[PATCH] D32401: [Devirtualization] insert placement new barrier with -O0

[Mehdi AMINI via Phabricator via cfe-commits]

mehdi_amini added a comment.

I tend to agree with @rjmccall on the principle. Howerever:

> The optimization design seems to rely on anticipating every case that should disable the optimization, hence this patch adding special-case logic to the frontend, and the 3 other patch

I believe this is patch is *removing* a special case logic that was wrong.

I agree that the whole system should be "safe". But I'm not sure it is unsafe as it is designed: if we are already issuing an error when merging two modules that have different settings IIUC.


https://reviews.llvm.org/D32401