OpenHarmony/third_party_mesa3d
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

intel/fs: Update SSBO & shared uniform block loads for Xe2

Browse Source 
Note: lower_lsc_block_logical_send() most likely stills needs some
related updates.

Ref: a358b97c58 ("intel/fs: optimize uniform SSBO & shared loads")
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Caio Oliveira <caio.oliveira@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25020>
This commit is contained in:
Jordan Justen
2023-04-13 15:55:07 -07:00
parent 9fb2b12c99
commit f1b9b7f955
1 changed files with 4 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
src/intel/compiler/brw_fs_nir.cpp
View File

@@ -5294,7 +5294,8 @@ fs_visitor::nir_emit_intrinsic(const fs_builder &bld, nir_intrinsic_instr *instr
srcs[SURFACE_LOGICAL_SRC_SURFACE] = fs_reg(brw_imm_ud(GFX7_BTI_SLM)); srcs[SURFACE_LOGICAL_SRC_SURFACE] = fs_reg(brw_imm_ud(GFX7_BTI_SLM));
} }
const unsigned total_dwords = ALIGN(instr->num_components, REG_SIZE / 4); const unsigned total_dwords = ALIGN(instr->num_components,
REG_SIZE * reg_unit(devinfo) / 4);
unsigned loaded_dwords = 0; unsigned loaded_dwords = 0;
const fs_builder ubld1 = bld.exec_all().group(1, 0); const fs_builder ubld1 = bld.exec_all().group(1, 0);
@@ -5325,7 +5326,8 @@ fs_visitor::nir_emit_intrinsic(const fs_builder &bld, nir_intrinsic_instr *instr
const fs_builder &ubld = block <= 8 ? ubld8 : ubld16; const fs_builder &ubld = block <= 8 ? ubld8 : ubld16;
ubld.emit(SHADER_OPCODE_UNALIGNED_OWORD_BLOCK_READ_LOGICAL, ubld.emit(SHADER_OPCODE_UNALIGNED_OWORD_BLOCK_READ_LOGICAL,
retype(byte_offset(packed_consts, loaded_dwords * 4), BRW_REGISTER_TYPE_UD), retype(byte_offset(packed_consts, loaded_dwords * 4), BRW_REGISTER_TYPE_UD),
srcs, SURFACE_LOGICAL_NUM_SRCS)->size_written = align(block_bytes, REG_SIZE); srcs, SURFACE_LOGICAL_NUM_SRCS)->size_written =
align(block_bytes, REG_SIZE * reg_unit(devinfo));
loaded_dwords += block; loaded_dwords += block;