Hi Simon,
On Tue, Apr 26, 2016 at 5:52 AM, Simon Horman <horms@xxxxxxxxxxxx> wrote:
> On Mon, Apr 25, 2016 at 04:08:29PM +0200, Geert Uytterhoeven wrote:
>> clk_get() on a disabled clock node will return -EPROBE_DEFER, which can
>> cause drivers to be deferred forever if such clocks are referenced in
>> their devices' clocks properties. Disabled clock nodes show this
>> behavior as of commit 3e5dd6f6e690048d ("clk: Ignore disabled DT clock
>> providers"), which is in v4.6-rc1.
>>
>> Commit ac6908b3049397b1 ("ARM: dts: r8a7791: Don't disable referenced
>> optional clocks") fixed this for the Porter development board, which
>> doesn't have the SCIF_CLK crystal populated. This patch series fixes
>> the same problem for the remaining Renesas DTSes, by updating disabled
>> external clock nodes to default to a frequency of 0, but not disabling
>> them.
>>
>> None of these should show up with in-tree supported boards, but Jürg
>> Billeter <j@xxxxxxxxx> reported seeing the problem with v4.6-rc5 on a
>> custom R-Car H3 board. Hence you may want to queue at least the arm64
>> fix for v4.6.
> I have tentatively done the following:
>
> * Queued up the portion of the arm64 patch that is relevant to v4.6
> (i.e. changes DT nodes that exist in v4.6) as a fix for v4.6.
> * Queued up the rest for v4.7
Thanks!
> Please check that what is in renesas-next-20160426-v4.6-rc1 looks sensible.
Looks all good to me...
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
