Matthias Schmidt wrote:
* Matthew Dillon wrote:
If you re-extract the tar into an empty filesystem does the problem
then show up again? If so I'd like to get a hold of that tarball
to reproduce the problem.
Does not work. Extracting fails:
# tar xfz /root/ports.tar.gz
[...]
p
Matthias Schmidt wrote:
>* Matthew Dillon wrote:
>>
>> If you re-extract the tar into an empty filesystem does the problem
>> then show up again? If so I'd like to get a hold of that tarball
>> to reproduce the problem.
>
>Does not work. Extracting fails:
>
># tar xfz /root/ports.t
:I can make the image files available if that helps ... 500MB each.
:
:Regards
:
: Matthias
If you could put that ports tar up somewhere for me to download... well,
just put it on your leaf account somewhere. I'm still trying to
reproduce the problem.
* Matthew Dillon wrote:
>
> Hmm. So far I can't reproduce it, untaring a tar of FreeBSD ports
> works. I will continue to try. To try to reproduce you environment:
>
> * How large (exactly) is the HAMMER filesystem.
500MB big files, mounted via vnconfig.
> * How much ram does
:Does not work. Extracting fails:
:
:# tar xfz /root/ports.tar.gz
:[...]
:ports/Mk/bsd.commands.mk: Can't create 'ports/Mk/bsd.commands.mk'
:ports/Mk/bsd.destdir.mk: Can't create 'ports/Mk/bsd.destdir.mk'
:ports/Mk/bsd.perl.mk: Can't create 'ports/Mk/bsd.perl.mk'
:tar: Error exit delayed from pre
* Matthew Dillon wrote:
>
> If you re-extract the tar into an empty filesystem does the problem
> then show up again? If so I'd like to get a hold of that tarball
> to reproduce the problem.
Does not work. Extracting fails:
# tar xfz /root/ports.tar.gz
[...]
ports/Mk/bsd.command
* Matthias Schmidt wrote:
>
> A final "ls" broke the machine. No clue if its related to hammer,
> because I have no physical access to the machine ATM. Will look
> at the crash later and post logs/dump if available.
OK, the machine is back up, but unfortunately I got no core dump.
The dump proc
* Matthew Dillon wrote:
>
> Hmm. There are no errors reported in the dmesg, just debugging output
> (the wrapped messages are just debugging output).
>
> Can you reproduce it trivially? e.g. on a newfs_hammer'd filesystem
> does the touch/ls -la issue still now show the file
:Yes. HEAD from today:
:
:pc12909!matthias:~ % uname -a
:DragonFly pc12909 1.13.0-DEVELOPMENT DragonFly 1.13.0-DEVELOPMENT #3:
:Wed Apr 9 16:23:21 CEST 2008
:[EMAIL PROTECTED]:/usr/obj/usr/src/sys/PC12909 i386
:
:
: Matthias
Hmm. There are no errors reported in the dmesg, just debug
* Matthew Dillon wrote:
>
> Are you sure you are doing all of that in HEAD? HAMMER is not going
> to work in 1.12-REL, I'm not keeping them in sync (not until it goes
> alpha anyway) and the version in 1.12 is very old.
Yes. HEAD from today:
pc12909!matthias:~ % uname -a
DragonFly
:Hi,
:
:I started playing around with HAMMER a bit and noticed some bugs. I
:know HAMMER is currently in alpha state, but I assume the report could
:be helpful :)
Are you sure you are doing all of that in HEAD? HAMMER is not going
to work in 1.12-REL, I'm not keeping them in sync (not u
11 matches
Mail list logo