Re: [PATCH] Fix shared interrupt handling of SA_INTERRUPT and SA_SAMPLE_RANDOM

[Russell King]

On Thu, Aug 26, 2004 at 04:10:54PM +0200, Takashi Iwai wrote:
> At Thu, 26 Aug 2004 15:04:04 +0100,
> Russell King wrote:
> > 
> > On Thu, Aug 26, 2004 at 02:50:52PM +0200, Takashi Iwai wrote:
> > > At Wed, 25 Aug 2004 13:41:12 -0700,
> > > Andrew Morton wrote:
> > > > 
> > > > Takashi Iwai <tiwai@xxxxxxx> wrote:
> > > > >
> > > > > Anyway, suppressing the unnecessary call of add_interrupt_randomness()
> > > > >  should be still valid.  The reduced patch is below.
> > > (snip)
> > > > 
> > > > Shouldn't that be `if (ret == IRQ_HANDLED)'?
> > > 
> > > Yes, it's more strict.
> > 
> > I don't think so.  Look at what's going on.  If "ret" is IRQ_HANDLED
> > all well and fine.  However, look at how "retval" is being used:
> > 
> > static void __report_bad_irq(int irq, irq_desc_t *desc, irqreturn_t action_ret)
> > {
> > ...
> >         if (action_ret != IRQ_HANDLED && action_ret != IRQ_NONE) {
> >                 printk(KERN_ERR "irq event %d: bogus return value %x\n",
> >                                 irq, action_ret);
> >         } else {
> >                 printk(KERN_ERR "irq %d: nobody cared!\n", irq);
> >         }
> > 
> > So, we're looking to see not only if a handler returned IRQ_HANDLED,
> > but also if a handler returned _some other value_ other than IRQ_HANDLED
> > or IRQ_NONE.
> 
> But obviously any other value is invalid as shown above, so we
> shouldn't take it seriously as the correct return value for
> triggering add_random_interrupt().

Point is: you're killing the method for detecting when IRQ handlers
return bogus return codes since you only look at them when they
respond with IRQ_HANDLED.

So, you're disabling the "bogus return value" check.  Completely.

The whole point of the check is to identify bad IRQ handlers so they
can be fixed up.

-- 
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:  2.6 PCMCIA      - http://pcmcia.arm.linux.org.uk/
                 2.6 Serial core
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/