| Message ID | CAOi56cXeRjGgE+=psk7K45oeAgZ27HFxbi+an-ELnA3-_Z_BTA@mail.gmail.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Wed, Jun 1, 2016 at 9:57 PM, Kevin Hilman <khilman@kernel.org> wrote: > Hi Krzysztof, > > On Sat, May 28, 2016 at 2:54 AM, Krzysztof Kozlowski <krzk@kernel.org> wrote: >> Just like other Exynos5 family SoCs, this one has four UARTs. Configure >> clocks for UART3 and enable it. >> >> Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org> >> Reviewed-by: Javier Martinez Canillas <javier@osg.samsung.com> > > This patch landed in linux-next[1] and kernelci.org found a boot > regression on the exynos5410-odroidxu[2], where it looks like it's > faulting in clk_prepare(). Bisecting pointed to this patch. I added > "status = disabled" back into the DTS on top of next-20160601[3] and > verified my odroid-XU to be booting fine again. Hi, Thanks for report. Indeed the patch brings up to the light an issue in serial driver. Fixed here: https://lkml.org/lkml/2016/5/31/570 I left the commit in my tree as it is correct (strictly speaking). Sorry for any trouble it makes. Best regards, Krzysztof
Hello Krzysztof, On 06/01/2016 04:06 PM, Krzysztof Kozlowski wrote: > On Wed, Jun 1, 2016 at 9:57 PM, Kevin Hilman <khilman@kernel.org> wrote: >> Hi Krzysztof, >> >> On Sat, May 28, 2016 at 2:54 AM, Krzysztof Kozlowski <krzk@kernel.org> wrote: >>> Just like other Exynos5 family SoCs, this one has four UARTs. Configure >>> clocks for UART3 and enable it. >>> >>> Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org> >>> Reviewed-by: Javier Martinez Canillas <javier@osg.samsung.com> >> >> This patch landed in linux-next[1] and kernelci.org found a boot >> regression on the exynos5410-odroidxu[2], where it looks like it's >> faulting in clk_prepare(). Bisecting pointed to this patch. I added >> "status = disabled" back into the DTS on top of next-20160601[3] and >> verified my odroid-XU to be booting fine again. > > Hi, > > Thanks for report. Indeed the patch brings up to the light an issue in > serial driver. Fixed here: > https://lkml.org/lkml/2016/5/31/570 > > I left the commit in my tree as it is correct (strictly speaking). > Sorry for any trouble it makes. > Kevin pointed me to this issue over IRC and I was able to reproduce it in my Peach Pi by removing the UART3 clocks from the Exynos5420 clock driver. Your patch indeed fixes the issue, I'll add a Reviewed-by on that thread. > Best regards, > Krzysztof > Best regards,
On 06/01/2016 04:52 PM, Javier Martinez Canillas wrote: > Hello Krzysztof, [snip] >> >> I left the commit in my tree as it is correct (strictly speaking). >> Sorry for any trouble it makes. >> > > Kevin pointed me to this issue over IRC and I was able to reproduce it in > my Peach Pi by removing the UART3 clocks from the Exynos5420 clock driver. > > Your patch indeed fixes the issue, I'll add a Reviewed-by on that thread. > err, I meant a Tested-by tag... but yes I've also reviewed the patch :) Best regards,
On 06/01/2016 10:54 PM, Javier Martinez Canillas wrote: > On 06/01/2016 04:52 PM, Javier Martinez Canillas wrote: >> Hello Krzysztof, > > [snip] > >>> >>> I left the commit in my tree as it is correct (strictly speaking). >>> Sorry for any trouble it makes. >>> >> >> Kevin pointed me to this issue over IRC and I was able to reproduce it in >> my Peach Pi by removing the UART3 clocks from the Exynos5420 clock driver. >> >> Your patch indeed fixes the issue, I'll add a Reviewed-by on that thread. >> > > err, I meant a Tested-by tag... but yes I've also reviewed the patch :) Sure, thanks! Best regards, Krzysztof
--- a/arch/arm/boot/dts/exynos5410.dtsi +++ b/arch/arm/boot/dts/exynos5410.dtsi @@ -238,6 +238,7 @@ }; &serial_3 { + status = "disabled"; clocks = <&clock CLK_UART3>, <&clock CLK_SCLK_UART3>; clock-names = "uart", "clk_uart_baud0"; };