Date: Wed, 2 Jan 2008 22:14:38 -0500 From: Nick Pope <lists@mipster.net> To: Travis Mikalson <bofh@terranova.net> Cc: freebsd-current@freebsd.org Subject: Re: ServerWorks/Broadcom HT1000 chipset errata saga Message-ID: <E3911800-6717-46AF-90E7-93EAAAA38882@mipster.net> In-Reply-To: <477B0B07.8080703@terranova.net> References: <1497D115-2534-4799-9D8E-18A267DF0B62@mipster.net> <9A374150-DC2A-439D-A205-E8867B663C5A@mipster.net> <477B0B07.8080703@terranova.net>
next in thread | previous in thread | raw e-mail | index | archive | help
I csup'ed to latest RELENG_7_0 (RC1) and built the GENERIC kernel. =20
I'm assuming Soren's patch has made it into RC1 because, without =20
commenting out the rr232x driver, the kernel mis-detects my Supermicro =20=
PCI-X SATA controller (MV88SX6081 chipset) as an rr232x.
--snip--
hptrr0: <sx508x> port 0x7800-0x78ff mem 0xfc300000-0xfc3fffff irq 28 =20
at device 3.0 on pci1
hptrr: adapter at PCI 1:3:0, IRQ 28
--snip--
Here's pciconf -lv:
[root@backup2 ~]# lspci -lv
-bash: lspci: command not found
[root@backup2 ~]# pciconf -lv
pcib1@pci0:0:6:0: class=3D0x060400 card=3D0x00000000 =
chip=3D0x74601022 =20
rev=3D0x07 hdr=3D0x01
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D 'AMD-8111 PCI Bridge'
class =3D bridge
subclass =3D PCI-PCI
isab0@pci0:0:7:0: class=3D0x060100 card=3D0x74681022 =
chip=3D0x74681022 =20
rev=3D0x05 hdr=3D0x00
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D 'AMD-8111 LPC Bridge'
class =3D bridge
subclass =3D PCI-ISA
atapci1@pci0:0:7:1: class=3D0x01018a card=3D0x74691022 =
chip=3D0x74691022 =20
rev=3D0x03 hdr=3D0x00
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D 'AMD-8111 UltraATA/133 Controller'
class =3D mass storage
subclass =3D ATA
none0@pci0:0:7:2: class=3D0x0c0500 card=3D0x746a1022 =
chip=3D0x746a1022 =20
rev=3D0x02 hdr=3D0x00
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D 'AMD-8111 SMBus 2.0 Controller'
class =3D serial bus
subclass =3D SMBus
none1@pci0:0:7:3: class=3D0x068000 card=3D0x746b1022 =
chip=3D0x746b1022 =20
rev=3D0x05 hdr=3D0x00
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D 'AMD-8111 ACPI System Management Controller'
class =3D bridge
pcib2@pci0:0:10:0: class=3D0x060400 card=3D0x00000000 =
chip=3D0x74501022 =20
rev=3D0x12 hdr=3D0x01
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D 'AMD-8131 PCI-X Bridge'
class =3D bridge
subclass =3D PCI-PCI
ioapic0@pci0:0:10:1: class=3D0x080010 card=3D0x36c01022 =
chip=3D0x74511022 =20
rev=3D0x01 hdr=3D0x00
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D 'AMD-8131 PCI-X IOAPIC'
class =3D base peripheral
subclass =3D interrupt controller
pcib3@pci0:0:11:0: class=3D0x060400 card=3D0x00000000 =
chip=3D0x74501022 =20
rev=3D0x12 hdr=3D0x01
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D 'AMD-8131 PCI-X Bridge'
class =3D bridge
subclass =3D PCI-PCI
ioapic1@pci0:0:11:1: class=3D0x080010 card=3D0x36c01022 =
chip=3D0x74511022 =20
rev=3D0x01 hdr=3D0x00
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D 'AMD-8131 PCI-X IOAPIC'
class =3D base peripheral
subclass =3D interrupt controller
hostb0@pci0:0:24:0: class=3D0x060000 card=3D0x00000000 =
chip=3D0x11001022 =20
rev=3D0x00 hdr=3D0x00
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D '(K8) Athlon 64/Opteron HyperTransport Technology =20=
Configuration'
class =3D bridge
subclass =3D HOST-PCI
hostb1@pci0:0:24:1: class=3D0x060000 card=3D0x00000000 =
chip=3D0x11011022 =20
rev=3D0x00 hdr=3D0x00
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D '(K8) Athlon 64/Opteron Address Map'
class =3D bridge
subclass =3D HOST-PCI
hostb2@pci0:0:24:2: class=3D0x060000 card=3D0x00000000 =
chip=3D0x11021022 =20
rev=3D0x00 hdr=3D0x00
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D '(K8) Athlon 64/Opteron DRAM Controller'
class =3D bridge
subclass =3D HOST-PCI
hostb3@pci0:0:24:3: class=3D0x060000 card=3D0x00000000 =
chip=3D0x11031022 =20
rev=3D0x00 hdr=3D0x00
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D '(K8) Athlon 64/Opteron Miscellaneous Control'
class =3D bridge
subclass =3D HOST-PCI
hostb4@pci0:0:25:0: class=3D0x060000 card=3D0x00000000 =
chip=3D0x11001022 =20
rev=3D0x00 hdr=3D0x00
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D '(K8) Athlon 64/Opteron HyperTransport Technology =20=
Configuration'
class =3D bridge
subclass =3D HOST-PCI
hostb5@pci0:0:25:1: class=3D0x060000 card=3D0x00000000 =
chip=3D0x11011022 =20
rev=3D0x00 hdr=3D0x00
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D '(K8) Athlon 64/Opteron Address Map'
class =3D bridge
subclass =3D HOST-PCI
hostb6@pci0:0:25:2: class=3D0x060000 card=3D0x00000000 =
chip=3D0x11021022 =20
rev=3D0x00 hdr=3D0x00
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D '(K8) Athlon 64/Opteron DRAM Controller'
class =3D bridge
subclass =3D HOST-PCI
hostb7@pci0:0:25:3: class=3D0x060000 card=3D0x00000000 =
chip=3D0x11031022 =20
rev=3D0x00 hdr=3D0x00
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D '(K8) Athlon 64/Opteron Miscellaneous Control'
class =3D bridge
subclass =3D HOST-PCI
ohci0@pci0:3:0:0: class=3D0x0c0310 card=3D0x74641022 =
chip=3D0x74641022 =20
rev=3D0x0b hdr=3D0x00
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D 'AMD-8111 USB OpenHCI Host Controller'
class =3D serial bus
subclass =3D USB
ohci1@pci0:3:0:1: class=3D0x0c0310 card=3D0x74641022 =
chip=3D0x74641022 =20
rev=3D0x0b hdr=3D0x00
vendor =3D 'Advanced Micro Devices (AMD)'
device =3D 'AMD-8111 USB OpenHCI Host Controller'
class =3D serial bus
subclass =3D USB
atapci0@pci0:3:5:0: class=3D0x018000 card=3D0x31141095 =
chip=3D0x31141095 =20
rev=3D0x02 hdr=3D0x00
vendor =3D 'Silicon Image Inc (Was: CMD Technology Inc)'
device =3D 'Sil 3114 SATALink/SATARaid Controller'
class =3D mass storage
vgapci0@pci0:3:6:0: class=3D0x030000 card=3D0x80081002 =
chip=3D0x47521002 =20
rev=3D0x27 hdr=3D0x00
vendor =3D 'ATI Technologies Inc'
device =3D 'Rage XL PCI'
class =3D display
subclass =3D VGA
re0@pci0:2:8:0: class=3D0x020000 card=3D0x816910ec chip=3D0x816910ec =20
rev=3D0x10 hdr=3D0x00
vendor =3D 'Realtek Semiconductor'
device =3D 'RTL8110SB Single-Chip Gigabit LOM Ethernet =20
Controller'
class =3D network
subclass =3D ethernet
bge0@pci0:2:9:0: class=3D0x020000 card=3D0x164414e4 =
chip=3D0x164814e4 =20
rev=3D0x03 hdr=3D0x00
vendor =3D 'Broadcom Corporation'
device =3D 'BCM5704 NetXtreme Dual Gigabit Adapter'
class =3D network
subclass =3D ethernet
bge1@pci0:2:9:1: class=3D0x020000 card=3D0x164414e4 =
chip=3D0x164814e4 =20
rev=3D0x03 hdr=3D0x00
vendor =3D 'Broadcom Corporation'
device =3D 'BCM5704 NetXtreme Dual Gigabit Adapter'
class =3D network
subclass =3D ethernet
ahd0@pci0:2:10:0: class=3D0x010000 card=3D0xffff9005 =
chip=3D0x801f9005 =20
rev=3D0x10 hdr=3D0x00
vendor =3D 'Adaptec Inc'
device =3D 'AIC-7902 Ultra320 SCSI Controller'
class =3D mass storage
subclass =3D SCSI
ahd1@pci0:2:10:1: class=3D0x010000 card=3D0xffff9005 =
chip=3D0x801f9005 =20
rev=3D0x10 hdr=3D0x00
vendor =3D 'Adaptec Inc'
device =3D 'AIC-7902 Ultra320 SCSI Controller'
class =3D mass storage
subclass =3D SCSI
hptrr0@pci0:1:3:0: class=3D0x010000 card=3D0x11ab11ab =
chip=3D0x608111ab =20
rev=3D0x09 hdr=3D0x00
vendor =3D 'Marvell Semiconductor (Was: Galileo Technology =
Ltd)'
device =3D 'MV88SX6081 8-port SATA II PCI-X Controller'
class =3D mass storage
subclass =3D SCSI
-Nick
On Jan 1, 2008, at 10:54 PM, Travis Mikalson wrote:
> Nick Pope wrote:
>> Installed the patch, now it thinks my Marvel SATA controller is an =20=
>> rr232x!!
>> If I comment out rr232x driver in the kernel config, the Marvel is =20=
>> detected no problem. However, as soon as I mount my zpool (using =20
>> the Marvel controller) the Ierrs start up again.
>> -Nick
>> On Dec 30, 2007, at 11:52 PM, <lists@mipster.net> =20
>> <lists@mipster.net> wrote:
>>> BTW, I've had PCI-X problems with a Marvel MV88SX6081 SATA =20
>>> controller on a M/B WITHOUT the HT1000 chipset (Tyan S2881UG2NR). =20=
>>> The M/B worked flawlessly under 7.0-BETA1 & BETA2. I added the =20
>>> Marvell controller and EVERYTHING on the PCI-X bus flaked out. =20
>>> I'm seeing Ierrs and Oerrs (netstat -i) on the built-in bge =20
>>> interfaces, ZFS checksum errors on disks attached to the Marvel =20
>>> controller, etc.
>>>
>>> I wasn't sure from reading the thread whether Soren thought the =20
>>> problem went beyond the HT1000 chipset or not. The problem seems =20=
>>> to manifest itself only when I use the disks attached to the =20
>>> Marvel controller. I'm applying the patch now to see if it fixes =20=
>>> the problem.
>>>
>>> -Nick
>
> Thanks for trying it out. I just waved this in front of S=F8ren.
>
> The apparently new mis-detection issue is certainly something to be =20=
> concerned about. You're absolutely sure it was detected as a Marvell =20=
> reliably (with rr232x still in there) before you applied that patch?
>
> The fact that the patch didn't do anything to solve your Marvell =20
> controller problem is too bad, also, but the mis-detection issue is =20=
> of more immediate concern since this HT1000/Marvell-fixing code is =20
> being vetted for inclusion in 7.0-R.
>
> Have you tried that hardware combination with any other OSes to see =20=
> if it's a FreeBSD-specific issue and not just a hardware-level issue?
>
> I haven't had a chance to confirm myself since my box with the =20
> Marvell controller in it is a long drive away from me, but S=F8ren did =
=20
> get his PCI-X 8-port Marvell working perfectly fine after making =20
> these fixes.
>
> --=20
> TerraNovaNet Internet Services - Key Largo, FL
> Voice: (305)453-4011 x101 Fax: (305)451-5991
> http://www.terranova.net/
> ----------------------------------------------
> Life's not fair, but the root password helps.
> _______________________________________________
> freebsd-current@freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-current
> To unsubscribe, send any mail to =
"freebsd-current-unsubscribe@freebsd.org=20
> "
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?E3911800-6717-46AF-90E7-93EAAAA38882>