Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 30 May 1998 23:37:20 -0500 (EST)
From:      "John S. Dyson" <dyson@FreeBSD.ORG>
To:        mbriggs@switchboard.net (Matthew R. Briggs)
Cc:        freebsd-current@FreeBSD.ORG
Subject:   Re: General SMP Instability SNAP-0523
Message-ID:  <199805310437.XAA00448@dyson.iquest.net>
In-Reply-To: <3570CD13.87BEDFA9@switchboard.net> from "Matthew R. Briggs" at "May 30, 98 11:22:59 pm"

next in thread | previous in thread | raw e-mail | index | archive | help
Matthew R. Briggs said:
> Hello,
> 	I'm new to FreeBSD, having come from the Debian GNU/Linux camp. I'm
> very impressed with what I have seen while working with the CURRENT
> snapshot from 0523, but stability has been an issue.  I know that the
> CURRENT-3.0 branch is under heavy development, but I only run SMP boxen
> and to try FreeBSD out I had to go to CURRENT.
> 	Here is the issue:  I am able to wedge the system fairly consistently
> (and often) with parallel makes of various pieces of software.  Not the
> kernel, but "make world -j4" will crash it, as will compilation of
> FileRunner from the ports collection (along with a few other things). 
> The system will lock hard and drop me to the kernel debugger.
> 	Since I'm new to this, what I'm asking for is a description of what
> information I should send to the right people to fix any bugs or to
> expose my own ignorance.  Here is the hardware I am using to test things
> out:
> 
I am working like a mad-man on the SMP context switching and other
like code.   I will have a new SMP upgrade snapshot in my
www.freebsd.org:~dyson directory at the end of the weekend.  Alot
of our diagnostic info is pretty rough, but feel free to send me/post
any info that you feel will be useful.  Useful kernel debugger
commands are "ps" and "trace".  Those will sometimes give a useful
pointer to some troubles.  However, with SMP, the system sometimes gets into
a state where the traceback is useless (in my private working sources,
and soon to be committed, or at least in my SMP update(s), stack
tracebacks will work better, due to maintaining frame pointer.)

One warning is that softupdates tends to make the system less stable
also.  SMP and softupdates can be challenging :-).

> 
> Thanks for any and all help...I'd really like to get this to work!
> 
I would like to be able to help, others will likely also try to chip
in.

-- 
John                  | Never try to teach a pig to sing,
dyson@freebsd.org     | it just makes you look stupid,
jdyson@nc.com         | and it irritates the pig.

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?199805310437.XAA00448>