Page MenuHomePhabricator

[AMDGPU] gfx1032 target
ClosedPublic

Authored by rampitec on Thu, Oct 15, 11:03 AM.

Diff Detail

Event Timeline

rampitec created this revision.Thu, Oct 15, 11:03 AM
Herald added a project: Restricted Project. · View Herald Transcript
rampitec requested review of this revision.Thu, Oct 15, 11:03 AM
This revision is now accepted and ready to land.Thu, Oct 15, 12:39 PM
tra accepted this revision.Thu, Oct 15, 12:41 PM
tra added a reviewer: yaxunl.
tra added a subscriber: tra.

LGTM.

clang/test/Driver/amdgpu-macros.cl
216

This test could use some refactoring.
Individual macro checks could be collapsed to

// RUN: ... --check-prefixes=...,FMA
//  FMA-DAG: #define FP_FAST_FMA 1
This revision was automatically updated to reflect the committed changes.
Herald added a project: Restricted Project. · View Herald TranscriptThu, Oct 15, 12:42 PM
Herald added a subscriber: cfe-commits. · View Herald Transcript
rampitec added inline comments.Thu, Oct 15, 12:43 PM
clang/test/Driver/amdgpu-macros.cl
216

Thanks! That makes sense.

foad added a subscriber: foad.Fri, Oct 16, 7:48 AM
foad added inline comments.Fri, Oct 16, 9:15 AM
llvm/docs/AMDGPUUsage.rst
280

xnack looks like a mistake here?

llvm/lib/Support/TargetParser.cpp
66

Use [] so we don't have to keep updating the number?

t-tye added inline comments.Fri, Oct 16, 10:02 AM
llvm/docs/AMDGPUUsage.rst
280

gfx1032 does not support XNACK. It should have the same target features as gfx1032.

835

Delete as the line above is using this reserved number.

rampitec marked 3 inline comments as done.Fri, Oct 16, 10:21 AM
rampitec added inline comments.
llvm/docs/AMDGPUUsage.rst
280
835
rampitec marked 3 inline comments as done.Fri, Oct 16, 10:45 AM
rampitec added inline comments.
llvm/lib/Support/TargetParser.cpp
66