On Tue, Jun 25, 2024 at 7:51 AM Oleg Nesterov <oleg@xxxxxxxxxx> wrote: > > On 06/25, Masami Hiramatsu wrote: > > > > On Mon, 24 Jun 2024 17:21:34 -0700 > > Andrii Nakryiko <andrii@xxxxxxxxxx> wrote: > > > > > Given unapply_uprobe() can call remove_breakpoint() which eventually > > > calls uprobe_write_opcode(), which can modify a set of memory pages and > > > expects mm->mmap_lock held for write, it needs to have writer lock. > > > > Oops, it is an actual bug, right? > > Why? > > So far I don't understand this change. Quite possibly I missed something, > but in this case the changelog should explain the problem more clearly. > I just went off of "Called with mm->mmap_lock held for write." comment in uprobe_write_opcode(), tbh. If we don't actually need writer mmap_lock, we should probably update at least that comment. There is a lot going on in uprobe_write_opcode(), and I don't understand all the requirements there. > Oleg. >
