summaryrefslogtreecommitdiff
path: root/src/amd/compiler/aco_scheduler.cpp
Commit message (Expand)AuthorAgeFilesLines
* amd: Cleanup old GS intrinsics code.Timur Kristóf2023-05-041-1/+1
* aco: don't move exec writes around exec writesRhys Perry2023-04-271-1/+1
* aco: don't move exec reads around exec writesRhys Perry2023-04-261-0/+8
* amd: fix typosHarri Nieminen2023-04-131-1/+1
* aco: fix out-of-bounds access when moving s_mem(real)time across SMEMRhys Perry2023-02-071-1/+1
* aco: add storage_gdsRhys Perry2022-10-281-2/+1
* aco: split the sendmsg enumeration into sendmsg_rtnSamuel Pitoiset2022-10-251-1/+1
* aco: add support for s_sendmsg_rtn_b{32,64}Samuel Pitoiset2022-10-251-2/+6
* aco: update sendmsg enum from LLVMRhys Perry2022-09-301-24/+19
* aco: add new pseudo instruction p_jump_to_epilogSamuel Pitoiset2022-07-181-1/+1
* aco: initialize scratch base registers on GFX9-GFX10.3Rhys Perry2022-07-081-1/+2
* aco: treat flat-like as vmem in some scheduling heuristicsRhys Perry2022-07-081-4/+5
* aco: move info pointer to a copy.Dave Airlie2022-05-111-2/+2
* aco: adjust num_waves for LDS before schedulingDaniel Schürmann2022-04-291-0/+1
* aco: Add storage class for Task Shader payload.Timur Kristóf2022-02-251-1/+2
* aco: try forming clauses even if reg_pressure exceedsDaniel Schürmann2021-09-271-3/+26
* aco: make clause-forming depend on the number of moved instructionsDaniel Schürmann2021-09-271-2/+2
* aco: stop scheduling if clause-forming failsDaniel Schürmann2021-09-271-0/+4
* aco/scheduler: allow moving down VMEM stores to below VMEM loadsRhys Perry2021-08-231-1/+1
* radv, aco, ac/nir: Tweak position export scheduling for NGG culling.Timur Kristóf2021-07-131-3/+16
* aco: Format.Daniel Schürmann2021-07-121-100/+137
* aco: add 'common/' and 'llvm/' prefix to #includesDaniel Schürmann2021-07-121-1/+2
* aco: reorder and cleanup #includesDaniel Schürmann2021-07-121-6/+7
* aco: move VMEM instructions below descriptor loadsRhys Perry2021-06-141-1/+2
* aco: don't move descriptor loads below buffer loadsRhys Perry2021-06-141-0/+3
* aco/scheduler: Move cursor handling state to dedicated interfacesTony Wasserka2021-06-071-113/+145
* aco/scheduler: Clean up register demand trackingTony Wasserka2021-06-071-25/+30
* aco/scheduler: Verify register demand invariants in debug modeTony Wasserka2021-05-131-2/+39
* aco/scheduler: Fix register demand computation for upwards movesTony Wasserka2021-05-131-6/+6
* aco/scheduler: Fix register demand computation for downwards movesTony Wasserka2021-05-131-3/+15
* aco: group loads from the same vertex binding into the same clauseRhys Perry2021-05-101-4/+2
* aco: add DeviceInfoRhys Perry2021-02-151-1/+1
* aco: don't decrease the vgpr_limit when encountering bpermuteDaniel Schürmann2021-02-121-0/+1
* aco: return references in instruction cast methodsRhys Perry2021-01-221-11/+11
* aco: use format-check methodsRhys Perry2021-01-221-11/+11
* aco: use instruction cast methodsRhys Perry2021-01-221-7/+5
* aco: fix num_waves on GFX10+Rhys Perry2021-01-201-6/+11
* aco: remove can_reorder semantic in get_sync_info_with_hackRhys Perry2021-01-111-1/+1
* aco: create VMEM clauses slightly more aggressiveDaniel Schürmann2020-12-221-1/+1
* aco: schedule position exports in the same pass as memory operationsDaniel Schürmann2020-12-221-17/+8
* aco: fix def-use distance calculation when scheduling.Daniel Schürmann2020-12-221-2/+5
* aco: allow to schedule SALU/SMEM through exec changesDaniel Schürmann2020-12-221-5/+12
* aco: Use program->num_waves as maximum in scheduler.Timur Kristóf2020-12-111-2/+2
* aco: Clean up symbol names and comments related to NGGTony Wasserka2020-10-211-1/+1
* aco: Use strong typing to model SW<->HW stage mappingsTony Wasserka2020-10-211-1/+2
* aco: remove unused radv_shader.h includesSamuel Pitoiset2020-10-141-1/+0
* aco: remove useless occurences of radv_nir_compiler_optionsSamuel Pitoiset2020-10-141-3/+1
* aco: don't move memory accesses to before control barriersRhys Perry2020-08-111-0/+6
* aco: implement subgroup shader_clock on GFX10.3Rhys Perry2020-08-041-1/+2
* aco: fix max_waves_per_simd on Polaris, VegaM and GFX10.3Rhys Perry2020-08-041-5/+9
View Lorry Controller Status