On Wed, Oct 18, 2023 at 11:27 AM Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> wrote: > > On Wed, 18 Oct 2023 at 10:14, Jeff Xu <jeffxu@xxxxxxxxxx> wrote: > This is simply not up for discussion. Any flag for "ignore sealing" is wrong. > > We do have one special "unmap" case, namely "unmap_vmas()' called at > last mmput() -> __mmput() -> exit_mmap(). > > And yes, that is called at munmap() time too, but that's after the > point of no return after we've already removed the vma's from the VM > lists. So it's long after any error cases have been checked. > Ah. I see. I didn't know there was no legit case, which is what I worried about before. this flag can be removed.
