From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: with ECARTIS (v1.0.0; list xfs); Mon, 30 Jun 2008 02:01:22 -0700 (PDT) Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with SMTP id m5U91FIA003187 for ; Mon, 30 Jun 2008 02:01:17 -0700 Received: from cxfsmac10.melbourne.sgi.com (cxfsmac10.melbourne.sgi.com [134.14.55.100]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id TAA24479; Mon, 30 Jun 2008 19:02:12 +1000 Message-ID: <4868A114.9080106@sgi.com> Date: Mon, 30 Jun 2008 19:02:12 +1000 From: Donald Douwsma MIME-Version: 1.0 Subject: Re: [PATCH] re-remove xfs custom bitops References: <480EB397.1040304@sandeen.net> <4829C360.5060500@sandeen.net> <482BF841.8050704@sgi.com> <48669A45.1050104@sandeen.net> In-Reply-To: <48669A45.1050104@sandeen.net> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: xfs-bounce@oss.sgi.com Errors-to: xfs-bounce@oss.sgi.com List-Id: xfs To: Eric Sandeen Cc: xfs-oss Eric Sandeen wrote: > Donald Douwsma wrote: >> Eric Sandeen wrote: >>> Eric Sandeen wrote: >>>> Once more, with feeling! >>>> >>>> This re-instates the reverted mod after the ppc panic of >>>> Feb '08. You guys do have ppc boxes in the test farm now right? :) >>>> >>>> This keeps xfs_lowbit64 as it was since there aren't good >>>> generic helpers there ... >>>> >>>> This should probably keep Dave's signed-off line, there's >>>> a bit of my (userspace) testing here but no original work. >>>> >>>> This exact patch isn't tested but it's based on a conglomeration >>>> of prior testing... >>> SGI guys, any takers on this one? >>> -Eric >> Sorry Eric, havent had chance to run this on all platforms yet. >> I want to test it on ppc as well as the usual x86_64/ia64 combinations, >> then I'll get it committed. >> >> Don > > Don, how's that all going then? :) > Good question, got sidetracked with product releases. First time round I hit an Oops on xfstests/177 while running the auto group on ppc32. I dont seem to hit it running the single test, its intermittent. I also hit this oops once back in Feb running x86_64, which was around the time we pulled the last version of this cleanup out, but it may be unrelated. Looping through the auto group overnight, I'll see If I hit it again. XFS mounting filesystem hdb4 Ending clean XFS mount for filesystem: hdb4 Device hdb4 - bad inode magic/vsn daddr 23607192 #0 (magic=0) ------------[ cut here ]------------ kernel BUG at fs/xfs/support/debug.c:54! Oops: Exception in kernel mode, sig: 5 [#1] PowerMac Modules linked in: iptable_filter ip_tables ip6table_filter ip6_tables x_tables ipv6 dm_mod ide_cd_mod cdrom uninorth_agp ohci1394 agpgart sungem sungem_phy natsemi ieee1394 ehci_hcd NIP: c01955c8 LR: c01955bc CTR: c0032cc4 REGS: c3359d10 TRAP: 0700 Not tainted (2.6.25-donaldd) MSR: 00029032 CR: 24002082 XER: 00000000 TASK = da1a4660[31911] 'xfssyncd' THREAD: c3358000 GPR00: 00000001 c3359dc0 da1a4660 00000041 00000001 00000000 00000000 00000001 GPR08: 0014ff97 c0500d48 d2aea3f2 c0500d48 00190834 00000000 0242db00 0242dc14 GPR16: 0242d860 02464870 0242dc38 c0448f8c c04512c8 00000000 c0160984 c3359e88 GPR24: 00000000 00000020 c3b1d1e0 00000000 c3359e90 00000000 c056cf34 00009032 NIP [c01955c8] cmn_err+0xd4/0xec LR [c01955bc] cmn_err+0xc8/0xec Call Trace: [c3359dc0] [c01955bc] cmn_err+0xc8/0xec (unreliable) [c3359e00] [c0160844] xfs_imap_to_bp+0x1a0/0x200 [c3359e80] [c0160984] xfs_itobp+0xe0/0x18c [c3359ed0] [c0162890] xfs_iflush+0x268/0x3f4 [c3359f10] [c017f98c] xfs_finish_reclaim+0xe8/0x190 [c3359f30] [c017facc] xfs_finish_reclaim_all+0x98/0xf4 [c3359f60] [c017ded0] xfs_syncsub+0x5c/0x290 [c3359f90] [c0192820] xfs_sync_worker+0x30/0x64 [c3359fa0] [c019404c] xfssyncd+0x118/0x16c [c3359fe0] [c004a790] kthread+0x4c/0x88 [c3359ff0] [c0012f9c] kernel_thread+0x44/0x60 Instruction dump: 7c1e19ae 3d20c03a 57a0103a 39291c84 3c60c047 7c89002e 7fc5f378 3863ae70 4bea0c49 7fe00124 7fa00034 5400d97e <0f000000> 80010044 bba10034 38210040 ---[ end trace f4636d0864e8be53 ]---