fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2025-12-20 11:40:10 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions 2

aco/optimizer: use new helpers for v_and opt

Browse source 
Foz-DB Navi48:
Totals from 465 (0.56% of 82419) affected shaders:
Instrs: 372721 -> 372083 (-0.17%); split: -0.18%, +0.01%
CodeSize: 2004568 -> 2003332 (-0.06%)
Latency: 3664162 -> 3660745 (-0.09%); split: -0.10%, +0.00%
InvThroughput: 892042 -> 890994 (-0.12%); split: -0.12%, +0.01%
Copies: 35552 -> 35549 (-0.01%)
VALU: 171781 -> 171333 (-0.26%); split: -0.28%, +0.02%
SALU: 87946 -> 87949 (+0.00%)
VOPD: 48 -> 49 (+2.08%)

Foz-DB Navi21:
Totals from 191 (0.23% of 82387) affected shaders:
Instrs: 139340 -> 139178 (-0.12%); split: -0.13%, +0.02%
CodeSize: 798660 -> 798284 (-0.05%)
Latency: 1672750 -> 1673194 (+0.03%); split: -0.06%, +0.08%
InvThroughput: 634847 -> 634651 (-0.03%); split: -0.06%, +0.03%
Copies: 16372 -> 16366 (-0.04%); split: -0.04%, +0.01%
VALU: 79668 -> 79506 (-0.20%); split: -0.23%, +0.03%
SALU: 38233 -> 38236 (+0.01%)

Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38530>
This commit is contained in:
Georg Lehmann 2025-03-27 15:51:41 +01:00 • committed by Marge Bot
parent 7bc6d8e2ad
commit adc55b1a1e
1 changed files with 3 additions and 48 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

51
src/amd/compiler/aco_optimizer.cpp
View file

@ -4078,52 +4078,6 @@ apply_load_extract(opt_ctx& ctx, aco_ptr<Instruction>& extract)
return true; return true;
} }
/* v_and(a, not(b)) -> v_bfi_b32(b, 0, a)
* v_or(a, not(b)) -> v_bfi_b32(b, a, -1)
*/
bool
combine_v_andor_not(opt_ctx& ctx, aco_ptr<Instruction>& instr)
{
if (instr->usesModifiers())
return false;
for (unsigned i = 0; i < 2; i++) {
Instruction* op_instr = follow_operand(ctx, instr->operands[i], true);
if (op_instr && !op_instr->usesModifiers() &&
(op_instr->opcode == aco_opcode::v_not_b32 ||
op_instr->opcode == aco_opcode::s_not_b32)) {
Operand ops[3] = {
op_instr->operands[0],
Operand::zero(),
instr->operands[!i],
};
if (instr->opcode == aco_opcode::v_or_b32) {
ops[1] = instr->operands[!i];
ops[2] = Operand::c32(-1);
}
if (!check_vop3_operands(ctx, 3, ops))
continue;
Instruction* new_instr = create_instruction(aco_opcode::v_bfi_b32, Format::VOP3, 3, 1);
if (op_instr->operands[0].isTemp())
ctx.uses[op_instr->operands[0].tempId()]++;
for (unsigned j = 0; j < 3; j++)
new_instr->operands[j] = ops[j];
new_instr->definitions[0] = instr->definitions[0];
new_instr->pass_flags = instr->pass_flags;
instr.reset(new_instr);
decrease_and_dce(ctx, op_instr->definitions[0].getTemp());
ctx.info[instr->definitions[0].tempId()].label = 0;
ctx.info[instr->definitions[0].tempId()].parent_instr = instr.get();
return true;
}
}
return false;
}
/* v_add_co(c, s_lshl(a, b)) -> v_mad_u32_u24(a, 1<<b, c) /* v_add_co(c, s_lshl(a, b)) -> v_mad_u32_u24(a, 1<<b, c)
* v_add_co(c, v_lshlrev(a, b)) -> v_mad_u32_u24(b, 1<<a, c) * v_add_co(c, v_lshlrev(a, b)) -> v_mad_u32_u24(b, 1<<a, c)
* v_sub(c, s_lshl(a, b)) -> v_mad_i32_i24(a, -(1<<b), c) * v_sub(c, s_lshl(a, b)) -> v_mad_i32_i24(a, -(1<<b), c)
@ -4602,8 +4556,6 @@ combine_instruction(opt_ctx& ctx, aco_ptr<Instruction>& instr)
combine_salu_n2(ctx, instr); combine_salu_n2(ctx, instr);
} else if (instr->opcode == aco_opcode::s_abs_i32) { } else if (instr->opcode == aco_opcode::s_abs_i32) {
combine_sabsdiff(ctx, instr); combine_sabsdiff(ctx, instr);
} else if (instr->opcode == aco_opcode::v_and_b32) {
combine_v_andor_not(ctx, instr);
} }
alu_opt_info info; alu_opt_info info;
@ -4819,6 +4771,9 @@ combine_instruction(opt_ctx& ctx, aco_ptr<Instruction>& instr)
add_opt(v_add_u32, v_add_lshl_u32, 0x2, "120", nullptr, true); add_opt(v_add_u32, v_add_lshl_u32, 0x2, "120", nullptr, true);
add_opt(s_add_u32, v_add_lshl_u32, 0x2, "120", nullptr, true); add_opt(s_add_u32, v_add_lshl_u32, 0x2, "120", nullptr, true);
add_opt(s_add_i32, v_add_lshl_u32, 0x2, "120", nullptr, true); add_opt(s_add_i32, v_add_lshl_u32, 0x2, "120", nullptr, true);
} else if (info.opcode == aco_opcode::v_and_b32) {
add_opt(v_not_b32, v_bfi_b32, 0x3, "10", insert_const_cb<1, 0>, true);
add_opt(s_not_b32, v_bfi_b32, 0x3, "10", insert_const_cb<1, 0>, true);
} }
if (match_and_apply_patterns(ctx, info, patterns)) { if (match_and_apply_patterns(ctx, info, patterns)) {