diff options
| author | Jakub Kuderski <kubak@google.com> | 2021-12-18 16:02:45 -0500 |
|---|---|---|
| committer | Jakub Kuderski <kubak@google.com> | 2021-12-18 16:03:09 -0500 |
| commit | d9ae852fcc97a6a1b48bf7516d6f3f03a85eed62 (patch) | |
| tree | 8f2ad5ed7ca1c16f8c00a2226a2a814ef301e9a7 /lldb/unittests/ScriptInterpreter/Python/PythonDataObjectsTests.cpp | |
| parent | 2da5af16acfe0081edb917100024260ab1100760 (diff) | |
| download | llvm-d9ae852fcc97a6a1b48bf7516d6f3f03a85eed62.zip llvm-d9ae852fcc97a6a1b48bf7516d6f3f03a85eed62.tar.gz llvm-d9ae852fcc97a6a1b48bf7516d6f3f03a85eed62.tar.bz2 | |
[AMDGPU] Fix data race in SIInsertWaitcnts
The race condition happened when two pass managers ran on two different modules but modified/read the global variables.
To address this, I considered using singletons and freestanding functions to allow getting/setting `HardwareLimits` and `RegisterEncoding`, or making it local to the pass. I chose the latter and made it a member of `WaitcntsBrackets`, to minimizes the amount of global state.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D115896
Diffstat (limited to 'lldb/unittests/ScriptInterpreter/Python/PythonDataObjectsTests.cpp')
0 files changed, 0 insertions, 0 deletions