Re: make test fails (Kernel 2.6.12.3)

This was the last message I had from Tom at RH about porting to ia64.
I also told him that I expected there to be more bugs after this, and
that somebody had better buy me an ia64 processer and motherboard if
they expected me to be able to do the port properly!


----- Forwarded message from Tom 'spot' Callaway -----

On Fri, 2005-07-15 at 21:10 +0200, Peter T. Breuer wrote:

> Meanwhile, in enbd_ack, do the comparison after casting, and fixup.
> 
> 
>                 xreq = list_entry (pos, struct request, queuelist);
>                 if (count++ > 1000)
>                         break;
> -               if (xreq == req)
> +               if ((typeof(reply.handle))xreq == (typeof(reply.handle))req) {
> -                       /* PTB found it */
> +                       /* PTB found it - fix up if necessary. Yell? */
> +                       if (req != xreq)
> +                               req = xreq;
>                         break;
> +               }

This did the trick. I'm now able to get to /dev/nda, and mount the
device on the client machine.

Thank you for your assistance!

~spot
-- 
Tom "spot" Callaway: Red Hat Senior Sales Engineer || GPG ID: 93054260
Fedora Extras Steering Committee Member (RPM Standards and Practices)
Aurora Linux Project Leader: http://auroralinux.org
Lemurs, llamas, and sparcs, oh my!

----- End of forwarded message from Tom 'spot' Callaway -----

(I believe that I added that patch to the driver code - you are seeing
userspace compilation problems).

Peter