Clear inherits the render targets from the current render pass. This
means we need to fill out the binding table after switching to meta
bindings. However, meta copies etc happen outside a render pass and
break when we try to fill in the render targets. This change fills the
render targets only for meta clear.
This just created extra upkeep and the push to move extern
C's into mesa code would mean a large number of extern's
in core Mesa driver interfaces. The Haiku Gallium renderers
are mostly insulated via the C-based Haiku state tracker.
As any future hardware support in Haiku will be gallium
based, lets just drop swrast.
Haiku has a Mesa 7.12 fork for gcc2 that uses swrast.
This commit fixes the last of the Haiku build issues.
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
We leave the block pool untracked so that reads/writes to freed blocks
will get caught and do the tracking at the state pool/stream level. We
have to do a few extra gymnastics for streams because valgrind works in
terms of poitners and we work in terms of separate map and offset.
Fortunately, the users of the state pool and stream should always be using
the map pointer provided in the anv_state structure. We just have to
track, per block, the map that was used when we initially got the block.
Then we can make sure we always use that map and valgrind should stay
happy.
GLSL IR vs. NIR shader-db results for SIMD8 vertex shaders on Broadwell:
total instructions in shared programs: 2742062 -> 2681339 (-2.21%)
instructions in affected programs: 1514770 -> 1454047 (-4.01%)
helped: 5813
HURT: 1120
The gained programs are ARB vertext programs that were previously going
through the vec4 backend. Now that we have prog_to_nir, ARB vertex
programs can go through the scalar backend so they show up as "gained" in
the shader-db results.
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Acked-by: Matt Turner <mattst88@gmail.com>
A fence can outlive the ctx, so we shouldn't deref the ctx to get at the
screen. We need some updates in libdrm_freedreno API to completely
handle fences properly, but this is at least an improvement.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Gen9 surface state is very similar to the previous generation. The important
changes here are aux mode, and the way clear colors work.
NOTE: There are some things intentionally left out of this decoding.
v2: Redo the string for the aux buffer type to address compressed variants.
v3: Use the shift for compression enable (instead of compression mode) (Topi)
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
0x00007da0: 0xc1da740e: SF_CLIP VP: guardband xmin = -27.306667
0x00007da4: 0x41da740e: SF_CLIP VP: guardband xmax = 27.306667
0x00007da4: 0x41da740e: SF_CLIP VP: guardband ymin = -23.405714
0x00007da8: 0xc1bb3ee7: SF_CLIP VP: guardband ymax = 23.405714
0x00007db0: 0x00000000: SF_CLIP VP: Min extents: 0.00x0.00
0x00007db8: 0x00000000: SF_CLIP VP: Max extents: 299.00x349.00
While here, fix the wrong offsets for the guardband (I didn't check if it used
to be valid on GEN4).
v2: Remove leftover GET_BITS which belongs later in the series. (Topi)
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
It's true that not all surfaces apply for every gen, but for the most part this
is what we want. (The unfortunate case is when we use a valid surface, but not
for the specific GEN).
This was automated with a vim macro.
v2: Shortened common forms such as R8G8B8A8->RGBA8. Note that this makes some of
the sample output in subsequent commits slightly incorrect.
v3: Use the name from the table (Ken). This requires declaring the surface
format array as extern, and declaring the struct in the .h file.
v4: Move the struct back and create a helper function to obtain the name (Ken)
Get rid of the now useless helper in the state_dump.c
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com> (v3)
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
We now always copy the entire struct unless pData is NULL and
unconditionally write back the struct size. It's not clear this is
useful if the structs may grow over time, but it seems to be the
expected behaviour for now.
Ivybridge and Baytrail can't use mach with 2Q quarter control, so just
do it without the accumulator. Stupid accumulator.
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
The next commit uses an add(16) with a UW destination with a stride of
2, which needs compression control since it's writing two registers. The
old code would have failed to set compression control correctly.
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Gen8+'s MUL instruction doesn't ignore the high 16-bits of one source
like on earlier platforms, so we can constant propagate into it without
worry. Integer multiplies (not into the accumulator, which is done for
imul_high) are lowered in lower_integer_multiplication(), so it's safe
there as well.
On Broadwell, fragment shaders only:
total instructions in shared programs: 4377769 -> 4377451 (-0.01%)
instructions in affected programs: 48064 -> 47746 (-0.66%)
helped: 156
On Broadwell, vertex shaders only:
total instructions in shared programs: 2858885 -> 2856313 (-0.09%)
instructions in affected programs: 26380 -> 23808 (-9.75%)
helped: 134
On Broadwell, vertex shaders only (with INTEL_USE_NIR=1):
total instructions in shared programs: 2911688 -> 2865984 (-1.57%)
instructions in affected programs: 1421715 -> 1376011 (-3.21%)
helped: 6186
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
32-bit x 32-bit integer multiplication requires multiple instructions
until Broadwell. This patch just lets us treat the MUL instruction in
the FS backend like it operates on Broadwell, and after optimizations
we lower it into a sequence of instructions on older platforms.
Doing this will allow us to some extra optimization on integer
multiplies.
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Patch marks uniforms inside UBO properly referenced by stages.
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Samuel Iglesias Gonsalvez <siglesias@igalia.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90397
Currently, when the MinFilter is GL_LINEAR or GL_NEAREST we hide the
actual miplevel count from the hardware (and we avoid re-creating
the miptree structure with all the levels), since we don't expect
levels other than the base level to be needed. Unfortunately,
GLSL's textureSize() function is an exception to this rule. This
function takes a lod parameter that we need to use to return the
size of the appropriate miplevel (if it exists). The spec only
requires that the miplevel exists, so even if the sampler is
configured with a linear or nearest MinFilter, as far as the user
has uploaded miplevels for the texture, textureSize() should return
the appropriate sizes.
This patch fixes this by exposing the actual miplevel count for all
sampling engine textures while keeping the original implementation
for render targets (for render targets textures we do not provide
the miplevel count but the actual LOD we are wrting to, so we
want to make sure that we make this the base level).
Fixes 28 dEQP tests in the following category:
dEQP-GLES3.functional.shaders.texture_functions.texturesize.*
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
