maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
283,818 Commits 23 Branches 0 Tags 2.2 GiB
Commit Graph

13 Commits

Author SHA1 Message Date
Xin Tong 8ba674e43b [MergeICmp] Fix a bug in MergeICmp that can lead to a block being processed more than once. 
Summary:
Fix a bug in MergeICmp that can lead to a BCECmp block being processed more than once and eventually lead to a broken LLVM module.
The problem is that if the non-constant value is not produced by the last block, the producer will be processed once when the its parent block
is processed and second time when the last block is processed.

We end up having 2 same BCECmpBlock in the merge queue. And eventually lead to a broken LLVM module.

Reviewers: courbet, davide

Reviewed By: courbet

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D43825

llvm-svn: 326318
 2018-02-28 12:08:00 +00:00
Clement Courbet a7a1746865 [MergeICmps] Handle chains with several complex BCE basic blocks. 
- Fix condition for detecting that a complex basic block was the first in
   the chain.
 - Add tests.

This was caught by buildbots when submitting rL324319.

llvm-svn: 324341
 2018-02-06 12:25:33 +00:00
Clement Courbet e2e8a5c496 Re-land "[MergeICmps] Disable mergeicmps if the target does not want to handle memcmp expansion." 
(fixed stability issues)

This reverts commit d6492333d3b478a1d88163315002022f8d5e58dc.

llvm-svn: 315281
 2017-10-10 08:00:45 +00:00
Clement Courbet 028d2eb671 [MergeICmp][NFC] Make test tuple-four-int8.ll more readable. 
llvm-svn: 315062
 2017-10-06 13:45:16 +00:00
Clement Courbet d12c189e2e Revert "[MergeICmps] Disable mergeicmps if the target does not want to handle memcmp expansion." 
Still a few stability issues on windows.

This reverts commit 67e3db9bc121ba244e20337aabc7cf341a62b545.

llvm-svn: 315058
 2017-10-06 13:02:24 +00:00
Clement Courbet 4e1bae8136 Re-land "[MergeICmps] Disable mergeicmps if the target does not want to handle memcmp expansion." 
(fixed unit tests by making comparisons stable)

This reverts commit 1b2d359ce256fd6737da4e93833346a0bd6d7583.

llvm-svn: 315056
 2017-10-06 12:12:35 +00:00
Clement Courbet 922e5bc698 Revert "Re-land "[MergeICmps] Disable mergeicmps if the target does not want to handle memcmp expansion.""" 
broken test on windows

This reverts commit c91479518344fd1fc071c5bd5848f6eb83e53dca.

llvm-svn: 314985
 2017-10-05 14:42:06 +00:00
Clement Courbet 4cafbb9b5e Re-land "[MergeICmps] Disable mergeicmps if the target does not want to handle memcmp expansion."" 
llvm-svn: 314980
 2017-10-05 12:39:57 +00:00
Clement Courbet 6603fc0e7b Revert "[MergeICmps] Disable mergeicmps if the target does not want to handle memcmp expansion." 
Breaks
clang-stage1-cmake-RA-incremental/llvm/test/Transforms/MergeICmps/X86/tuple-four-int8.ll

This reverts commit 3038c459d67f8898ffa295d54a013b280690abfa.

llvm-svn: 314972
 2017-10-05 08:03:39 +00:00
Clement Courbet 902eef32eb [MergeICmps] Disable mergeicmps if the target does not want to handle memcmp expansion. 
Summary: This is to avoid e.g. merging two cheap icmps if the target is not going to expand to something nice later.

Reviewers: dberlin, spatel

Subscribers: davide, nemanjai

Differential Revision: https://reviews.llvm.org/D38232

llvm-svn: 314970
 2017-10-05 07:49:09 +00:00
Clement Courbet 65130e2d8d Reland rL312315: [MergeICmps] MergeICmps is a new optimization pass that turns chains of integer 
Add missing header.

This reverts commit 86dd6335cf7607af22f383a9a8e072ba929848cf.

llvm-svn: 312322
 2017-09-01 10:56:34 +00:00
Clement Courbet 316212575b Revert "[MergeICmps] MergeICmps is a new optimization pass that turns chains of integer" 
Break build

This reverts commit d07ab866f7f88f81e49046d691a80dcd32d7198b.

llvm-svn: 312317
 2017-09-01 09:43:08 +00:00
Clement Courbet 9473c01e96 [MergeICmps] MergeICmps is a new optimization pass that turns chains of integer 
comparisons into memcmp.

Thanks to recent improvements in the LLVM codegen, the memcmp is typically
inlined as a chain of efficient hardware comparisons.
This typically benefits C++ member or nonmember operator==().

For now this is disabled by default until:
 - https://bugs.llvm.org/show_bug.cgi?id=33329 is complete
 - Benchmarks show that this is always useful.

Differential Revision:
https://reviews.llvm.org/D33987

llvm-svn: 312315
 2017-09-01 09:07:05 +00:00