Re: [Lsf-pc] [LSF/MM/BPF TOPIC] Reclamation interactions with RCU

[Kent Overstreet <kent.overstreet@linux.dev>]

On Fri, Mar 01, 2024 at 04:54:55PM +1100, Dave Chinner wrote:
> On Fri, Mar 01, 2024 at 01:16:18PM +1100, NeilBrown wrote:
> > While we are considering revising mm rules, I would really like to
> > revised the rule that GFP_KERNEL allocations are allowed to fail.
> > I'm not at all sure that they ever do (except for large allocations - so
> > maybe we could leave that exception in - or warn if large allocations
> > are tried without a MAY_FAIL flag).
> > 
> > Given that GFP_KERNEL can wait, and that the mm can kill off processes
> > and clear cache to free memory, there should be no case where failure is
> > needed or when simply waiting will eventually result in success.  And if
> > there is, the machine is a gonner anyway.
> 
> Yes, please!
> 
> XFS was designed and implemented on an OS that gave this exact
> guarantee for kernel allocations back in the early 1990s.  Memory
> allocation simply blocked until it succeeded unless the caller
> indicated they could handle failure. That's what __GFP_NOFAIL does
> and XFS is still heavily dependent on this behaviour.

I'm not saying we should get rid of __GFP_NOFAIL - actually, I'd say
let's remove the underscores and get rid of the silly two page limit.
GFP_NOFAIL|GFP_KERNEL is perfectly safe for larger allocations, as long
as you don't mind possibly waiting a bit.

But it can't be the default because, like I mentioned to Neal, there are
a _lot_ of different places where we allocate memory in the kernel, and
they have to be able to fail instead of shoving everything else out of
memory.

> This is the sort of thing I was thinking of in the "remove
> GFP_NOFS" discussion thread when I said this to Kent:
> 
> 	"We need to start designing our code in a way that doesn't require
> 	extensive testing to validate it as correct. If the only way to
> 	validate new code is correct is via stochastic coverage via error
> 	injection, then that is a clear sign we've made poor design choices
> 	along the way."
> 
> https://lore.kernel.org/linux-fsdevel/ZcqWh3OyMGjEsdPz@dread.disaster.area/
> 
> If memory allocation doesn't fail by default, then we can remove the
> vast majority of allocation error handling from the kernel. Make the
> common case just work - remove the need for all that code to handle
> failures that is hard to exercise reliably and so are rarely tested.
> 
> A simple change to make long standing behaviour an actual policy we
> can rely on means we can remove both code and test matrix overhead -
> it's a win-win IMO.

We definitely don't want to make GFP_NOIO/GFP_NOFS allocations nofail by
default - a great many of those allocations have mempools in front of
them to avoid deadlocks, and if you do that you've made the mempools
useless.