http://bugzilla.xensource.com/bugzilla/show_bug.cgi?id=522

------- Additional Comments From vinc...@cl.cam.ac.uk 2006-02-13 14:58
------- This seems normal behavior, when running out of memory & swap, the kernel start killing task. now the attached files contains lots of different errors, lots of them seem not related to the specific problem.

This is very weird since this occurs in filemap_nopage (make sense lmbench does lots of mmap), which is supposed to give only cache page .. memory that can be easily reclaim.

How many memory do you have in domains that goes out-of-memory (seems it is 256M but I'm not sure) ? Is that a new behavior ? if yes, since when ?

BTW, you forget to label which trace correspond to dom0 and domU (with there respective configs would be even better)