Re: Slow pthread_create() under high load

From: Juergen Kreileder (kreilede@issan.cs.uni-dortmund.de)
Date: Wed Mar 29 2000 - 16:05:55 EST


>>>>> Christopher Smith writes:

    Christopher> On Tue, Mar 28, 2000 at 03:33:00PM -0700, Richard Gooch wrote:
>> - do we actually need efficient (POSIX compliant) pthreads for Linux

    Christopher> I think yes. Perhaps it doesn't have to be the most
    Christopher> efficient threading API on Linux, but it should be
    Christopher> reasonably efficient. There are lots of
    Christopher> developers/companies who build their software
    Christopher> specifically to POSIX API's, and I think there is a
    Christopher> value add to having their software run reasonably
    Christopher> well (and reliably) on Linux. A case in point is the
    Christopher> blackdown group, who didn't so much have problems
    Christopher> with performance, but they did find that pthreads
    Christopher> behavior was different enough from standard POSIX
    Christopher> that we had to wait years before there was a stable
    Christopher> native-threads capable version of blackdown-java on
    Christopher> Linux.

That wasn't the main problem. The real problem were race conditions
and other bugs in the LinuxThreads library, most of them have been
fixed in glibc-2.1.3.
We don't need real Pthreads (the original code uses Sun's thread
extensions), I would be quite happy with a similar library if it would
provide better performance than the current LinuxThreads library.

        Juergen

-- 
Juergen Kreileder, Blackdown Java-Linux Team
http://www.blackdown.org/java-linux.html
JVM '01: http://www.usenix.org/events/jvm01

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Fri Mar 31 2000 - 21:00:25 EST