dolfin team mailing list archive

[Chris Richardson <615250@xxxxxxxxxxxxxxxxxx>]

On 18/02/11 14:51, Johannes Ring wrote:
> I recently got a report about the same bug from one of the system
> administrators at Simula. After that I tried to reproduce it on several
> different virtual machines (sid, natty, maverick and lucid) and with
> different configurations of DOLFIN and I found that it no longer happens
> in the latest development version. The problem always occurred when
> using DOLFIN 0.9.9, either when using the Debian/Ubuntu packages or when
> everything is built from source with Dorsal.
>
> @Chris: can you confirm that this behavior is gone in the development
> version?
>

Yes, it seems to be OK now.

-- 
You received this bug notification because you are a member of DOLFIN
Team, which is subscribed to DOLFIN.
https://bugs.launchpad.net/bugs/615250

Title:
  Failed X connection corrupts root filesystem

Status in DOLFIN:
  New

Bug description:
  I noticed this problem a while ago, and just fell victim to it again
  this morning.

  I was doing a dolfin install from bzr as root on Ubuntu Lucid 10.4 64-bit server.
  When running a demo, if the X forwarding is not correctly set, then the code crashes in a plot() call, with a PETSC error. Subsequently, the root file system is severely damaged, and the only option seems to be a complete reinstall of the OS.

  I know, this can be avoided by not running the demos as root, but it is an easy mistake to make after doing a system-wide install.
  I presume it is a bug in viper.

  Any ideas?

  Thanks

  Chris Richardson