This is an automated email from the git hooks/post-receive script.
unknown user pushed a change to branch master in repository llvm.
from f885500 [ADMGPU] SDWA peephole optimization pass. new b5adc93 [GlobalISel] Fix shufflevector tests new 4b7cf6c [AMDGPU] Iterative scheduling infrastructure + minimal regist [...]
The 2 revisions listed above as "new" are entirely new to this repository and will be described in separate emails. The revisions listed as "adds" were already present in the repository and have only been added to this reference.
Summary of changes: lib/Target/AMDGPU/AMDGPUSubtarget.h | 6 + lib/Target/AMDGPU/AMDGPUTargetMachine.cpp | 25 + lib/Target/AMDGPU/CMakeLists.txt | 3 + lib/Target/AMDGPU/GCNIterativeScheduler.cpp | 528 +++++++++++++++++++++ lib/Target/AMDGPU/GCNIterativeScheduler.h | 118 +++++ lib/Target/AMDGPU/GCNMinRegStrategy.cpp | 266 +++++++++++ lib/Target/AMDGPU/GCNRegPressure.cpp | 355 ++++++++++++++ lib/Target/AMDGPU/GCNRegPressure.h | 170 +++++++ lib/Target/AMDGPU/GCNSchedStrategy.cpp | 4 +- lib/Target/AMDGPU/GCNSchedStrategy.h | 2 + .../AArch64/GlobalISel/arm64-irtranslator.ll | 34 +- test/CodeGen/AMDGPU/schedule-regpressure-limit.ll | 2 + test/CodeGen/AMDGPU/schedule-regpressure-limit2.ll | 288 +++++++++++ test/CodeGen/ARM/GlobalISel/arm-irtranslator.ll | 50 +- 14 files changed, 1806 insertions(+), 45 deletions(-) create mode 100644 lib/Target/AMDGPU/GCNIterativeScheduler.cpp create mode 100644 lib/Target/AMDGPU/GCNIterativeScheduler.h create mode 100644 lib/Target/AMDGPU/GCNMinRegStrategy.cpp create mode 100644 lib/Target/AMDGPU/GCNRegPressure.cpp create mode 100644 lib/Target/AMDGPU/GCNRegPressure.h create mode 100644 test/CodeGen/AMDGPU/schedule-regpressure-limit2.ll