mirror of
https://gitlab.freedesktop.org/mesa/mesa.git
synced 2026-05-04 22:49:13 +02:00
intel/compiler: use nir_shader_instructions_pass in brw_nir_opt_peephole_ffma
No functional changes. Signed-off-by: Marcin Ślusarz <marcin.slusarz@intel.com> Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13189>
This commit is contained in:
Marcin Ślusarz
Marge Bot
parent
9e6acd801d
commit
71bec85db0
1 changed files with 80 additions and 111 deletions
|
|
@ -163,138 +163,107 @@ any_alu_src_is_a_constant(nir_alu_src srcs[])
|
|||
}
|
||||
|
||||
static bool
|
||||
brw_nir_opt_peephole_ffma_block(nir_builder *b, nir_block *block)
|
||||
brw_nir_opt_peephole_ffma_instr(nir_builder *b,
|
||||
nir_instr *instr,
|
||||
UNUSED void *cb_data)
|
||||
{
|
||||
bool progress = false;
|
||||
if (instr->type != nir_instr_type_alu)
|
||||
return false;
|
||||
|
||||
nir_foreach_instr_safe(instr, block) {
|
||||
if (instr->type != nir_instr_type_alu)
|
||||
continue;
|
||||
nir_alu_instr *add = nir_instr_as_alu(instr);
|
||||
if (add->op != nir_op_fadd)
|
||||
return false;
|
||||
|
||||
nir_alu_instr *add = nir_instr_as_alu(instr);
|
||||
if (add->op != nir_op_fadd)
|
||||
continue;
|
||||
assert(add->dest.dest.is_ssa);
|
||||
if (add->exact)
|
||||
return false;
|
||||
|
||||
assert(add->dest.dest.is_ssa);
|
||||
if (add->exact)
|
||||
continue;
|
||||
assert(add->src[0].src.is_ssa && add->src[1].src.is_ssa);
|
||||
|
||||
assert(add->src[0].src.is_ssa && add->src[1].src.is_ssa);
|
||||
/* This, is the case a + a. We would rather handle this with an
|
||||
* algebraic reduction than fuse it. Also, we want to only fuse
|
||||
* things where the multiply is used only once and, in this case,
|
||||
* it would be used twice by the same instruction.
|
||||
*/
|
||||
if (add->src[0].src.ssa == add->src[1].src.ssa)
|
||||
return false;
|
||||
|
||||
/* This, is the case a + a. We would rather handle this with an
|
||||
* algebraic reduction than fuse it. Also, we want to only fuse
|
||||
* things where the multiply is used only once and, in this case,
|
||||
* it would be used twice by the same instruction.
|
||||
*/
|
||||
if (add->src[0].src.ssa == add->src[1].src.ssa)
|
||||
continue;
|
||||
nir_alu_instr *mul;
|
||||
uint8_t add_mul_src, swizzle[4];
|
||||
bool negate, abs;
|
||||
for (add_mul_src = 0; add_mul_src < 2; add_mul_src++) {
|
||||
for (unsigned i = 0; i < 4; i++)
|
||||
swizzle[i] = i;
|
||||
|
||||
nir_alu_instr *mul;
|
||||
uint8_t add_mul_src, swizzle[4];
|
||||
bool negate, abs;
|
||||
for (add_mul_src = 0; add_mul_src < 2; add_mul_src++) {
|
||||
for (unsigned i = 0; i < 4; i++)
|
||||
swizzle[i] = i;
|
||||
negate = false;
|
||||
abs = false;
|
||||
|
||||
negate = false;
|
||||
abs = false;
|
||||
mul = get_mul_for_src(&add->src[add_mul_src],
|
||||
add->dest.dest.ssa.num_components,
|
||||
swizzle, &negate, &abs);
|
||||
|
||||
mul = get_mul_for_src(&add->src[add_mul_src],
|
||||
add->dest.dest.ssa.num_components,
|
||||
swizzle, &negate, &abs);
|
||||
|
||||
if (mul != NULL)
|
||||
break;
|
||||
}
|
||||
|
||||
if (mul == NULL)
|
||||
continue;
|
||||
|
||||
unsigned bit_size = add->dest.dest.ssa.bit_size;
|
||||
|
||||
nir_ssa_def *mul_src[2];
|
||||
mul_src[0] = mul->src[0].src.ssa;
|
||||
mul_src[1] = mul->src[1].src.ssa;
|
||||
|
||||
/* If any of the operands of the fmul and any of the fadd is a constant,
|
||||
* we bypass because it will be more efficient as the constants will be
|
||||
* propagated as operands, potentially saving two load_const instructions.
|
||||
*/
|
||||
if (any_alu_src_is_a_constant(mul->src) &&
|
||||
any_alu_src_is_a_constant(add->src)) {
|
||||
continue;
|
||||
}
|
||||
|
||||
b->cursor = nir_before_instr(&add->instr);
|
||||
|
||||
if (abs) {
|
||||
for (unsigned i = 0; i < 2; i++)
|
||||
mul_src[i] = nir_fabs(b, mul_src[i]);
|
||||
}
|
||||
|
||||
if (negate)
|
||||
mul_src[0] = nir_fneg(b, mul_src[0]);
|
||||
|
||||
nir_alu_instr *ffma = nir_alu_instr_create(b->shader, nir_op_ffma);
|
||||
ffma->dest.saturate = add->dest.saturate;
|
||||
ffma->dest.write_mask = add->dest.write_mask;
|
||||
|
||||
for (unsigned i = 0; i < 2; i++) {
|
||||
ffma->src[i].src = nir_src_for_ssa(mul_src[i]);
|
||||
for (unsigned j = 0; j < add->dest.dest.ssa.num_components; j++)
|
||||
ffma->src[i].swizzle[j] = mul->src[i].swizzle[swizzle[j]];
|
||||
}
|
||||
nir_alu_src_copy(&ffma->src[2], &add->src[1 - add_mul_src]);
|
||||
|
||||
assert(add->dest.dest.is_ssa);
|
||||
|
||||
nir_ssa_dest_init(&ffma->instr, &ffma->dest.dest,
|
||||
add->dest.dest.ssa.num_components,
|
||||
bit_size, NULL);
|
||||
nir_ssa_def_rewrite_uses(&add->dest.dest.ssa,
|
||||
&ffma->dest.dest.ssa);
|
||||
|
||||
nir_builder_instr_insert(b, &ffma->instr);
|
||||
assert(list_is_empty(&add->dest.dest.ssa.uses));
|
||||
nir_instr_remove(&add->instr);
|
||||
|
||||
progress = true;
|
||||
if (mul != NULL)
|
||||
break;
|
||||
}
|
||||
|
||||
return progress;
|
||||
}
|
||||
if (mul == NULL)
|
||||
return false;
|
||||
|
||||
static bool
|
||||
brw_nir_opt_peephole_ffma_impl(nir_function_impl *impl)
|
||||
{
|
||||
bool progress = false;
|
||||
unsigned bit_size = add->dest.dest.ssa.bit_size;
|
||||
|
||||
nir_builder builder;
|
||||
nir_builder_init(&builder, impl);
|
||||
nir_ssa_def *mul_src[2];
|
||||
mul_src[0] = mul->src[0].src.ssa;
|
||||
mul_src[1] = mul->src[1].src.ssa;
|
||||
|
||||
nir_foreach_block(block, impl) {
|
||||
progress |= brw_nir_opt_peephole_ffma_block(&builder, block);
|
||||
/* If any of the operands of the fmul and any of the fadd is a constant,
|
||||
* we bypass because it will be more efficient as the constants will be
|
||||
* propagated as operands, potentially saving two load_const instructions.
|
||||
*/
|
||||
if (any_alu_src_is_a_constant(mul->src) &&
|
||||
any_alu_src_is_a_constant(add->src)) {
|
||||
return false;
|
||||
}
|
||||
|
||||
if (progress) {
|
||||
nir_metadata_preserve(impl, nir_metadata_block_index |
|
||||
nir_metadata_dominance);
|
||||
} else {
|
||||
nir_metadata_preserve(impl, nir_metadata_all);
|
||||
b->cursor = nir_before_instr(&add->instr);
|
||||
|
||||
if (abs) {
|
||||
for (unsigned i = 0; i < 2; i++)
|
||||
mul_src[i] = nir_fabs(b, mul_src[i]);
|
||||
}
|
||||
|
||||
return progress;
|
||||
if (negate)
|
||||
mul_src[0] = nir_fneg(b, mul_src[0]);
|
||||
|
||||
nir_alu_instr *ffma = nir_alu_instr_create(b->shader, nir_op_ffma);
|
||||
ffma->dest.saturate = add->dest.saturate;
|
||||
ffma->dest.write_mask = add->dest.write_mask;
|
||||
|
||||
for (unsigned i = 0; i < 2; i++) {
|
||||
ffma->src[i].src = nir_src_for_ssa(mul_src[i]);
|
||||
for (unsigned j = 0; j < add->dest.dest.ssa.num_components; j++)
|
||||
ffma->src[i].swizzle[j] = mul->src[i].swizzle[swizzle[j]];
|
||||
}
|
||||
nir_alu_src_copy(&ffma->src[2], &add->src[1 - add_mul_src]);
|
||||
|
||||
assert(add->dest.dest.is_ssa);
|
||||
|
||||
nir_ssa_dest_init(&ffma->instr, &ffma->dest.dest,
|
||||
add->dest.dest.ssa.num_components,
|
||||
bit_size, NULL);
|
||||
nir_ssa_def_rewrite_uses(&add->dest.dest.ssa, &ffma->dest.dest.ssa);
|
||||
|
||||
nir_builder_instr_insert(b, &ffma->instr);
|
||||
assert(list_is_empty(&add->dest.dest.ssa.uses));
|
||||
nir_instr_remove(&add->instr);
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
bool
|
||||
brw_nir_opt_peephole_ffma(nir_shader *shader)
|
||||
{
|
||||
bool progress = false;
|
||||
|
||||
nir_foreach_function(function, shader) {
|
||||
if (function->impl)
|
||||
progress |= brw_nir_opt_peephole_ffma_impl(function->impl);
|
||||
}
|
||||
|
||||
return progress;
|
||||
return nir_shader_instructions_pass(shader, brw_nir_opt_peephole_ffma_instr,
|
||||
nir_metadata_block_index |
|
||||
nir_metadata_dominance,
|
||||
NULL);
|
||||
}
|
||||
|
|
|
|||
Loading…
Add table
Reference in a new issue