| Age | Commit message (Collapse) | Author |
|---|
|
Bug #25194.
|
|
Add a GLbitfield64 type and several macros to operate on 64-bit
fields. The OutputsWritten field of gl_program is changed to use that
type. This results in a fair amount of fallout in drivers that use
programs.
No changes are strictly necessary at this point as all bits used are
below the 32-bit boundary. Fairly soon several bits will be added for
clip distances written by a vertex shader. This will cause several
bits used for varyings to be pushed above the 32-bit boundary. This
will affect any drivers that support GLSL.
At this point, only the i965 driver has been modified to support this
eventuality.
I did this as a "squash" merge. There were several places through the
outputswritten64 branch where things were broken. I foresee this
causing difficulties later for bisecting. The history is still
available in the branch.
Conflicts:
src/mesa/drivers/dri/i965/brw_wm.h
|
|
|
|
This should fix TXB on G45 and older in the GLSL case.
|
|
New comments should explain some of the confusion about how this message
works.
|
|
|
|
|
|
Fixing this is a prereq for avoiding flagging all state at new
batch time. Eliminating that still causes problems, though (notably
glean logicOp fails on my GM965).
|
|
|
|
|
|
Saves CPU time, resulting in a 2.5% FPS win on ETQW.
|
|
|
|
They're the same regardless of execution width for 8, 4x2, and 16.
|
|
This is a 2.9% (+/-.3%) performance win for my GL demo, which hits MAD
sequences for matrix transforms.
|
|
The cube map array index arg is always present.
|
|
Instead of doing math based on the (broken for VBO && offset != 0)
input->count number, just use the BO size. Fixes assertion failure in ETQW.
|
|
This didn't work for quad/quadstrips at all, and for all other primitive types
it only worked when they were unclipped.
Fix up the former in gs stage (could probably do without these changes and
instead set QuadsFollowProvokingVertexConvention to false), and the rest in
clip stage.
|
|
Previously, we'd load linearly from ParameterValues[0] for the constants,
though ParameterValues[1] may not equal ParameterValues[0] + 4. Additionally,
the STATE_VAL type paramters didn't get updated.
Fixes piglit vp-constant-array-huge.vpfp and ET:QW object locations.
Bug #23226.
|
|
Fixes piglit vp-sge-alias test, and the googleearth ground shader. \o/
Bug #22228
(cherry picked from commit 56ab92bad8f1d05bc22b8a8471d5aeb663f220de)
|
|
Fixes piglit arl.vp.
(cherry picked from commit d52d78b4bcd6d4c0578f972c0b8ebac09e632196)
|
|
|
|
For an app that's blowing out the state cache, like sauerbraten, the
memset of the giant arrays ended up taking 11% of the CPU even when only a
"few" of the entries got used. With this, the WM program compile drops back
down to 1% of CPU time.
Bug #24981 (bisected to BRW_WM_MAX_INSN increase).
|
|
|
|
Fixes piglit vp-sge-alias test, and the googleearth ground shader. \o/
Bug #22228
|
|
Fixes piglit arl.vp.
|
|
No statistically significant performance difference at n=3 with either
openarena or my GL demo, but cutting program size seems like a good
thing to be doing for the hypothetical app that has a working set near
icache size.
|
|
|
|
This should fix issues with antialiased lines in GLSL.
|
|
The PINTERP code should be faster for brw_wm_glsl.c now since brw_wm_emit.c's
had been improved, and pixel_w should no longer stomp on a neighbor to dst.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
This drops support for get_src_reg_imm in these, but the prospect of getting
brw_wm_pass*.c onto our GLSL path is well worth some temporary pain.
|
|
This matches brw_wm_emit.c, which we'll be using shortly. There's a
possible penalty here in that we'll allocate registers for unused channels,
since we aren't doing ref tracking like brw_wm_pass*.c does. However, my
measurements on GM965 don't show any for either OA or UT2004 with the GLSL
path forced.
|
|
This keeps the individual state files from having to export their
structures for brw_state_cache initialization.
|
|
|
|
This should do all the things that MI_FLUSH did, but it can be pipelined
so that further rendering isn't blocked on the flush completion unless
necessary.
|
|
Before, if we just called glXMakeCurrent() and didn't render anything we'd
still trigger a flushFrontBuffer() call.
Now only set the intel->front_buffer_dirty field at state validation time
just before we draw something.
NOTE: additional calls to intel_check_front_buffer_rendering() might be
needed if I missed some rendering paths.
|
|
|
|
Part of fixing bug #24355.
|
|
Pass a gl_format to intel_create_renderbuffer() instead of GLenum.
Add cases for MESA_FORMAT_XRGB8888 textures and renderbuffers.
However, we don't yet create any renderbuffers or textures with that
format. It seems the default alpha value is zero instead of one.
Need to investigate that first.
|
|
|
|
Shrinks size of key to 8 bytes from 12.
Note that progs/demos/spriteblast.c is still broken.
|
|
GLushort is big enough for the swizzle and origin fields.
The key could probably be made smaller still by re-ordering things.
I'll hold off on that until after the outputswritten64 branch is merged.
The key will get a little larger again with the GLbitfield64 fields.
|
|
If the fragment shader doesn't use FRAG_ATTRIB_WPOS (gl_FragCoord) we
don't need to worry about the window size and origin in
brw_wm_populate_key().
This avoids re-generating the i965 shader code when a window is resized.
Issue spotted by Keith Whitwell.
|
|
i965 might support more than 4 color draw buffers. But if not, this protects
from breakage if the Mesa limit is raised.
|
|
|
