On 19/01/2015 10:52, Rémy Maucherat wrote: > 2015-01-17 2:41 GMT+01:00 Mark Thomas <ma...@apache.org>: > >> The other thought I had (but haven't had time to test for performance >> impact) is that the readPending semaphore is really only useful when the >> previous or current read is non-blocking. A blocking read followed by a >> blocking read doesn't need it. > > I tested it (commenting it out in flushBlocking), the impact is rather > marginal with ab.
Thanks for testing that. I won't investigate this idea further then. Mark --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org