Re: [RFC] Free up a page flag

[David Hildenbrand]

On 09.03.22 23:07, Linus Torvalds wrote:
> On Wed, Mar 9, 2022 at 12:50 PM Matthew Wilcox <willy@xxxxxxxxxxxxx> wrote:
>>
>> We're always running out of page flags.  Here's an attempt to free one
>> up for the next time somebody wants one.
> 
> Ugh. This is too ugly for words.
> 
> I wouldn't mind something along the conceptual lines of "these bits
> are only used for this type", but I think it would need to be much
> more organized and explicit, not this kind of randomness.
> 
> For example, quite a few of the page bits really only make sense for
> the "page cache and anonymous pages" kind.
> 
> I think this includes some really fundamental bits like the lock bit
> (and the associated waiters bit), along with a lot of the "owner" aka
> "this can be used by the filesystem" bits.
> 
> I think it _also_ includes all the LRU and workingset bits etc.
> 
> So if we consider that kind of case the "normal" case, the not-normal
> case is likely (a) slab, (b) reserved pages and (c) zspages.,
> 
> Which is pretty close to your "xyzzy" bit (I think you came to the
> same list of "slab or reserved" conclusion because of the fundamental
> issues above), but my point is that I think this approach is
> acceptable if we make it much less random, and make it a lot more
> explicit and thought through.
> 
> And we'd probably need to actually *verify* that we don't do things
> like lock (or LRU) those non-normal pages.
> 

Looking at isolate_movable_page(), I think we can easily end up locking
random pages temporarily. And especially non-lru page migration makes
use of it as well.

Personally, I'd not try restricting PG_lock and PG_waiter for specific
page types, it IMHO a way to generic infrastructure.

Other page flags are different and we already reuse them in different
context: e.g., reusing PG_uptodate for buddy pages via PG_reported.

-- 
Thanks,

David / dhildenb