On Sun, 2009-03-15 at 10:01 -0700, Joe Eykholt wrote:
> 
> Sivaram Kannan wrote:
> > Hi all,
> > 
> > I have created a patch for Open-Fcoe-Target which could compile against the 
> > latest kernel test
> branch(linux-2.6.29-rc8). The same patch will also compile against the latest 
> stable version 2.6.28. I exposed the
> target and when I triggered the initiator, I could see sequence of SCSI 
> packets at the wireshark. I did not understand
> the messages completely, I am decoding it. One problem that I noticed during 
> the testing is with the initiator. For
> initiator I compiled the latest 2.6.29-rc8 in Fedora 10, and when I give the 
> command "echo "ethX" >
> /sys/module/fcoe/parameters/create", I am getting a kernel opps. There is no 
> trouble at the target end, wireshark was
> still printing SCSI packets. Would attaching the wireshark trace from target 
> end would be helpful to identify what is
> the problem?
> 
> Wireshark might help, but more useful would be the serial console output from 
> the crashing

Also, if you turn on libfc debugging for certain sub-modules you might
get more information about the library's state when it fails. To do so,
just change-

static int fc_lport_debug;

to

static int fc_lport_debug = 1;

(in this case in libfc/fc_lport.c)

recompile and reproduce and you should have more verbose output.

Since you're doing point-to-point mode, I'd probably start by turning on
fc_lport_debug and fc_rport_debug, but turning on more of them is
probably _not_ overkill and could be sorted through.


_______________________________________________
devel mailing list
[email protected]
http://www.open-fcoe.org/mailman/listinfo/devel

Reply via email to