* Running SPEC2K on the Snowball board. A fresh kernel with HIGHMEM enabled made it possible to run the tests. Great variations in the results indicates that something strange is going on. Turning off one of the CPU:s gives stable result (but slow), so my current guess is that the variations are caused by a known bug that makes one cpu run slower.
http://igloocommunity.org/bugzilla3/show_bug.cgi?id=1
The patch for this bug was not included in my kernel. Will have another go with a kernel where the patch is included, as a background activity.

* Planned and started working on the "Adding browsing benchmarks to our current set of tests"-activity. I will try to keep documentation up to date here:
https://wiki.linaro.org/AsaSandahl/Sandbox/BrowsingBenchmarks
Experimenting with building Firefox in different ways, so far for x86.

Best Regards
Åsa