diff options
| author | Min-Yih Hsu <min.hsu@sifive.com> | 2024-01-26 15:47:26 -0800 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2024-01-26 15:47:26 -0800 |
| commit | 3bece3d72d1323af6b2418b13281e66b583f25b3 (patch) | |
| tree | 2cbc01f5ba45d9241fc7f73e4dfe96ff489b6fc7 /lldb/packages/Python/lldbsuite/test/lldbgdbclient.py | |
| parent | 9816863dd439cd60c96d5764c4c21ca1a7f3d8da (diff) | |
| download | llvm-3bece3d72d1323af6b2418b13281e66b583f25b3.zip llvm-3bece3d72d1323af6b2418b13281e66b583f25b3.tar.gz llvm-3bece3d72d1323af6b2418b13281e66b583f25b3.tar.bz2 | |
[Exegesis] Do not assume the size and layout of the assembled snippet (#79636)
Currently llvm-exegesis assumes that there will only be 3 symbols in the
snippet object, in which the benchmarking function 'foo' is always the
last symbol.
These assumptions do not hold for object file formats of other targets
we support downstream. I think it would be more ideal to generalize this
part of the logics into a simple search on all symbols, as proposed by
this patch.
Diffstat (limited to 'lldb/packages/Python/lldbsuite/test/lldbgdbclient.py')
0 files changed, 0 insertions, 0 deletions