[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAAa6QmR6KeCCMHGyR9jkwk-KCwb_FTcEfnKeHRJ2U7+3SVz7hg@mail.gmail.com>
Date: Tue, 17 Jan 2023 15:00:30 -0800
From: "Zach O'Keefe" <zokeefe@...gle.com>
To: David Hildenbrand <david@...hat.com>
Cc: Yang Shi <shy828301@...il.com>, Hugh Dickins <hughd@...gle.com>,
Andrew Morton <akpm@...ux-foundation.org>,
Jann Horn <jannh@...gle.com>, Song Liu <songliubraving@...com>,
linux-kernel@...r.kernel.org, linux-mm@...ck.org
Subject: Re: [PATCH] mm/khugepaged: fix collapse_pte_mapped_thp() to allow anon_vma
On Mon, Jan 9, 2023 at 12:50 AM David Hildenbrand <david@...hat.com> wrote:
>
> >>>>>>
> >>>>>> Side note: set_huge_pmd() wins the award of "ugliest mm function of early
> >>>>>> 2023". I was briefly concerned how do_set_pmd() decides whether the PMD can be
> >>>>>> writable or not. Turns out it's communicated via vm_fault->flags. Just
> >>>>>> horrible.
> >
Hey David,
Sorry for the late response here.
> > My first Linux award! :) At least it's not "worst mm security issue of
> > early 2023". I'll take it!
>
> Good that you're not taking my words the wrong way.
>
> MADV_COLLAPSE is a very useful feature (especially also for THP tests
> [1]). I wish I could have looked at some of the patches earlier. But we
> cannot wait forever to get something merged, otherwise we'd never get
> bigger changes upstream.
>
> ... so there is plenty of time left in 2023 to cleanup khugepaged.c :P
>
>
> [1] https://lkml.kernel.org/r/20230104144905.460075-1-david@redhat.com
Yes, thank you for these tests. I have them open in another tab along
with a mental TODO to take a closer look at them, and response
half-written. In-place collapse of anonymous memory *is* something
that I was interested in exploring later (I have a use-case for it;
hugepage-aware malloc() implementations). I'm taking so long on it
(sorry) b/c I need to review your point (2) (all PTE's mapping
exclusively). Hopefully I can get to it shortly.
> [...]
>
>
> >> For example: why even *care* about the complexity of installing a PMD in
> >> collapse_pte_mapped_thp() using set_huge_pmd() just for MADV_COLLAPSE?
> >>
> >> Sure, we avoid a single page fault afterwards, but is this *really*
> >> worth the extra code here? I mean, after we installed the PMD, the page
> >> could just get reclaimed either way, so there is no guarantee that we
> >> have a PMD mapped once we return to user space IIUC.
> >
> > A valid question. The first reason is just semantic symmetry for
> > MADV_COLLAPSE called on anon vs file/shmem memory. It would be nice to
> > say that "on success, the memory range provided will be backed by
> > PMD-mapped hugepages", rather than special-casing file/shmem.
>
> But there will never be such a guarantee, right? We could even see a
> split before just before we return to user space IIRC.
Absolutely. But at least we are *attempting* for symmetry here; though
admittedly, even a successful return code provides no guarantees.
Perhaps this is a weak argument by itself, though.
> >
> > The second reason has a more practical use case. In userfaultfd-based
> > live migration (using UFFDIO_REGISTER_MODE_MINOR) pages are migrated
> > at 4KiB granularity, and it may take a long (O(many minutes)) for the
> > transfer of all pages to complete. To avoid severe performance
> > degradation on the target guest, the vmm wants to MADV_COLLAPSE
> > hugepage-sized regions as they fill up. Since the guest memory is
> > still uffd-registered, requiring refault post-MADV_COLLAPSE won't
> > work, since the uffd machinery will intercept the fault, and no PMD
> > will be mapped. As such, either uffd needs to be taught to install PMD
> > mappings, or the PMD mapping already must be in-place.
>
> That's an interesting point, thanks. I assume we'd get another minor
> fault and when resolving that, we'll default to a PTE mapping.
Yes-ish; I think it depends on how userspace decides to deal with the
event. At least in my own test cases, IIRC (hazy memory here), we
ended up in some loop of:
done faulting all 512 pages -> MADV_COLLAPSE -> fault -> copy page ->
done faulting all 512 pages -> ...
Thanks,
Zach
> --
> Thanks,
>
> David / dhildenb
>
Powered by blists - more mailing lists