[openib-general] Re: testing amso1100

[Tom Tucker]

The firmware on the adapter is down level.  Sorry, the message needs to
just say that.  "Invalid PCI regs magic" isn't particularly descriptive.
There's a new firmware load mechanism that we have been using. You can
download this mechanism from our website at
http://www.opengridcomputing.com/downloads.html. Select Ammasso 1100
OPENIB Device Support Kit.

On Fri, 2006-03-17 at 16:27 -0500, Pete Wyckoff wrote:
> I'm trying to get your iwarp branch working with our Ammasso NICs.
> Starting from a stock linux-2.6.16-rc6-git9 and with svn rev 5878, I
> replaced two kernel dirs with links into svn:
> 
>     <linux>/include/rdma ->
> 	<svn>/gen2/branches/iwarp/src/linux-kernel/infiniband/include/rdma
>     <linux>/drivers/infiniband ->
> 	<svn>/gen2/branches/iwarp/src/linux-kernel/infiniband
> 
> and a little patch in net/ipv4/fib_frontend.c to re-export
> ip_dev_find().  CONFIG_INFINIBAND_AMSO1100_DEBUG is on.
> 
> The iw_c2 module isn't happy about the state of the NIC.  It says
> via dmesg (on three different identical machines):
> 
>     c2: AMSO1100 Gigabit Ethernet driver v1.1 loaded
>     ACPI: PCI Interrupt 0000:09:08.0[A] -> GSI 24 (level, low) -> IRQ 209
>     c2: BAR0 size = 0x4000 bytes
>     c2: BAR2 size = 0x200000 bytes
>     c2: BAR4 size = 0x20000000 bytes
>     c2: Invalid PCI regs magic [1/7: got 0x42, exp 0x43]
>     c2: Adapter not claimed
>     ACPI: PCI interrupt for device 0000:09:08.0 disabled
>     c2: probe of 0000:09:08.0 failed with error -5
> 
> I made it print out the magic regs:  BOOTLDR rather than CEPHEUS.
> 
> lspci says:
> 
>     am28# lspci -v -xxxx -s 09:08.0
>     09:08.0 Co-processor: Ammasso AMSO 1100 iWARP/RDMA Gigabit Ethernet Coprocessor (rev 01)
> 	    Subsystem: Ammasso AMSO 1100 iWARP/RDMA Gigabit Ethernet Coprocessor
> 	    Flags: slow devsel, IRQ 209
> 	    Memory at b2600000 (64-bit, prefetchable) [size=16K]
> 	    Memory at b2800000 (64-bit, prefetchable) [size=2M]
> 	    Memory at c0000000 (64-bit, prefetchable) [size=512M]
> 	    [virtual] Expansion ROM at b2610000 [disabled] [size=64K]
> 	    Capabilities: [40] Power Management version 2
> 	    Capabilities: [48] Message Signalled Interrupts: 64bit+ Queue=0/0 Enable-
> 	    Capabilities: [58] PCI-X non-bridge device.
>     00: b8 18 01 b0 13 01 10 04 01 00 40 0b 10 40 00 00
>     10: 0c 00 60 b2 00 00 00 00 0c 00 80 b2 00 00 00 00
>     20: 0c 00 00 c0 00 00 00 00 00 00 00 00 b8 18 01 b0
>     30: 00 00 00 00 40 00 00 00 00 00 00 00 0b 01 40 00
>     40: 01 48 0a 48 00 00 00 00 05 58 80 00 00 00 00 00
>     50: 00 00 00 00 00 00 00 00 07 00 02 00 40 09 03 00
>     60: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>     70: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>     80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>     90: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>     a0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>     b0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>     c0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>     d0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>     e0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>     f0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
> 
> This box is a
> 
>     Linux am28 2.6.16-rc6-git9 #2 SMP Fri Mar 17 15:46:39 EST 2006 x86_64 x86_64 x86_64 GNU/Linux
>     fc4 updated
>     Tyan S2891 mobo 2-cpu
> 
> Any idea what's wrong here?  Thanks,
> 
> 		    -- Pete