Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 27 Jun 2006 11:30:31 -0400
From:      John Baldwin <jhb@freebsd.org>
To:        Andrew Reilly <andrew-freebsd@areilly.bpc-users.org>
Cc:        freebsd-amd64@freebsd.org
Subject:   Re: SMP system not running SMP
Message-ID:  <200606271130.32300.jhb@freebsd.org>
In-Reply-To: <20060626232753.GB92989@duncan.reilly.home>
References:  <200606231908.k5NJ8DTB009354@guild.plethora.net> <200606260940.48404.jhb@freebsd.org> <20060626232753.GB92989@duncan.reilly.home>

next in thread | previous in thread | raw e-mail | index | archive | help
On Monday 26 June 2006 19:27, Andrew Reilly wrote:
> On Mon, Jun 26, 2006 at 09:40:48AM -0400, John Baldwin wrote:
> > Ok.  That sounds like an interrupt routing issue.   It could be that the
> > interrupt routing info in the MP Table is incorrect.  Let's stick with
> > i386 for now (amd64 has the same code).
> 
> For another, perhaps unrelated datum, my Athlon64-X2 box has
> become wonderfully stable and seemingly fully functional ever
> since I took all of the PCI cards out of it.  I wasn't trying to
> run SCSI disks (just on-board SATA), but I was trying to use PCI
> network adaptors, since there was no driver for the on-board
> NVidia MCP9 at the time.  (I'm now using the experimental nfe
> driver, and it seems to work great.)  I tried a bog-standard crappy
> RTL8139 and a good-as-gold Intel 21143, and they both displayed
> symptoms consistent with missing interrupts: failure to detect
> carrier, multiple up/downs from watchdog timeouts, etc.
> 
> Now, my Gigabyte Nvidia4 board has a totally different motherboard
> chipset from the MSI K8D Master-F in question, but the similar
> behaviour just seemed interesting.
> 
> There isn't any vm86-mode setup (bios calls?) involved in
> routing interrupts, is there?  Any similarity in BIOS vendors
> between the two boards? (Maybe firmware bug?)

It's all done via table lookups and ACPI calls, no vm86 stuff.  The tables and 
ACPI code is provided by the BIOS though, so a horked BIOS can always break 
things.

-- 
John Baldwin



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200606271130.32300.jhb>