From mboxrd@z Thu Jan 1 00:00:00 1970 From: Vladimir Saveliev Subject: Re: Possible bug: clog length 0 Date: Mon, 25 Oct 2004 18:49:47 +0400 Message-ID: <1098715787.3475.118.camel@tribesman.namesys.com> References: <662c77ee04102420101ae1a40@mail.gmail.com> <417C85A2.9030701@namesys.com> Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Return-path: list-help: list-unsubscribe: list-post: Errors-To: flx@namesys.com In-Reply-To: <417C85A2.9030701@namesys.com> List-Id: Content-Type: text/plain; charset="us-ascii" To: Michael Barry Cc: reiserfs-list@namesys.com Hello On Mon, 2004-10-25 at 08:48, Hans Reiser wrote: > Vitaly, can you fix this? If not, give it to vs to fix. > > Hans > > Michael Barry wrote: > > >I would like to being by thanking anyone for thier time who can help > >me with this. > > > >I am installing php currently. When the installation gets to a certain > >step the php process freezes in state D(Disk Sleep). This process > >cannot be killed with kill -9. When I checked my log files I saw "clog > >length 0" repeated many times. I traced this text to the reiserfs4 > >code. I then compiled my kernel 2.6.9-mm1 (note: this also occurs with > >2.6.8.1-mm4) with reiserfs4 with all debug options. > > > >Note: This possible bug is repoduceable and always freezes at the same point. > > > >When I recompile php, now I get an oops earlier in the installation > >process. The following is my log file output: > > > > > > > >clog length 0 > >reiser4[php(17428)]: object_lookup (fs/reiser4/search.c:397)[nikita-2104]: > >code: -503 at fs/reiser4/plugin/file/file.c:803 > >reiser4 panicked cowardly: assertion failed: > >lock_stack_isclean(get_current_lock_stack()) > >context: trace_flags: 0 > >pid: 17428, comm: php > > lock stack: > >.... nr_signaled 0 > >.... curpri high > >.... current locks: > >...... write: c000000000007bb2 > >...... write: c000000000007bb2 > >...... write: c000000000007bb3 > > atom: refcount: 2 id: 1 flags: 0 txnh_count: 1 capture_count: > >10050 stage: 1 start: 4294698070, flushed: 0 > >------------[ cut here ]------------ > >kernel BUG at fs/reiser4/debug.c:131! > >invalid operand: 0000 [#1] > >SMP > >Modules linked in: sg sr_mod > >CPU: 0 > >EIP: 0060:[] Not tainted VLI > >EFLAGS: 00010246 (2.6.9-mm1) > >EIP is at reiser4_do_panic+0x230/0x2d8 > >eax: c050cb08 ebx: c050cb00 ecx: c050cb00 edx: c050cb00 > >esi: c996c000 edi: c996d4b4 ebp: c996d42c esp: c996d26c > >ds: 007b es: 007b ss: 0068 > >Process php (pid: 17428, threadinfo=c996c000 task=c9bf2020) > >Stack: cec03200 00000001 c03f193b c996d298 c0401527 00000323 c996d2c0 c7cb4188 > > c01dfb79 c03f1926 c0403858 c03f8573 c03baf67 c03f8483 0000018d 00000002 > > c5bc6500 c01b799f c018f142 c01f058e c996d398 00000000 00000000 00000000 > >Call Trace: > > [] object_lookup+0x1a9/0x451 > > [] coord_is_existing_item+0xd9/0x1e8 > > [] get_current_trace_flags+0xc/0xe4 > > [] reiser4_blocknr_is_sane_for+0x27/0xf8 > > [] get_current_trace_flags+0xc/0xe4 > > [] num_of_items_node40+0x24/0x1b5 > > [] estimate_cde+0x8c/0x192 > > [] get_current_trace_flags+0xc/0xe4 > > [] get_current_trace_flags+0xc/0xe4 > > [] coord_is_existing_item+0xd9/0x1e8 > > [] get_current_trace_flags+0xc/0xe4 > > [] xmemcpy+0x21/0x25 > > [] return_err+0x1c/0xa0 > > [] key_at_node40+0x5d/0x141 > > [] get_current_trace_flags+0xc/0xe4 > > [] get_current_trace_flags+0xc/0xe4 > > [] xmemset+0x22/0x26 > > [] hint_validate+0x53/0x181 > > [] xmemset+0x22/0x26 > > [] find_file_item+0x312/0x440 > > [] readpage_unix_file+0x19e/0xc02 > > [] num_of_items_node40+0x24/0x1b5 > > [] get_current_trace_flags+0xc/0xe4 > > [] reiser4_readpage+0xdf/0x2cd > > [] num_of_items_node40+0x24/0x1b5 > > [] get_current_trace_flags+0xc/0xe4 > > [] page_cache_read+0xa1/0xd5 > > [] filemap_nopage+0x32f/0x3a6 > > [] unix_file_filemap_nopage+0x56/0xd8 > > [] unix_file_filemap_nopage+0x0/0xd8 > > [] do_no_page+0xba/0x30f > > [] vsprintf+0x27/0x2b > > [] handle_mm_fault+0xf4/0x16c > > [] do_page_fault+0x1c4/0x60a > > [] try_capture_block+0x253/0x719 > > [] add_to_page_cache+0xbe/0xc3 > > [] check_lock_object+0x17/0x186 > > [] get_current_trace_flags+0xc/0xe4 > > [] get_current_trace_flags+0xc/0xe4 > > [] znode_is_any_locked+0x7c/0x87 > > [] num_of_items_node40+0x24/0x1b5 > > [] schedulable+0x8/0x79 > > [] get_current_trace_flags+0xc/0xe4 > > [] schedulable+0x8/0x79 > > [] get_current_trace_flags+0xc/0xe4 > > [] coord_is_existing_item+0xd9/0x1e8 > > [] get_current_trace_flags+0xc/0xe4 > > [] num_of_items_node40+0x24/0x1b5 > > [] num_of_items_node40+0x24/0x1b5 > > [] length_by_coord_node40+0x52/0x2c7 > > [] coord_is_existing_item+0xd9/0x1e8 > > [] do_page_fault+0x0/0x60a > > [] error_code+0x2d/0x38 > > [] load_aout_binary+0x456/0x711 > > [] __copy_user_zeroing_intel+0x16/0xac > > [] __copy_from_user_ll+0x68/0x6c > > [] paste_tail+0x159/0x40f > > [] paste_tail+0x0/0x40f > > [] create_item_node40+0x867/0xa1b > > [] carry_insert_flow+0x225/0x363 > > [] carry_on_level+0x98/0x487 > > [] lock_carry_level+0x66/0x153 > > [] perthread_pages_reserve+0x64/0x133 > > [] carry+0x1dc/0x490 > > [] insert_flow+0x115/0x1e4 > > [] get_current_trace_flags+0xc/0xe4 > > [] num_of_items_node40+0x24/0x1b5 > > [] inode_add_bytes+0x28/0x74 > > [] write_tail+0x1a8/0x232 > > [] should_have_notail+0x40/0x4a > > [] write_tail+0x0/0x232 > > [] write_tail+0x0/0x232 > > [] append_and_or_overwrite+0x516/0x94c > > [] check_block_counters+0x6d/0x19a > > [] flow_by_inode_unix_file+0x62/0x165 > > [] write_flow+0xb8/0x107 > > [] grabbed2free+0x67/0xdc > > [] write_file+0x7a/0xe5 > > [] write_unix_file+0x271/0x3c0 > > [] get_current_trace_flags+0xc/0xe4 > > [] reiser4_write+0x191/0x44b > > [] reiser4_write+0x0/0x44b > > [] vfs_write+0xb0/0x119 > > [] sys_write+0x51/0x80 > > [] syscall_call+0x7/0xb > >Code: 44 ff ff ff 8b 11 85 d2 0f 84 3a ff ff ff 8b 00 39 c8 0f 85 30 > >ff ff ff 39 42 04 0f 85 27 ff ff ff b8 08 cb 50 c0 e8 cc 93 21 > >00 <0f> 0b 83 00 3c 19 3f c0 c7 04 24 5c af 41 c0 c7 44 24 04 a0 c6 > > This looks like a problem we are debugging a fix for. Michael, we will let you know when we are done. As a temporary workaround, using reiser4 made without tails (mkfs.reiser4 -o formatting=extent40) should protect against this problem. > > > >Michael > > > > > > > > > >