On Wed, Aug 21, 2019 at 11:57:03AM -0700, 'Ira Weiny' wrote: > On Wed, Aug 21, 2019 at 03:13:43PM -0300, Jason Gunthorpe wrote: > > On Wed, Aug 21, 2019 at 11:02:00AM -0700, Ira Weiny wrote: > > > On Tue, Aug 20, 2019 at 08:55:15AM -0300, Jason Gunthorpe wrote: > > > > On Tue, Aug 20, 2019 at 11:12:10AM +1000, Dave Chinner wrote: > > > > > On Mon, Aug 19, 2019 at 09:38:41AM -0300, Jason Gunthorpe wrote: > > > > > > On Mon, Aug 19, 2019 at 07:24:09PM +1000, Dave Chinner wrote: > > > > > > > > > > > > > So that leaves just the normal close() syscall exit case, where the > > > > > > > application has full control of the order in which resources are > > > > > > > released. We've already established that we can block in this > > > > > > > context. Blocking in an interruptible state will allow fatal signal > > > > > > > delivery to wake us, and then we fall into the > > > > > > > fatal_signal_pending() case if we get a SIGKILL while blocking. > > > > > > > > > > > > The major problem with RDMA is that it doesn't always wait on close() for the > > > > > > MR holding the page pins to be destoyed. This is done to avoid a > > > > > > deadlock of the form: > > > > > > > > > > > > uverbs_destroy_ufile_hw() > > > > > > mutex_lock() > > > > > > [..] > > > > > > mmput() > > > > > > exit_mmap() > > > > > > remove_vma() > > > > > > fput(); > > > > > > file_operations->release() > > > > > > > > > > I think this is wrong, and I'm pretty sure it's an example of why > > > > > the final __fput() call is moved out of line. > > > > > > > > Yes, I think so too, all I can say is this *used* to happen, as we > > > > have special code avoiding it, which is the code that is messing up > > > > Ira's lifetime model. > > > > > > > > Ira, you could try unraveling the special locking, that solves your > > > > lifetime issues? > > > > > > Yes I will try to prove this out... But I'm still not sure this fully solves > > > the problem. > > > > > > This only ensures that the process which has the RDMA context (RDMA FD) is safe > > > with regard to hanging the close for the "data file FD" (the file which has > > > pinned pages) in that _same_ process. But what about the scenario. > > > > Oh, I didn't think we were talking about that. Hanging the close of > > the datafile fd contingent on some other FD's closure is a recipe for > > deadlock.. > > The discussion between Jan and Dave was concerning what happens when a user > calls > > fd = open() > fnctl(...getlease...) > addr = mmap(fd...) > ib_reg_mr() <pin> > munmap(addr...) > close(fd) > > Dave suggested: > > "I'm of a mind to make the last close() on a file block if there's an > active layout lease to prevent processes from zombie-ing layout > leases like this. i.e. you can't close the fd until resources that > pin the lease have been released." > > -- Dave https://lkml.org/lkml/2019/8/16/994 I think this may all be easier if there was a way to block a dup() if it comes from an SCM_RIGHTS. Does anyone know if that is easy? I assume it would just mean passing some flag through the dup() call chain. Jason, if we did that would it break RDMA use cases? I personally don't know of any. We could pass data back from vaddr_pin indicating that a file pin has been taken and predicate the blocking of SCM_RIGHTS on that? Of course if the user called: fd = open() fnctl(...getlease...) addr = mmap(fd...) ib_reg_mr() <pin> munmap(addr...) close(fd) fork() <=== in another thread because close is hanging Would that dup() "fd" above into the child? Or maybe that would be part of the work to make close() hang? Ensure the fd/file is still in the FD table so it gets dupped??? Ira > > > > > IMHO the pin refcnt is held by the driver char dev FD, that is the > > object you need to make it visible against. > > I'm sorry but what do you mean by "driver char dev FD"? > > > > > Why not just have a single table someplace of all the layout leases > > with the file they are held on and the FD/socket/etc that is holding > > the pin? Make it independent of processes and FDs? > > If it is independent of processes how will we know which process is blocking > the truncate? Using a global table is an interesting idea but I still believe > the users are going to want to track this to specific processes. It's not > clear to me how that would be done with a global table. > > I agree the XDP/socket case is bothersome... I was thinking that somewhere the > fd of the socket could be hooked up in this case. But taking a look at it > reveals that is not going to be easy. And I assume XDP has the same issue WRT > SCM_RIGHTS and the ability to share the xdp context? > > Ira > > > > > Jason
