On Tue, 2006-10-24 at 16:50 -0400, Chris Mason wrote: > On Tue, Oct 24, 2006 at 01:28:31PM -0700, Badari Pulavarty wrote: > > On Tue, 2006-10-24 at 15:34 -0400, Chris Mason wrote: > > > Hello, > > > > > > Here is a new cut of my O_DIRECT locking rework. It has a much lower > > > cpu cost than the last set, and simple benchmarks no longer show a > > > regression here in system time. But, the complexity for inserting > > > placeholder pages has gone up. > > > > > > I've also changed the way I test for a place holder page (from > > > mm/filemap.c): > > > > > > static struct address_space placeholder_address_space; > > > #define PagePlaceHolder(page) ((page)->mapping == &placeholder_address_space) > > > > > > This is more stable than the last one but I'm just starting to run race > > > and load testing on it. > > > > > > -chris > > > > Gave it a spin with simple fsx tests and ran into .. > > Thanks, I missed an i_mutex change. Let me audit mutexes coming in and > out again and resend. > > I think the new rules should be: > > reads: don't need i_mutex at all > writes: don't need i_mutex for io inside i_size. > > Oh, and don't mess with i_mutex under DIO_OWN_LOCKING. No problem. I ran into one more nasty looking bug .. ------------[ cut here ]------------ kernel BUG at mm/filemap.c:818! invalid opcode: 0000 [1] SMP last sysfs file: /block/sda/sda1/size CPU 0 Modules linked in: autofs4 hidp rfcomm l2cap bluetooth sunrpc af_packet xt_state ip_conntrack nfnetlink xt_tcpudp ip6table_filter ip6_tables x_tables ipv6 freq_table processor binfmt_misc parport_pc lp parport ide_cd cdrom generic tsdev floppy piix e752x_edac edac_mc ehci_hcd shpchp uhci_hcd i2c_i801 pci_hotplug i2c_core serio_raw usbcore dm_snapshot dm_zero dm_mirror dm_mod ide_disk ide_core Pid: 2669, comm: fsx-linux Not tainted 2.6.19-rc2-mm1-smp #3 RIP: 0010:[<ffffffff802667cf>] [<ffffffff802667cf>] remove_placeholder_pages+0x6e/0x129 RSP: 0018:ffff8101090cba28 EFLAGS: 00010093 RAX: 0000000000000002 RBX: 0000000000000004 RCX: 0000000000000006 RDX: 0000000000000040 RSI: 0000000000000002 RDI: 0000000000000040 RBP: ffff8101090cba88 R08: ffff810109163fa0 R09: 0000000000001000 R10: 0000000000000002 R11: ffff810109163d88 R12: 0000000000000000 R13: 000000000000003f R14: 0000000000000000 R15: ffff810120739718 FS: 00002ab9e503b860(0000) GS:ffffffff806b6000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 0000000000608000 CR3: 00000001096af000 CR4: 00000000000006e0 Process fsx-linux (pid: 2669, threadinfo ffff8101090ca000, task ffff8101091d3080) Stack: 0000000000000002 0000000000000040 0000000000000043 ffff8101186506f8 ffff810118650750 0000000000000002 0000000200000000 ffff810118650608 0000000000000000 0000000000000000 0000000000000000 00000000fffffff1 Call Trace: [<ffffffff802b4d1f>] dio_complete+0x9f/0xd7 [<ffffffff802b57ee>] __blockdev_direct_IO+0xa97/0xb37 [<ffffffff80301813>] ext3_direct_IO+0xf4/0x192 [<ffffffff8026794a>] generic_file_direct_IO+0xa2/0xeb [<ffffffff802679f7>] generic_file_direct_write+0x64/0x104 [<ffffffff8026838e>] __generic_file_aio_write_nolock+0x2e4/0x417 [<ffffffff80268525>] generic_file_aio_write+0x64/0xc0 [<ffffffff802fce06>] ext3_file_write+0x1e/0x9b [<ffffffff8028f463>] do_sync_write+0xe2/0x126 [<ffffffff8028fcc7>] vfs_write+0xcf/0x175 [<ffffffff802902af>] sys_write+0x47/0x70 [<ffffffff80209cee>] system_call+0x7e/0x83 DWARF2 unwinder stuck at system_call+0x7e/0x83 Leftover inexact backtrace: Code: 0f 0b eb fe 49 8b 1e 48 81 7b 18 40 2d c1 80 75 18 48 3b 5d RIP [<ffffffff802667cf>] remove_placeholder_pages+0x6e/0x129 RSP <ffff8101090cba28> <3>BUG: sleeping function called from invalid context at kernel/rwsem.c:20 in_atomic():0, irqs_disabled():1 Call Trace: [<ffffffff8020b0d7>] dump_trace+0xaa/0x3ed [<ffffffff8020b456>] show_trace+0x3c/0x52 [<ffffffff8020b481>] dump_stack+0x15/0x17 [<ffffffff8022a2ad>] __might_sleep+0xb2/0xb4 [<ffffffff80248d8e>] down_read+0x1d/0x4a [<ffffffff8023f867>] blocking_notifier_call_chain+0x1b/0x40 [<ffffffff80234195>] profile_task_exit+0x15/0x17 [<ffffffff80235c1a>] do_exit+0x25/0x8a5 [<ffffffff8020b75c>] kernel_math_error+0x0/0x96 [<ffffffff804a323c>] do_trap+0xdf/0xee [<ffffffff8020bd87>] do_invalid_op+0xac/0xb8 [<ffffffff804a2b1d>] error_exit+0x0/0x96 DWARF2 unwinder stuck at error_exit+0x0/0x96 Leftover inexact backtrace: [<ffffffff802667cf>] remove_placeholder_pages+0x6e/0x129 [<ffffffff802667b8>] remove_placeholder_pages+0x57/0x129 [<ffffffff802b4d1f>] dio_complete+0x9f/0xd7 [<ffffffff802b57ee>] __blockdev_direct_IO+0xa97/0xb37 [<ffffffff80301813>] ext3_direct_IO+0xf4/0x192 [<ffffffff802ffc10>] ext3_get_block+0x0/0xec [<ffffffff8026794a>] generic_file_direct_IO+0xa2/0xeb [<ffffffff802679f7>] generic_file_direct_write+0x64/0x104 [<ffffffff8026838e>] __generic_file_aio_write_nolock+0x2e4/0x417 [<ffffffff804a0c6d>] mutex_lock+0x2a/0x2e [<ffffffff80268525>] generic_file_aio_write+0x64/0xc0 [<ffffffff802fce06>] ext3_file_write+0x1e/0x9b [<ffffffff8028f463>] do_sync_write+0xe2/0x126 [<ffffffff80246220>] autoremove_wake_function+0x0/0x38 [<ffffffff8028fcc7>] vfs_write+0xcf/0x175 [<ffffffff802902af>] sys_write+0x47/0x70 [<ffffffff80209cee>] system_call+0x7e/0x83 NMI Watchdog detected LOCKUP on CPU 0 CPU 0 Modules linked in: autofs4 hidp rfcomm l2cap bluetooth sunrpc af_packet xt_state ip_conntrack nfnetlink xt_tcpudp ip6table_filter ip6_tables x_tables ipv6 freq_table processor binfmt_misc parport_pc lp parport ide_cd cdrom generic tsdev floppy piix e752x_edac edac_mc ehci_hcd shpchp uhci_hcd i2c_i801 pci_hotplug i2c_core serio_raw usbcore dm_snapshot dm_zero dm_mirror dm_mod ide_disk ide_core Pid: 106, comm: pdflush Not tainted 2.6.19-rc2-mm1-smp #3 RIP: 0010:[<ffffffff8036d445>] [<ffffffff8036d445>] __read_lock_failed +0x5/0x14RSP: 0018:ffff8101210d9b38 EFLAGS: 00000097 RAX: ffff810120c480c0 RBX: ffff810120739730 RCX: 0000000000000002 RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff810120739730 RBP: ffff8101210d9b50 R08: 0000000000000002 R09: 0000000000000001 R10: ffffffff80266222 R11: ffff810120c480c0 R12: 0000000000000000 R13: 000000000000000e R14: ffff8101210d9c30 R15: ffff810120739730 FS: 0000000000000000(0000) GS:ffffffff806b6000(0000) knlGS:0000000000000000 CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b CR2: 00002b8ea4469b20 CR3: 000000011576d000 CR4: 00000000000006e0 Process pdflush (pid: 106, threadinfo ffff8101210d8000, task ffff810120c480c0) Stack: ffffffff8036e17d ffff810120c47988 ffff810120739730 ffff8101210d9b70 ffffffff804a25f2 ffff8101210d9c30 ffff810120739720 ffff8101210d9bb0 ffffffff80266222 ffff8101210d9ca8 ffff8101210d9c20 ffff8101210d9e20 Call Trace: Inexact backtrace: [<ffffffff8036e17d>] _raw_read_lock+0x2f/0x33 [<ffffffff804a25f2>] _read_lock_irq+0x36/0x3b [<ffffffff80266222>] find_get_pages_tag+0x31/0xce [<ffffffff8026ed64>] pagevec_lookup_tag+0x22/0x2b [<ffffffff8026c920>] generic_writepages+0x309/0x36d [<ffffffff8030131d>] ext3_ordered_writepage+0x0/0x199 [<ffffffff8026c9b6>] do_writepages+0x32/0x3a [<ffffffff802ab6e1>] __writeback_single_inode+0x1ab/0x37d [<ffffffff802abcbf>] generic_sync_sb_inodes+0x1e8/0x2e2 [<ffffffff802abdd6>] sync_sb_inodes+0x1d/0x1f [<ffffffff802ac181>] writeback_inodes+0x8f/0xde [<ffffffff8026d2c9>] pdflush+0x0/0x1f5 [<ffffffff8026cec6>] wb_kupdate+0x9f/0x113 [<ffffffff8026d413>] pdflush+0x14a/0x1f5 [<ffffffff8026ce27>] wb_kupdate+0x0/0x113 [<ffffffff802460ee>] kthread+0xea/0x120 [<ffffffff8020ab25>] child_rip+0xa/0x15 [<ffffffff804a2622>] _spin_unlock_irq+0x2b/0x31 [<ffffffff8020a29c>] restore_args+0x0/0x30 [<ffffffff80242659>] run_workqueue+0x19/0xf5 [<ffffffff80246004>] kthread+0x0/0x120 [<ffffffff8020ab1b>] child_rip+0x0/0x15 Code: 83 3f 01 78 f9 f0 ff 0f 0f 88 ed ff ff ff c3 55 31 c9 48 89 - To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
