From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeff Mahoney Subject: Re: Un-long listable files (strace+Oops) Date: Tue, 09 Sep 2003 11:00:36 -0400 Message-ID: <3F5DEB14.1000606@suse.com> References: <200308061431.13695.Michael.James@csiro.au> <200308070823.40130.Michael.James@csiro.au> <3F3188C5.7050606@suse.com> <200309091512.11914.Michael.James@csiro.au> <20030909144742.GA3177@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: <20030909144742.GA3177@namesys.com> List-Id: Content-Type: text/plain; charset="us-ascii"; format="flowed" To: Oleg Drokin Cc: Michael.James@csiro.au, reiserfs-list@namesys.com -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I guess I never sent my final update on the last time this happened. I worked with Michael for a few days and determined that it *was not* xattr related, that just happened to be the system call that the access hung on. When xattrs are disabled, like they were, the operations are effectively a no-op. There's nothing that could cause the hang. When I reviewed the logs, the filesystem had already oopsed, thus causing the hang during the next access. - -Jeff Oleg Drokin wrote: | Hello! | | On Tue, Sep 09, 2003 at 03:12:11PM +1000, Michael.James@csiro.au wrote: | | |>I'm getting un-long listable files cropping up occasionally. |>(on a Suse8.2 system, Reiserfs 3.6) | | | Yes, I remember that case. (it was related to xattrs last time) | | |>Please get in touch directly if you are interested, |> particularily if there is some other test I could perform. |>Generally I find having these files around an irritating form of bitrot, |> so I get rid of them, which can hinder investigations |> as I can't create them repeatably. | | | Hm, but do you get an oops each time before you see such files? | | |>Here's an strace of what happens when I "ls -l it. |>suse:/home/pi/jam176/tmp # cat strace.ls-l | | [...] | |>lstat64("KNOPPIX_V3.2-2003-07-26-EN.iso", {st_mode=S_IFREG|0644, st_size=290107392, ...}) = 0 |>getxattr("KNOPPIX_V3.2-2003-07-26-EN.iso", "system.posix_acl_access"s | | | Ok, and in hung at xattrs stuff again. | May be Jeff will be interested in metadata snapshot of corrupted fs. | | |>And there it hangs, waiting for heat-death-of-universe. | | |>Sep 3 12:39:13 suse kernel: Unable to handle kernel paging request at virtual address 00001d08 |>Sep 3 12:39:13 suse kernel: EIP: 0010:[__find_lock_page_helper+23/96] Tainted: PF | | | Hm, you have binary modules loaded. What are those? | Can you reproduce without those modules ever loaded? | | |>Sep 3 12:39:15 suse kernel: Call Trace: [generic_file_write_nolock+705/2000] [generic_file_write+56/80] [reiserfs:__insmod_reiserfs_S.text_L165788+51778/139792] [sys_write+120/256] [system_call+51/64] | | | Otherwise the oops does not even looks like reiserfs related, just some in-memory corruption. | | Bye, | Oleg | - -- jeffm@suse.com jeffm@csh.rit.edu -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.2-rc1-SuSE (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQE/XesTLPWxlyuTD7IRAvMtAKCSD3GFS6UueUaSjDXW79j/yoJUpACbBFww 2y4N42+BRTd3pTeLrJnSkOY= =urK1 -----END PGP SIGNATURE-----