Details
- Reviewers
devnexen krytarowski eugenis morehouse - Commits
- rG78457e68f436: [sanitizer] Move mmap interceptors into sanitizer_common
rL241362: Revert "Refactored ARMTargetInfo in order to use the API of…
rL326851: [sanitizer] Move mmap interceptors into sanitizer_common
rCRT326851: [sanitizer] Move mmap interceptors into sanitizer_common
Diff Detail
- Build Status
Buildable 15749 Build 15749: arc lint + arc unit
Event Timeline
General question of curiosity, do we need to be admins to be able to launch a build based on a diff ?
Is it a builedbot ?
...
Build Status
Buildable 15701
Build 15701: arc lint + arc unit
Usually you don't need any login to trigger build on buildbot. Just put SVN revision into "Revision to build" and click build.
Also there is http://green.lab.llvm.org/green/, not sure if you can rebuild anything there.
compiler-rt/lib/hwasan/hwasan_interceptors.cc | ||
---|---|---|
266–267 | Aren't the above 3 lines handled by COMMON_INTERCEPTOR_ENTER? | |
472 | Why does INTERCEPT_FUNCTION(mmap) stay here but not for other sanitizers? | |
compiler-rt/lib/msan/msan_interceptors.cc | ||
973–974 | Above 3 lines handled by COMMON_INTERCEPTOR_ENTER? | |
compiler-rt/lib/sanitizer_common/sanitizer_common_interceptors.inc | ||
276 | Purpose of if (0 && ... )? | |
compiler-rt/lib/tsan/rtl/tsan_interceptors.cc | ||
2292 | Looks like this is already called by COMMON_INTERCEPTOR_ENTER. |
compiler-rt/lib/sanitizer_common/sanitizer_common_interceptors.inc | ||
---|---|---|
276 | Does it make sense to condition this on common_flags()->intercept_intrin? Isn't that flag intended to affect only memset/memcpy/memmove? | |
277 | Does it make sense to call COMMON_INTERCEPTOR_WRITE_RANGE at all? Doesn't mmap just create a mapping rather than writing to memory? |
No COMMON_INTERCEPTOR_WRITE_RANGE
compiler-rt/lib/sanitizer_common/sanitizer_common_interceptors.inc | ||
---|---|---|
277 | sure, we don't need this. |
Hi, this broke a Darwin bot: http://green.lab.llvm.org/green/job/clang-stage1-configure-RA/43370/console. Can you fix this so that we don't try to intercept mmap64 on Darwin?
lib/sanitizer_common/sanitizer_common_interceptors.inc | ||
---|---|---|
6818 ↗ | (On Diff #137290) | This should use OFF64_T. OFF_T is 32-bit and we pick garbage from the stack as the higher half of the 64-bit offset. This affects linux-i386 with FILE_OFFSET_BITS set to 64. |
lib/sanitizer_common/sanitizer_common_interceptors.inc | ||
---|---|---|
6818 ↗ | (On Diff #137290) | Thanks, fixed in r327596 |
Aren't the above 3 lines handled by COMMON_INTERCEPTOR_ENTER?