From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jiri Slaby Subject: Re: 2.6.21-mm1: many processes end up in D state Date: Fri, 04 May 2007 16:58:13 +0200 Message-ID: <463B4A05.7060504@gmail.com> References: <46360DA7.6040003@gmail.com> <46371128.4060101@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <46371128.4060101@gmail.com> Sender: linux-kernel-owner@vger.kernel.org Cc: stable@kernel.org, Andrew Morton , Linux Kernel Mailing List , Mikael Pettersson , linux-ide@vger.kernel.org, Jeff Garzik List-Id: linux-ide@vger.kernel.org Cc: Mikael Pettersson Cc: linux-ide@vger.kernel.org Cc: Jeff Garzik Jiri Slaby napsal(a): > Cc: stable@kernel.org > > Jiri Slaby napsal(a): >> Hi, >> >> I have a problem with higher disk loads (e.g. running git-log or yum update). >> Many processes end up in D state and system is unusable -- I'm not able to run >> anything but smooth mouse moving when this happens. >> >> If I wait for a 20-30sec it becomes usable. This happens in 2.6.21-rc7-mm2 and >> also in 2007-04-28-05-06 broken-out snapshot. I think 2.6.21-rc6-mm1 worked >> fine, but I'm uncertain. If it is important, let me know to re-test. >> >> sysrq-t: >> >> yum D 0000004C 0 2976 2955 (NOTLB) >> c455bd00 00200082 8c68ea9e 0000004c 00000000 00000000 c455bcdc c455bc9c >> c455bc9c c2d6a580 8c68ea9e 0000004c c455bcc0 c2d6a694 c1809980 00000000 >> c455bcd0 c4640c80 000036a0 00000001 c455bcf0 c02a7132 c3682164 00001cde >> Call Trace: >> [] io_schedule+0x21/0x2e >> [] sync_page+0x38/0x43 >> [] __wait_on_bit_lock+0x40/0x63 >> [] __lock_page+0x54/0x5c >> [] do_generic_mapping_read+0x1dc/0x57e >> [] generic_file_aio_read+0xc9/0x1b5 >> [] do_sync_read+0xd0/0x106 >> [] vfs_read+0x89/0x11d >> [] sys_pread64+0x64/0x68 >> [] syscall_call+0x7/0xb >> [<47176012>] 0x47176012 >> ======================= >> xterm S 00000051 0 2980 1 (NOTLB) >> c4678ae0 00000082 b2fbf0ce 00000051 00000000 00000000 00000000 00000000 >> c4678b30 c469d540 b2fbf0ce 00000051 c4678ab0 c469d654 c1813980 00000001 >> c4678ad0 c47f0e40 c4678b00 00000286 c4678ae0 c012c077 00000001 00003286 >> Call Trace: >> [] schedule_timeout+0x44/0xa4 >> [] do_select+0x46d/0x55b >> [] core_sys_select+0x19e/0x2c0 >> [] sys_select+0xd3/0x18d >> [] syscall_call+0x7/0xb >> [<47180f98>] 0x47180f98 >> ======================= >> bash S 0000004A 0 2982 2980 (NOTLB) >> c50b3e80 00000086 ccf4cc7e 0000004a 00000000 c021bed9 00000000 c45c0030 >> c50b3e30 c45c0030 ccf4cc7e 0000004a c50b3e50 c45c0144 c1809980 00000000 >> c1809e44 c4517900 00000000 000000b3 c50b3e90 c014f841 c477cb40 c2d79000 >> Call Trace: >> [] schedule_timeout+0x6a/0xa4 >> [] read_chan+0x1b9/0x5b9 >> [] tty_read+0x75/0xaa >> [] vfs_read+0x89/0x11d >> [] sys_read+0x3d/0x64 >> [] syscall_call+0x7/0xb >> [<4717883e>] 0x4717883e >> ======================= >> metacity S 00000050 0 3005 2429 (NOTLB) >> c2fd5bb0 00000086 5205656d 00000050 00000000 00000000 00000000 c441c51c >> 00000246 c45c0540 5205656d 00000050 c2fd5b80 c45c0654 c1809980 00000000 >> c1813e44 c27f4900 00002cb8 00000246 c2fd5ba0 c0135a80 c2fd5ba8 00002cb2 >> Call Trace: >> [] schedule_timeout+0x6a/0xa4 >> [] do_sys_poll+0x33f/0x46d >> [] sys_poll+0x41/0x43 >> [] syscall_call+0x7/0xb >> [<4717e3a6>] 0x4717e3a6 >> ======================= >> ccpd D 0000004C 0 3009 2010 (NOTLB) >> c50abbb0 00000082 f08ef585 0000004c 00000000 00000000 c180abc0 c22fc788 >> c50abba0 c505d070 f08ef585 0000004c 00000000 c505d184 c1809980 00000000 >> c21167f0 c2fc6740 c5794e3c c31eff74 c5794310 c22fc788 00000002 00001e64 >> Call Trace: >> [] do_get_write_access+0x310/0x4f5 >> [] journal_get_write_access+0x1b/0x2a >> [] __ext3_journal_get_write_access+0x19/0x3f >> [] ext3_reserve_inode_write+0x53/0x6c >> [] ext3_mark_inode_dirty+0x20/0x37 >> [] ext3_dirty_inode+0x6b/0x6d >> [] __mark_inode_dirty+0x2a/0x16d >> [] touch_atime+0x84/0xd8 >> [] do_generic_mapping_read+0x463/0x57e >> [] generic_file_aio_read+0xc9/0x1b5 >> [] do_sync_read+0xd0/0x106 >> [] vfs_read+0x89/0x11d >> [] kernel_read+0x36/0x48 >> [] prepare_binprm+0xb2/0xdf >> [] do_execve+0xe6/0x1e1 >> [] sys_execve+0x2e/0x7f >> [] syscall_call+0x7/0xb >> [<471474da>] 0x471474da >> ======================= >> bash D 0000004D 0 3011 2859 (NOTLB) >> c471dc90 00000082 08f8373c 0000004d 00000000 00000000 c471dc30 c471dc2c >> c471dc2c c4670540 08f8373c 0000004d c471dc50 c4670654 c1809980 00000000 >> c471dc60 c2f3f040 00000005 00000001 c471dc80 c02a7132 c591af64 00001eea >> Call Trace: >> [] io_schedule+0x21/0x2e >> [] sync_page+0x38/0x43 >> [] __wait_on_bit_lock+0x40/0x63 >> [] __lock_page+0x54/0x5c >> [] do_generic_mapping_read+0x1dc/0x57e >> [] generic_file_aio_read+0xc9/0x1b5 >> [] do_sync_read+0xd0/0x106 >> [] vfs_read+0x89/0x11d >> [] kernel_read+0x36/0x48 >> [] prepare_binprm+0xb2/0xdf >> [] do_execve+0xe6/0x1e1 >> [] sys_execve+0x2e/0x7f >> [] syscall_call+0x7/0xb >> [<471474da>] 0x471474da >> ======================= >> bash D 0000004D 0 3012 2884 (NOTLB) >> c50d5a80 00000082 a21bfc35 0000004d 00000000 00000000 c180abc0 c2033dd4 >> c50d5a70 c2f4f070 a21bfc35 0000004d 00000000 c2f4f184 c1809980 00000000 >> c21167bc c47f0c80 c2033310 c22fc788 c5794e3c c2033dd4 00000002 0000216d >> Call Trace: >> [] do_get_write_access+0x310/0x4f5 >> [] journal_get_write_access+0x1b/0x2a >> [] __ext3_journal_get_write_access+0x19/0x3f >> [] ext3_reserve_inode_write+0x53/0x6c >> [] ext3_mark_inode_dirty+0x20/0x37 >> [] ext3_dirty_inode+0x6b/0x6d >> [] __mark_inode_dirty+0x2a/0x16d >> [] touch_atime+0x84/0xd8 >> [] __link_path_walk+0x893/0xca4 >> [] link_path_walk+0x46/0xc3 >> [] do_path_lookup+0x86/0x1b0 >> [] __path_lookup_intent_open+0x44/0x7f >> [] path_lookup_open+0x21/0x27 >> [] open_exec+0x27/0xa2 >> [] load_elf_binary+0x1482/0x1acf >> [] search_binary_handler+0x6e/0x19f >> [] do_execve+0x14a/0x1e1 >> [] sys_execve+0x2e/0x7f >> [] syscall_call+0x7/0xb >> [<471474da>] 0x471474da >> ======================= >> >> Note that yum works on lvm on raid0 and git too, but on the another md volume. >> Both ext3s. Drivers are sata_promise and ata_piix (sata disk); CFQ scheduler. >> Using noop is no change (but seems to be harder to reproduce with it). I figured >> out that it probably happens when 2+ processes are on both "processors" (HT on >> P4) and are IO wait (multiload-applet shows red above the half). >> >> Swap usage is 0 all the time. > > I'm able to reproduce this in the latest git (dc87c3985e9b442c). Going to bisect. Is it possible, that Commit: 176efb054422bffe6b5a34194ffad134366c7f7e Author: Mikael Pettersson Wed, 14 Mar 2007 09:51:35 +0100 sata_promise: decode and report error reasons This patch adds much needed error reason decoding and reporting to sata_promise. It's simplistic but should log all relevant error info the controller provides. Signed-off-by: Mikael Pettersson Signed-off-by: Jeff Garzik causes this? When I change this: diff --git a/drivers/ata/sata_promise.c b/drivers/ata/sata_promise.c index f56549b..a0024ae 100644 --- a/drivers/ata/sata_promise.c +++ b/drivers/ata/sata_promise.c @@ -668,10 +668,8 @@ static inline unsigned int pdc_host_intr( struct ata_port *ap, else err_mask &= ~PDC2_ERR_MASK; port_status = readl(port_mmio + PDC_GLOBAL_CTL); - if (unlikely(port_status & err_mask)) { + if (unlikely(port_status & err_mask)) pdc_error_intr(ap, qc, port_status, err_mask); - return 1; - } switch (qc->tf.protocol) { case ATA_PROT_DMA: the problem disappears. regards, -- http://www.fi.muni.cz/~xslaby/ Jiri Slaby faculty of informatics, masaryk university, brno, cz e-mail: jirislaby gmail com, gpg pubkey fingerprint: B674 9967 0407 CE62 ACC8 22A0 32CC 55C3 39D4 7A7E