Date: Tue, 08 Jul 1997 04:47:22 -0700 From: David Greenman <dg@root.com> To: Simon Shapiro <Shimon@i-Connect.Net> Cc: FreeBSD-Hackers@FreeBSD.ORG Subject: Re: PCI Bridge Question Message-ID: <199707081147.EAA08802@implode.root.com> In-Reply-To: Your message of "Mon, 07 Jul 1997 23:29:34 PDT." <XFMail.970707232934.Shimon@i-Connect.Net>
next in thread | previous in thread | raw e-mail | index | archive | help
>> >> an mbuf cluster had been freed onto the mclfree list. In any case,
>> this
>> >> appears to be a much more generic problem - not specific to the fxp
>> >> device
>> >> driver.
>> >
>> >Most likely. The fxp is where I see it on this system.
>>
>> I can't reproduce the problem here, but I haven't tried very hard. Is
>> NFS
>> static in your kernel, or is it getting loaded as an LKM?
>
>* Setup BONDING PPP (128Kbps)
>* NFS mount a large file system (say, FreeBSD/packages-current) directly
> form the host you PPP to
>* cd to the NFS mount point
>* find . | cpio -dmpv /somewhere/else
>* Go about your business
>* Wait about 20-30 minutes
>
>[ I don't really expect you to do that. This is the setup I am using to
> ``prove'' the problem ]
It would be useful if you could examine the longword pointed to by mclfree;
it is apparantly outside of the mb_map map range. Knowing it's value might
give a clue as to the cause of the problem (then again, maybe not).
>> Use -current kernel with -current userland sources; start with 2.2.2
>> and
>> upgrade to -current ("3.0") with a "make world" if necessary.
>
>I understand the procedure and it works well. I was inquiring about 2.2
>userland and -current kernel. Guess your answer is ``no''. the more I play
>with it, the less I like it (the hybrid idea).
Right, I'm saying it's a bad idea and will cause problems.
-DG
David Greenman
Core-team/Principal Architect, The FreeBSD Project
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199707081147.EAA08802>