipp Klaus Krause)
Brian Paul [Mon, 15 Aug 2005 22:57:30 +0000 (22:57 +0000)]
test the standard stencil op modes too (Philipp Klaus Krause)

18 years agoIn _mesa_TexEnvfv, do not accept values for GL_OPERAND2_ALPHA and GL_OPERAND2_RGB...
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.

18 years agoPut back '-I/usr/X11R6/include'.
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.

18 years agoAdd Egberts fixes for 64bit architectures
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).

18 years agoAdd a check which compares the passed in size of the DDX driver private DRI
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.

18 years agoEnable support for EXT_stencil_wrap. The code was already there, but the
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.

18 years agoadded FREXPF() macro (bug 4060)
Brian Paul [Fri, 12 Aug 2005 18:56:56 +0000 (18:56 +0000)]
added FREXPF() macro (bug 4060)

18 years agoRemove the logic that determines at compile time whether or not HAVE_ALIAS
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.

18 years agoRemove support for the R200_NO_VTXFMT environment variable. Instead use
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.

18 years agoMakefile cleanups suggested by George Fufutos (bugzilla #3899).
Ian Romanick [Thu, 11 Aug 2005 19:37:55 +0000 (19:37 +0000)]
Makefile cleanups suggested by George Fufutos (bugzilla #3899).

18 years agoDon't link with libGL, libEGL, or libEGLdri. On EGL setups libGL isn't
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.

18 years agoSince Brian's recent commit, "-I/usr/X11R6/include/X11/extensions" is no
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.

18 years agofix X #includes so they're relative to X11/ directory (bug 4047)
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)

18 years agoMake ffb driver build on x86-64. The fix is two fold. First, use the
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.

18 years agoRemove _glapi_check_multithread from the interface exported by the loader to
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

18 years agoMesa uses the glVertexAttrib*NV functions to implement fixed-function
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.

18 years agothe PIXEL_MAP_I_TO_I table should be GLfloat, not GLint
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

18 years agofix static assertion problem for gcc (bug 4022)
Brian Paul [Wed, 10 Aug 2005 14:42:21 +0000 (14:42 +0000)]
fix static assertion problem for gcc (bug 4022)

18 years agodisable the pointer size assertions (bug 4021)
Brian Paul [Wed, 10 Aug 2005 14:40:05 +0000 (14:40 +0000)]
disable the pointer size assertions (bug 4021)

18 years agofix indentation
Brian Paul [Wed, 10 Aug 2005 14:38:39 +0000 (14:38 +0000)]
fix indentation

18 years agoCheck for prim/vertex store non-null before messing with refcounts.
Keith Whitwell [Wed, 10 Aug 2005 14:13:45 +0000 (14:13 +0000)]
Check for prim/vertex store non-null before messing with refcounts.

18 years agoAdd rules and target to build src/mesa/x86-64/glapi_x86-64.S.
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.

18 years agoDecrement the context's refcounts on part-used storage structs on
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.

18 years agoactually free the renderbuffer in xmesa_delete_renderbuffer(), fixes a 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

18 years agodelete display lists upon exit
Brian Paul [Tue, 9 Aug 2005 15:14:51 +0000 (15:14 +0000)]
delete display lists upon exit

18 years agofree display lists and destroy window upon exit
Brian Paul [Tue, 9 Aug 2005 15:13:32 +0000 (15:13 +0000)]
free display lists and destroy window upon exit

18 years ago_mesa_free_framebuffer_data() call frees the renderbuffers, no need to call the Delet...
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

18 years agofree context structure (bug 1252316) and s/free/_mesa_free/
Brian Paul [Tue, 9 Aug 2005 14:23:28 +0000 (14:23 +0000)]
free context structure (bug 1252316) and s/free/_mesa_free/

18 years agoMake the linux-dri-x86 builds work on x86-64 again. mklib now
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.

18 years agos/Disaptch/Dispatch/ (bug 3918)
Brian Paul [Mon, 8 Aug 2005 21:42:51 +0000 (21:42 +0000)]
s/Disaptch/Dispatch/ (bug 3918)

18 years agoadded directfbgl.h to tarball
Brian Paul [Mon, 8 Aug 2005 21:30:12 +0000 (21:30 +0000)]
added directfbgl.h to tarball

18 years agoSince glcontextmodes.[ch] are no longer used in the DRI drivers, removed the
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.

18 years agofix up color tiling in miniglx radeon server code
Dave Airlie [Mon, 8 Aug 2005 04:34:35 +0000 (04:34 +0000)]
fix up color tiling in miniglx radeon server code

18 years agofix miniglx setting virtual size of 832 for 800x600
Dave Airlie [Mon, 8 Aug 2005 04:17:52 +0000 (04:17 +0000)]
fix miniglx setting virtual size of 832 for 800x600

18 years agoRemove platform specific bits from radeon_egl.c
Jon Smirl [Mon, 8 Aug 2005 03:58:26 +0000 (03:58 +0000)]
Remove platform specific bits from radeon_egl.c

18 years agoDike out the linux-specific EGL bits unless it's a linux build.
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.

18 years agoSince this isn't a bash script, don't use function and instead inline the usage
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.

18 years agoAllow R300 driver to build on x86-64 and PowerPC.
Vladimir Dergachev [Sun, 7 Aug 2005 21:50:33 +0000 (21:50 +0000)]
Allow R300 driver to build on x86-64 and PowerPC.

18 years agoAdd missing egldri.h header file
Jon Smirl [Sun, 7 Aug 2005 02:30:32 +0000 (02:30 +0000)]
Add missing egldri.h header file

18 years agoFix up some of the sample to work with the r200 EGL driver
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

18 years agoEnable EGL support in 'linux-dri' config
Jon Smirl [Sun, 7 Aug 2005 02:13:14 +0000 (02:13 +0000)]
Enable EGL support in 'linux-dri' config

18 years agoFollow suggestion by Aapo Tahkola to fix giant memory leak from forgetting to free...
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).

18 years agoFix card_extensions by following corresponding r200 code.
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.

18 years agoFixes bugzilla #1730.
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.

18 years agoFixes bugzillla #812.
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.

18 years agoremove the drmAddress
Alan Hourihane [Fri, 5 Aug 2005 20:47:59 +0000 (20:47 +0000)]
remove the drmAddress

18 years agoIn order to create the set of available extensions, we have to know if
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.

18 years agoFix recent problems with display lists and other parts of the code.
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.

18 years agoRemove EGL from linux-solo makefile
Jon Smirl [Fri, 5 Aug 2005 03:48:15 +0000 (03:48 +0000)]
Remove EGL from linux-solo makefile

18 years agoMore EGL prep. No impact on anything outside of EGL
Jon Smirl [Fri, 5 Aug 2005 03:33:59 +0000 (03:33 +0000)]
More EGL prep. No impact on anything outside of EGL

18 years agoNew files for radeon egl driver.
Jon Smirl [Thu, 4 Aug 2005 22:23:34 +0000 (22:23 +0000)]
New files for radeon egl driver.
Still a work in progress.

18 years agoConvert all instances of XTHREADS to USE_XTHREADS. This fixes one of
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.

18 years agoAdd color tiling support to miniglx for radeon
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)

18 years agoFix build problems related to finding Xthreads.h.
Ian Romanick [Mon, 1 Aug 2005 16:30:24 +0000 (16:30 +0000)]
Fix build problems related to finding Xthreads.h.

18 years agoRemove unused structures and data fields. These had been left in because
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.

18 years agoAdd a new config called linux-indirect. This build the GLX libGL just like
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.

18 years agoLeave the AGP_BASE code in. It is pointless to bump the interface
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.

18 years agoRemove OUTREG(RADEON_AGP_BASE, info->ringHandle); from DRI driver
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.

18 years agoCommit Ian's fixes from Bug 3877
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.

18 years agoFix a == vs !- typo for glx_enable_extension
Jon Smirl [Fri, 29 Jul 2005 00:19:36 +0000 (00:19 +0000)]
Fix a == vs !- typo for glx_enable_extension

18 years agoUse the routines supplied in the __DRIinterfaceMethods structure to create a
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.

18 years agoglxEnableExtension is a platform-specific function exported by the loader.
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.

18 years agoRemove '#if 0' block that was accidentally left in.
Ian Romanick [Thu, 28 Jul 2005 20:55:26 +0000 (20:55 +0000)]
Remove '#if 0' block that was accidentally left in.

18 years agoSPARC, DRI interface changes for 6.3.1
Brian Paul [Thu, 28 Jul 2005 02:46:26 +0000 (02:46 +0000)]
SPARC, DRI interface changes for 6.3.1

18 years agoThe dri/*/depend files don't exist on a raw check-out, and that makes 'make
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.

18 years agoMajor rip-up of internal function insertion interface. The old
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)

18 years agoFix SPARC assembly for 64-bit
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.

18 years agoThis precious BeOS include file was lost since some released MesaLibs. Fixed. ;-)
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. ;-)

18 years agotest for the proper drm version, at least 1.17 is required
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

18 years agotypo
Alan Hourihane [Tue, 26 Jul 2005 23:09:50 +0000 (23:09 +0000)]
typo

18 years agosupport GLX_DIRECT_COLOR in 1.4.0
Alan Hourihane [Tue, 26 Jul 2005 23:06:10 +0000 (23:06 +0000)]
support GLX_DIRECT_COLOR in 1.4.0

18 years agoRemove the last remnants of GLX_BUILT_IN_XMESA. This allows the removal of
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.

18 years agoBump the required ddx to 1.4.0 as drmAddress has been removed so that 32/64bit
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.

18 years agoNo, really, get rid of all the remaining references to DRI_USE_NEW_INTERFACE.
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. :)

18 years agoQuote $(CC) and $(CXX) so that 'CC=ccache gcc' and 'CXX=ccache g++' will
Ian Romanick [Tue, 26 Jul 2005 05:55:56 +0000 (05:55 +0000)]
Quote $(CC) and $(CXX) so that 'CC=ccache gcc' and 'CXX=ccache g++' will
work again.

18 years agoFixes the glXGetProcAddress portion of the interface. Most of the functions
Ian Romanick [Tue, 26 Jul 2005 02:44:01 +0000 (02:44 +0000)]
Fixes the glXGetProcAddress portion of the interface.  Most of the functions
that are currently obtained via glXGetProcAddress and all of the XF86DRI
functions are replaced with a funciton table.  This table will be passed to
__driCreateNewScreen.

One of the functions in the table is getProcAddress.  This allows some
loaders to expose functionality not in all loaders.  This will be immediatly
used for glxEnableExtension (formerly known to drivers as
__glXScrEnableExtension).  libGL (and in the future libglx) expose this
function so that drivers can enable GLX extensions.  libEGL should exposed
eglEnableExtension to enable EGL extensions.  The same function cannot be
used for both because the extensions have different names and (possibly)
different semantics.  Drivers can optionally use one, both, or neither.

The key parts are in the __DRIinterfaceMethodsRec structure in
dri_interface.h.  A pointer to one of these structures is passed into
__driCreateNewScreen.  Because of this, the version of the API is bumped to
20050725.  Since the previous version(s) were never in a release, their
existance is erased.

I was actually a little surprised by how much code this cuts from the
drivers.  A lot of glXGetProcAddress calls disappear, and a lot of
version checks go with them.  Nice.

The one thing I'm not sure of is removing __glXInitialize.  For some
reason that function was in the glXGetProcAddress table, but *nothing*
in the Mesa tree used it.  Did something with DRI conf. use this
function?  It seems odd...

18 years agoimprove comments about win[0..3] values (bug 2554)
Brian Paul [Tue, 26 Jul 2005 01:04:53 +0000 (01:04 +0000)]
improve comments about win[0..3] values (bug 2554)

18 years agoFix bug 1858 - array subscript out of range in gamma driver.
Brian Paul [Tue, 26 Jul 2005 00:10:53 +0000 (00:10 +0000)]
Fix bug 1858 - array subscript out of range in gamma driver.
Set ctx->Const.MaxTextureLevels = GAMMA_TEX_MAXLEVELS = 12
Remove bogus WRITE(level 12).
Max gamma texture size is 2kx2k  (note: 2k = 1 << (MaxLevels - 1))

18 years agoAdd license/copyright.
Brian Paul [Mon, 25 Jul 2005 23:49:50 +0000 (23:49 +0000)]
Add license/copyright.
Add -h/--help option for usage.
Minor clean-ups.

18 years agoAdded -linker option to mklib, used to specify a particular program for
Brian Paul [Mon, 25 Jul 2005 22:59:58 +0000 (22:59 +0000)]
Added -linker option to mklib, used to specify a particular program for
linking, if relevant.
Updated Makefiles to use -linker option instead of setting CC, CXX env vars.

18 years agoGut a few more dead bits. Replace uses of the CreateNewScreenFunc typedef
Ian Romanick [Sun, 24 Jul 2005 07:38:23 +0000 (07:38 +0000)]
Gut a few more dead bits.  Replace uses of the CreateNewScreenFunc typedef
(from glxclient.h) with PFNCREATENEWSCREEN (from dri_interface.h).

Remove the prototype for __driCreateScreen and fix the prototype for
__driCreateNewScreen (append the API version) in dri_interface.h.

18 years agoAll elements of pre-DRI_NEW_INTERFACE_ONLY are removed. This allows
Ian Romanick [Sun, 24 Jul 2005 06:29:14 +0000 (06:29 +0000)]
All elements of pre-DRI_NEW_INTERFACE_ONLY are removed.  This allows
1,402 lines of code to be removed from Mesa (drivers and libGL).  The
big winner is dri_util.c.

Primary changes are:

1. Remove all "deprecated" entry-points from the various structures in
dri_interface.h.

2. Rename the remaining fields to removed "version numbers."  So,
bindContext3 becomes bindContext.  Functions with "New" in the name
(e.g., CreateNewContext) were *not* changed, but that is an option.
Having "New" in the name is less annoying to me than having "3" in the name.

3. Remove all compatibility code that handles cases where the driver or
the loader is too old to support the latest interfaces.

4. Append the API version to the __driCreateNewScreen function name.
This is currently done by hand.  In the future (i.e., the next time we
make an incompatible change to the interface) we'll want to come up with
a better way to do this.  This prevents old loaders from being able to load
new (incompatible) drivers.

5. Bump the API version to 20050722.  All drivers (by way of dri_util.c)
require this version.

6. All drivers are *required* to expose GLX_SGIX_fbconfig and
GLX_OML_swap_method (or the moral equivalents).  Support for these
functions in implicit in the use of the "new" interface.

7. Some cases still exist that need to be compiled differently in a loader
or core Mesa versus in a driver.  These are identified by the define
IN_DRI_DRIVER.

18 years agoAdd two aliases for some PointParameters entrypoints for libglx compatibility.
Adam Jackson [Sat, 23 Jul 2005 02:09:58 +0000 (02:09 +0000)]
Add two aliases for some PointParameters entrypoints for libglx compatibility.
Doing this here rather than on the GLX side since the X monolith hasn't had
Mesa 6.3 imported yet.

idr, I know this is touching a generated file, but I haven't figured out what
needs changing in the generator scripts to make this change happen.  If you
could add that I'd appreciate it.

18 years agoBreak the long list of MAIN_FILES and DEMO_FILES into subgroups.
Brian Paul [Fri, 22 Jul 2005 15:52:46 +0000 (15:52 +0000)]
Break the long list of MAIN_FILES and DEMO_FILES into subgroups.
rm/touch the depend files when making tarballs so they don't include
RedHat-specific paths (bug 3842).

18 years ago ----------------------------------------------------------------------
Jouk Jansen [Fri, 22 Jul 2005 09:30:54 +0000 (09:30 +0000)]
 ----------------------------------------------------------------------
 Committing in .

 _mesa_sprintf crashes on OpenVMS -> replaced it by the OS-supplied sprintf

 Modified Files:
  Mesa/src/mesa/drivers/x11/fakeglx.c
 ----------------------------------------------------------------------

18 years agosome GLuint idx -> GLint idx changes
Brian Paul [Fri, 22 Jul 2005 02:55:36 +0000 (02:55 +0000)]
some GLuint idx -> GLint idx changes

18 years agoadded vp negative addressing bug fix
Brian Paul [Fri, 22 Jul 2005 02:54:14 +0000 (02:54 +0000)]
added vp negative addressing bug fix

18 years agoMake the vertex program source register Index field a signed int since
Brian Paul [Fri, 22 Jul 2005 02:53:38 +0000 (02:53 +0000)]
Make the vertex program source register Index field a signed int since
relative addressing can be negative.
Change some GLuint indexes to GLint in the t_vp_build.c file.
Added PROGRAM_UNDEFINED token for initializing the register File field to
avoid a gcc 4.0 warning.

18 years agobump version to 6.3.1
Brian Paul [Thu, 21 Jul 2005 18:46:49 +0000 (18:46 +0000)]
bump version to 6.3.1

18 years ago6.3.1 updates
Brian Paul [Thu, 21 Jul 2005 18:45:54 +0000 (18:45 +0000)]
6.3.1 updates

18 years agoupdated with DRI building instructions
Brian Paul [Thu, 21 Jul 2005 18:45:44 +0000 (18:45 +0000)]
updated with DRI building instructions

18 years agobump version to 6.3.1 and include DRI/GLX files in tarballs
Brian Paul [Thu, 21 Jul 2005 18:44:52 +0000 (18:44 +0000)]
bump version to 6.3.1 and include DRI/GLX files in tarballs

18 years agomake i810PrintSetupFlags() static, first param const
Brian Paul [Thu, 21 Jul 2005 18:11:24 +0000 (18:11 +0000)]
make i810PrintSetupFlags() static, first param const

18 years agofirst 6.3.1 bug fixes
Brian Paul [Thu, 21 Jul 2005 15:58:16 +0000 (15:58 +0000)]
first 6.3.1 bug fixes

18 years agos/2004/2005/
Brian Paul [Thu, 21 Jul 2005 15:57:29 +0000 (15:57 +0000)]
s/2004/2005/

18 years agocheck for FEATURE_EXT_framebuffer_object to enable EXT_framebuffer_object
Brian Paul [Thu, 21 Jul 2005 15:48:39 +0000 (15:48 +0000)]
check for FEATURE_EXT_framebuffer_object to enable EXT_framebuffer_object

18 years agoremove ^M chars, disable shading language extensions
Brian Paul [Thu, 21 Jul 2005 15:41:56 +0000 (15:41 +0000)]
remove ^M chars, disable shading language extensions