[PATCH] D45888: [AMDGPU][Waitcnt] Add debug options

[Mark Searles via Phabricator via llvm-commits]

msearles created this revision.
msearles added reviewers: kzhuravl, arsenm.
msearles added a project: AMDGPU.
Herald added subscribers: t-tye, tpr, dstuttard, yaxunl, nhaehnle, wdng.

- Add "amdgpu-waitcnt-forcezero" to force all waitcnt instrs to be emitted as s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)

- Add debug counters to control force emit of s_waitcnt instrs; debug counters: si-insert-waitcnts-forceexp: force emit s_waitcnt expcnt(0) instrs si-insert-waitcnts-forcevm: force emit s_waitcnt lgkmcnt(0) instrs si-insert-waitcnts-forcelgkm: force emit s_waitcnt vmcnt(0) instrs

- Add some debug statements

- Code cleanup: naming consistency s/SWaitcnt/Waitcnt s/WaitCnt/Waitcnt

Note that a variant of this patch was previously attempted but never successfully landed (committed/reverted). I'm creating a new review to shake-off the bad mojo of the prior attempt :) .


https://reviews.llvm.org/D45888

Files:
  lib/Target/AMDGPU/SIInsertWaitcnts.cpp
  test/CodeGen/AMDGPU/waitcnt-debug.mir

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D45888.143334.patch
Type: text/x-patch
Size: 19531 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20180420/f1dbf26e/attachment.bin>