summaryrefslogtreecommitdiffstats
path: root/llvm/test/Transforms/SeparateConstOffsetFromGEP/AMDGPU
Commit message (Collapse)AuthorAgeFilesLines
* [lit] Delete empty lines at the end of lit.local.cfg NFCFangrui Song2019-06-171-1/+0
| | | | llvm-svn: 363538
* Revert "Temporarily Revert "Add basic loop fusion pass.""Eric Christopher2019-04-172-0/+142
| | | | | | | | The reversion apparently deleted the test/Transforms directory. Will be re-reverting again. llvm-svn: 358552
* Temporarily Revert "Add basic loop fusion pass."Eric Christopher2019-04-172-142/+0
| | | | | | | | As it's causing some bot failures (and per request from kbarton). This reverts commit r358543/ab70da07286e618016e78247e4a24fcb84077fda. llvm-svn: 358546
* [AMDGPU] Change constant addr space to 4Yaxun Liu2018-02-131-32/+32
| | | | | | Differential Revision: https://reviews.llvm.org/D43170 llvm-svn: 325030
* [SeparateConstOffsetFromGEP] Fix up addrspace in the AMDGPU testMarek Olsak2018-01-311-6/+6
| | | | llvm-svn: 323913
* [SeparateConstOffsetFromGEP] Preserve metadata when splitting GEPsMarek Olsak2018-01-311-0/+45
| | | | | | | | | | | | | | Summary: !amdgpu.uniform needs to be preserved for AMDGPU, otherwise bad things happen. Reviewers: arsenm, nhaehnle, jingyue, broune, majnemer, bjarke.roune, dblaikie Subscribers: wdng, tpr, llvm-commits Differential Revision: https://reviews.llvm.org/D42744 llvm-svn: 323907
* AMDGPU: Mark all unspecified CC functions in tests as amdgpu_kernelMatt Arsenault2017-03-211-3/+3
| | | | | | | | | | | | Currently the default C calling convention functions are treated the same as compute kernels. Make this explicit so the default calling convention can be changed to a non-kernel. Converted with perl -pi -e 's/define void/define amdgpu_kernel void/' on the relevant test directories (and undoing in one place that actually wanted a non-kernel). llvm-svn: 298444
* [ReassociateGEP] Update tests to allow missing "inbounds" on certain GEPs.Jingyue Wu2016-07-151-4/+3
| | | | | | | | With r275532 fixing miscompilation of GVN, "inbounds" on certain GEPs in these tests cannot be preserved any more. Left a TODO in the tests for future reference. llvm-svn: 275596
* XFAIL two SeparateConstOffsetFromGEP testsDavid Majnemer2016-07-151-0/+1
| | | | | | | | They appear to have relied on bugs hidden in copyIRFlags/andIRFlags. This has been filed as PR28564. llvm-svn: 275533
* [SeparateConstOffsetFromGEP] strengthen the inbounds attributeJingyue Wu2015-08-131-7/+7
| | | | | | | | | | | | | | We used to be over-conservative about preserving inbounds. Actually, the second GEP (which applies the constant offset) can inherit the inbounds attribute of the original GEP, because the resultant pointer is equivalent to that of the original GEP. For example, x = GEP inbounds a, i+5 => y = GEP a, i // inbounds removed x = GEP inbounds y, 5 // inbounds preserved llvm-svn: 244937
* AMDGPU: Fix some places missed in renameMatt Arsenault2015-06-192-0/+97
llvm-svn: 240143
OpenPOWER on IntegriCloud