Re: [PATCH] ARM: dts: add rk3288-firefly-reload

[Heiko Stuebner]

Hi Shawn,

Am Sonntag, 7. August 2016, 11:40:21 schrieb Shawn Lin:
> On 2016/7/19 3:46, Heiko Stübner wrote:
> > Am Montag, 18. Juli 2016, 23:32:32 schrieb Randy Li:
> >> The Firefly RK3288 Reload is a combination Firefly rk3288 core board
> >> with the Reload baseboard. Add a dtsi for the Firefly rk3288 core
> >> which can be included into the dts for the various baseboards
> >> in the future and dts for Reload base board.
> >> 
> >> Currently supported are serial console, wired networking, eMMC and
> >> SD storage, SPFIF, IR receiver, LEDs, SDIO wifi and USB. But only
> >> the OTG could work on the host mode now, the other USB host can't
> >> work now, additional patches are required.
> >> 
> >> Signed-off-by: Randy Li <ayaka@xxxxxxxxxxx>
> >> Acked-by: Rob Herring <robh@xxxxxxxxxx>
> > 
> > applied [0], after some minor reordering - please double check.
> > 
> > Also, while the branch is named 4.8 this will only go into 4.9, as the
> > merge- window supposedly opens on the weekend.
> 
> Sorry for the noise here:), but I find there is some mmc stuff
> which confuse me. Maybe Randy could elaborate more?
> 
> Your sdmmc claims to support UHS-I mode like sd-uhs-sdr104, etc, which
> will ask dw_mmc to switch vqmmc from 3V3 to 1V8 after sending CMD11.
> 
> But the vqmmc is vccio_sd which is fixed as 3V3? :)

so far we have only been able to achieve UHS speeds on devices based on the 
rk808 pmic (I had sucess on veyron and popmetal). The act8846-based boards 
seem to have some issue when trying to do the switch to 1.8V, resulting in 
mmc errors and the mmc-core switching back to 3.3V.

I think this might be caused by a bug in the act8846 driver or some dts 
mistake. But so far nobody had the time and enthusiasm to investigate :-)


Heiko