* [PATCH AUTOSEL 5.11 26/52] clk: qcom: gdsc: Implement NO_RET_PERIPH flag
[not found] <20210302115534.61800-1-sashal@kernel.org>
@ 2021-03-02 11:55 ` Sasha Levin
2021-03-02 23:02 ` Stephen Boyd
0 siblings, 1 reply; 3+ messages in thread
From: Sasha Levin @ 2021-03-02 11:55 UTC (permalink / raw)
To: linux-kernel, stable
Cc: AngeloGioacchino Del Regno, Stephen Boyd, Sasha Levin, linux-clk
From: AngeloGioacchino Del Regno <angelogioacchino.delregno@somainline.org>
[ Upstream commit 785c02eb35009a4be6dbc68f4f7d916e90b7177d ]
In some rare occasions, we want to only set the RETAIN_MEM bit, but
not the RETAIN_PERIPH one: this is seen on at least SDM630/636/660's
GPU-GX GDSC, where unsetting and setting back the RETAIN_PERIPH bit
will generate chaos and panics during GPU suspend time (mainly, the
chaos is unaligned access).
For this reason, introduce a new NO_RET_PERIPH flag to the GDSC
driver to address this corner case.
Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@somainline.org>
Link: https://lore.kernel.org/r/20210113183817.447866-8-angelogioacchino.delregno@somainline.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
drivers/clk/qcom/gdsc.c | 10 ++++++++--
drivers/clk/qcom/gdsc.h | 3 ++-
2 files changed, 10 insertions(+), 3 deletions(-)
diff --git a/drivers/clk/qcom/gdsc.c b/drivers/clk/qcom/gdsc.c
index af26e0695b86..51ed640e527b 100644
--- a/drivers/clk/qcom/gdsc.c
+++ b/drivers/clk/qcom/gdsc.c
@@ -183,7 +183,10 @@ static inline int gdsc_assert_reset(struct gdsc *sc)
static inline void gdsc_force_mem_on(struct gdsc *sc)
{
int i;
- u32 mask = RETAIN_MEM | RETAIN_PERIPH;
+ u32 mask = RETAIN_MEM;
+
+ if (!(sc->flags & NO_RET_PERIPH))
+ mask |= RETAIN_PERIPH;
for (i = 0; i < sc->cxc_count; i++)
regmap_update_bits(sc->regmap, sc->cxcs[i], mask, mask);
@@ -192,7 +195,10 @@ static inline void gdsc_force_mem_on(struct gdsc *sc)
static inline void gdsc_clear_mem_on(struct gdsc *sc)
{
int i;
- u32 mask = RETAIN_MEM | RETAIN_PERIPH;
+ u32 mask = RETAIN_MEM;
+
+ if (!(sc->flags & NO_RET_PERIPH))
+ mask |= RETAIN_PERIPH;
for (i = 0; i < sc->cxc_count; i++)
regmap_update_bits(sc->regmap, sc->cxcs[i], mask, 0);
diff --git a/drivers/clk/qcom/gdsc.h b/drivers/clk/qcom/gdsc.h
index bd537438c793..5bb396b344d1 100644
--- a/drivers/clk/qcom/gdsc.h
+++ b/drivers/clk/qcom/gdsc.h
@@ -42,7 +42,7 @@ struct gdsc {
#define PWRSTS_ON BIT(2)
#define PWRSTS_OFF_ON (PWRSTS_OFF | PWRSTS_ON)
#define PWRSTS_RET_ON (PWRSTS_RET | PWRSTS_ON)
- const u8 flags;
+ const u16 flags;
#define VOTABLE BIT(0)
#define CLAMP_IO BIT(1)
#define HW_CTRL BIT(2)
@@ -51,6 +51,7 @@ struct gdsc {
#define POLL_CFG_GDSCR BIT(5)
#define ALWAYS_ON BIT(6)
#define RETAIN_FF_ENABLE BIT(7)
+#define NO_RET_PERIPH BIT(8)
struct reset_controller_dev *rcdev;
unsigned int *resets;
unsigned int reset_count;
--
2.30.1
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH AUTOSEL 5.11 26/52] clk: qcom: gdsc: Implement NO_RET_PERIPH flag
2021-03-02 11:55 ` [PATCH AUTOSEL 5.11 26/52] clk: qcom: gdsc: Implement NO_RET_PERIPH flag Sasha Levin
@ 2021-03-02 23:02 ` Stephen Boyd
2021-03-04 21:30 ` Sasha Levin
0 siblings, 1 reply; 3+ messages in thread
From: Stephen Boyd @ 2021-03-02 23:02 UTC (permalink / raw)
To: Sasha Levin, linux-kernel, stable
Cc: AngeloGioacchino Del Regno, Sasha Levin, linux-clk
Quoting Sasha Levin (2021-03-02 03:55:07)
> From: AngeloGioacchino Del Regno <angelogioacchino.delregno@somainline.org>
>
> [ Upstream commit 785c02eb35009a4be6dbc68f4f7d916e90b7177d ]
>
> In some rare occasions, we want to only set the RETAIN_MEM bit, but
> not the RETAIN_PERIPH one: this is seen on at least SDM630/636/660's
> GPU-GX GDSC, where unsetting and setting back the RETAIN_PERIPH bit
> will generate chaos and panics during GPU suspend time (mainly, the
> chaos is unaligned access).
>
> For this reason, introduce a new NO_RET_PERIPH flag to the GDSC
> driver to address this corner case.
Is there a patch that's going to use this in stable trees? On its own
this patch doesn't make sense to backport.
>
> Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@somainline.org>
> Link: https://lore.kernel.org/r/20210113183817.447866-8-angelogioacchino.delregno@somainline.org
> Signed-off-by: Stephen Boyd <sboyd@kernel.org>
> Signed-off-by: Sasha Levin <sashal@kernel.org>
> ---
> drivers/clk/qcom/gdsc.c | 10 ++++++++--
> drivers/clk/qcom/gdsc.h | 3 ++-
> 2 files changed, 10 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/clk/qcom/gdsc.c b/drivers/clk/qcom/gdsc.c
> index af26e0695b86..51ed640e527b 100644
> --- a/drivers/clk/qcom/gdsc.c
> +++ b/drivers/clk/qcom/gdsc.c
> @@ -183,7 +183,10 @@ static inline int gdsc_assert_reset(struct gdsc *sc)
> static inline void gdsc_force_mem_on(struct gdsc *sc)
> {
> int i;
> - u32 mask = RETAIN_MEM | RETAIN_PERIPH;
> + u32 mask = RETAIN_MEM;
> +
> + if (!(sc->flags & NO_RET_PERIPH))
> + mask |= RETAIN_PERIPH;
>
> for (i = 0; i < sc->cxc_count; i++)
> regmap_update_bits(sc->regmap, sc->cxcs[i], mask, mask);
> @@ -192,7 +195,10 @@ static inline void gdsc_force_mem_on(struct gdsc *sc)
> static inline void gdsc_clear_mem_on(struct gdsc *sc)
> {
> int i;
> - u32 mask = RETAIN_MEM | RETAIN_PERIPH;
> + u32 mask = RETAIN_MEM;
> +
> + if (!(sc->flags & NO_RET_PERIPH))
> + mask |= RETAIN_PERIPH;
>
> for (i = 0; i < sc->cxc_count; i++)
> regmap_update_bits(sc->regmap, sc->cxcs[i], mask, 0);
> diff --git a/drivers/clk/qcom/gdsc.h b/drivers/clk/qcom/gdsc.h
> index bd537438c793..5bb396b344d1 100644
> --- a/drivers/clk/qcom/gdsc.h
> +++ b/drivers/clk/qcom/gdsc.h
> @@ -42,7 +42,7 @@ struct gdsc {
> #define PWRSTS_ON BIT(2)
> #define PWRSTS_OFF_ON (PWRSTS_OFF | PWRSTS_ON)
> #define PWRSTS_RET_ON (PWRSTS_RET | PWRSTS_ON)
> - const u8 flags;
> + const u16 flags;
> #define VOTABLE BIT(0)
> #define CLAMP_IO BIT(1)
> #define HW_CTRL BIT(2)
> @@ -51,6 +51,7 @@ struct gdsc {
> #define POLL_CFG_GDSCR BIT(5)
> #define ALWAYS_ON BIT(6)
> #define RETAIN_FF_ENABLE BIT(7)
> +#define NO_RET_PERIPH BIT(8)
> struct reset_controller_dev *rcdev;
> unsigned int *resets;
> unsigned int reset_count;
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH AUTOSEL 5.11 26/52] clk: qcom: gdsc: Implement NO_RET_PERIPH flag
2021-03-02 23:02 ` Stephen Boyd
@ 2021-03-04 21:30 ` Sasha Levin
0 siblings, 0 replies; 3+ messages in thread
From: Sasha Levin @ 2021-03-04 21:30 UTC (permalink / raw)
To: Stephen Boyd; +Cc: linux-kernel, stable, AngeloGioacchino Del Regno, linux-clk
On Tue, Mar 02, 2021 at 03:02:25PM -0800, Stephen Boyd wrote:
>Quoting Sasha Levin (2021-03-02 03:55:07)
>> From: AngeloGioacchino Del Regno <angelogioacchino.delregno@somainline.org>
>>
>> [ Upstream commit 785c02eb35009a4be6dbc68f4f7d916e90b7177d ]
>>
>> In some rare occasions, we want to only set the RETAIN_MEM bit, but
>> not the RETAIN_PERIPH one: this is seen on at least SDM630/636/660's
>> GPU-GX GDSC, where unsetting and setting back the RETAIN_PERIPH bit
>> will generate chaos and panics during GPU suspend time (mainly, the
>> chaos is unaligned access).
>>
>> For this reason, introduce a new NO_RET_PERIPH flag to the GDSC
>> driver to address this corner case.
>
>Is there a patch that's going to use this in stable trees? On its own
>this patch doesn't make sense to backport.
Right, I'll also grab a59c16c80bd7 ("clk: qcom: gpucc-msm8998: Add
resets, cxc, fix flags on gpu_gx_gdsc") for 5.10 and 5.11. Thanks!
--
Thanks,
Sasha
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2021-03-04 21:32 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
[not found] <20210302115534.61800-1-sashal@kernel.org>
2021-03-02 11:55 ` [PATCH AUTOSEL 5.11 26/52] clk: qcom: gdsc: Implement NO_RET_PERIPH flag Sasha Levin
2021-03-02 23:02 ` Stephen Boyd
2021-03-04 21:30 ` Sasha Levin
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox