projects / gem5.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 63c7644) | patch
gpu-compute, arch-gcn3: Change how waitcnts are implemented
authorTony Gutierrez <anthony.gutierrez@amd.com>
Fri, 10 Aug 2018 19:01:12 +0000 (15:01 -0400)
committerAnthony Gutierrez <anthony.gutierrez@amd.com>
Fri, 17 Jul 2020 16:36:23 +0000 (16:36 +0000)
commit4d737462c29149342c9f8a5bce560bebc2e196d4
treeb673b7cd7ff4caf002bc24553b5d2ee32f5928e6tree
parent63c76448ebd1411af02ff6c0fce5a6afccfb9c87commit | diff
gpu-compute, arch-gcn3: Change how waitcnts are implemented

Use single counters per memory operation type and increment
them upon issue, not execute.

Change-Id: I6afc0b66b21882538ef90a14a57a3ab3cc7bd6f3
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/29973
Reviewed-by: Anthony Gutierrez <anthony.gutierrez@amd.com>
Maintainer: Anthony Gutierrez <anthony.gutierrez@amd.com>
Tested-by: kokoro <noreply+kokoro@google.com>
src/arch/gcn3/insts/instructions.cc diff | blob | history
src/gpu-compute/global_memory_pipeline.cc diff | blob | history
src/gpu-compute/gpu_dyn_inst.cc diff | blob | history
src/gpu-compute/local_memory_pipeline.cc diff | blob | history
src/gpu-compute/scalar_memory_pipeline.cc diff | blob | history
src/gpu-compute/schedule_stage.cc diff | blob | history
src/gpu-compute/wavefront.cc diff | blob | history
src/gpu-compute/wavefront.hh diff | blob | history