[PATCH] D37788: [ARM] builtins: Do not abort in clear_cache.

[Peter Smith via Phabricator via llvm-commits]

peter.smith added a comment.

A colleague mentioned that on Arm removing write and adding exec should cause mprotect() to flush the at least some of the cache hierarchy. This might explain why we don't see lots of failures, however it doesn't explain why we are seeing an abort when trying to clear the cache. The only thing that I can think of is that there is some dirty data in _a_ cache that can't be written back to the memory as the mprotect() has just marked it read-only. However I'm struggling to work out how this situation could arise.


https://reviews.llvm.org/D37788