diff options
| author | Rob Clark <[email protected]> | 2019-12-13 14:09:39 -0800 |
|---|---|---|
| committer | Marge Bot <[email protected]> | 2020-02-01 02:40:22 +0000 |
| commit | 0f78c32492ed096649b015a4967d6d56c18dd14a (patch) | |
| tree | 1912b19e653107713582762f3116fa8992a7eb73 /src/freedreno/ir3/ir3.h | |
| parent | 3369406e44b0226295e7475e189da2e42efd7f22 (diff) | |
freedreno/ir3: post-RA sched pass
After RA, we can schedule to increase parallelism (reduce nop's) without
worrying about increasing register pressure. This pass lets us cut down
the instruction count ~10%, and prioritize bary.f, kill, etc, which
would tend to increase register pressure if we tried to do that before
RA.
It should be more useful if RA round-robin'd register choices.
Signed-off-by: Rob Clark <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/merge_requests/3569>
Diffstat (limited to 'src/freedreno/ir3/ir3.h')
| -rw-r--r-- | src/freedreno/ir3/ir3.h | 3 |
1 files changed, 3 insertions, 0 deletions
diff --git a/src/freedreno/ir3/ir3.h b/src/freedreno/ir3/ir3.h index b10f2f0da79..f78a7d6f364 100644 --- a/src/freedreno/ir3/ir3.h +++ b/src/freedreno/ir3/ir3.h @@ -1138,6 +1138,9 @@ void ir3_sun(struct ir3 *ir); void ir3_sched_add_deps(struct ir3 *ir); int ir3_sched(struct ir3 *ir); +struct ir3_context; +int ir3_postsched(struct ir3_context *ctx); + bool ir3_a6xx_fixup_atomic_dests(struct ir3 *ir, struct ir3_shader_variant *so); /* register assignment: */ |