Testing needed for new pipe code (in master)

Matthew Dillon dillon at apollo.backplane.com
Tue Jul 28 11:34:32 PDT 2009


:> 
:>     The latest code is now able to fully saturated all available cpus
:>     when running multiple copies of the pipe2 test (from
:>     /usr/src/test/sysperf).
:
:Recently I occasionally see processes stuck on `piperd' state, probably
:after it's been turned on by default.  The stuck processes can be
:interrupted and re-issued, though.  It happened on both vkernel and
:the real kernel.  The first time I saw it was when `make upgrade' was
:performing `./MAKEDEV upgrade', so I thought it was temporary back then.

    Ok, I'll start investigating.  If you can get a kernel core with one
    or more processes stuck in that state it would help a lot.  I am
    assuming that it is a write/read-block or close/read-block race of
    some sort somewhere.  I had a few like that when I was developing the
    code but I thought I had closed them all.

					-Matt
					Matthew Dillon 
					<dillon at backplane.com>





More information about the Kernel mailing list