[Date Prev][Date Next]
Re: spurious connection_read()s
> This is solaris, right? I think it is a side effect of the solaris file
> descriptor kernel bug.
Wasn't there a workaround in place for that?
Not that I think a workaround is such a terrific idea. If I should be
yelling at Sun, I'll yell at Sun. But is there something in the world a
bit less complex than slapd that can prove wrong Solaris behavior? It
would be good to hand them a test case that's smaller than 150KLOC. Or if
its a known issue that needs votes, a Solaris bug id would be more useful
than a test case.