Hi Vincent!

Thanks for your report!

You wrote:

> I was doing some tests of quotatool and found that on some combinations
> of package versions it would fail because it was trying to do a quotactl()
> call to a block device name with an extra octal byte appended, e.g.
>   quotactl(Q_GETFMT|USRQUOTA, "/dev/loop0\362", ...) = -1 ENOENT
> 
> The specific test I was doing was running a simple dump:
>   /usr/sbin/quotatool -u <user> -d /local/mnt
> For speed, these tests are being done with an ext3 filesystem image,
> which is why /dev/loop0 is involved.
> 
> The problem occurs on two separate systems running etch, but not on a third.
> The only relevant difference I can see is that for the third host I was
> using a host that is a local user, not one in the NIS user map. If I try
> using a local user (username 'user', uid=1000) then quotatool works ok.
> 
> I also tested on a sarge system and a sarge+backports system, which
> are in the same autofs+NIS environment as the etch boxes above.
> On those hosts, quotatool (1.4.7-1) works.

Ok, that's pretty weird.  I'm particularly amazed that 1.4.7 does work,
as the changes with 1.4.9 are very minor.  So that really suggests some
kind of glibc or yp bug.  Could you send me the versions of the glibc6
and nis on both of these systems?

Could you please also send me the full debugging output (add -v -v -v to
the quotatool command) and a full strace (with -s 1024 or so) for the
case that it's going wrong?

Thanks!

-- 
Kind regards,
+--------------------------------------------------------------------+
| Bas Zoetekouw              | GPG key: 0644fab7                     |
|----------------------------| Fingerprint: c1f5 f24c d514 3fec 8bf6 |
| [EMAIL PROTECTED], [EMAIL PROTECTED] |              a2b1 2bae e41f 0644 fab7 |
+--------------------------------------------------------------------+ 


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Reply via email to