Re: [PATCH] mfd: ti_am335x_tscadc: fix spin lock and reg_cache

From: Lee Jones
Date: Tue Oct 22 2013 - 13:06:31 EST


On Tue, 22 Oct 2013, Sebastian Andrzej Siewior wrote:

> On 10/22/2013 06:05 PM, Lee Jones wrote:
> >> I added reg_se_cache to cache the content of REG_SE once and
> >> synchronize it among TSC & ADC access. REG_SE is set to 0 by the HW
> >> after "work" has been done. So you need to know the old value or TSC may
> >> disable ADC and the other way around.
> >
> > Yep, it's initialised as '0'.
> >
> > 12.5.1.15 STEPENABLE Register (offset = 54h) [reset = 0h]
>
> Ehm yes but!. After init it is set to 0, correct. The value was never
> read from the HW. It was always set via am335x_tsc_se_set() to "cache
> | argument" and written to HW from both sides (TSC, ADC). This
> initialization is done at ->probe() time in both drivers.
>
> The value remains (remained) constant over the whole time
> so both drivers only called am335x_tsc_se_update() to set the value
> (the enabled steps of both sides) back to the register (because after
> the conversation the value was 0 according to my memory) and since
> 32bit reads are atomic I didn't use a lock here.

Hmm.. I'm starting to see what you mean.

So what's the point of the read before write then?

Why don't you use the cache all of the time?

--
Lee Jones
Linaro STMicroelectronics Landing Team Lead
Linaro.org â Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/