[PATCH v5 2/2] OMAP:IOMMU:flush L1 and L2 caches

Gupta, Ramesh grgupta at ti.com
Thu Sep 13 00:34:09 EDT 2012


Hi Russell,


On Wed, Sep 12, 2012 at 10:11 PM, Russell King - ARM Linux
<linux at arm.linux.org.uk> wrote:
> On Wed, Sep 12, 2012 at 08:49:16PM +0530, Gupta, Ramesh wrote:
>> Thanks to the RMK's suggestions.
>
> I should've made clear the distinction between _range and _area.
> A _range function takes start and end.  An _area function takes a start
> and size.  So...
>
>> -static void flush_iopgd_range(u32 *first, u32 *last)
>> +static void flush_iopgd_range(u32 *first, size_t size)
>
> This should change to flush_iopgd_area().

Looks like I missed this, I will fix and send the updated patch.

>
>>  {
>> -     /* FIXME: L2 cache should be taken care of if it exists */
>> -     do {
>> -             asm("mcr        p15, 0, %0, c7, c10, 1 @ flush_pgd"
>> -                 : : "r" (first));
>> -             first += L1_CACHE_BYTES / sizeof(*first);
>> -     } while (first <= last);
>> +     phys_addr_t phys = virt_to_phys(first);
>> +
>> +     iommu_flush_area(first, size);
>> +     outer_flush_range(phys, phys + size);
>>  }
>>
>> -static void flush_iopte_range(u32 *first, u32 *last)
>> +static void flush_iopte_range(u32 *first, size_t size)
>> +
>
> flush_iopte_area().  (And there shouldn't be a blank line between this
> and the open curley.)

I will fix it.

> Otherwise, looks fine.

thank you.


Best regards
Ramesh Gupta G



More information about the linux-arm-kernel mailing list