Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 5 May 2010 10:04:07 -0600
From:      Scott Long <scottl@samsco.org>
To:        ulf@alameda.net
Cc:        freebsd-scsi@freebsd.org
Subject:   Re: kern/146287: [ciss] ciss(4) cannot see more than one SmartArray controller
Message-ID:  <F4B88810-2F17-4902-8325-432B4FC08F8E@samsco.org>
In-Reply-To: <20100505155946.GL30353@evil.alameda.net>
References:  <201005041800.o44I08d4017511@freefall.freebsd.org> <E860C0FC-EB64-4001-9BE4-C74075DB0D57@samsco.org> <4BE06F10.5080904@intersonic.se> <32244F4E-DAF3-4CC0-B178-32B30076A0D5@samsco.org> <4BE083B8.6070402@intersonic.se> <20100505155946.GL30353@evil.alameda.net>

next in thread | previous in thread | raw e-mail | index | archive | help
On May 5, 2010, at 9:59 AM, Ulf Zimmermann wrote:
> On Tue, May 04, 2010 at 10:29:44PM +0200, Per olof Ljungmark wrote:
>> Scott Long wrote:
>>> Ah, I see a problem where the ACPI driver is rejecting the bridge
>>> that the 2 extra controllers are on.  This machine is likely 7-8
>>> years old, yes?  Disabling ACPI from the freebsd boot menu or from
>>> the BIOS.  Also check if there are any BIOS/firmware upgrades
>>> available for the machine.
>>>=20
>>> Scott
>>=20
>> A G4 should be around five years I think, the last U320 equipped =
ProLiant's.
>=20
> DL380 g4 was retired January 31 2007.
>=20
>>=20
>> Moving the 6400 card to the slot closest to the MB fixed it. There is =
a
>> great deal of confusion in HP's docs on slot numbering and such and
>> although I did not find such information I suspect the controller =
should
>> go inte this slot even if it's pictured in slot two in the manual.
>=20
> There were 3 different cages for the expanasion cards available. =
Non-hotswap
> PCI-X, Hotswap PCI-X and PCI-Express. I don't remember on the G4, but =
on the G3
> the cage was bottom slot 133Mhz and not hotswap, while the top two =
were
> slower, but hotswap. That could mean different PCI bridge chip.
>=20

Right, there's a PCIX bridge that is failing to attach, leaving the bus =
and devices behind it invisible.  I believe that the failure is because =
the bridge is seen in the PCI register scan, but not in the ACPI AML.  =
Because it's not in the AML, the acpi_pcib driver vetos the attach.  I =
believe that if the machine was booted with ACPI disabled, all of the =
bridges+buses+devices would be seen regardless of what slots they are =
in.

Scott




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?F4B88810-2F17-4902-8325-432B4FC08F8E>