On Wed, 28 Sep 2005, Hal Rosenstock wrote:
halr> That seems better but I still see the following: <snip> The failure is as expected, but at least it was graceful. halr> Also, I don't understand why: halr> kdapltest -T T -s <IP> -D mthca0a -d -t 2 -w 8 -i 20 client SR server SR halr> would work and halr> kdapltest -T T -s <IP> -D mthca0a -d -i 10000 -w 8 client SR server SR halr> would fail. It seems the former is more strenuous (everything same but 2 halr> threads and less iterations). My hypothesis is that each iteration allocates memory that isn't free'd until the test is over. I haven't been able to reproduce this yet, even when I turn on DEBUG_PAGEALLOC. What are the contents of /etc/cpuinfo and /etc/meminfo on this system? james _______________________________________________ openib-general mailing list [email protected] http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
