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;
From:   Minchan Kim <minchan.kernel@gmail.com>
Date:   Tue, 15 May 2018 17:56:10 +0900
To:     Martijn Coenen <maco@android.com>
Cc:     Minchan Kim <minchan.kernel@gmail.com>,
        Joel Fernandes <joel@joelfernandes.org>,
        LKML <linux-kernel@vger.kernel.org>,
        Ganesh Mahendran <opensource.ganesh@gmail.com>,
        Joe Perches <joe@perches.com>,
        Arve =?iso-8859-1?B?SGr4bm5lduVn?= <arve@android.com>,
        Todd Kjos <tkjos@google.com>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Subject: Re: [PATCH v6] ANDROID: binder: change down_write to down_read
Message-ID: <20180515085610.GA98128@rodete-desktop-imager.corp.google.com>
References: <20180507141537.4855-1-minchan@kernel.org>
 <20180507172829.GA66161@joelaf.mtv.corp.google.com>
 <20180508105101.GB8209@rodete-desktop-imager.corp.google.com>
 <20180508230813.GA102094@joelaf.mtv.corp.google.com>
 <20180509064023.GD8209@rodete-desktop-imager.corp.google.com>
 <20180509231941.GA4790@joelaf.mtv.corp.google.com>
 <20180515062732.GA183759@rodete-desktop-imager.corp.google.com>
 <CAB0TPYHseQ-Lx4guzykjxY9xP2yP5Kga+oZMTmC7Bb7XXRAFFQ@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <CAB0TPYHseQ-Lx4guzykjxY9xP2yP5Kga+oZMTmC7Bb7XXRAFFQ@mail.gmail.com>
User-Agent: Mutt/1.9.2 (2017-12-15)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Tue, May 15, 2018 at 09:46:01AM +0200, Martijn Coenen wrote:

< snip >

> >> About the unmap at runtime part, your commit message was a bit confusing. You
> >> said "every binder buffers should be mapped in advance by binder_mmap." but I
> >> think the new binder shrinker mechanism doesn't make that true anymore.
> >
> > It's good point. I didn't know know that.
> > When I see binder_vm_fault, it emits SIGBUS. That means shrinker cannot zap pages
> > process is using, I think. IOW, every pages for binder are mapped at mmap time
> > and is never mapped in runtime by page fault. Right?
> 
> Right - the address range is allocated once, and an initial amount of
> pages is mapped into it. For every transaction into that process, we
> will see if there's enough pages, and if not allocate so that we have
> enough of them - so this is not done by page fault. The shrinker won't
> touch pages for which a transaction is in progress. Of course a
> process itself could still try to read from an unallocated address,
> but in that case returning SIGBUS and having that process crash seems
> fine.

Thanks for the confirmation.

> 
> I'm also not sure the read lock is needed, but I would need to read a
> whole lot more code to convince myself it's not.

For page zapping, we shouldn't need mmap_sem write lock.
We should replace it with down_read/write, too.

Thanks.