Re: [PATCH v20 02/20] mm/memcg: bail early from swap accounting if memcg disabled

From: Johannes Weiner
Date: Fri Oct 30 2020 - 10:07:02 EST


On Fri, Oct 30, 2020 at 10:27:51AM +0800, Alex Shi wrote:
>
>
> 在 2020/10/29 下午9:46, Johannes Weiner 写道:
> >> ? release_pages+0x1ae/0x410
> >> shmem_alloc_and_acct_page+0x77/0x1c0
> >> shmem_getpage_gfp+0x162/0x910
> >> shmem_fault+0x74/0x210
> >> ? filemap_map_pages+0x29c/0x410
> >> __do_fault+0x37/0x190
> >> handle_mm_fault+0x120a/0x1770
> >> exc_page_fault+0x251/0x450
> >> ? asm_exc_page_fault+0x8/0x30
> >> asm_exc_page_fault+0x1e/0x30
> >>
> >> Signed-off-by: Alex Shi <alex.shi@xxxxxxxxxxxxxxxxx>
> >> Reviewed-by: Roman Gushchin <guro@xxxxxx>
> >> Acked-by: Michal Hocko <mhocko@xxxxxxxx>
> >> Acked-by: Hugh Dickins <hughd@xxxxxxxxxx>
> >> Cc: Johannes Weiner <hannes@xxxxxxxxxxx>
> >> Cc: Michal Hocko <mhocko@xxxxxxxxxx>
> >> Cc: Vladimir Davydov <vdavydov.dev@xxxxxxxxx>
> >> Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> >> Cc: cgroups@xxxxxxxxxxxxxxx
> >> Cc: linux-mm@xxxxxxxxx
> >> Cc: linux-kernel@xxxxxxxxxxxxxxx
> > Acked-by: Johannes Weiner <hannes@xxxxxxxxxxx>
> >
> > This should go in before the previous patch that adds the WARN for it.
>
> Right, but than the long ops may not weird. Should I remove the ops and resend the whole patchset?

You mean the warning in the changelog? I think that's alright. You can
just say that you're about to remove the !page->memcg check in the
next patch because the original reasons for having it are gone, and
memcg being disabled is the only remaining exception, so this patch
makes that check explicit in preparation for the next.

Sorry, it's all a bit of a hassle, I just wouldn't want to introduce a
known warning into the kernel between those two patches (could confuse
bisection runs, complicates partial reverts etc.)