Message-ID: <1429550126.7346.268.camel@edumazet-glaptop2.roam.corp.google.com>
Subject: Re: [RFC PATCH] fs: use a sequence counter instead of file_lock in
 fd_install
From: Eric Dumazet <eric.dumazet@gmail.com>
To: Mateusz Guzik <mguzik@redhat.com>
Cc: Al Viro <viro@ZenIV.linux.org.uk>,
        Andrew Morton <akpm@linux-foundation.org>,
        "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>,
        Yann Droneaud <ydroneaud@opteya.com>,
        Konstantin Khlebnikov <khlebnikov@yandex-team.ru>,
        linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org
Date: Mon, 20 Apr 2015 10:15:26 -0700
In-Reply-To: <20150420151054.GD2513@mguzik>
References: <20150416121628.GA20615@mguzik>
	 <1429307216.7346.255.camel@edumazet-glaptop2.roam.corp.google.com>
	 <20150417221646.GA15589@mguzik> <20150417230252.GE889@ZenIV.linux.org.uk>
	 <20150420130633.GA2513@mguzik> <20150420134326.GC2513@mguzik>
	 <20150420151054.GD2513@mguzik>
Content-Type: text/plain; charset="UTF-8"
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1222
Lines: 30

On Mon, 2015-04-20 at 17:10 +0200, Mateusz Guzik wrote:

> Sorry for spam but I came up with another hack. :)
> 
> The idea is that we can have a variable which would signify the that
> given thread is playing with fd table in fd_install (kind of a lock
> embedded into task_struct). We would also have a flag in files struct
> indicating that a thread would like to resize it.
> 
> expand_fdtable would set the flag and iterate over all threads waiting
> for all of them to have the var set to 0.

The opposite : you have to block them in some way and add a rcu_sched()
or something.

Another way would be to expand the table leaving the old one in place,
thanks to a new vrealloc() api. This would require all file tables to at
least use one page.

(Instead of use a new set of pages for the new vmalloc()ed area, reuse
the pages that are already mapped into previous vmalloc() area.

Right now, expanding 4M slots to 8M slots is taking a long time and is a
latency killer.

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/