On Thu, Mar 30, 2017 at 8:25 AM, Michal Hocko <mhocko@xxxxxxxxxx> wrote:
> On Wed 29-03-17 16:25:18, Ilya Dryomov wrote:
>> On Wed, Mar 29, 2017 at 1:16 PM, Michal Hocko <mhocko@xxxxxxxxxx> wrote:
>> > On Wed 29-03-17 13:10:01, Ilya Dryomov wrote:
>> >> On Wed, Mar 29, 2017 at 12:55 PM, Michal Hocko <mhocko@xxxxxxxxxx> wrote:
>> >> > On Wed 29-03-17 12:41:26, Michal Hocko wrote:
>> >> > [...]
>> >> >> > ceph_con_workfn
>> >> >> > mutex_lock(&con->mutex) # ceph_connection::mutex
>> >> >> > try_write
>> >> >> > ceph_tcp_connect
>> >> >> > sock_create_kern
>> >> >> > GFP_KERNEL allocation
>> >> >> > allocator recurses into XFS, more I/O is issued
>> >> >
>> >> > One more note. So what happens if this is a GFP_NOIO request which
>> >> > cannot make any progress? Your IO thread is blocked on con->mutex
>> >> > as you write below but the above thread cannot proceed as well. So I am
>> >> > _really_ not sure this acutally helps.
>> >>
>> >> This is not the only I/O worker. A ceph cluster typically consists of
>> >> at least a few OSDs and can be as large as thousands of OSDs. This is
>> >> the reason we are calling sock_create_kern() on the writeback path in
>> >> the first place: pre-opening thousands of sockets isn't feasible.
>> >
>> > Sorry for being dense here but what actually guarantees the forward
>> > progress? My current understanding is that the deadlock is caused by
>> > con->mutext being held while the allocation cannot make a forward
>> > progress. I can imagine this would be possible if the other io flushers
>> > depend on this lock. But then NOIO vs. KERNEL allocation doesn't make
>> > much difference. What am I missing?
>>
>> con->mutex is per-ceph_connection, osdc->request_mutex is global and is
>> the real problem here because we need both on the submit side, at least
>> in 3.18. You are correct that even with GFP_NOIO this code may lock up
>> in theory, however I think it's very unlikely in practice.
>
> No, it would just make such a bug more obscure. The real problem seems
> to be that you rely on locks which cannot guarantee a forward progress
> in the IO path. And that is a bug IMHO.
>
>> We got rid of osdc->request_mutex in 4.7, so these workers are almost
>> independent in newer kernels and should be able to free up memory for
>> those blocked on GFP_NOIO retries with their respective con->mutex
>> held. Using GFP_KERNEL and thus allowing the recursion is just asking
>> for an AA deadlock on con->mutex OTOH, so it does make a difference.
>
> You keep saying this but so far I haven't heard how the AA deadlock is
> possible. Both GFP_KERNEL and GFP_NOIO can stall for an unbounded amount
> of time and that would cause you problems AFAIU.
>
>> I'm a little confused by this discussion because for me this patch was
>> a no-brainer...
>
> No, it is a brainer. Because recursion prevention should be carefully
> thought through. The lack of this approach has caused that we have
> thousands of GFP_NOFS uses all over the kernel without a clear or proper
> justification. Adding more on top doesn't help long term
> maintainability.
>
>> Locking aside, you said it was the stack trace in the changelog that
>> got your attention
>
> No, it is the usage of the scope GFP_NOIO API usage without a proper
> explanation which caught my attention.
>
>> are you saying it's OK for a block
>> device to recurse back into the filesystem when doing I/O, potentially
>> generating more I/O?
>
> No, block device has to make a forward progress guarantee when
> allocating and so use mempools or other means to achieve the same.
OK, let me put this differently. Do you agree that a block device
cannot make _any_ kind of progress guarantee if it does a GFP_KERNEL
allocation in the I/O path?
Thanks,
Ilya