From: Mike Kravetz <mike.kravetz@oracle.com>
To: Peter Xu <peterx@redhat.com>
Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org,
James Houghton <jthoughton@google.com>,
Jann Horn <jannh@google.com>,
Andrew Morton <akpm@linux-foundation.org>,
Andrea Arcangeli <aarcange@redhat.com>,
Rik van Riel <riel@surriel.com>,
Nadav Amit <nadav.amit@gmail.com>,
Miaohe Lin <linmiaohe@huawei.com>,
Muchun Song <songmuchun@bytedance.com>,
David Hildenbrand <david@redhat.com>
Subject: Re: [PATCH 03/10] mm/hugetlb: Document huge_pte_offset usage
Date: Mon, 5 Dec 2022 13:47:59 -0800 [thread overview]
Message-ID: <Y45nD5HjHYtBjJ5V@monkey> (raw)
In-Reply-To: <Y4d9opX7wok4GCdb@x1n>
On 11/30/22 10:58, Peter Xu wrote:
> Hi, Mike,
>
> On Tue, Nov 29, 2022 at 08:55:21PM -0800, Mike Kravetz wrote:
> > > + * (2) For shared mappings: pmd unsharing is possible (so the PUD-ranged
> > > + * pgtable page can go away from under us! It can be done by a pmd
> > > + * unshare with a follow up munmap() on the other process), then we
> > > + * need either:
> > > + *
> > > + * (2.1) hugetlb vma lock read or write held, to make sure pmd unshare
> > > + * won't happen upon the range (it also makes sure the pte_t we
> > > + * read is the right and stable one), or,
> > > + *
> > > + * (2.2) hugetlb mapping i_mmap_rwsem lock held read or write, to make
> > > + * sure even if unshare happened the racy unmap() will wait until
> > > + * i_mmap_rwsem is released.
> >
> > Is that 100% correct? IIUC, the page tables will be released via the
> > call to tlb_finish_mmu(). In most cases, the tlb_finish_mmu() call is
> > performed when holding i_mmap_rwsem. However, in the final teardown of
> > a hugetlb vma via __unmap_hugepage_range_final, the tlb_finish_mmu call
> > is done outside the i_mmap_rwsem lock. In this case, I think we are
> > still safe because nobody else should be walking the page table.
> >
> > I really like the documentation. However, if i_mmap_rwsem is not 100%
> > safe I would prefer not to document it here. I don't think anyone
> > relies on this do they?
>
> I think i_mmap_rwsem is 100% safe.
>
> It's not in tlb_finish_mmu(), but when freeing the pgtables we need to
> unlink current vma from the vma list first:
>
> free_pgtables
> unlink_file_vma
> i_mmap_lock_write
> tlb_finish_mmu
Thanks!
Sorry, I was thinking about page freeing not page table freeing.
Agree that is 100% safe.
--
Mike Kravetz
next prev parent reply other threads:[~2022-12-05 21:51 UTC|newest]
Thread overview: 60+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-11-29 19:35 [PATCH 00/10] mm/hugetlb: Make huge_pte_offset() thread-safe for pmd unshare Peter Xu
2022-11-29 19:35 ` [PATCH 01/10] mm/hugetlb: Let vma_offset_start() to return start Peter Xu
2022-11-30 10:11 ` David Hildenbrand
2022-11-29 19:35 ` [PATCH 02/10] mm/hugetlb: Don't wait for migration entry during follow page Peter Xu
2022-11-30 4:37 ` Mike Kravetz
2022-11-30 10:15 ` David Hildenbrand
2022-11-29 19:35 ` [PATCH 03/10] mm/hugetlb: Document huge_pte_offset usage Peter Xu
2022-11-30 4:55 ` Mike Kravetz
2022-11-30 15:58 ` Peter Xu
2022-12-05 21:47 ` Mike Kravetz [this message]
2022-11-30 10:21 ` David Hildenbrand
2022-11-30 10:24 ` David Hildenbrand
2022-11-30 16:09 ` Peter Xu
2022-11-30 16:11 ` David Hildenbrand
2022-11-30 16:25 ` Peter Xu
2022-11-30 16:31 ` David Hildenbrand
2022-11-29 19:35 ` [PATCH 04/10] mm/hugetlb: Move swap entry handling into vma lock when faulted Peter Xu
2022-12-05 22:14 ` Mike Kravetz
2022-12-05 23:36 ` Peter Xu
2022-11-29 19:35 ` [PATCH 05/10] mm/hugetlb: Make userfaultfd_huge_must_wait() safe to pmd unshare Peter Xu
2022-11-30 16:08 ` David Hildenbrand
2022-12-05 22:23 ` Mike Kravetz
2022-11-29 19:35 ` [PATCH 06/10] mm/hugetlb: Make hugetlb_follow_page_mask() " Peter Xu
2022-11-30 16:09 ` David Hildenbrand
2022-12-05 22:29 ` Mike Kravetz
2022-11-29 19:35 ` [PATCH 07/10] mm/hugetlb: Make follow_hugetlb_page() " Peter Xu
2022-11-30 16:09 ` David Hildenbrand
2022-12-05 22:45 ` Mike Kravetz
2022-11-29 19:35 ` [PATCH 08/10] mm/hugetlb: Make walk_hugetlb_range() " Peter Xu
2022-11-30 16:11 ` David Hildenbrand
2022-12-05 23:33 ` Mike Kravetz
2022-12-05 23:52 ` John Hubbard
2022-12-06 16:45 ` Peter Xu
2022-12-06 18:50 ` Mike Kravetz
2022-12-06 21:03 ` John Hubbard
2022-12-06 21:51 ` Peter Xu
2022-12-06 22:31 ` John Hubbard
2022-12-07 0:07 ` Peter Xu
2022-12-07 2:38 ` John Hubbard
2022-12-07 14:58 ` Peter Xu
2022-11-29 19:35 ` [PATCH 09/10] mm/hugetlb: Make page_vma_mapped_walk() " Peter Xu
2022-11-30 16:18 ` David Hildenbrand
2022-11-30 16:32 ` Peter Xu
2022-11-30 16:39 ` David Hildenbrand
2022-12-05 23:52 ` Mike Kravetz
2022-12-06 17:10 ` Mike Kravetz
2022-12-06 17:39 ` Peter Xu
2022-12-06 17:43 ` Peter Xu
2022-12-06 19:58 ` Mike Kravetz
2022-11-29 19:35 ` [PATCH 10/10] mm/hugetlb: Introduce hugetlb_walk() Peter Xu
2022-11-30 5:18 ` Eric Biggers
2022-11-30 15:37 ` Peter Xu
2022-12-06 0:21 ` Mike Kravetz
2022-11-29 20:49 ` [PATCH 00/10] mm/hugetlb: Make huge_pte_offset() thread-safe for pmd unshare Andrew Morton
2022-11-29 21:19 ` Peter Xu
2022-11-29 21:26 ` Andrew Morton
2022-11-29 20:51 ` Andrew Morton
2022-11-29 21:36 ` Peter Xu
2022-11-30 9:46 ` David Hildenbrand
2022-11-30 16:23 ` Peter Xu
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=Y45nD5HjHYtBjJ5V@monkey \
--to=mike.kravetz@oracle.com \
--cc=aarcange@redhat.com \
--cc=akpm@linux-foundation.org \
--cc=david@redhat.com \
--cc=jannh@google.com \
--cc=jthoughton@google.com \
--cc=linmiaohe@huawei.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=nadav.amit@gmail.com \
--cc=peterx@redhat.com \
--cc=riel@surriel.com \
--cc=songmuchun@bytedance.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).