On Thu, 7 Jun 2012 17:52:52 -0700 Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> wrote: > On Thu, Jun 7, 2012 at 5:24 PM, Dave Jones <davej@xxxxxxxxxx> wrote: > > I just started seeing crashes while doing simple things, like logging on a console.. > > I'm looking at it right now, and the sync_mm_rss() patch is pure > garbage. In many ways. > > You can't do sync_mm_rss() from mmdrop(), because there's no reason to > believe that the task that does mmdrop() does it on its own active_mm. > And even if you *could* do it there, it's still horribly wrong, > because it does it at the end *after* it already freed the mm! > > Does it go away if you revert that (commit 40af1bbdca47). I wish I > hadn't merged it, or that I had noticed how horrible it was before I > pushed out. > It appears this is due to me fat-fingering conflict resolution last week. That hunk is supposed to be in mm_release(), not mmput(). It's probably best to throw the patch away for now - we'll try again. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>