Greetings,
While load testing 2.5.43-mm2, I noticed some peculiar context switch
storms. (The test load is a make -j30 bzImage. Box is a PIII-500 w. 128MB
ram, using gcc 2.95.3) I do not see storms running the same load under
2.5.44.virgin.
Trying narrow down what the heck it could be, I wrote attached
diff.txt. Twice, it produced the cleaned/processed output in trace1.txt
(near start of build!). One running task with > 100000 context
switches/sec seems unlikely, so I suspect I missed the event, and wonder if
there's a better way to photograph the little beasty. The output in
trace2.txt is pretty typical.. seems to be much scheduling with no place to go.
Ideas on how to better focus my camera?
-Mike
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Wed Oct 23 2002 - 22:00:47 EST