[PATCH 1/3] U6/U6715 ARM architecture files

[Philippe Langlais]

Hi

On 07/22/10 10:40, Russell King - ARM Linux wrote:
> On Mon, Jul 19, 2010 at 11:24:53AM +0200, Philippe Langlais wrote:
>    
>> +/*
>> + * Macros to group U6xxx into cpu classes.
>> + * These can be used in most places.
>> + * cpu_is_u67xx():	True for U67XX
>> + */
>> +
>> +#define GET_U6XXX_CLASS		((inl(SCON_SYSVER_REG)&0x0F00)>>8)
>> +#define GET_U6XXX_SUBCLASS	((inl(SCON_SYSVER_REG)&0x0FF0)>>4)
>>      
> Bah.  Still finding ISA/PCI IO macros in your source.  Can you review
> your own patches and eliminate all {in,out}s?[bwl] calls please?
> (that's inb, inw, inl, insb, insw, insl, and the out variants too...)
>
>    
OK, I beg your pardon, these are the last occurrences.
>> +static void u6_clockevent_init_mmtu(void)
>> +{
>> +	printk(PKMOD "clockevent_init_mmtu\n");
>> +
>> +	/* prescale 13Mhz ->  1Mhz */
>> +#ifndef U6_MMTU_CLOCK_SOURCE
>> +	u6_mmtu_timer_init(0, 0, (MMTU_SYS_FRQ / MMTU_ROOT_FRQ), 0);
>> +#endif
>> +
>> +/* issue it is shorter than reality and generates spurious irq */
>> +/*      clockevent_mmtu.mult = div_sc(MMTU_ROOT_FRQ, NSEC_PER_SEC,
>> + *      clockevent_mmtu.shift) + 1;*/
>>      
> Shouldn't these comments be indented?
>
>    
>> +	clockevent_mmtu.mult =
>> +	    div_sc(MMTU_ROOT_FRQ, NSEC_PER_SEC, clockevent_mmtu.shift);
>> +
>> +/*      clockevent_mmtu.max_delta_ns = div_sc(RELOAD_COUNTER_MMTU,
>> + *      clockevent_mmtu.mult, clockevent_mmtu.shift);*/
>> +/*  In fact it is wider than the 32bits variable !!! */
>>      
> Ditto.
>
>    
>> +	clockevent_mmtu.max_delta_ns = 0xFFFFFFFF;
>> +
>> +/*  MMTU HW limitation: match register can't be set w/ tcval+1 */
>> +/*      clockevent_mmtu.min_delta_ns = div_sc(1, clockevent_mmtu.mult,
>> + *      clockevent_mmtu.shift)+1;*/
>>      
> Ditto.
>
>    
OK
I attach the patch with all fixes.
The last iteration ?

Regards