I have recently create the mesa_7_4_idr_staging branch. This contains a giant pile of cherry-picked patches from master that I believe are candidates for the mesa_7_4_branch. I wanted to give people a chance to veto any of these changes before they actually get to mesa_7_4_branch.
You can see the complete list of changes with:
git log mesa_7_4_branch..mesa_7_4_idr_staging
In addition to the changes that I have already brought over, there are
a bunch more changes that I think may also be candidates. However,
I'd like some review / commentary before doing so. There has been a
lot of FBO, GLSL, and VBO bug fixing in master, and it would be really
nice to have as much of that as is appropriate in 7.4. I have broken
these commits into a few related groups to make review easier.
If there are any candidate commits not mentioned below, please speak
up!
There have been giant piles of changes to the framebuffer object code.
This area has been notorious for bugs, especially in the Intel
drivers. Shoud any of these be pulled over? There are also a bunch
of changes related to the handling of depth / stencil FBO rendering
that are not listed below. Should any of those be pulled over?
commit 317687cd94d510ed72e32b0aee2782fbc123576b
Author: Michel Dänzer <[email protected]>
Date: Mon Mar 9 13:27:54 2009 +0100
Fix up another instance of _mesa_unreference_framebuffer().
commit 0f04a1d3f8989b0a391e6dad80abf06ce151d1f1
Author: Brian Paul <[email protected]>
Date: Sat Mar 7 11:32:18 2009 -0700
mesa: remove last of _mesa_unreference_framebuffer() calls
commit ee41bb2ed01b9480e5370d67257334b81ec0c90e
Author: Brian Paul <[email protected]>
Date: Thu Mar 5 17:04:27 2009 -0700
i965: fix screen depth test in intel_validate_framebuffer)_
front_region may be null.
commit 4f8ed56d168e9175e76bc42d8b924c7bcaa59dea
Author: Brian Paul <[email protected]>
Date: Thu Feb 26 18:42:00 2009 -0700
intel: no-op the intel_finish_render_texture() function
It doesn't have to do anything. See comments for more details.
commit f77b720cde981d441e482bbbd68115634b3041ce
Author: Brian Paul <[email protected]>
Date: Thu Feb 26 16:51:50 2009 -0700
intel: check texture formats in intel_validate_framebuffer()
We can't render into any texture format; only certain formats.
Check that render-to-texture's format is renderable in the
intel_validate_framebuffer()
There seems to be a bug somewhere that causes rendering to rgb565 textures
to be corrupted so disallow that for now. This will be revisted.
commit 645f220710923f72e80281fa75e577e133b5a53c
Author: Brian Paul <[email protected]>
Date: Thu Feb 26 16:44:42 2009 -0700
intel: updated comment, some debug code (disabled)
commit cdc63901df6af3b370935bd4997e3c9c4eb4b933
Author: root <[email protected]>
Date: Thu Feb 26 15:04:20 2009 -0700
i965: rename draw_regions -> color_regions
Be a little more specific about what these are.
commit 16144632354cb310f090c8713a11d3c65696969e
Author: Brian Paul <[email protected]>
Date: Thu Feb 26 14:49:24 2009 -0700
mesa: avoid extraneous calls to ctx->Driver.BindFramebuffer()
Only call this driver function when we really need to bind different
buffers.
commit 7eb5a2892d3d89f4b1800311babb40af1ec7d4e5
Author: Brian Paul <[email protected]>
Date: Mon Feb 9 08:31:28 2009 -0700
mesa: replace _mesa_unreference_framebuffer() with
_mesa_reference_framebuffer()
commit d52294439c80c99a4bfa2c7ecc65efa67f095a3f
Author: Brian Paul <[email protected]>
Date: Mon Feb 9 08:30:55 2009 -0700
mesa: replace _mesa_unreference_framebuffer() with
_mesa_reference_framebuffer()
commit 49002698193784dacffcbf993de26fcdd42c523e
Author: Brian Paul <[email protected]>
Date: Mon Feb 9 07:55:08 2009 -0700
mesa: rework _mesa_reference_framebuffer() to look like
_mesa_reference_renderbuffer()
_mesa_unreference_framebuffer() is deprecated since
_mesa_reference_framebuffer(ptr, NULL)
can be used instead.
commit bd944ef78397fd96dc2b239f542066643b06274a
Author: Brian Paul <[email protected]>
Date: Thu Jan 29 15:40:43 2009 -0700
intel: remove unused RenderToTexture field
commit 4b37b1129ea7fed915d353c39d4b74dbdbb7f0a5
Author: Brian Paul <[email protected]>
Date: Thu Jan 29 15:40:21 2009 -0700
i915: updated render to texture/fbo test
There has also been a lot of churn related to stencil operations
without stencil buffer. Quite a few of the changes were specific to
the metaops code, but these three stood out. Should they be pulled
over?
commit 8fec37c0f11e624644da48d612c60e736861a212
Author: Brian Paul <[email protected]>
Date: Mon Mar 2 11:49:19 2009 -0700
mesa: remove warning/short-circuit of stencil enable w/ no stencil buffer
With FBOs one could enable stencil before a stencil buffer is later bound.
commit 91e61f435a71436c209934a0ece165b540aba3e0
Author: Brian Paul <[email protected]>
Date: Mon Mar 2 11:47:52 2009 -0700
mesa: use Stencil._Enabled field instead of Stencil.Enabled
commit f352a80aec10c3faf2d84e0b35d59b4edc0395ef
Author: Brian Paul <[email protected]>
Date: Mon Mar 2 11:46:49 2009 -0700
mesa: add ctx->Stencil._Enabled field
Only true if stenciling is enabled, and there's a stencil buffer.
In addition to the FBO changes, there were a number of VBO changes.
This portion of the code is a bit more stable than the FBO code, but
it is still know to have problems. Should any of these changes be
pulled over?
commit c6bde8873fbda6d8467600b7491d8543c75b0509
Author: Brian Paul <[email protected]>
Date: Fri Feb 27 13:13:52 2009 -0700
intel: remove some unneeded buffer unmap calls
Core mesa now unmaps the buffers if needed in these cases.
commit 67025f789324163a69771436e852975d3acbcd86
Author: Brian Paul <[email protected]>
Date: Fri Feb 27 13:10:45 2009 -0700
mesa: set bufObj->Pointer = NULL after unmapping
Also, ctx->Driver.UnmapBuffer can never be null, so remove conditional.
commit a7f434b486187129ae8d5507170c42a9ce750258
Author: Brian Paul <[email protected]>
Date: Fri Feb 27 13:04:38 2009 -0700
mesa: if a buffer object is mapped when glDeleteBuffers() is called, unmap
it
commit 395bcad8c095e78621e7aca18af1dab71fe69813
Author: Brian Paul <[email protected]>
Date: Fri Feb 27 12:41:11 2009 -0700
mesa: updated comments
commit 75e3ccf6a5b639834bcda0ff6f9035b148fca8f1
Author: Brian Paul <[email protected]>
Date: Fri Feb 27 12:30:21 2009 -0700
mesa: fix incorrect error handling in glBufferDataARB()
If glBufferDataARB() is called while a buffer object is currently mapped
we're supposed to unmap the current buffer, then replace it. Don't generate
an error.
These two commits mention regressions. Regressions since when? Are
these bugs present in mesa_7_4_branch? Are they caused by commits
that I have cherry-picked?
commit e232471baa3a7cfa7a76b9cfe55dfb7da79fedbe
Author: Brian Paul <[email protected]>
Date: Mon Mar 2 15:19:53 2009 -0700
mesa: fix texture enable regression
Need to clear the _ReallyEnabled field before possibly continuing the loop.
Also, set _Current pointer to NULL if the unit is no longer enabled.
Fixes piglit lodbias regression
commit 7e0f47c1348273ed3bb39d00aa4a9498a28b8d6a
Author: Brian Paul <[email protected]>
Date: Mon Mar 2 14:27:16 2009 -0700
mesa: use _mesa_reference_texobj() when setting texUnit->Current pointer
Fixes piglit copytexsubimage regression.
These last few seem appropriate, but I'd like to hear someone speak up
for them. I would just bring the vec4_texp_rect fix over, but I don't
have a test case for it.
commit 20f49252e1fe2e72bb620c26292f33d5315452a1
Author: Brian Paul <[email protected]>
Date: Thu Mar 5 15:08:36 2009 -0700
i965: init dest reg CondMask = COND_TR (the proper default)
Plus fix up a debug printf.
commit ad2cfa41992d0676881440596c43ab6021c1b025
Author: Brian Paul <[email protected]>
Date: Fri Feb 20 11:42:28 2009 -0700
glsl: fix vec4_texp_rect IR code (need projective version)
commit f085147258713741895945dcb81fdb251bb6c9cc
Author: Eric Anholt <[email protected]>
Date: Thu Mar 5 08:25:22 2009 -0800
intel: Remove a gratuitous MI_FLUSH after clearing with a blit.
The 3D destination shares the same cache so we don't have any trouble with
the later commands needing the writes flushed inside of the same
batchbuffer.
pgpwgfAM99MwS.pgp
Description: PGP signature
------------------------------------------------------------------------------ Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are powering Web 2.0 with engaging, cross-platform capabilities. Quickly and easily build your RIAs with Flex Builder, the Eclipse(TM)based development software that enables intelligent coding and step-through debugging. Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com
_______________________________________________ Mesa3d-dev mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/mesa3d-dev
