maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/CodeGen/AMDGPU/GlobalISel/inst-select-bitcast.mir

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

22 lines
493 B
Plaintext
Raw Normal View History

AMDGPU/GlobalISel: Prepare some tests for store selection Mostsly these would fail due to trying to use SI with a flat operation. Implementing global loads with MUBUF is more work than flat, so these won't be handled in the initial load selection. Others fail because store of s64 won't initially work, as the current set of patterns expect everything to be turned into v2i32. llvm-svn: 365493
2019-07-09 22:30:57 +08:00
# RUN: llc -march=amdgcn -run-pass=instruction-select -verify-machineinstrs -o - %s | FileCheck %s -check-prefixes=GCN
AMDGPU/GlobalISel: Implement select() for G_BITCAST s32 <--> <2 x s16> Reviewers: arsenm, nhaehnle Subscribers: kzhuravl, wdng, yaxunl, rovka, kristof.beyls, dstuttard, tpr, t-tye, llvm-commits Differential Revision: https://reviews.llvm.org/D45881 llvm-svn: 332042
2018-05-11 05:20:10 +08:00
---
name: bitcast
legalized: true
regBankSelected: true
AMDGPU/GlobalISel: Prepare some tests for store selection Mostsly these would fail due to trying to use SI with a flat operation. Implementing global loads with MUBUF is more work than flat, so these won't be handled in the initial load selection. Others fail because store of s64 won't initially work, as the current set of patterns expect everything to be turned into v2i32. llvm-svn: 365493
2019-07-09 22:30:57 +08:00
tracksRegLiveness: true
AMDGPU/GlobalISel: Implement select() for G_BITCAST s32 <--> <2 x s16> Reviewers: arsenm, nhaehnle Subscribers: kzhuravl, wdng, yaxunl, rovka, kristof.beyls, dstuttard, tpr, t-tye, llvm-commits Differential Revision: https://reviews.llvm.org/D45881 llvm-svn: 332042
2018-05-11 05:20:10 +08:00
# GCN-LABEL: name: bitcast
# GCN: [[A:%[0-9]+]]:vgpr_32 = COPY $vgpr0
AMDGPU/GlobalISel: Prepare some tests for store selection Mostsly these would fail due to trying to use SI with a flat operation. Implementing global loads with MUBUF is more work than flat, so these won't be handled in the initial load selection. Others fail because store of s64 won't initially work, as the current set of patterns expect everything to be turned into v2i32. llvm-svn: 365493
2019-07-09 22:30:57 +08:00
# GCN: S_ENDPGM 0, implicit [[A]]
AMDGPU/GlobalISel: Implement select() for G_BITCAST s32 <--> <2 x s16> Reviewers: arsenm, nhaehnle Subscribers: kzhuravl, wdng, yaxunl, rovka, kristof.beyls, dstuttard, tpr, t-tye, llvm-commits Differential Revision: https://reviews.llvm.org/D45881 llvm-svn: 332042
2018-05-11 05:20:10 +08:00
body: |
bb.0:
AMDGPU/GlobalISel: Prepare some tests for store selection Mostsly these would fail due to trying to use SI with a flat operation. Implementing global loads with MUBUF is more work than flat, so these won't be handled in the initial load selection. Others fail because store of s64 won't initially work, as the current set of patterns expect everything to be turned into v2i32. llvm-svn: 365493
2019-07-09 22:30:57 +08:00
liveins: $vgpr0
AMDGPU/GlobalISel: Implement select() for G_BITCAST s32 <--> <2 x s16> Reviewers: arsenm, nhaehnle Subscribers: kzhuravl, wdng, yaxunl, rovka, kristof.beyls, dstuttard, tpr, t-tye, llvm-commits Differential Revision: https://reviews.llvm.org/D45881 llvm-svn: 332042
2018-05-11 05:20:10 +08:00
%0:vgpr(s32) = COPY $vgpr0
AMDGPU/GlobalISel: Prepare some tests for store selection Mostsly these would fail due to trying to use SI with a flat operation. Implementing global loads with MUBUF is more work than flat, so these won't be handled in the initial load selection. Others fail because store of s64 won't initially work, as the current set of patterns expect everything to be turned into v2i32. llvm-svn: 365493
2019-07-09 22:30:57 +08:00
%1:vgpr(<2 x s16>) = G_BITCAST %0
%2:vgpr(s32) = G_BITCAST %1
S_ENDPGM 0, implicit %2
AMDGPU/GlobalISel: Implement select() for G_BITCAST s32 <--> <2 x s16> Reviewers: arsenm, nhaehnle Subscribers: kzhuravl, wdng, yaxunl, rovka, kristof.beyls, dstuttard, tpr, t-tye, llvm-commits Differential Revision: https://reviews.llvm.org/D45881 llvm-svn: 332042
2018-05-11 05:20:10 +08:00
...