[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#522726: kernel problem after a simple 'rm' command: RESERVE_SPACE(805) failed in function encode_lookup



Hi Ben,

No, I haven't got a chance to check if the bug exists in newer version.
We changed our NFS server from Linux to OpenSolaris.

But it was a major problem. It re-occurred every time a user would
attempt a filesystem operation where the filename was very long (e.g.
500 characters). Any fs write operation (rm, create new file) would
cause the kernel panic.

The crash happened several times a year. In all cases it was when
someone would antecedently pass data instead of a filename to a peace
of code that expects filenames.

Alex


On Mon, Apr 5, 2010 at 3:45 PM, Ben Hutchings <ben@decadent.org.uk> wrote:
> On Sun, 2009-04-05 at 23:08 -0700, Aleksandr Levchuk wrote:
>> Package: nfs-kernel-server
>> Version: 1:1.0.10-6+etch.1
>> Severity: important
>>
>> My very stable server crashed as a result of a 'rm' command in an
>> NFS-mounted home directory. The 'rm' command was a file name (with
>> newlines) but that file did not exist.
> [...]
>
> Sorry for the delay in replying to this.  The nfs-kernel-server package
> only contains supporting scripts, but the bug is clearly in the kernel
> itself (linux-image-* packages).
>
> The system you reported this bug from was apparently running Linux
> 2.6.22.  I assume that is the same version in which you saw this bug.
> Have you seen the bug reoccur in any more recent kernel version?
>
> Ben.
>
> --
> Ben Hutchings
> Once a job is fouled up, anything done to improve it makes it worse.
>



-- 
---------------------------------------------------------------------------------
Aleksandr Levchuk
Administrator of Bioinformatic Systems and Databases

Homepage: http://biocluster.ucr.edu/~alevchuk/
Cell Phone: (951) 368-0004
Lab Phone: (951) 905-5232

Institute for Integrative Genome Biology
University of California, Riverside
---------------------------------------------------------------------------------



Reply to: