Date: Mon, 18 Feb 2002 23:18:46 -0800 (PST) From: Matthew Dillon <dillon@apollo.backplane.com> To: Peter Wemm <peter@wemm.org> Cc: current@FreeBSD.ORG Subject: Re: Patch sets to date and timing tests with Giant out of userret. Message-ID: <200202190718.g1J7Ik959010@apollo.backplane.com> References: <20020219071214.B21BE3A9A@overcee.wemm.org>
next in thread | previous in thread | raw e-mail | index | archive | help
:Regarding the instrumentation of Giant for *trivial* stuff like this: I'm
:one of the people you called "bozos" that disagrees with you about the
:usefulness of bloating the source with this stuff that only needs to be
:removed again later.
:
:Cheers,
:-Peter
:--
:Peter Wemm - peter@FreeBSD.org; peter@yahoo-inc.com; peter@netplex.com.au
You have a right to your opinion, but please keep in mind that the
instrumentation is there to compartmentalize work being done on entire
subsystems. Even though I think Giant instrumentation isn't
necessarily necessary for something like getuid() I finally decided that
it had to be instrumented because people are working on so many subsystems
in parallel that the ONLY way one can debug them properly is by having
the ability to put everyone else's subsystem work back under Giant.
If the ucred subsystem were the only thing we were working on I probably
wouldn't bother. But it ISN'T the only thing we are working on. We
are working on ucred, PROC locking, file descriptor locking, and soon
we will be working on file ops, vnops, the I/O subsystem, the network
stack..... you see where this is leading?
-Matt
Matthew Dillon
<dillon@backplane.com>
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200202190718.g1J7Ik959010>