From: Andi Kleen Subject: reiserfs unstable on large systems in 2.6.13-git9 Date: Mon, 12 Sep 2005 00:30:46 +0200 Message-ID: <20050911223045.GA11071@wotan.suse.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: linux-fsdevel@vger.kernel.org Return-path: Received: from ns.suse.de ([195.135.220.2]:60889 "EHLO mx1.suse.de") by vger.kernel.org with ESMTP id S1750985AbVIKWar (ORCPT ); Sun, 11 Sep 2005 18:30:47 -0400 To: reiserfs-dev@namesys.com, mason@suse.de, jeffm@suse.de Content-Disposition: inline Sender: linux-fsdevel-owner@vger.kernel.org List-Id: linux-fsdevel.vger.kernel.org When I run even relatively minor stress on 8 or 16 core Opterons I get deadlocks like this: At one point I also had a deadlock on a semaphore with all processes that did disk access going into D (backtrace lost on that one unfortunately but all the traces went through reiserfs_set_acl) Watchdog detected LOCKUP on CPU 10 CPU 10 Modules linked in: Pid: 21498, comm: reaim Not tainted 2.6.13-git9 #4 RIP: 0010:[] {.text.lock.spinlock+22} RSP: 0018:ffff81013b59bc40 EFLAGS: 00000086 RAX: 0000000000000000 RBX: ffffffff804c2ba0 RCX: 00000000c0000100 RDX: 0000000000000000 RSI: ffff81023b6f30c0 RDI: ffffffff804c2ba8 RBP: 0000000000000282 R08: ffff81013b59a000 R09: 0000000000000000 R10: 00000000ffffffff R11: 0000000000000246 R12: ffffffff804c2ba8 R13: ffff81023b6f30c0 R14: ffff81013b59bc50 R15: 00000000000001ff FS: 00002aaaaaf3b0a0(0000) GS:ffffffff80603d00(0000) knlGS:00000000401c9c60 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 00007ffffffe048c CR3: 000000013dbdb000 CR4: 00000000000006a0 Process reaim (pid: 21498, threadinfo ffff81013b59a000, task ffff81023b6f30c0) Stack: 0000000000000282 ffffffff80418ad0 0000000000000001 ffff81023b6f30c0 ffffffff801313b0 ffff810134a0bc68 ffffffff804c2bb0 ffff81013b59bdd8 ffff81013b59bdd8 ffff8102b586c678 Call Trace:{__down+160} {default_wake_function+0} {__down_failed+53} {.text.lock.kernel_lock+25} {reiserfs_setattr+44} {__down_write+51} {notify_change+340} {do_truncate+65} {may_open+468} {open_namei+734} {thread_return+0} {filp_open+39} {get_unused_fd+219} {do_sys_open+81} {system_call+126} Code: 80 3f 00 7e f9 e9 90 fd ff ff f3 90 80 3f 00 7e f9 e9 9c fd console shuts up ... NMI Watchdog detected LOCKUP on CPU 6 ... -Andi