Date: Tue, 6 Jul 1999 13:30:02 -0700 (PDT) From: Matthew Dillon <dillon@apollo.backplane.com> To: Ted Faber <faber@ISI.EDU> Cc: hackers@FreeBSD.ORG Subject: Re: kern/11222 Message-ID: <199907062030.NAA86024@apollo.backplane.com> References: <199907062015.NAA17844@boreas.isi.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi Ted. I did look at it and it seems ok, though I didn't run it
by the gurus. Unfortunately, without commit privs my hands are
tied and I am just not willing to spend the hours of testing necessary
to ensure that obvious bugs have been avoided when other people get
the credit and the comments in the commit message.
I was hoping the issue would be resolved at USENIX but core is unwilling
to discuss the issue in a frank and open manner, instead opting for closed,
one-sided, and jaded discussions. Even now what they say on open forums
belies their actual actions - which so far has been zippo. It is a sad
state of affairs and although I'm sure I'm not making very many friends by
keeping the pressure up, at this point I don't have anything to lose.
Core has been entirely unresponsive for several months now and I'm
through being nice about it.
Perhaps someone else can look at it and commit it.
There are several items on my hot list submitted by other people which
I have not been able to move on due to the lack of resolution to my
committer's status, including some very cool looking optimizations that
would cut NFS stat traffic in half.
I find it rather humorous that I always spend hours testing things that
other committers would commit without any testing whatsoever.
-Matt
Matthew Dillon
<dillon@backplane.com>
:-----BEGIN PGP SIGNED MESSAGE-----
:
:Hi, folks. I was hoping that someone would take a look at kern/11222,
:which has been open since I submitted it in April. The problem it
:addresses is that MFS buffers that are dirty when the reboot syscall
:is made are never cleaned because the reboot call never calls sleep,
:so the user-level process backing the MFS can never run to mark the
:buffers as clean. There are two patches with the pr, one from the
:date it was submitted which specifically detects MFS buffers and calls
:sleep if they're present, and one which replaces the DELAY calls in
:reboot with tsleep calls. The second was at the suggestion of Matt
:Dillon, who looked at the patch and PR in late April, but apparently
:hasn't had more time to look at it (yes, I pinged him).
:
:Is there someone else who is knowledgable about the reboot sequence
:who's willing to have a look at these patches? I've been running the
:second patch (removes DELAY) on my 3 3.1-RELEASE and 3.2-RELEASE
:machines (2 desktops and a laptop) without incident since April. I
:think they're pretty stable. Of course I'm happy to discuss the
:patches in tdetail with anyone willing to look at them.
:
:Thanks!
:
:- ----------------------------------------------------------------------
:Ted Faber faber@isi.edu
:USC/ISI Computer Scientist http://www.isi.edu/~faber
:(310) 822-1511 x190 PGP Key: http://www.isi.edu/~faber/pubkey.asc
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199907062030.NAA86024>