Re: [PATCH] mm,shmem: Fix a typo in shmem_swapin_page()

[Huang, Ying]

Hi, Andrew,

huang ying <huang.ying.caritas@xxxxxxxxx> writes:
>>
>> Thanks for catching that; and as David says, it's worse than a typo.
>>
>> But this is not the right fix:
>> 2efa33fc7f6e ("mm/shmem: fix shmem_swapin() race with swapoff")
>> needs to be reverted.
>>
>> It's been on my pile to look at for weeks: now I look at it and see
>> it's just a bad patch.  Over-enthusiastic stablehands already rushed
>> it out, I was wary, and reverts are already in -rc for 5.13 and 5.10,
>> phew, but 5.12.19 EOL is stuck with it unfortunately, oh well.
>>
>> I was wary because, if the (never observed) race to be fixed is in
>> swap_cluster_readahead(), why was shmem_swapin_page() being patched?
>
> When we get a swap entry from the page table or shmem xarray, and no
> necessary lock is held to prevent the swap device to be swapoff (e.g.
> page table lock, page lock, etc.), it's possible that the swap device
> has been swapoff when we operate on the swap entry (e.g. swapin).  So
> we need to find a way to prevent the swap device to be swapoff,
> get_swap_device() based on percpu_ref is used for that.  To avoid to
> call get_swap_device() here and there (e.g. now it is called in many
> different places), I think it's better to call get_swap_device() when
> we just get a swap entry without holding the necessary lock, that is,
> in do_swap_page() and shmem_swapin_page(), etc.  So that we can delete
> the get_swap_device() call in lookup_swap_cache(),
> __read_swap_cache_async(), etc.  This will make it easier to
> understand when to use get_swap_device() and clean up the code.  Do
> you agree?
>
>> Not explained in its commit message, probably a misunderstanding of
>> how mm/shmem.c already manages races (and prefers not to be involved
>> in swap_info_struct stuff).
>
> Yes.  The commit message isn't clean enough about why we do that.
>
>> But why do I now say it's bad?  Because even if you correct the EINVAL
>> to -EINVAL, that's an unexpected error: -EEXIST is common, -ENOMEM is
>> not surprising, -ENOSPC can need consideration, but -EIO and anything
>> else just end up as SIGBUS when faulting (or as error from syscall).
>
> Yes.  -EINVAL isn't a good choice.  If it's the swapoff race, then
> retrying can fix the race, so -EAGAIN may be a choice.  But if the
> swap entry is really invalid (almost impossible in theory), we may
> need something else, for example, WARN_ON_ONCE() and SIGBUS?  This
> reminds me that we may need to distinguish the two possibilities in
> get_swap_device()?

As Hugh pointed out, EINVAL isn't an appropriate error code for race
condition.  After checking the code, I found that EEXIST is the error
code used for race condition.  So I revise the patch as below.  If Hugh
doesn't object, can you help to replace the patch with the below one?

Best Regards,
Huang, Ying

-----------------------------8<---------------------------------------