Dirk Behme wrote:
> Tony Lindgren wrote:
>
>>> diff -Nru linux-2.6.14-omap2/arch/arm/plat-omap/gpio.c
>>> linux-2.6.14-omap2-palmtt2/arch/arm/plat-omap/gpio.c 
>>> --- linux-2.6.14-omap2/arch/arm/plat-omap/gpio.c        Sat Dec 10 
>>> 21:05:46 2005      +++ 
>>> linux-2.6.14-omap2-palmtt2/arch/arm/plat-omap/gpio.c        Sat Dec 
>>> 10 19:57:30 2005      @@ -758,6 +758,10 @@
>>>    while(1) {            isr = __raw_readl(isr_reg);
>>> +#ifdef CONFIG_ARCH_OMAP15XX   +       if (bank->method == 
>>> METHOD_MPUIO)
>>> +           isr&=0xffff;   +#endif !
>>>
>>> _enable_gpio_irqbank(bank, isr, 0);
>>> _clear_gpio_irqbank(bank, isr);            
>>> _enable_gpio_irqbank(bank, isr, 1);
>>
>>
>> Can't read this patch because of the bad formatting... Can you please
>> repost? Maybe sending it as an attachment works better with a web
>> interface :)
>
> arch/arm/plat-omap/gpio.c has changed since 2.6.14-omap2, so patch above
> will not apply to recent git even if formatting is correct. I tried to
> adapt it to current version of gpio.c. I can't test it, so is patch
> below okay?
>
>
> ------------------------------------------------------------------------
>
> --- ./arch/arm/plat-omap/gpio.c_orig	2006-01-06 19:37:32.602154520 +0100
> +++ ./arch/arm/plat-omap/gpio.c	2006-01-06 19:44:03.917665520 +0100
> @@ -767,6 +767,9 @@ static void gpio_irq_handler(unsigned in
>  					OMAP24XX_GPIO_LEVELDETECT0) |
>  				__raw_readl(bank->base +
>  					OMAP24XX_GPIO_LEVELDETECT1);
> +		if(cpu_is_omap15xx())
> +			if(bank->method == METHOD_MPUIO)
> +				level_mask = 0xffff0000; 
>  
>  		/* clear edge sensitive interrupts before handler(s) are
>  		called so that we don't miss any interrupt occurred while
>
>
>
>   
Patch to fixed MPUIO interrupt.
Tested on linux-2.6.15-omap1.