Re: [PATCH V2 RESEND] clk: mxs: Fix invalid 32-bit access to frac registers

[Mike Turquette]

Quoting Marek Vasut (2015-02-10 14:07:51)
> On Tuesday, February 10, 2015 at 10:54:52 PM, Fabio Estevam wrote:
> > Hi Stefan,
> > 
> > On Tue, Feb 10, 2015 at 7:24 PM, Stefan Wahren <stefan.wahren@xxxxxxxx> wrote:
> > > thanks this is very helpful. I built the linux-next for my Duckbill and
> > > add the SSP2 section from imx28-evk.dts into imx28-duckbill.dts.
> > > 
> > > Without my patch i get the following for HW_CLKCTRL_FRAC0:
> > > 
> > > ./memwatch -a 0x800401B0
> > > 
> > > 0x800401b0: 0x5e5b5513
> > > 
> > > With my patch i get:
> > > 
> > > ./memwatch -a 0x800401B0
> > > 
> > > 0x800401b0: 0x5e1b5513
> > > 
> > > So it looks like a problem in my patch.
> > 
> > Yes, let's try to get the same value HW_CLKCTRL_FRAC0.
> > 
> > > I orded such a flash chip, but it will take some time until i can use it
> > > with my hardware.
> > 
> > Thanks for doing this.
> > 
> > Maybe if you find out a way to fix the calculation of
> > HW_CLKCTRL_FRAC0, then I can test it on my board.
> 
> Hi,
> 
> the difference is this 0x40 bit, right ? That's _STABLE bit, so it means
> the clock are not stable, right ? Why would that happen in the first place?

Can you dump all the registers related to all of the clocks provided by
your clock driver, both with and without your patch? Then just diff
them. Might be more deltas than just this one register.

Regards,
Mike

> 
> Best regards,
> Marek Vasut
--
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/