Re: [PATCH v7 00/21] mm/zsmalloc: add zpdesc memory descriptor for zswap.zpool

[Hyeonggon Yoo]

On Tue, Dec 3, 2024 at 10:58 PM Alex Shi <seakeel@xxxxxxxxx> wrote:
>
>
>
> On 12/3/24 04:05, Vishal Moola wrote:
> > On Wed, Sep 04, 2024 at 01:04:11PM -0700, Vishal Moola wrote:
> >> On Wed, Sep 04, 2024 at 02:54:14PM +0800, Alex Shi wrote:
> >>>
> >>>
> >>> On 9/2/24 3:21 PM, alexs@xxxxxxxxxx wrote:
> >>>> From: Alex Shi <alexs@xxxxxxxxxx>
> >>>>
> >>> ...
> >>>
> >>>>
> >>>> This patchset abstracts the memory descriptor used in zsmalloc by zswap/zram.
> >>>> The descriptor still overlays the struct page; nothing has changed
> >>>> in that regard. What this patchset accomplishes is the use of folios in
> >>>> to save some code size, and the introduction of a new concept, zpdesc.
> >>>> This patchset is just an initial step; it does not bias the potential
> >>>> changes to kmem_alloc or larger zspage modifications.
> >>>>
> >>> ...
> >>>>
> >>>> Thanks a lot for comments and suggestion from Yosry, Yoo, Sergey, Willy
> >>>> and Vishal!
> >>>>
> >>>
> >>> This patchset could save 6.3% code size, and it's a nice abstract of zsmalloc
> >>> memory usage.
> >>> Is there any more comments, or mind to give a reviewed-by?
> >>
> >> Please CC me on future versions. Most of the zsmalloc conversions seem
> >> ok, but I'd hold off on further iterations of the descriptor patches until
> >> the maintainers decide on what/how this descriptor will be used
> >> (i.e. our end goals).
> >
> > I apologize for leaving this in limbo for this long. This patchset is a
> > prerequisite to shrinking struct page, so we should get this memdesc in.
> >
> > I think it's safe to assume (since we've heard no definitive goal from
> > the maintainers) that in our memdesc world we want zsmalloc + zspage to be
> > similar to how it currently looks today.
> >
> > Would you like to rebase this on the current mm-unstable? I'll re-review
> > it in case anything changed (and can then give you my reviewed-by).
> Hi, Vishal,
>
> Thanks a lot for the review and kindness, I will try to do the rebase in this weekend.

For your information, there is a patch series (not merged yet)
invalidating the assumption
that zsmalloc always allocates order-0 pages [1].

I'm sorry I wasn't able to dedicate more time to this project this year.
Thank you for pushing this forward, Alex!

[1] https://lore.kernel.org/linux-mm/20241121222521.83458-1-21cnbao@xxxxxxxxx/

Best,
Hyeonggon