On Tue, 7 Jun 2011, Soho Soho123 wrote:
> Dear Alan,,
>
> Could you kindly help to explain more detail about "spin_unlock_irq "
> you mentioned?
Well, netdev->hard_xmit() is called from somewhere, right? The place
where it gets called from might be running with interrupts disabled.
Now see what happens if hard_xmit() uses spin_lock_irq() and
spin_unlock_irq():
some_other_function() {
...
disable interrupts
...
call netdev->hard_xmit()
hard_xmit() {
spin_lock_irq();
...
spin_unlock_irq();
/* Now interrupts are enabled again */
}
...
/* An unexpected interrupt occurs -- CRASH! */
...
enable interrupts
}
The problem is that spin_unlock_irq() always enables interrupts, even
if interrupts were disabled when spin_lock_irq() was called.
Alan Stern
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
