On Wed, 2 Aug 2023 at 11:09, Suren Baghdasaryan <surenb@xxxxxxxxxx> wrote: > > Ok, IOW the vma would be already locked before mmap() is called... Yup. > Just to confirm, you are suggesting to remove vma_start_write() call > from hfi1_file_mmap() and let vm_flags_reset() generate an assertion > if it's ever called with an unlocked vma, correct? Correct. Linus