[PATCH v3 04/20] mm: VMA sequence count

[Laurent Dufour]

On 14/09/2017 10:13, Sergey Senozhatsky wrote:
> Hi,
> 
> On (09/14/17 09:55), Laurent Dufour wrote:
> [..]
>>> so if there are two CPUs, one doing write_seqcount() and the other one
>>> doing read_seqcount() then what can happen is something like this
>>>
>>> 	CPU0					CPU1
>>>
>>> 						fs_reclaim_acquire()
>>> 	write_seqcount_begin()
>>> 	fs_reclaim_acquire()			read_seqcount_begin()
>>> 	write_seqcount_end()
>>>
>>> CPU0 can't write_seqcount_end() because of fs_reclaim_acquire() from
>>> CPU1, CPU1 can't read_seqcount_begin() because CPU0 did write_seqcount_begin()
>>> and now waits for fs_reclaim_acquire(). makes sense?
>>
>> Yes, this makes sense.
>>
>> But in the case of this series, there is no call to
>> __read_seqcount_begin(), and the reader (the speculative page fault
>> handler), is just checking for (vm_seq & 1) and if this is true, simply
>> exit the speculative path without waiting.
>> So there is no deadlock possibility.
> 
> probably lockdep just knows that those locks interleave at some
> point.
> 
> 
> by the way, I think there is one path that can spin
> 
> find_vma_srcu()
>  read_seqbegin()
>   read_seqcount_begin()
>    raw_read_seqcount_begin()
>     __read_seqcount_begin()


That's right, but here this is the  sequence counter mm->mm_seq, not the
vm_seq one.

This one is held to protect walking the VMA list "locklessly"...

Cheers,
Laurent.