Hi Brice,
> Le 12/07/2010 00:08, Jirka Hladky a écrit :
> > $./lstopo --xml /tmp/2010-Jul-10_22h14m_results/2.6.32-44.el6.ppc64_OS-
> > indexing.xml a.txt
> > Segmentation fault (core dumped)
>
> This was a crash in a drawing code (in the merge() function) that Samuel
> fixed in trunk r2234 (and backported in 1.0.x). The log doesn't speak
> about a crash, looks like we were lucky... gdb log says:
Great, thanks!
> > Please notice that hwloc-distrib is
> > also not working correctly - check CPU_AFFINITY/0008.log for example.
>
> The problem is that one of the NUMA nodes has an empty cpuset (it could
> be a BIOS bug by the way). hwloc-distrib should probably ignore such
> object and not distribute among them.
Oh, you are right! I didn't notice it!
It seems to be indeed BIOS bug. numactl shows the same information.
Luckily,
[root_at_ibm-js22-vios-02-lp2 hwloc-1.0.1]# lstopo --ignore numanode a.xml
[root_at_ibm-js22-vios-02-lp2 hwloc-1.0.1]# hwloc-distrib --xml a.xml --single 4
can be used to WA the issue.
I'm looking forward to have --ignore switch available in hwloc-distrib once
hwloc 1.1 is released.
Thanks
Jirka
|