Re: lockdep and threaded IRQs (was: ...)

[David Brownell]

On Monday 02 March 2009, Ingo Molnar wrote:
> > 
> > p.s. Note that those changes would still leave the
> >      lockdep bug around ... it will still be breaking
> >      various drivers that use normal IRQs, by forcibly
> >      enabling IRQF_DISABLED.
> 
> it's not a bug - and i think Peter explained that already. 

No.  But I did get a non-response that didn't include any
explanation, and relied totally on unfounded assertions
combined with the presumption that someday IRQF_DISABLED
will be forced on in all drivers.


> It's  
> not really breaking things either - we've had this for more than 
> 2 years.

I happened across two MMC host adapter drivers which it
broke.  They preceded the patch whereby lockdep changes
request_irq() semantics.  The drivers break only with
lockdep ... so platforms relying on those can't use lockdep.
No reason to think there aren't more such bugs lurking.

We also wasted quite a few months tracking down USB bugs
that were masked by that patch ... the indeterminacy of
IRQF_DISABLED behavior given IRQF_SHARED was hidden by
enabling lockdep.  Led to a lot of totally crapulous oops
traces, I assure you.

So ... it has most certainly broken things.  There has
seemed to be a bit of stick-fingers-in-ears going on, that
prevents hearing about such problems though.

- Dave



--
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/