Re: Known and unfixed active data loss bug in MM + XFS with large folios since Dec 2021 (any kernel from 6.1 upwards)

[Jens Axboe]

On 9/12/24 3:55 PM, Matthew Wilcox wrote:
> On Thu, Sep 12, 2024 at 11:18:34PM +0200, Christian Theune wrote:
>> This bug is very hard to reproduce but has been known to exist as a
>> ?fluke? for a while already. I have invested a number of days trying
>> to come up with workloads to trigger it quicker than that stochastic
>> ?once every few weeks in a fleet of 1.5k machines", but it eludes
>> me so far. I know that this also affects Facebook/Meta as well as
>> Cloudflare who are both running newer kernels (at least 6.1, 6.6,
>> and 6.9) with the above mentioned patch reverted. I?m from a much
>> smaller company and seeing that those guys are running with this patch
>> reverted (that now makes their kernel basically an untested/unsupported
>> deviation from the mainline) smells like desparation. I?m with a
>> much smaller team and company and I?m wondering why this isn?t
>> tackled more urgently from more hands to make it shallow (hopefully).
> 
> This passive-aggressive nonsense is deeply aggravating.  I've known
> about this bug for much longer, but like you I am utterly unable to
> reproduce it.  I've spent months looking for the bug, and I cannot.

What passive aggressiveness?! There's a data corruption bug where we
know what causes it, yet we continue to ship it. That's aggravating.

People are aware of the bug, and since there's no good reproducer, it's
hard to fix. That part is fine and understandable. What seems amiss here
is the fact that large folio support for xfs hasn't just been reverted
until the issue is understood and resolved.

When I saw Christian's report, I seemed to recall that we ran into this
at Meta too. And we did, and hence have been reverting it since our 5.19
release (and hence 6.4, 6.9, and 6.11 next). We should not be shipping
things that are known broken.

-- 
Jens Axboe