Re: [RFC PATCH v2 00/19] RDMA/FS DAX truncate proposal V1,000,002 ;-)

[John Hubbard <jhubbard@nvidia.com>]

On 8/21/19 11:13 AM, 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..
> 
> IMHO the pin refcnt is held by the driver char dev FD, that is the
> object you need to make it visible against.


If you do that, it might make it a lot simpler to add lease support
to drivers like XDP, which is otherwise looking pretty difficult to
set up with an fd. (It's socket-based, and not immediately clear where
to connect up the fd.)


thanks,
-- 
John Hubbard
NVIDIA

> 
> 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?
> 
> Jason
>