On Fri, Jun 28, 2019 at 9:37 AM Matthew Wilcox <willy@xxxxxxxxxxxxx> wrote: > > On Thu, Jun 27, 2019 at 07:39:37PM -0700, Dan Williams wrote: > > On Thu, Jun 27, 2019 at 12:59 PM Matthew Wilcox <willy@xxxxxxxxxxxxx> wrote: > > > On Thu, Jun 27, 2019 at 12:09:29PM -0700, Dan Williams wrote: > > > > > This bug feels like we failed to unlock, or unlocked the wrong entry > > > > > and this hunk in the bisected commit looks suspect to me. Why do we > > > > > still need to drop the lock now that the radix_tree_preload() calls > > > > > are gone? > > > > > > > > Nevermind, unmapp_mapping_pages() takes a sleeping lock, but then I > > > > wonder why we don't restart the lookup like the old implementation. > > > > > > If something can remove a locked entry, then that would seem like the > > > real bug. Might be worth inserting a lookup there to make sure that it > > > hasn't happened, I suppose? > > > > Nope, added a check, we do in fact get the same locked entry back > > after dropping the lock. > > Okay, good, glad to have ruled that out. > > > The deadlock revolves around the mmap_sem. One thread holds it for > > read and then gets stuck indefinitely in get_unlocked_entry(). Once > > that happens another rocksdb thread tries to mmap and gets stuck > > trying to take the mmap_sem for write. Then all new readers, including > > ps and top that try to access a remote vma, then get queued behind > > that write. > > > > It could also be the case that we're missing a wake up. > > That was the conclusion I came to; that one thread holding the mmap sem > for read isn't being woken up when it should be. Just need to find it ... > obviously it's something to do with the PMD entries. Can you explain to me one more time, yes I'm slow on the uptake on this, the difference between xas_load() and xas_find_conflict() and why it's ok for dax_lock_page() to use xas_load() while grab_mapping_entry() uses xas_find_conflict()?
