If we don't reset ctx.vm_cnt/gpr_map/etc, this will spam a lot of
s_waitcnt instructions.
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17207>
By default, std::stack uses std::deque to allocate its elements, which has
poor cache efficiency. std::vector makes appending elements more expensive
(due to potential reallocations), but in the changed contexts the element
count should always be low anyway.
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11925>
clang generates a warning if there's no explicit break or fall-through
annotation. The latter would be kind of silly in this case, and not
robust against any future changes turning the fall-through invalid.
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Juan A. Suarez <jasuarez@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10220>
Instead of assuming WGP mode on GFX10+ in different places, add a member
to Program that can be used instead.
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8761>
When there are no param exports in an NGG (or legacy VS) shader,
the NO_PC_EXPORT=1 is set, which means PS waves can launch before
the current stage finishes.
If the current stage has any stores, we need to make sure to wait for
those before we allow PS waves to start, so that PS can read what
these instructions stored.
Fossil DB results on Navi 10:
Totals from 45 (0.03% of 136420) affected shaders:
CodeSize: 87224 -> 87404 (+0.21%)
Instrs: 16750 -> 16795 (+0.27%)
Cycles: 69580 -> 69760 (+0.26%)
VMEM: 8022 -> 8167 (+1.81%)
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7868>
Use auto to avoid mistyping the constness of the pair key, which
triggers implicit conversions rather than compilation errors.
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7346>
Iterate over maps by reference to avoid copies.
Replace find/insert with insert to avoid double search.
Use range-based for loop, avoiding copies by reference. Delete comment.
Erase by iterator instead of key to avoid repeat search.
Iterators unneeded to modify unwaited_instrs. Use range-based for loop.
Reviewed-by: Tony Wasserka <tony.wasserka@gmx.de>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7285>
Sounds useful for debugging missing wait-states and for improving
detection of the faulty instruction in case of memory violations.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6386>
If one VMEM instruction uses a sampler and the other doesn't, we can't do
this optimization.
Totals from 47 (0.04% of 127638) affected shaders:
CodeSize: 271744 -> 271656 (-0.03%); split: -0.04%, +0.01%
Instrs: 52783 -> 52761 (-0.04%); split: -0.05%, +0.01%
Cycles: 5547040 -> 5546952 (-0.00%); split: -0.00%, +0.00%
VMEM: 10022 -> 9887 (-1.35%)
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4949>
This isn't perfect (for example, changes might not be too meaningful when
comparing shaders with different control flow) but it should be useful for
evaluating scheduler changes.
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Acked-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/2965>
Clear the workgroup size for all supported shader stages.
Also, unify the workgroup size calculation accross various places.
As a result, insert_waitcnt can use the proper workgroup size
which means that some waits can be dropped from tessellation
shaders. Also, in cases where the previous calculation was wrong,
we now insert s_barrier instructions.
Totals from affected shaders (GFX10):
Code Size: 340116 -> 338484 (-0.48 %) bytes
Fixes: a8d15ab6da
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4165>
Seems required also on GFX8-GFX9 to achieve correct behaviour. This
is an undocumented behaviour but it makes real sense to me.
pipeline-db on GFX9:
Totals from affected shaders:
SGPRS: 1018 -> 1018 (0.00 %)
VGPRS: 516 -> 516 (0.00 %)
Code Size: 40516 -> 40636 (0.30 %) bytes
Max Waves: 280 -> 280 (0.00 %)
This fixes some sort of sun flickering with Assassins Creed Origins.
Closes: https://gitlab.freedesktop.org/mesa/mesa/issues/2488
Cc: <mesa-stable@lists.freedesktop.org>
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Tested-by: Marge Bot <https://gitlab.freedesktop.org/mesa/mesa/merge_requests/3750>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/merge_requests/3750>
For all VMEM instructions, the resource constant is now
in operands[0]. For MIMG instructions, the sampler shares
operands[1] with write data in case this instruction writes memory.
Moving the VADDR to be the last operand for MIMG is the first step to
support Navi NSA encoding.
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/merge_requests/3602>
v2: implement GFX10
v3: rebase
v7: rebase after shader args MR
v8: fix gs_vtx_offset usage on GFX9/GFX10
v8: use unreachable() instead of printing intrinsic
v8: rename output_state to ge_output_state
v8: fix formatting around nir_foreach_variable()
v8: rename some helpers in the scheduler
v8: rename p_memory_barrier_all to p_memory_barrier_common
v8: fix assertion comparing ctx.stage against vertex_geometry_gs
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/merge_requests/2421>
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Fixes: d1b9deee ('aco: improve waitcnt insertion around loops')
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Do this by repeating processing of loops until no progress is made.
Totals from affected shaders:
SGPRS: 162576 -> 162576 (0.00 %)
VGPRS: 145228 -> 145228 (0.00 %)
Spilled SGPRs: 668 -> 668 (0.00 %)
Spilled VGPRs: 0 -> 0 (0.00 %)
Private memory VGPRs: 0 -> 0 (0.00 %)
Scratch size: 0 -> 0 (0.00 %) dwords per thread
Code Size: 15778640 -> 15771336 (-0.05 %) bytes
LDS: 146 -> 146 (0.00 %) blocks
Max Waves: 6087 -> 6087 (0.00 %)
v2: use block_kind_loop_header/block_kind_loop_exit to repeat at the end
of loops instead of at each continue
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Currently if you have an SMEM store followed by an SMEM load that
loads the same location as was written, it won't work because the
store isn't finished before the load is executed. This is NOT
mitigated by an s_nop instruction on GFX10.
Since we currently don't have proper alias analysis, this commit adds
a workaround which will insert an s_waitcnt lgkmcnt(0) before each
SSBO load if they follow a store. We should further refine this in
the future when we can make sure to only add the wait when we load the
same thing as has been stored.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
