Hello, there is daemon which has about 30 threads using same RW lock. most of the time rw lock is read locked by several threads. it worx as expected in all setups except one. customer with 16 ultra IV cpus had terrible performance problems with this daemon. investigation revealed that half of threads spend 90% of time in read lock call.
using temporary binary wo read lock - problem solved. we have also several customers with T2000 machines (and solaris 10) - no such reports. is there any difference in RW lock implementation between sol9 and sol10? is it expected behavior for sol9 that read RW lock being used by 10+ threads simultaneously has very bad performance? looking on pstacks, it looks like on sol9 read lock has contention on some internal mutex. This message posted from opensolaris.org _______________________________________________ perf-discuss mailing list perf-discuss@opensolaris.org
