Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 21 Apr 2003 23:56:47 -0700 (PDT)
From:      Don Lewis <truckman@FreeBSD.org>
To:        mb@imp.ch
Cc:        freebsd-openoffice@FreeBSD.org
Subject:   Re: openoffice-devel problems on recent -current
Message-ID:  <200304220656.h3M6ulXB025967@gw.catspoiler.org>
In-Reply-To: <20030422083532.M95995@cvs.imp.ch>

next in thread | previous in thread | raw e-mail | index | archive | help
On 22 Apr, Martin Blapp wrote:
> 
> Hi,
> 
>> The first problem is that setup bombs out near, after populating most of
>> ~/OpenOffice.org644/ and ~/.kde/, but before the "registering
>> components" stage.
>>
>> % setup
>> crash_report: not found
>> Abort (core dumped)
> 
> Hmm. The crash_report: not found is normal, but it should not
> core after that.
> 
>> If I try to run scalc, I get an alert window that says:
>> 	The application cannot be started.
>> 	Unable to create instance of com.sun.star.office.SplashScreen
>> and it exits when I hit the "Ok" button.
> 
> That happens if you don't start it from the wrappers. Either you
> use a wrapper or you cd into the program dir.

Ok, that works.  But setup still aborts if I run openoffice_1.1-setup


>> Another problem I ran across is that if I blow away ~/OpenOffice.org644
>> without removing the entry from ~/.sversionrc, that when I attempt to
>> start the software it attempts to "repair" the missing directory by
>> doing the much larger "Local Installation" instead of the smaller
>> "Workstation Installation".  It doesn't ask which I'd prefer, but just
>> starts writing.  Unfortunately my home directory is NFS mounted and this
>> triggered a bug in the NFS client code, which require a reboot.
> 
> That's really bad. Can you get repeat the panic and get a backtrace ?

It doesn't panic.  The NFS mount point just locks up with a server not
responding message.  If I try to kill the process, I get a kernel lock
order reversal message and a stack trace printed to the console, but the
system is still running and the process is still wedged.  I can still
run other stuff on the system as long as it doesn't touch the frozen
mount point.

I'm running OpenOffice remotely with X11 forwarding over ssh.  The other
ugly thing is that the "repair" task takes over the whole screen, so I
have to switch to another screen.

I think I have a fix for the lock order reversal, but think there is
still another NFS bug that happens first.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200304220656.h3M6ulXB025967>