RE: FC-AL: FC: Cannot enque FLOGI/REPORT_MAP packet on socal0 port A

Lou,

Thanks for your reply.

E5000 is a sun4u. Machine was very popular in 1999 (when the driver was
developed) and in the developer's notes the EX500 series (which share the
same architecture, just a faster backplane) is explicitely mentioned. If the
driver was not written for Sun4u, Sun4m or Sun4d, what was it written for ?

Regards

Raoul

-----Original Message-----
From: Lou Picciano [mailto:[email protected]]
Sent: Monday, April 29, 2002 5:11 PM
To: SuSE SPARC LIST
Cc: Solarisexpert.com; Matthias Schulz
Subject: Re: [suse-sparc] FC-AL: FC: Cannot enque FLOGI/REPORT_MAP
packet on socal0 port A


Hello!

Your problem is due to architecture-dependent configuration and source files
in the Linux distribution.  This is exactly the behavior I - and others -
have now seen on sun4m architecture using the soc, pluto and fcal modules.
It walks and talks like an IRQ error.  The good news (if we can spin it a
little!) is that the compile _does_ work on other architectures, so there's
hope...

The following is thanks to Matthias Schulz - on this list:

'in the directory /usr/src/linux/arch/sparc/kernel there are at
least different files for sun4m and sun4d architecture. One of the
files deals with interrupts.'

I think we can safely surmise that Jakub Jelinek - who authored the soc
modules and these architecture files - did most of his work on a
newer-architecture machine.  We've tried to email him, but as yet no reply.
Matthias has gotten the soc module working, on a new Linux kernel compile,
on a (E1000?).

One of you guys much more brilliant than me would probably be able to figure
out, and correct this file, for sun4m, in an hour.  For the time being,
though, I am completely dead in the water regarding use of my SS10 with a
fibre interface...

Any help??

Lou Picciano

> I have tried them all:
>
> I used the both the 2.2.20 and the 2.4.18 kernel (which I had to compile
> myself since there is no downloadable version anywhere).
>
> I insmod fc4 first and socal righ after (I have also tried modprobe fcal)
>
> This is what  get on console
>
> socal.c: SOC+ driver v1.1 9/Feb/99 Jakub Jelinek ([email protected])
> FC: Cannot enque FLOGI/REPORT_MAP packet on socal0 port A
> FC: Cannot enque FLOGI/REPORT_MAP packet on socal0 port B
> FC: Cannot enque FLOGI/REPORT_MAP packet on socal0 port A
> FC: Cannot enque FLOGI/REPORT_MAP packet on socal0 port B
> FC: Cannot enque FLOGI/REPORT_MAP packet on socal0 port A
> FC: Cannot enque FLOGI/REPORT_MAP packet on socal0 port B
> FC: Cannot enque FLOGI/REPORT_MAP packet on socal0 port A
> FC: Cannot enque FLOGI/REPORT_MAP packet on socal0 port B
> FC: Cannot enque FLOGI/REPORT_MAP packet on socal0 port A
> FC: Cannot enque FLOGI/REPORT_MAP packet on socal0 port B
> FC: Cannot enque FLOGI/REPORT_MAP packet on socal0 port A
> FC: Cannot enque FLOGI/REPORT_MAP packet on socal0 port B
>
> This is the output of "lsmod"
>
> socal                   6824   0  (autoclean) (unused)
> fc4                    11336   0  [socal]
>
> The machine is an E5000 attache to an A5000 array. Boot channels are
> attached to the array from the built-in SOC+ ports of the I/O board.
>
> ANY IDEAS ? What is the relationship between the drivers. FC4 sets up the
> loop (that was my understanding, but why is it not seeing anything) ?
>
> RV
>
>



--
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]


-- 
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]