summaryrefslogtreecommitdiff
path: root/lldb/test/API/python_api/global_module_cache/TestGlobalModuleCache.py
diff options
context:
space:
mode:
authorAlexey Bataev <a.bataev@outlook.com>2024-10-08 16:43:10 -0400
committerGitHub <noreply@github.com>2024-10-08 16:43:10 -0400
commita65a5feb1a20581c85ee817dae8826f65fef62af (patch)
treea76a8364290e930d84493451f637d3343866faee /lldb/test/API/python_api/global_module_cache/TestGlobalModuleCache.py
parent36fc291b6ec6d4c8e3d956d6855631ab2db7b05a (diff)
[SLP]Improve masked loads vectorization, attempting gathered loads
If the vector of loads can be vectorized as masked gather and there are several other masked gather nodes, compiler can try to attempt to check, if it possible to gather such nodes into big consecutive/strided loads node, which provide better performance. Reviewers: RKSimon Reviewed By: RKSimon Pull Request: https://github.com/llvm/llvm-project/pull/110151
Diffstat (limited to 'lldb/test/API/python_api/global_module_cache/TestGlobalModuleCache.py')
0 files changed, 0 insertions, 0 deletions