Dear Fabio Estevam,
> On Wed, Apr 4, 2012 at 9:48 PM, Marek Vasut wrote:
> > I have some obscure card now and I'm seeing breakage (DMA transfer
> > timeout).
>
> Can you post more details, please? How do you reproduce the timeouts
> and where exactly do they occur?
Yes, can you remind me on apr
On Wed, Apr 4, 2012 at 9:48 PM, Marek Vasut wrote:
>
> I have some obscure card now and I'm seeing breakage (DMA transfer timeout).
Can you post more details, please? How do you reproduce the timeouts
and where exactly do they occur?
___
U-Boot mailing
Dear Fabio Estevam,
> On Wed, Apr 4, 2012 at 9:48 PM, Marek Vasut wrote:
> >> Someone may want to enable cache, but still uses MMC driver in PIO mode.
> >
> > You don't, it's slow as crap then.
>
> Well, this is how we have been using the MMC driver for the last 4 months.
With DCache disabled
On Wed, Apr 4, 2012 at 9:48 PM, Marek Vasut wrote:
>> Someone may want to enable cache, but still uses MMC driver in PIO mode.
>
> You don't, it's slow as crap then.
Well, this is how we have been using the MMC driver for the last 4 months.
I like the fact that your patch let PIO and DMA mode c
Dear Fabio Estevam,
> On Wed, Apr 4, 2012 at 6:51 PM, Marek Vasut wrote:
> > +/* Enable DMA transfers only if DCache is on. */
> > +#ifndef CONFIG_SYS_DCACHE_OFF
> > +#define CONFIG_MXS_MMC_DMA
> > +#endif
>
> Your patch makes sense, but I would prefer not to couple
> CONFIG_MXS_MMC_DMA with DCA
On Wed, Apr 4, 2012 at 6:51 PM, Marek Vasut wrote:
> +/* Enable DMA transfers only if DCache is on. */
> +#ifndef CONFIG_SYS_DCACHE_OFF
> +#define CONFIG_MXS_MMC_DMA
> +#endif
Your patch makes sense, but I would prefer not to couple
CONFIG_MXS_MMC_DMA with DCACHE support.
Someone may want to en
6 matches
Mail list logo
