Return-Path: linux-nfs-owner@vger.kernel.org Received: from bombadil.infradead.org ([198.137.202.9]:53941 "EHLO bombadil.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752235AbbAXStM (ORCPT ); Sat, 24 Jan 2015 13:49:12 -0500 Date: Sat, 24 Jan 2015 10:49:11 -0800 From: Christoph Hellwig To: Peng Tao Cc: Trond Myklebust , Linux NFS Mailing list , Tom Haynes Subject: Re: [PATCH v5 00/51] Add Flexfile Layout Module Message-ID: <20150124184911.GA8159@infradead.org> References: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: Sender: linux-nfs-owner@vger.kernel.org List-ID: On Sat, Jan 24, 2015 at 10:37:09PM +0800, Peng Tao wrote: > oops, sorry for editing the mail subject... It was supposed to reply > to your email "Re: [PATCH v5 00/51] Add Flexfile Layout Module" With your patch it surives a little longer, but then this happens: (you could just try my receipt for the reproducer, it really is that simple..): generic/013 26s ...[ 192.023340] NMI watchdog: BUG: soft lockup - CPU#0 stuck for 22s! [fsstress:9440] [ 192.024091] Modules linked in: [ 192.024489] irq event stamp: 93792 [ 192.024845] hardirqs last enabled at (93791): [] restore_args+0x0/0x30 [ 192.025701] hardirqs last disabled at (93792): [] apic_timer_interrupt+0x6d/0x80 [ 192.026606] softirqs last enabled at (93790): [] __do_softirq+0x1e3/0x2d0 [ 192.026651] softirqs last disabled at (93785): [] irq_exit+0x9e/0xc0 [ 192.026651] CPU: 0 PID: 9440 Comm: fsstress Not tainted 3.19.0-rc5+ #1555 [ 192.026651] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2007 [ 192.026651] task: ffff88007a4e82d0 ti: ffff88007d7fc000 task.ti: ffff88007d7fc000 [ 192.026651] RIP: 0010:[] [] __radix_tree_lookup+0x5c/0xb0 [ 192.026651] RSP: 0018:ffff88007d7ffa58 EFLAGS: 00000293 [ 192.026651] RAX: 0000000000000002 RBX: ffffffff81dfd9f7 RCX: 0000000000000000 [ 192.026651] RDX: 0000000000000000 RSI: 0000000000000028 RDI: 0000000000000002 [ 192.026651] RBP: ffff88007d7ffa58 R08: 000000000000002b R09: 0000000000000000 [ 192.026651] R10: ffff880078426d48 R11: 0000000000000000 R12: ffff88007d7ff9c8 [ 192.026651] R13: ffff88007a4e82d0 R14: ffff88007d7fc000 R15: ffff88007a4e82d0 [ 192.026651] FS: 00007ff90c7b1700(0000) GS:ffff88007fc00000(0000) knlGS:0000000000000000 [ 192.026651] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 192.026651] CR2: 00007ff904020000 CR3: 000000007a12a000 CR4: 00000000000006f0 [ 192.026651] Stack: [ 192.026651] ffff88007d7ffa68 ffffffff8178afcd ffff88007d7ffa98 ffffffff81178d6a [ 192.026651] ffff88007adf4140 ffff88007d7ffd48 ffff88007adf4140 ffff88007879b518 [ 192.026651] ffff88007d7ffaf8 ffffffff8139c8b6 ffffffff8139c878 0000000000000000 [ 192.026651] Call Trace: [ 192.026651] [] radix_tree_lookup+0xd/0x10 [ 192.026651] [] page_cache_next_hole+0x3a/0x50 [ 192.026651] [] bl_pg_init_write+0xf6/0x140 [ 192.026651] [] ? bl_pg_init_write+0xb8/0x140 [ 192.026651] [] __nfs_pageio_add_request+0xf0/0x470 [ 192.026651] [] nfs_pageio_add_request+0x93/0x1c0 [ 192.026651] [] ? find_get_pages_tag+0x1f/0x190 [ 192.026651] [] nfs_do_writepage+0x120/0x1e0 [ 192.026651] [] nfs_writepages_callback+0x18/0x30 [ 192.026651] [] write_cache_pages+0x1e6/0x490 [ 192.026651] [] ? nfs_writepage_locked+0xa0/0xa0 [ 192.026651] [] nfs_writepages+0x94/0x130 [ 192.026651] [] do_writepages+0x1c/0x40 [ 192.026651] [] __filemap_fdatawrite_range+0x51/0x60 [ 192.026651] [] filemap_write_and_wait_range+0x35/0x80 [ 192.026651] [] nfs4_file_fsync+0x5f/0xb0 [ 192.026651] [] ? nfs4_file_llseek+0x60/0x60 [ 192.026651] [] vfs_fsync+0x24/0x40 [ 192.026651] [] nfs_file_flush+0x82/0xc0 [ 192.026651] [] filp_close+0x35/0x80 [ 192.026651] [] __close_fd+0x8a/0xb0 [ 192.026651] [] SyS_close+0x1e/0x50 [ 192.026651] [] system_call_fastpath+0x12/0x17 [ 192.026651] Code: 07 48 83 c7 08 49 89 3b 5d c3 0f 1f 44 00 00 49 89 c2 49 83 e2 fe 41 8b 3a 81 e7 ff 0f 00 00 89 f8 48 3b 34 c5 80 b3 58 82 77 4c <8d> 04 7f 8d 4c 00 fa eb 06 0f 1f 00 49 89 c2 49 89 f0 49 d3 e8 and never recovers from it, printing that backtrace again and again.