> 5 & 11) Nope, I had not read Doc/networking/netdevices.txt, so I have a
question. What does being inside rtnl_lock() imply other than the sleep
> issues?
>
> The calls to cli() and save_flags() were wrong from the beginning.
They were imported by the last maintainer since this driver is a modified
version > of the olympic token ring driver. The current spin_lock() and
spin_unlock() calls protect the srb_queued variable. If we were to set it
to one and then > get interrupted before we actually write() to the srb,
the interrupt function will try to service whatever junk happens to be in
the srb. If going into the
> interrupt function is covered by rtnl_lock(), this would be covered, but
I guess its not (?)....
actually I should be using spin_lock_irqsave() in open() and close()
since the lock is taken inside the interrupt function, no?
Kent
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/