It was changed from 0 to allow shader outputs at 0 that are
different from position.
Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
One of the assertion made no sense for buffer rendertargets
(due to the union), so drop it. (The same assertion is present already in
the path for texture surfaces later.).
v2: make assertion completely accurate (suggested by Jose).
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
The overallocation was very bad especially for things like 1d array
textures which got blown up by a factor of 64. (Even ordinary smallish
2d textures benefit a lot from this, a mipmapped 64x64 rgba8 texture
previously used 7*16kB = 112kB instead of now ~22kB.)
4x4 is chosen because this is the size the jit functions run on, so
making it smaller is going to be a bit more complicated.
It is actually not strictly 4x4 pixel, since we'd want to avoid situations
where different threads are rendering to the same cacheline so we keep
cacheline size alignment in x direction (often 64bytes).
To make this work introduce new task width/height parameters and make
sure clears don't clear the whole tile if it's a partial tile. Likewise,
the rasterizer may produce fragments outside the 4x4 blocks present in a
tile, so don't call the jit function for them.
This does not yet fix rendering to buffers (which cannot have any y
alignment at all), and 1d/1d array textures are still overallocated by a
factor of 4.
v2: replace magic number 4 with LP_RASTER_BLOCK_SIZE, fix size of buffers
allocated (needed in case we render to them).
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
These were mostly just a waste of memory and cache pressure, and were
really only used for debugging.
This change reduces instruction count (as measured by callgrind's Ir
event) of gnome-shell-perf-tool on Ivybridge by 3.5% ± 0.015% (n=20).
Signed-off-by: Adam Jackson <ajax@redhat.com>
Remove hash function on shader variants. Nature of variants limits them to a
small number and thus its more efficient to just do a memory compare of the
actual shader structures rather than compute and compare hashes.
We need to clamp to make sure invalid shader doesn't crash our
driver. The spec says to return 0-th index for everything that's
out of bounds.
Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: José Fonseca<jfonseca@vmware.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
draw_find_shader_output like most of the code in draw used to
depend on position always being at output slot 0. which meant
that any other attribute being at 0 could signify an error.
unfortunately position can be at any of the output slots, thus
other attributes can occupy slot 0 and we need to mark the ones
which were not found by something else. This commit changes
draw_find_shader_output so that it returns -1 if it can't
find the given attribute and adjust the code that depended
on it returning >0 whenever it correctly found an attrib.
Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: José Fonseca<jfonseca@vmware.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Largely related to making sure the rasterizer can correctly
pick out the correct scissor box for the current viewport.
Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: José Fonseca<jfonseca@vmware.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Gallium supported only a single viewport/scissor combination. This
commit changes the interface to allow us to add support for multiple
viewports/scissors.
Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Reviewed-by: José Fonseca<jfonseca@vmware.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Eliminate the rest of the no longer needed layout logic.
(It is possible some code could be simplified a bit further still.)
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
Set up CB_SHADER_MASK register according to pixel shader exports, and enable
some minimal state for colour buffer 1 in case dual source blending is used.
This patch improves handling of unconditional KILL instructions inside
the conditional blocks, uncovering more opportunities for if-conversion.
Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
PRED_SET instructions that update exec mask should be scheduled immediately
prior to the "if-then-else" block, because any instruction that is
inserted after alu clause with PRED_SET and before conditional block is
also conditionally executed by hw (exec mask is already updated at that
moment).
Propbably it's better to make PRED_SET a part of conditional
"if-then-else" block in the IR to handle this more cleanly,
but for now this temporary solution should prevent the problem.
Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
For compute shaders we need to let the backend know that
GPRs 0 and 1 are preloaded with some compute-specific input
values, otherwise any use of these regs without previous
definition is considered as undefined value and usually
is simply replaced with 0.
Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
This allows GVN rewrite pass to propagate non-const (register)
values to FETCH source operands, helping to eliminate unnecessary
copies in some cases.
Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
We have to assume that all GPRs in compute shader can be indirectly
addressed because LLVM backend doesn't provide any indirect array info.
That's why for compute shaders GPR array is created that covers all used
GPRs (0..r600_bytecode::ngpr-1), but this seriously restricts register
allocation in sb.
This patch checks for actual use of indirect access in the shader and
if it's not used then GPR array is not created, so that regalloc is not
unnecessarily restricted.
Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
need_flush was uninitialized if hw3d->new_batch was true.
Fixes "Uninitialized scalar variable" defect reported by Coverity.
Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
'type' was not fully initialized when calling lp_build_context_init.
Fixes "Uninitialized scalar variable" defect reported by Coverity.
NOTE: This is a candidate for the stable branches.
Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Initially we had NUM_TEX_TILE_ENTRIES of 50, however this was using too much
memory (mostly because the tile cache is operating on fixed max current
sampler views which could be fixed but that's another topic). So it was
decreased to 4. However this is a ridiculously low number which can't
actually really work (the number of tiles needed for as little as
a single quad with linear_mipmap_linear is 2 to 8 for a 2d texture, and
4 to 16 for a 3d texture), as it just about guarantees there will be
cache thrashing sometimes (just about always for 3d textures in fact, since
while there are 4 entries the cache is direct mapped).
So increase that number to 16 (which is still on the low side for direct
mapped cache though I guess using something like 4-way associativity would
be more effective than increasing this further) which has at least some good
chance to avoid thrashing. Since we don't want to increase memory requirements
however in turn decrease the tile size accordingly from 64 to 32 (as a bonus
point this also decreases the cost of texture thrashing which might still
happen sometimes).
I've seen performance improvement in the order of factor ~200 (specifically,
drawing the first frame from the replay from bug 41787 needs "only" ~10s
instead of ~30min, meaning I can actually compare the output with other
drivers...) with this.
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
