Feedback-ID: ie3994620:Fastmail
Date:   Tue, 24 Jan 2023 03:51:02 +0300
From:   "Kirill A. Shutemov" <kirill@shutemov.name>
To:     David Hildenbrand <david@redhat.com>, g@box.shutemov.name
Cc:     Jann Horn <jannh@google.com>,
        Andrew Morton <akpm@linux-foundation.org>, linux-mm@kvack.org,
        "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>,
        Zach O'Keefe <zokeefe@google.com>,
        linux-kernel@vger.kernel.org, Yang Shi <shy828301@gmail.com>
Subject: Re: [PATCH] mm/khugepaged: Fix ->anon_vma race
Message-ID: <20230124005102.ow33h7tfmj5eqyr3@box.shutemov.name>
References: <20230111133351.807024-1-jannh@google.com>
 <20230112085649.gvriasb2t5xwmxkm@box.shutemov.name>
 <CAG48ez3434wZBKFFbdx4M9j6eUwSUVPd4dxhzW_k_POneSDF+A@mail.gmail.com>
 <20230115190654.mehtlyz2rxtg34sl@box.shutemov.name>
 <CAG48ez2zeQ4+g1=B4eyrrvZRYMr1S1xKBh2_eAhCjVjhj7Lpfg@mail.gmail.com>
 <20230116123403.fiyv22esqgh7bzp3@box.shutemov.name>
 <5a7fdfa7-5b25-0ed4-2479-661d387b397b@redhat.com>
 <20230116134710.n4dgtrutt6rqif62@box.shutemov.name>
 <1b8696ec-e2be-7b7b-705c-e2dcabb2e8e5@redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <1b8696ec-e2be-7b7b-705c-e2dcabb2e8e5@redhat.com>
Precedence: bulk

On Mon, Jan 23, 2023 at 12:07:05PM +0100, David Hildenbrand wrote:
> > > commit bedf03416913d88c796288f9dca109a53608c745
> > > Author: Yang Shi <shy828301@gmail.com>
> > > Date:   Wed Sep 7 11:01:44 2022 -0700
> > > 
> > >      powerpc/64s/radix: don't need to broadcast IPI for radix pmd collapse flush
> > >      The IPI broadcast is used to serialize against fast-GUP, but fast-GUP will
> > >      move to use RCU instead of disabling local interrupts in fast-GUP.  Using
> > >      an IPI is the old-styled way of serializing against fast-GUP although it
> > >      still works as expected now.
> > >      And fast-GUP now fixed the potential race with THP collapse by checking
> > >      whether PMD is changed or not.  So IPI broadcast in radix pmd collapse
> > >      flush is not necessary anymore.  But it is still needed for hash TLB.
> > 
> > Okay. But I think tlb_remove_table_sync_one() belongs inside
> > pmdp_collapse_flush(). Collapsing pmd table into huge page without
> > serialization is a bug. They should not be separate.
> 
> Agreed. But I wonder if it should be moved into a generic
> pmdp_collapse_flush(), that calls an arch specific __pmdp_collapse_flush().

Um? There's no __pmdp_collapse_flush(). Architectural code overrides
pmdp_collapse_flush() altogether. It has know how to serialize properly
for the arch.

-- 
  Kiryl Shutsemau / Kirill A. Shutemov