This is a pure refactoring change. It just moves code that is
related to filesystem operations from sanitizer_common.{cc,h} to
sanitizer_file.{cc,h}. This makes it cleaner to disable the
filesystem-related code for a new port that doesn't want it.
Details
- Reviewers
kcc eugenis alekseyshl - Commits
- rGd4abe9e7d366: [sanitizer_common] Move filesystem-related code out of sanitizer_common.cc
rG89e85d17173b: [sanitizer_common] Move filesystem-related code out of sanitizer_common.cc
rCRT308819: [sanitizer_common] Move filesystem-related code out of sanitizer_common.cc
rCRT308640: [sanitizer_common] Move filesystem-related code out of sanitizer_common.cc
rL308819: [sanitizer_common] Move filesystem-related code out of sanitizer_common.cc
rL308640: [sanitizer_common] Move filesystem-related code out of sanitizer_common.cc
Diff Detail
- Repository
- rL LLVM
Event Timeline
Reverted r308648.
This breaks windows build.
http://lab.llvm.org:8011/builders/sanitizer-windows/builds/14444/steps/build%20compiler-rt/logs/stdio
I've added the #include's that I think from inspection might fix the Windows and Mac builds, but I'm not in a position to test those builds easily.
If someone can test those builds so this can re-land, that would be great.
I have no access to Win or Mac either but I'll ask around. Or we can just land it and revert if something goes wrong again...
phosek is setting up a Windows box and said he'd test it for me when he gets that in shape. (I think he has a Mac too. :-)
But I'm also fairly confident that my fix will have covered the build breakage, so landing again optimistically will most likely be just fine.
Thanks,
Roland