On Wed, Oct 22, 2014 at 2:33 AM, Qais Yousef <qais.yousef@xxxxxxxxxx> wrote:
> On 10/20/2014 08:03 PM, Andrew Bresticker wrote:
>>
>> Stop using the REG macros from gic.h and instead use proper iomem
>> accessors.
>>
>> Signed-off-by: Andrew Bresticker <abrestic@xxxxxxxxxxxx>
>> ---
>> arch/mips/mti-sead3/sead3-int.c | 7 +++----
>> 1 file changed, 3 insertions(+), 4 deletions(-)
>>
>> diff --git a/arch/mips/mti-sead3/sead3-int.c
>> b/arch/mips/mti-sead3/sead3-int.c
>> index 69ae185..995c401 100644
>> --- a/arch/mips/mti-sead3/sead3-int.c
>> +++ b/arch/mips/mti-sead3/sead3-int.c
>> @@ -20,16 +20,15 @@
>> #define SEAD_CONFIG_BASE 0x1b100110
>> #define SEAD_CONFIG_SIZE 4
>> -static unsigned long sead3_config_reg;
>> +static void __iomem *sead3_config_reg;
>> void __init arch_init_irq(void)
>> {
>> if (!cpu_has_veic)
>> mips_cpu_irq_init();
>> - sead3_config_reg = (unsigned
>> long)ioremap_nocache(SEAD_CONFIG_BASE,
>> - SEAD_CONFIG_SIZE);
>> - gic_present = (REG32(sead3_config_reg) &
>> SEAD_CONFIG_GIC_PRESENT_MSK) >>
>> + sead3_config_reg = ioremap_nocache(SEAD_CONFIG_BASE,
>> SEAD_CONFIG_SIZE);
>> + gic_present = (readl(sead3_config_reg) &
>> SEAD_CONFIG_GIC_PRESENT_MSK) >>
>> SEAD_CONFIG_GIC_PRESENT_SHF;
>> pr_info("GIC: %spresent\n", (gic_present) ? "" : "not ");
>> pr_info("EIC: %s\n",
>
>
> I think you need to use the __raw_readl() variant here and for all other
> similar changes.
Thanks, will do.
-Andrew
