[PATCH v4 05/13] pm: at91: move the copying the sram function to the sram initializationi phase

[Alexandre Belloni]

Hi,

On 29/01/2015 at 11:28:00 +0000, Russell King - ARM Linux wrote :
> On Wed, Jan 28, 2015 at 09:43:16AM +0800, Wenyou Yang wrote:
> > -#ifdef CONFIG_AT91_SLOW_CLOCK
> > -				/* copy slow_clock handler to SRAM, and call it */
> > -				memcpy(slow_clock, at91_slow_clock, at91_slow_clock_sz);
> > -#endif
> >  				slow_clock(at91_pmc_base, at91_ramc_base[0],
> >  					   at91_ramc_base[1],
> >  					   at91_pm_data.memctrl);
> > @@ -272,6 +268,9 @@ static void __init at91_pm_sram_init(void)
> >  	sram_pbase = gen_pool_virt_to_phys(sram_pool, sram_base);
> >  	slow_clock = __arm_ioremap_exec(sram_pbase, at91_slow_clock_sz, false);
> >  
> > +	/* Copy the slow_clock handler to SRAM */
> > +	memcpy(slow_clock, at91_slow_clock, at91_slow_clock_sz);
> > +
> 
> Why is this code not using the fncpy() support for copying functions.

Indeed, this was done in the original version of the patch that I acked.

> Why is it not checking the return code from __arm_ioremap_exec() or
> gen_pool_virt_to_phys() for failure?

gen_pool_virt_to_phys() will not fail as the chunk is allocated just
before so it will necessarily be found in the list.

We need to reintroduce a check for slow_clock != NULL before fncpy()
since it is moved out of its original if block.

-- 
Alexandre Belloni, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com