summaryrefslogtreecommitdiff
path: root/src/amd/compiler/aco_reduce_assign.cpp
Commit message (Expand)AuthorAgeFilesLines
* aco: drop leftover variableDavid Heidelberg2023-03-221-2/+0
* aco: end reduce tmp after control flow, when used within control flowRhys Perry2023-02-011-21/+24
* aco: Emulate Wave64 bpermute on GFX11.Timur Kristóf2022-12-141-3/+6
* aco/gfx11: fix FS input loads in quad-divergent control flowRhys Perry2022-11-011-26/+38
* amd: change chip_class naming to "enum amd_gfx_level gfx_level"Marek Olšák2022-05-131-3/+3
* aco: Format.Daniel Schürmann2021-07-121-15/+21
* aco: reorder and cleanup #includesDaniel Schürmann2021-07-121-1/+3
* aco: return references in instruction cast methodsRhys Perry2021-01-221-2/+2
* aco: use instruction cast methodsRhys Perry2021-01-221-2/+2
* aco: keep track of temporaries' regclasses in the ProgramRhys Perry2020-09-211-2/+2
* aco: fix value numbering of reductionsRhys Perry2020-09-091-28/+0
* aco/gfx10: Refactor of GFX10 wave64 bpermute.Timur Kristóf2020-06-021-5/+0
* aco: allocate a temp VGPR for some 8-bit/16-bit reduction ops on GFX10Samuel Pitoiset2020-05-291-1/+4
* aco: use a temporary SGPR for 8-bit/16-bit literal reduction identitiesSamuel Pitoiset2020-05-211-3/+5
* aco: fix gfx10_wave64_bpermuteRhys Perry2020-02-061-1/+6
* aco: implement (clustered) reductions for SI/CIDaniel Schürmann2019-12-071-0/+2
* aco: don't use a scalar temporary for reductions on GFX10Daniel Schürmann2019-12-071-1/+1
* aco/wave32: Use lane mask regclass for exec/vcc.Timur Kristóf2019-12-041-1/+1
* aco: implement 64-bit integer reductionsRhys Perry2019-11-191-2/+12
* aco: Implement subgroup shuffle in GFX10 wave64 mode.Timur Kristóf2019-10-281-1/+3
* aco: Fix reductions on GFX10.Rhys Perry2019-10-281-7/+12
* aco: Initial commit of independent AMD compilerDaniel Schürmann2019-09-191-0/+164