From mboxrd@z Thu Jan 1 00:00:00 1970 From: =?iso-8859-1?Q?Llu=EDs?= Batlle i Rossell Subject: More reiserfs hangs, easy to reproduce (Sheevaplug) Date: Sun, 10 Jun 2012 16:27:25 +0200 Message-ID: <20120610142724.GA2039@vicerveza.homeunix.net> Mime-Version: 1.0 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Content-Disposition: inline Sender: reiserfs-devel-owner@vger.kernel.org List-ID: Content-Type: text/plain; charset="iso-8859-1" To: reiserfs-devel@vger.kernel.org Cc: linux-kernel@vger.kernel.org Hello, I had since the latest 2.6 kernels lockups in a sheevaplug computer (wi= thout swap), when it reached OOMK conditions. The machine was in such a deadl= ock that it only answered to sysrq request through the serial console. Almost ev= ery OOMK caused this, so once you reproduce an OOMK, I think it will almost sure= hang the system. The lock looks related to swapping in read-only pages of progra= ms. As I recently experienced reiserfs lockups in a desktop pc (http://www.spinics.net/lists/reiserfs-devel/msg03157.html), I wondered= if those OOMK lockups could happen due to the reiserfs of /. So I moved / to ex= t4, and all locks disappeared. OOMK works fine now. So I'm definitely leaving reiserfs due to the deadlocks; a pity, becaus= e it served very well to me for many years. I'd be glad to use it again with= fixed deadlock troubles. If it's of help of anyone, here is the log (since just before the OOMK)= taken by netconsole, linux 3.4.1, and with some sysrq dumps. Since the OOMK mess= age appears, the machine answers to nothing but sysrq requests. The log: http://viric.name/tmp/hang/sheevaplug-hang1.txt.gz Regards, Llu=EDs. -- To unsubscribe from this list: send the line "unsubscribe reiserfs-deve= l" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html