KonstantinSeurer
/
mesa
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

intel/nir: Move 64-bit lowering later 

Now that we have a loop unrolling cost function and loop unrolling isn't
going to kill us the moment we have a 64-bit op in a loop, we can go
ahead and move 64-bit lowering later.  This gives us the opportunity to
do more optimizations and actually let the full optimizer run even on
64-bit ops rather than hoping one round of opt_algebraic will fix
everything.  This substantially reduces both fp64 shader compile times
and the resulting code size.  On the vs-isnan-dvec test from piglit:

Before this commit:

    1684.63s user 17.29s system 99% cpu 28:28.24 total
    101479 instructions. 0 loops. 802452 cycles. 79:369 spills:fills.
    Peak memory usage (according to massif): 1.435 GB

After this commit:

    179.64s user 7.75s system 99% cpu 3:07.92 total
    57316 instructions. 0 loops. 459287 cycles. 0:0 spills:fills.
    Peak memory usage (according to massif): 531.0 MB

Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
This commit is contained in:
Jason Ekstrand 2019-03-04 16:11:57 -06:00 committed by Jason Ekstrand
parent e02959f442
commit 656ace3dd8
1 changed files with 13 additions and 21 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

34
src/intel/compiler/brw_nir.c
View File

@ -665,27 +665,6 @@ brw_preprocess_nir(const struct brw_compiler *compiler, nir_shader *nir,
OPT(nir_lower_alu_to_scalar);
}
/* Run opt_algebraic before int64 lowering so we can hopefully get rid
* of some int64 instructions.
*/
OPT(nir_opt_algebraic);
/* Lower 64-bit operations before nir_optimize so that loop unrolling sees
* their actual cost.
*/
bool lowered_64bit_ops = false;
do {
progress = false;
OPT(nir_lower_int64, nir->options->lower_int64_options);
OPT(nir_lower_doubles, softfp64, nir->options->lower_doubles_options);
/* Necessary to lower add -> sub and div -> mul/rcp */
OPT(nir_opt_algebraic);
lowered_64bit_ops |= progress;
} while (progress);
if (nir->info.stage == MESA_SHADER_GEOMETRY)
OPT(nir_lower_gs_intrinsics);
@ -714,6 +693,19 @@ brw_preprocess_nir(const struct brw_compiler *compiler, nir_shader *nir,
nir = brw_nir_optimize(nir, compiler, is_scalar, true);
bool lowered_64bit_ops = false;
do {
progress = false;
OPT(nir_lower_int64, nir->options->lower_int64_options);
OPT(nir_lower_doubles, softfp64, nir->options->lower_doubles_options);
/* Necessary to lower add -> sub and div -> mul/rcp */
OPT(nir_opt_algebraic);
lowered_64bit_ops |= progress;
} while (progress);
/* This needs to be run after the first optimization pass but before we
* lower indirect derefs away
*/