Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Sun, 3 Mar 2019 02:27:02 -0500 (EST)
From:   Jan Stancek <jstancek@redhat.com>
To:     Andrea Arcangeli <aarcange@redhat.com>, peterz@infradead.org
Cc:     linux-mm@kvack.org, akpm@linux-foundation.org, willy@infradead.org,
        riel@surriel.com, mhocko@suse.com,
        ying huang <ying.huang@intel.com>,
        jrdr linux <jrdr.linux@gmail.com>, jglisse@redhat.com,
        aneesh kumar <aneesh.kumar@linux.ibm.com>, david@redhat.com,
        raquini@redhat.com, rientjes@google.com, kirill@shutemov.name,
        mgorman@techsingularity.net, linux-kernel@vger.kernel.org
Message-ID: <701776300.4537344.1551598022408.JavaMail.zimbra@redhat.com>
In-Reply-To: <20190302185144.GD31083@redhat.com>
References: <20190302171043.GP11592@bombadil.infradead.org> <a5234d11b8cc158352a2f97fc33aa9ad90bb287b.1551550112.git.jstancek@redhat.com> <20190302185144.GD31083@redhat.com>
Subject: Re: [PATCH v2] mm/memory.c: do_fault: avoid usage of stale
 vm_area_struct
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
Thread-Topic: mm/memory.c: do_fault: avoid usage of stale vm_area_struct
Thread-Index: pAmRXCN5CetzZ5PIfXitNyEm+tfURQ==
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


----- Original Message -----
> Hello Jan,
> 
> On Sat, Mar 02, 2019 at 07:19:39PM +0100, Jan Stancek wrote:
> > +	struct mm_struct *vm_mm = READ_ONCE(vma->vm_mm);
> 
> The vma->vm_mm cannot change under gcc there, so no need of
> READ_ONCE. The release of mmap_sem has release semantics so the
> vma->vm_mm access cannot be reordered after up_read(mmap_sem) either.
> 
> Other than the above detail:
> 
> Reviewed-by: Andrea Arcangeli <aarcange@redhat.com>

Thank you for review, I dropped READ_ONCE and sent v3 with your
Reviewed-by included. I also successfully re-ran tests over-night.

> Would this not need a corresponding WRITE_ONCE() in vma_init() ?

There's at least 2 context switches between, so I think it wouldn't matter.
My concern was gcc optimizing out vm_mm, and vma->vm_mm access happening only
after do_read_fault().