[PATCH] clk: meson: axg: mark fdiv2 and fdiv3 as critical

[PATCH] clk: meson: axg: mark fdiv2 and fdiv3 as critical

[Jerome Brunet]

Similar to gxbb and gxl platforms, axg SCPI Cortex-M co-processor
uses the fdiv2 and fdiv3 to, among other things, provide the cpu
clock.

Until clock hand-off mechanism makes its way to CCF and the generic
SCPI claims platform specific clocks, these clocks must be marked as
critical to make sure they are never disabled when needed by the
co-processor.

Fixes: 05f814402d61 ("clk: meson: add fdiv clock gates")
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
---

Hi Stephen,

If you can put this one in clk-fixes as well, it would be awesome.
It is basically the same thing as the change you took this Tuesday.

Since then, we had reports the same problem with SCPI was happening
on the axg, calling for the same fixup.

Cheers
Jerome

 drivers/clk/meson/axg.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)

diff --git a/drivers/clk/meson/axg.c b/drivers/clk/meson/axg.c
index c981159b02c0..792735d7e46e 100644
--- a/drivers/clk/meson/axg.c
+++ b/drivers/clk/meson/axg.c
@@ -325,6 +325,7 @@ static struct clk_regmap axg_fclk_div2 = {
 		.ops = &clk_regmap_gate_ops,
 		.parent_names = (const char *[]){ "fclk_div2_div" },
 		.num_parents = 1,
+		.flags = CLK_IS_CRITICAL,
 	},
 };
 
@@ -349,6 +350,18 @@ static struct clk_regmap axg_fclk_div3 = {
 		.ops = &clk_regmap_gate_ops,
 		.parent_names = (const char *[]){ "fclk_div3_div" },
 		.num_parents = 1,
+		/*
+		 * FIXME:
+		 * This clock, as fdiv2, is used by the SCPI FW and is required
+		 * by the platform to operate correctly.
+		 * Until the following condition are met, we need this clock to
+		 * be marked as critical:
+		 * a) The SCPI generic driver claims and enable all the clocks
+		 *    it needs
+		 * b) CCF has a clock hand-off mechanism to make the sure the
+		 *    clock stays on until the proper driver comes along
+		 */
+		.flags = CLK_IS_CRITICAL,
 	},
 };
 
-- 
2.19.1

Re: [PATCH] clk: meson: axg: mark fdiv2 and fdiv3 as critical

[Neil Armstrong]

On 08/11/2018 10:31, Jerome Brunet wrote:
> Similar to gxbb and gxl platforms, axg SCPI Cortex-M co-processor
> uses the fdiv2 and fdiv3 to, among other things, provide the cpu
> clock.
> 
> Until clock hand-off mechanism makes its way to CCF and the generic
> SCPI claims platform specific clocks, these clocks must be marked as
> critical to make sure they are never disabled when needed by the
> co-processor.
> 
> Fixes: 05f814402d61 ("clk: meson: add fdiv clock gates")
> Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
> ---
> 
> Hi Stephen,
> 
> If you can put this one in clk-fixes as well, it would be awesome.
> It is basically the same thing as the change you took this Tuesday.
> 
> Since then, we had reports the same problem with SCPI was happening
> on the axg, calling for the same fixup.
> 
> Cheers
> Jerome
> 
>  drivers/clk/meson/axg.c | 13 +++++++++++++
>  1 file changed, 13 insertions(+)
> 
> diff --git a/drivers/clk/meson/axg.c b/drivers/clk/meson/axg.c
> index c981159b02c0..792735d7e46e 100644
> --- a/drivers/clk/meson/axg.c
> +++ b/drivers/clk/meson/axg.c
> @@ -325,6 +325,7 @@ static struct clk_regmap axg_fclk_div2 = {
>  		.ops = &clk_regmap_gate_ops,
>  		.parent_names = (const char *[]){ "fclk_div2_div" },
>  		.num_parents = 1,
> +		.flags = CLK_IS_CRITICAL,
>  	},
>  };
>  
> @@ -349,6 +350,18 @@ static struct clk_regmap axg_fclk_div3 = {
>  		.ops = &clk_regmap_gate_ops,
>  		.parent_names = (const char *[]){ "fclk_div3_div" },
>  		.num_parents = 1,
> +		/*
> +		 * FIXME:
> +		 * This clock, as fdiv2, is used by the SCPI FW and is required
> +		 * by the platform to operate correctly.
> +		 * Until the following condition are met, we need this clock to
> +		 * be marked as critical:
> +		 * a) The SCPI generic driver claims and enable all the clocks
> +		 *    it needs
> +		 * b) CCF has a clock hand-off mechanism to make the sure the
> +		 *    clock stays on until the proper driver comes along
> +		 */
> +		.flags = CLK_IS_CRITICAL,
>  	},
>  };
>  
> 

Acked-by: Neil Armstrong <narmstrong@baylibre.com>

Re: [PATCH] clk: meson: axg: mark fdiv2 and fdiv3 as critical

[Stephen Boyd]

Quoting Jerome Brunet (2018-11-08 01:31:23)
> Similar to gxbb and gxl platforms, axg SCPI Cortex-M co-processor
> uses the fdiv2 and fdiv3 to, among other things, provide the cpu
> clock.
> 
> Until clock hand-off mechanism makes its way to CCF and the generic
> SCPI claims platform specific clocks, these clocks must be marked as
> critical to make sure they are never disabled when needed by the
> co-processor.
> 
> Fixes: 05f814402d61 ("clk: meson: add fdiv clock gates")
> Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
> ---
> 
> Hi Stephen,
> 
> If you can put this one in clk-fixes as well, it would be awesome.
> It is basically the same thing as the change you took this Tuesday.
> 
> Since then, we had reports the same problem with SCPI was happening
> on the axg, calling for the same fixup.
> 

Ok. I suppose someone needs to work on clk handoff................ me?

Re: [PATCH] clk: meson: axg: mark fdiv2 and fdiv3 as critical

[Neil Armstrong]

Hi Stable team,

Le 08/11/2018 10:31, Jerome Brunet a écrit :
> Similar to gxbb and gxl platforms, axg SCPI Cortex-M co-processor
> uses the fdiv2 and fdiv3 to, among other things, provide the cpu
> clock.
> 
> Until clock hand-off mechanism makes its way to CCF and the generic
> SCPI claims platform specific clocks, these clocks must be marked as
> critical to make sure they are never disabled when needed by the
> co-processor.
> 
> Fixes: 05f814402d61 ("clk: meson: add fdiv clock gates")
> Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>

Could this fix go into the next 4.18 and 4.19 stable releases since it hit linus master with commit id d6ee1e7e9004d3d246cdfa14196989e0a9466c16 ?

Thanks,
Neil


> ---
> 
> Hi Stephen,
> 
> If you can put this one in clk-fixes as well, it would be awesome.
> It is basically the same thing as the change you took this Tuesday.
> 
> Since then, we had reports the same problem with SCPI was happening
> on the axg, calling for the same fixup.
> 
> Cheers
> Jerome
> 
>  drivers/clk/meson/axg.c | 13 +++++++++++++
>  1 file changed, 13 insertions(+)
> 
> diff --git a/drivers/clk/meson/axg.c b/drivers/clk/meson/axg.c
> index c981159b02c0..792735d7e46e 100644
> --- a/drivers/clk/meson/axg.c
> +++ b/drivers/clk/meson/axg.c
> @@ -325,6 +325,7 @@ static struct clk_regmap axg_fclk_div2 = {
>  		.ops = &clk_regmap_gate_ops,
>  		.parent_names = (const char *[]){ "fclk_div2_div" },
>  		.num_parents = 1,
> +		.flags = CLK_IS_CRITICAL,
>  	},
>  };
>  
> @@ -349,6 +350,18 @@ static struct clk_regmap axg_fclk_div3 = {
>  		.ops = &clk_regmap_gate_ops,
>  		.parent_names = (const char *[]){ "fclk_div3_div" },
>  		.num_parents = 1,
> +		/*
> +		 * FIXME:
> +		 * This clock, as fdiv2, is used by the SCPI FW and is required
> +		 * by the platform to operate correctly.
> +		 * Until the following condition are met, we need this clock to
> +		 * be marked as critical:
> +		 * a) The SCPI generic driver claims and enable all the clocks
> +		 *    it needs
> +		 * b) CCF has a clock hand-off mechanism to make the sure the
> +		 *    clock stays on until the proper driver comes along
> +		 */
> +		.flags = CLK_IS_CRITICAL,
>  	},
>  };
>  
> 

Re: [PATCH] clk: meson: axg: mark fdiv2 and fdiv3 as critical

[Sasha Levin]

On Tue, Nov 13, 2018 at 05:43:25PM +0100, Neil Armstrong wrote:
>Hi Stable team,
>
>Le 08/11/2018 10:31, Jerome Brunet a écrit :
>> Similar to gxbb and gxl platforms, axg SCPI Cortex-M co-processor
>> uses the fdiv2 and fdiv3 to, among other things, provide the cpu
>> clock.
>>
>> Until clock hand-off mechanism makes its way to CCF and the generic
>> SCPI claims platform specific clocks, these clocks must be marked as
>> critical to make sure they are never disabled when needed by the
>> co-processor.
>>
>> Fixes: 05f814402d61 ("clk: meson: add fdiv clock gates")
>> Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
>
>Could this fix go into the next 4.18 and 4.19 stable releases since it hit linus master with commit id d6ee1e7e9004d3d246cdfa14196989e0a9466c16 ?

Queued for 4.19 and 4.18, thank you.

--
Thanks,
Sasha