Re: [FIXED] Cannot install onto mpt-driven drive

From: Joe Marcus Clarke <marcus_at_marcuscom.com>
Date: Sat, 21 Aug 2004 18:11:10 -0400
On Sat, 2004-08-21 at 00:12, Joe Marcus Clarke wrote:
> On Fri, 2004-08-20 at 23:17, Joe Marcus Clarke wrote:
> > I just got a new Dell Precision 360 workstation with an LSILogic SCSI
> > controller:
> > 
> > mpt0: <LSILogic 1030 Ultra4 Adapter> port 0xcd00-0xcdff mem
> > 0xfe780000-0xfe79ffff,0xfe760000-0xfe77ffff irq 10 at device 3.0 on pci2
> > mpt1: <LSILogic 1030 Ultra4 Adapter> port 0xce00-0xceff mem
> > 0xfe7x0000-0xfe7dffff,0xfe7a0000-0xfe7bffff irq 9 at device 3.1 on pci2
> > 
> > I'm trying to install 5.2.1-RELEASE on this machine, then upgrade it to
> > RELENG_5.  The controller and 36 GB SCSI harddrive are detected just
> > fine at boot time, but as soon as I go to write out my partitions to
> > this drive, I get a timeout:
> > 
> > mpt1: time out on request index = 0xfe sequence = 0x000001a0
> > mpt: Status 00000001, Mask 00000001, Doorbell 24000000
> > request state On Chip
> > SCSI IO Request _at_ 0xcde1fc30
> > 	Chain Offset 0x10
> > 	Msg Flags 0x00
> > 	MsgContext 0x000000fe
> > 	Bus: 0
> > 	TargetID 0
> > 	SenseBufferLength 32
> > 	LUN: 0x0
> > 	Control 0x01000000 WRITE SIMPLEQ
> > 	DataLength 0x00010000
> > 	SenseBufferAddr 0x00391de0
> > 	CDB[0:6] 0a 01 3a 65 80 00
> > 	SE32 0xd6937c30: Addr=0x9aa7000 FlagsLength=0x14001000
> > 	 HOST_TO_IOC
> > 	SE32 0xd6937c38: Addr=0xa128000 FlagsLength=0x94001000
> > 	 HOST_TO_IOC LAST_ELEMENT
> > 	CE32 0xd6937c40: Addr=0x391c48 NxtChn0=0x16 Flags=0x30 Len=0x50
> > 	SE32 0xd6937c48: Addr=0xaa69000 FlagsLength=0x14001000
> > 	 HOST_TO_IOC
> > 
> > This trace continues for a while, but I'm transcribing it from the
> > screen.  I can certainly do the entire thing if someone is interested. 
> > I can even set someone up with serial console access if they'd like. 
> > Also note: this same model controller and drive work in another Dell
> > Precision machine under -STABLE.   I've tried booting 5.2.1 under normal
> > sand safe modes with the same result.  I looked in CVS, and there didn't
> > appear to be any relevant driver changes since 5.2.1-RELEASE.  Hoping
> > someone can help me out.  Thanks.
> 
> Sorry to follow-up to myself, but it looks like this problem can be
> fixed by the patch in
> http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/66479.  I saw the
> original NetBSD PR that reported this:
> http://www.netbsd.org/cgi-bin/query-pr-single.pl?number=24967, and my
> symptoms look the same.
> 
> I'll try and build a release with this patch to see if it does the
> trick.  In the meantime, if someone could have a look at this PR, that
> would be great.

Okay, one last follow-up to myself.  I built a custom 5.2.1-RELEASE with
the patch from the above PR, and it worked perfectly.  I was able to
install 5.2.1 on this system, and then boot the resulting install.  It
would be really great if this could make it into 5.3-RELEASE, or at
least into HEAD.  Thanks.

Joe

> 
> Joe
> 
> > 
> > Joe
-- 
PGP Key : http://www.marcuscom.com/pgp.asc

Received on Sat Aug 21 2004 - 20:11:16 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:07 UTC