On Wed, 2005-03-09 at 09:35 +0100, Erik Mouw wrote:
> On Wed, Mar 09, 2005 at 10:22:55AM +0530, Mandeep Sandhu wrote:
> > On Tue, 2005-03-08 at 11:48 +0100, Erik Mouw wrote:
> > > Each driver must request_irq() and free_irq() its own interrupt. IOW:
> > > if the serial driver is unloaded, the interrupt is free. Besides,
> > > you're supposed to call free_irq() with a pointer to your own device
> > > structure, not a NULL pointer.
> > I think NULL can be passed on to free_irq...if the irq line is not
> > shared. dev_id would be usefull only in case of shared interrupts (like
> > those of PCI devices maybe...). correct me if i'm wrong.
>
> Yes, you're right, dev_id is not required in the case of a non-shared
> IRQ (but it certainly doesn't hurt).
>
yeah OK. I was going through free_irq under ../arch/i386/kernel/irq.c
(kernel 2.4.20) to confirm what I said. but the following lines have me
confused...! :P
...
...
for (;;) {
struct irqaction * action = *p;
if (action) {
struct irqaction **pp = p;
p = &action->next;
if (action->dev_id != dev_id) <<<<<*************** (1)
continue;
/* Found it - now remove it from the list of entries */
*pp = action->next;
if (!desc->action) {
desc->status |= IRQ_DISABLED;
....
....
In case I pass NULL as the dev_id for a non-shared interrupt (in request_irq)
will free_irq be able to free it.
Will the condition if(action->dev_id != dev_id) indicated by (1) be FALSE
if both dev_id's are NULL??? just got a little confused by the NULL comparison
thing!
thanks,
-mandeep
>
> Erik
>
--
Kernelnewbies: Help each other learn about the Linux kernel.
Archive: http://mail.nl.linux.org/kernelnewbies/
FAQ: http://kernelnewbies.org/faq/
