Re: [RFC PATCH v2 00/21] mm/zsmalloc: Split zsdesc from struct page

[Yosry Ahmed]

<snip>
>
> > > > It seems to me though the sizeof(zsdesc) is actually 56 bytes (on
> > > > 64-bit), so sizeof(zsdesc) + sizeof(memdesc) would be equal to the
> > > > current size of struct page. If that's true, then there is no loss,
> > >
> > > Yeah, zsdesc would be 56 bytes on 64 bit CPUs as memcg_data field is
> > > not used in zsmalloc.
> > > More fields in the current struct page might not be needed in the
> > > future, although it's hard to say at the moment.
> > > but it's not a loss.
> >
> > Is page->memcg_data something that we can drop? Aren't there code
> > paths that will check page->memcg_data even for kernel pages (e.g.
> > __folio_put() -> __folio_put_small() -> mem_cgroup_uncharge() ) ?
>
> zsmalloc pages are not accounted for via __GFP_ACCOUNT,

Yeah, but the code in the free path above will check page->memcg_data
nonetheless to check if it is charged. I think to drop memcg_data we
need to enlighten the code that some pages do not even have memcg_data
at all, no?

> and IIUC the current implementation of zswap memcg charging does not
> use memcg_data
> either - so I think it can be dropped.

My question is more about the generic mm code expecting to see
page->memcg_data in every page, even if it is not actually used
(zero).

>
> I think we don't want to increase memdesc to 16 bytes by adding memcg_data.
> It should be in use-case specific descriptors if it can be charged to memcg?
>
<snip>