summaryrefslogtreecommitdiffstats
path: root/src/amd/compiler
Commit message (Expand)AuthorAgeFilesLines
* aco: Introduce vgpr_limit to keep track of available VGPRs.Timur Kristóf2019-10-286-5/+12
* aco: Implement subgroup shuffle in GFX10 wave64 mode.Timur Kristóf2019-10-286-16/+113
* aco: Remove dead code in reduction lowering.Rhys Perry2019-10-281-16/+14
* aco: Fix reductions on GFX10.Rhys Perry2019-10-283-18/+95
* aco: Refactor hazard mitigations, separate pass for GFX10.Timur Kristóf2019-10-251-70/+113
* aco/gfx10: Fix mitigation of VMEMtoScalarWriteHazard.Timur Kristóf2019-10-251-10/+20
* aco/gfx10: Mitigate LdsBranchVmemWARHazard.Timur Kristóf2019-10-252-0/+66
* aco/gfx10: Mitigate SMEMtoVectorWriteHazard.Timur Kristóf2019-10-252-0/+70
* aco/gfx10: Mitigate VcmpxExecWARHazard.Timur Kristóf2019-10-252-0/+59
* aco/gfx10: Mitigate VcmpxPermlaneHazard.Timur Kristóf2019-10-252-0/+28
* aco/gfx10: Add notes about some GFX10 hazards.Timur Kristóf2019-10-252-2/+37
* aco: take LDS into account when calculating num_wavesRhys Perry2019-10-234-7/+42
* aco: increase accuracy of SGPR limitsRhys Perry2019-10-236-28/+95
* Revert "aco: only emit waitcnt on loop continues if we there was some load or...Rhys Perry2019-10-221-1/+1
* aco: add missing bld.scc()Rhys Perry2019-10-221-1/+1
* aco: keep can_reorder/barrier when combining addition into SMEMRhys Perry2019-10-221-0/+2
* aco: add a few missing checks in value numberingRhys Perry2019-10-221-1/+4
* aco: use ds_read2_b64/ds_write2_b64Rhys Perry2019-10-221-7/+24
* aco: properly combine additions into ds_write2_b64/ds_read2_b64Rhys Perry2019-10-221-1/+2
* aco: fix sparse store_lds()Rhys Perry2019-10-221-8/+40
* aco: create load_lds/store_lds helpersRhys Perry2019-10-221-176/+195
* aco: fix 64-bit p_extract_vector on 32-bit p_create_vectorRhys Perry2019-10-221-1/+2
* aco: small stage correctionsRhys Perry2019-10-222-11/+13
* aco: refactor value numberingDaniel Schürmann2019-10-221-55/+53
* aco: run opt_algebraic in a loopRhys Perry2019-10-211-3/+8
* aco: use nir_lower_idiv_preciseRhys Perry2019-10-211-1/+1
* nir/lower_idiv: add new llvm-based pathRhys Perry2019-10-211-1/+1
* aco: ensure that uniform booleans are computed in WQM if their uses happen in...Daniel Schürmann2019-10-211-1/+2
* aco/gfx10: Update constant addresses in fix_branches_gfx10.Timur Kristóf2019-10-211-1/+12
* aco/gfx10: Fix PS exports for SPI_SHADER_32_AR.Timur Kristóf2019-10-211-1/+7
* aco/gfx10: Wait for pending SMEM stores before loadsTimur Kristóf2019-10-212-1/+33
* aco: don't combine minmax3 if there is a neg or abs modifier in betweenDaniel Schürmann2019-10-171-29/+1
* aco: emit_split_vector() s_memtime resultsRhys Perry2019-10-161-0/+1
* aco: don't CSE s_memtimeRhys Perry2019-10-161-1/+1
* aco: fix scheduling with s_memtime/s_memrealtimeRhys Perry2019-10-161-0/+8
* aco: don't use p_as_uniform for vgpr sampler/image indicesRhys Perry2019-10-111-1/+3
* aco: implement divergent vulkan_resource_indexRhys Perry2019-10-112-4/+14
* aco: readfirstlane vgpr pointers in convert_pointer_to_64_bit()Rhys Perry2019-10-111-0/+2
* aco: use can_accept_constant in valu_can_accept_literalRhys Perry2019-10-111-7/+8
* aco: don't apply sgprs/constants to read/write lane instructionsRhys Perry2019-10-111-1/+11
* aco: update print_irRhys Perry2019-10-101-0/+39
* aco: rework scratch resource codeRhys Perry2019-10-101-22/+28
* aco/gfx10: disable GFX9 1D texture workaroundsRhys Perry2019-10-101-5/+5
* aco/gfx10: fix inline uniform blocksRhys Perry2019-10-101-3/+9
* aco: Clean up usages of PhysReg::reg from aco_assembler.Timur Kristóf2019-10-101-27/+27
* aco: Add extra assertion for number of FS input VGPRs.Timur Kristóf2019-10-101-0/+7
* aco: Fix s_dcache_wb on GFX10.Timur Kristóf2019-10-102-0/+13
* aco: Have s_waitcnt_vscnt write to NULL.Rhys Perry2019-10-101-2/+3
* aco: Use the VOP3-only add/sub GFX10 instructions if needed.Rhys Perry2019-10-101-1/+15
* aco: Initial work to avoid GFX10 hazards.Rhys Perry2019-10-102-36/+117
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-27 05:38:47 +0000