Re: [PATCH] mmc-atmel : fix kunmap wrong usage

[Marc Pignat <marc.pignat@hevs.ch>]

On Friday 22 June 2007 15:34, Nicolas Ferre wrote:
> Hugh Dickins :
> > Aren't you just guessing there?  Those kunmap_atomics in at91_mci.c
> > may look wrong to you, but they're not incorrect (so long as sg->offset
> > falls within the page, as it must do here to make sense of the page).
> > Especially not on ARM, where kunmap_atomic actually has no interest
> > in the argument passed.  And the oops was in the flush_dcache_page.
> > 
> > If you actually reproduced Nicolas' problem on ARM, and verified
> > that your patch then fixes it, please let us know: that will be
> > remarkably interesting.
> 
> Patch tested without success. Indeed, I always see the Oops with 
> Marc's patch.
So, it's really interesting, it worked really for me (oops without patch) and
no oops with it. My hardware is a custom board with an at91rm9200.

I had a look the the kunmap_atomic function, and I _really_ don't understand
how this patch can do something :-/

And last but not least, my patch is completly wrong...
void *kmap_atomic(...);
unsigned int *buffer = kmap_atomic(...) + sg->offset; // addition in u8*
kunmap_atomic(buffer - sg->offset); // <- sub in u32*

-> please forget my patch

Regards

Marc