[PATCH V2 00/13] soc: imx: gpcv2: support i.MX8MM

[PATCH V2 00/13] soc: imx: gpcv2: support i.MX8MM

[Peng Fan (OSS)]

From: Peng Fan <peng.fan@nxp.com>


V2:
 - Add R-b/A-b tag
 - Merge V1 patch 13 to V2 patch 6
 - Drop V1 patch 15
 - Merge V1 patch 16 to V2 patch 5 and add comments in patch 5 to explain
 details
 - Add explaination in patch 8 for "why the resets are not defined"

This patchset is a pick up Lucas's gpcv2 work for i.MX8MM and several
minor changes from me to make it could work with i.MX BLK-CTL driver.

Thanks for Lucas's work and suggestion, Frieder Schrempf for collecting
all the patches, Jacky Bai on help debug issues.

Lucas Stach (12):
  soc: imx: gpcv2: move to more ideomatic error handling in probe
  soc: imx: gpcv2: move domain mapping to domain driver probe
  soc: imx: gpcv2: switch to clk_bulk_* API
  soc: imx: gpcv2: split power up and power down sequence control
  soc: imx: gpcv2: wait for ADB400 handshake
  soc: imx: gpcv2: add runtime PM support for power-domains
  soc: imx: gpcv2: allow domains without power-sequence control
  dt-bindings: imx: gpcv2: add support for optional resets
  soc: imx: gpcv2: add support for optional resets
  dt-bindings: power: add defines for i.MX8MM power domains
  soc: imx: gpcv2: add support for i.MX8MM power domains
  soc: imx: gpcv2: Add support for missing i.MX8MM VPU/DISPMIX power
    domains

Peng Fan (1):
  soc: imx: gpcv2: move reset assert after requesting domain power up

 .../bindings/power/fsl,imx-gpcv2.yaml         |   9 +
 drivers/soc/imx/gpcv2.c                       | 542 ++++++++++++++----
 include/dt-bindings/power/imx8mm-power.h      |  22 +
 3 files changed, 458 insertions(+), 115 deletions(-)
 create mode 100644 include/dt-bindings/power/imx8mm-power.h

-- 
2.30.0


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH V2 01/13] soc: imx: gpcv2: move to more ideomatic error handling in probe

[Peng Fan (OSS)]

From: Lucas Stach <l.stach@pengutronix.de>

Switch to "goto out..." error handling in domain driver probe to
avoid repeating all the error paths.

Reviewed-by: Marek Vasut <marex@denx.de>
Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Tested-by: Adam Ford <aford173@gmail.com>
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
---
 drivers/soc/imx/gpcv2.c | 13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)

diff --git a/drivers/soc/imx/gpcv2.c b/drivers/soc/imx/gpcv2.c
index db7e7fc321b1..512e6f4acafd 100644
--- a/drivers/soc/imx/gpcv2.c
+++ b/drivers/soc/imx/gpcv2.c
@@ -502,18 +502,23 @@ static int imx_pgc_domain_probe(struct platform_device *pdev)
 	ret = pm_genpd_init(&domain->genpd, NULL, true);
 	if (ret) {
 		dev_err(domain->dev, "Failed to init power domain\n");
-		imx_pgc_put_clocks(domain);
-		return ret;
+		goto out_put_clocks;
 	}
 
 	ret = of_genpd_add_provider_simple(domain->dev->of_node,
 					   &domain->genpd);
 	if (ret) {
 		dev_err(domain->dev, "Failed to add genpd provider\n");
-		pm_genpd_remove(&domain->genpd);
-		imx_pgc_put_clocks(domain);
+		goto out_genpd_remove;
 	}
 
+	return 0;
+
+out_genpd_remove:
+	pm_genpd_remove(&domain->genpd);
+out_put_clocks:
+	imx_pgc_put_clocks(domain);
+
 	return ret;
 }
 
-- 
2.30.0


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH V2 02/13] soc: imx: gpcv2: move domain mapping to domain driver probe

[Peng Fan (OSS)]

From: Lucas Stach <l.stach@pengutronix.de>

As long as the power domain driver is active we want power control
over the domain (which is what the mapping bit requests), so there
is no point in whacking it for every power control action, simply
set the bit in driver probe and clear it when the driver is removed.

Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
---
 drivers/soc/imx/gpcv2.c | 22 +++++++++++++---------
 1 file changed, 13 insertions(+), 9 deletions(-)

diff --git a/drivers/soc/imx/gpcv2.c b/drivers/soc/imx/gpcv2.c
index 512e6f4acafd..552d3e6bee52 100644
--- a/drivers/soc/imx/gpcv2.c
+++ b/drivers/soc/imx/gpcv2.c
@@ -140,14 +140,11 @@ static int imx_gpc_pu_pgc_sw_pxx_req(struct generic_pm_domain *genpd,
 	int i, ret = 0;
 	u32 pxx_req;
 
-	regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
-			   domain->bits.map, domain->bits.map);
-
 	if (has_regulator && on) {
 		ret = regulator_enable(domain->regulator);
 		if (ret) {
 			dev_err(domain->dev, "failed to enable regulator\n");
-			goto unmap;
+			return ret;
 		}
 	}
 
@@ -203,9 +200,7 @@ static int imx_gpc_pu_pgc_sw_pxx_req(struct generic_pm_domain *genpd,
 		/* Preserve earlier error code */
 		ret = ret ?: err;
 	}
-unmap:
-	regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
-			   domain->bits.map, 0);
+
 	return ret;
 }
 
@@ -499,10 +494,13 @@ static int imx_pgc_domain_probe(struct platform_device *pdev)
 	if (ret)
 		return dev_err_probe(domain->dev, ret, "Failed to get domain's clocks\n");
 
+	regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
+			   domain->bits.map, domain->bits.map);
+
 	ret = pm_genpd_init(&domain->genpd, NULL, true);
 	if (ret) {
 		dev_err(domain->dev, "Failed to init power domain\n");
-		goto out_put_clocks;
+		goto out_domain_unmap;
 	}
 
 	ret = of_genpd_add_provider_simple(domain->dev->of_node,
@@ -516,7 +514,9 @@ static int imx_pgc_domain_probe(struct platform_device *pdev)
 
 out_genpd_remove:
 	pm_genpd_remove(&domain->genpd);
-out_put_clocks:
+out_domain_unmap:
+	regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
+			   domain->bits.map, 0);
 	imx_pgc_put_clocks(domain);
 
 	return ret;
@@ -528,6 +528,10 @@ static int imx_pgc_domain_remove(struct platform_device *pdev)
 
 	of_genpd_del_provider(domain->dev->of_node);
 	pm_genpd_remove(&domain->genpd);
+
+	regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
+			   domain->bits.map, 0);
+
 	imx_pgc_put_clocks(domain);
 
 	return 0;
-- 
2.30.0


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH V2 03/13] soc: imx: gpcv2: switch to clk_bulk_* API

[Peng Fan (OSS)]

From: Lucas Stach <l.stach@pengutronix.de>

Use clk_bulk API to simplify the code a bit. Also add some error
checking to the clk_prepare_enable calls.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
---
 drivers/soc/imx/gpcv2.c | 60 +++++++++--------------------------------
 1 file changed, 12 insertions(+), 48 deletions(-)

diff --git a/drivers/soc/imx/gpcv2.c b/drivers/soc/imx/gpcv2.c
index 552d3e6bee52..4222b6e87e7c 100644
--- a/drivers/soc/imx/gpcv2.c
+++ b/drivers/soc/imx/gpcv2.c
@@ -100,13 +100,11 @@
 
 #define GPC_PGC_CTRL_PCR		BIT(0)
 
-#define GPC_CLK_MAX		6
-
 struct imx_pgc_domain {
 	struct generic_pm_domain genpd;
 	struct regmap *regmap;
 	struct regulator *regulator;
-	struct clk *clk[GPC_CLK_MAX];
+	struct clk_bulk_data *clks;
 	int num_clks;
 
 	unsigned int pgc;
@@ -149,8 +147,12 @@ static int imx_gpc_pu_pgc_sw_pxx_req(struct generic_pm_domain *genpd,
 	}
 
 	/* Enable reset clocks for all devices in the domain */
-	for (i = 0; i < domain->num_clks; i++)
-		clk_prepare_enable(domain->clk[i]);
+	ret = clk_bulk_prepare_enable(domain->num_clks, domain->clks);
+	if (ret) {
+		dev_err(domain->dev, "failed to enable reset clocks\n");
+		regulator_disable(domain->regulator);
+		return ret;
+	}
 
 	if (enable_power_control)
 		regmap_update_bits(domain->regmap, GPC_PGC_CTRL(domain->pgc),
@@ -187,8 +189,7 @@ static int imx_gpc_pu_pgc_sw_pxx_req(struct generic_pm_domain *genpd,
 				   GPC_PGC_CTRL_PCR, 0);
 
 	/* Disable reset clocks for all devices in the domain */
-	for (i = 0; i < domain->num_clks; i++)
-		clk_disable_unprepare(domain->clk[i]);
+	clk_bulk_disable_unprepare(domain->num_clks, domain->clks);
 
 	if (has_regulator && !on) {
 		int err;
@@ -438,41 +439,6 @@ static const struct imx_pgc_domain_data imx8m_pgc_domain_data = {
 	.reg_access_table = &imx8m_access_table,
 };
 
-static int imx_pgc_get_clocks(struct imx_pgc_domain *domain)
-{
-	int i, ret;
-
-	for (i = 0; ; i++) {
-		struct clk *clk = of_clk_get(domain->dev->of_node, i);
-		if (IS_ERR(clk))
-			break;
-		if (i >= GPC_CLK_MAX) {
-			dev_err(domain->dev, "more than %d clocks\n",
-				GPC_CLK_MAX);
-			ret = -EINVAL;
-			goto clk_err;
-		}
-		domain->clk[i] = clk;
-	}
-	domain->num_clks = i;
-
-	return 0;
-
-clk_err:
-	while (i--)
-		clk_put(domain->clk[i]);
-
-	return ret;
-}
-
-static void imx_pgc_put_clocks(struct imx_pgc_domain *domain)
-{
-	int i;
-
-	for (i = domain->num_clks - 1; i >= 0; i--)
-		clk_put(domain->clk[i]);
-}
-
 static int imx_pgc_domain_probe(struct platform_device *pdev)
 {
 	struct imx_pgc_domain *domain = pdev->dev.platform_data;
@@ -490,9 +456,10 @@ static int imx_pgc_domain_probe(struct platform_device *pdev)
 				      domain->voltage, domain->voltage);
 	}
 
-	ret = imx_pgc_get_clocks(domain);
-	if (ret)
-		return dev_err_probe(domain->dev, ret, "Failed to get domain's clocks\n");
+	domain->num_clks = devm_clk_bulk_get_all(domain->dev, &domain->clks);
+	if (domain->num_clks < 0)
+		return dev_err_probe(domain->dev, domain->num_clks,
+				     "Failed to get domain's clocks\n");
 
 	regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
 			   domain->bits.map, domain->bits.map);
@@ -517,7 +484,6 @@ static int imx_pgc_domain_probe(struct platform_device *pdev)
 out_domain_unmap:
 	regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
 			   domain->bits.map, 0);
-	imx_pgc_put_clocks(domain);
 
 	return ret;
 }
@@ -532,8 +498,6 @@ static int imx_pgc_domain_remove(struct platform_device *pdev)
 	regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
 			   domain->bits.map, 0);
 
-	imx_pgc_put_clocks(domain);
-
 	return 0;
 }
 
-- 
2.30.0


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH V2 03/13] soc: imx: gpcv2: switch to clk_bulk_* API

[Frieder Schrempf]

On 06.05.21 03:04, Peng Fan (OSS) wrote:
> From: Lucas Stach <l.stach@pengutronix.de>
> 
> Use clk_bulk API to simplify the code a bit. Also add some error
> checking to the clk_prepare_enable calls.
> 
> Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
> Signed-off-by: Peng Fan <peng.fan@nxp.com>

Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>

> ---
>  drivers/soc/imx/gpcv2.c | 60 +++++++++--------------------------------
>  1 file changed, 12 insertions(+), 48 deletions(-)
> 
> diff --git a/drivers/soc/imx/gpcv2.c b/drivers/soc/imx/gpcv2.c
> index 552d3e6bee52..4222b6e87e7c 100644
> --- a/drivers/soc/imx/gpcv2.c
> +++ b/drivers/soc/imx/gpcv2.c
> @@ -100,13 +100,11 @@
>  
>  #define GPC_PGC_CTRL_PCR		BIT(0)
>  
> -#define GPC_CLK_MAX		6
> -
>  struct imx_pgc_domain {
>  	struct generic_pm_domain genpd;
>  	struct regmap *regmap;
>  	struct regulator *regulator;
> -	struct clk *clk[GPC_CLK_MAX];
> +	struct clk_bulk_data *clks;
>  	int num_clks;
>  
>  	unsigned int pgc;
> @@ -149,8 +147,12 @@ static int imx_gpc_pu_pgc_sw_pxx_req(struct generic_pm_domain *genpd,
>  	}
>  
>  	/* Enable reset clocks for all devices in the domain */
> -	for (i = 0; i < domain->num_clks; i++)
> -		clk_prepare_enable(domain->clk[i]);
> +	ret = clk_bulk_prepare_enable(domain->num_clks, domain->clks);
> +	if (ret) {
> +		dev_err(domain->dev, "failed to enable reset clocks\n");
> +		regulator_disable(domain->regulator);
> +		return ret;
> +	}
>  
>  	if (enable_power_control)
>  		regmap_update_bits(domain->regmap, GPC_PGC_CTRL(domain->pgc),
> @@ -187,8 +189,7 @@ static int imx_gpc_pu_pgc_sw_pxx_req(struct generic_pm_domain *genpd,
>  				   GPC_PGC_CTRL_PCR, 0);
>  
>  	/* Disable reset clocks for all devices in the domain */
> -	for (i = 0; i < domain->num_clks; i++)
> -		clk_disable_unprepare(domain->clk[i]);
> +	clk_bulk_disable_unprepare(domain->num_clks, domain->clks);
>  
>  	if (has_regulator && !on) {
>  		int err;
> @@ -438,41 +439,6 @@ static const struct imx_pgc_domain_data imx8m_pgc_domain_data = {
>  	.reg_access_table = &imx8m_access_table,
>  };
>  
> -static int imx_pgc_get_clocks(struct imx_pgc_domain *domain)
> -{
> -	int i, ret;
> -
> -	for (i = 0; ; i++) {
> -		struct clk *clk = of_clk_get(domain->dev->of_node, i);
> -		if (IS_ERR(clk))
> -			break;
> -		if (i >= GPC_CLK_MAX) {
> -			dev_err(domain->dev, "more than %d clocks\n",
> -				GPC_CLK_MAX);
> -			ret = -EINVAL;
> -			goto clk_err;
> -		}
> -		domain->clk[i] = clk;
> -	}
> -	domain->num_clks = i;
> -
> -	return 0;
> -
> -clk_err:
> -	while (i--)
> -		clk_put(domain->clk[i]);
> -
> -	return ret;
> -}
> -
> -static void imx_pgc_put_clocks(struct imx_pgc_domain *domain)
> -{
> -	int i;
> -
> -	for (i = domain->num_clks - 1; i >= 0; i--)
> -		clk_put(domain->clk[i]);
> -}
> -
>  static int imx_pgc_domain_probe(struct platform_device *pdev)
>  {
>  	struct imx_pgc_domain *domain = pdev->dev.platform_data;
> @@ -490,9 +456,10 @@ static int imx_pgc_domain_probe(struct platform_device *pdev)
>  				      domain->voltage, domain->voltage);
>  	}
>  
> -	ret = imx_pgc_get_clocks(domain);
> -	if (ret)
> -		return dev_err_probe(domain->dev, ret, "Failed to get domain's clocks\n");
> +	domain->num_clks = devm_clk_bulk_get_all(domain->dev, &domain->clks);
> +	if (domain->num_clks < 0)
> +		return dev_err_probe(domain->dev, domain->num_clks,
> +				     "Failed to get domain's clocks\n");
>  
>  	regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
>  			   domain->bits.map, domain->bits.map);
> @@ -517,7 +484,6 @@ static int imx_pgc_domain_probe(struct platform_device *pdev)
>  out_domain_unmap:
>  	regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
>  			   domain->bits.map, 0);
> -	imx_pgc_put_clocks(domain);
>  
>  	return ret;
>  }
> @@ -532,8 +498,6 @@ static int imx_pgc_domain_remove(struct platform_device *pdev)
>  	regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
>  			   domain->bits.map, 0);
>  
> -	imx_pgc_put_clocks(domain);
> -
>  	return 0;
>  }
>  
> 

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH V2 04/13] soc: imx: gpcv2: split power up and power down sequence control

[Peng Fan (OSS)]

From: Lucas Stach <l.stach@pengutronix.de>

The current mixed function to control both power up and power down
sequences is very hard to follow and already contains some sequence
errors like triggering the ADB400 handshake at the wrong time due to
this. Split the function into two, which results in slightly more
code, but is way easier to get right.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
---
 drivers/soc/imx/gpcv2.c | 141 ++++++++++++++++++++++++----------------
 1 file changed, 86 insertions(+), 55 deletions(-)

diff --git a/drivers/soc/imx/gpcv2.c b/drivers/soc/imx/gpcv2.c
index 4222b6e87e7c..bcf1f338b0bf 100644
--- a/drivers/soc/imx/gpcv2.c
+++ b/drivers/soc/imx/gpcv2.c
@@ -125,20 +125,19 @@ struct imx_pgc_domain_data {
 	const struct regmap_access_table *reg_access_table;
 };
 
-static int imx_gpc_pu_pgc_sw_pxx_req(struct generic_pm_domain *genpd,
-				      bool on)
+static inline struct imx_pgc_domain *
+to_imx_pgc_domain(struct generic_pm_domain *genpd)
 {
-	struct imx_pgc_domain *domain = container_of(genpd,
-						      struct imx_pgc_domain,
-						      genpd);
-	unsigned int offset = on ?
-		GPC_PU_PGC_SW_PUP_REQ : GPC_PU_PGC_SW_PDN_REQ;
-	const bool enable_power_control = !on;
-	const bool has_regulator = !IS_ERR(domain->regulator);
-	int i, ret = 0;
-	u32 pxx_req;
-
-	if (has_regulator && on) {
+	return container_of(genpd, struct imx_pgc_domain, genpd);
+}
+
+static int imx_pgc_power_up(struct generic_pm_domain *genpd)
+{
+	struct imx_pgc_domain *domain = to_imx_pgc_domain(genpd);
+	u32 reg_val;
+	int ret;
+
+	if (!IS_ERR(domain->regulator)) {
 		ret = regulator_enable(domain->regulator);
 		if (ret) {
 			dev_err(domain->dev, "failed to enable regulator\n");
@@ -150,69 +149,101 @@ static int imx_gpc_pu_pgc_sw_pxx_req(struct generic_pm_domain *genpd,
 	ret = clk_bulk_prepare_enable(domain->num_clks, domain->clks);
 	if (ret) {
 		dev_err(domain->dev, "failed to enable reset clocks\n");
-		regulator_disable(domain->regulator);
-		return ret;
+		goto out_regulator_disable;
 	}
 
-	if (enable_power_control)
-		regmap_update_bits(domain->regmap, GPC_PGC_CTRL(domain->pgc),
-				   GPC_PGC_CTRL_PCR, GPC_PGC_CTRL_PCR);
-
-	if (domain->bits.hsk)
-		regmap_update_bits(domain->regmap, GPC_PU_PWRHSK,
-				   domain->bits.hsk, on ? domain->bits.hsk : 0);
-
-	regmap_update_bits(domain->regmap, offset,
+	/* request the domain to power up */
+	regmap_update_bits(domain->regmap, GPC_PU_PGC_SW_PUP_REQ,
 			   domain->bits.pxx, domain->bits.pxx);
-
 	/*
 	 * As per "5.5.9.4 Example Code 4" in IMX7DRM.pdf wait
 	 * for PUP_REQ/PDN_REQ bit to be cleared
 	 */
-	ret = regmap_read_poll_timeout(domain->regmap, offset, pxx_req,
-				       !(pxx_req & domain->bits.pxx),
+	ret = regmap_read_poll_timeout(domain->regmap, GPC_PU_PGC_SW_PUP_REQ,
+				       reg_val, !(reg_val & domain->bits.pxx),
 				       0, USEC_PER_MSEC);
 	if (ret) {
 		dev_err(domain->dev, "failed to command PGC\n");
-		/*
-		 * If we were in a process of enabling a
-		 * domain and failed we might as well disable
-		 * the regulator we just enabled. And if it
-		 * was the opposite situation and we failed to
-		 * power down -- keep the regulator on
-		 */
-		on = !on;
+		goto out_clk_disable;
 	}
 
-	if (enable_power_control)
-		regmap_update_bits(domain->regmap, GPC_PGC_CTRL(domain->pgc),
-				   GPC_PGC_CTRL_PCR, 0);
+	/* disable power control */
+	regmap_clear_bits(domain->regmap, GPC_PGC_CTRL(domain->pgc),
+			  GPC_PGC_CTRL_PCR);
+
+	/* request the ADB400 to power up */
+	if (domain->bits.hsk)
+		regmap_update_bits(domain->regmap, GPC_PU_PWRHSK,
+				   domain->bits.hsk, domain->bits.hsk);
 
 	/* Disable reset clocks for all devices in the domain */
 	clk_bulk_disable_unprepare(domain->num_clks, domain->clks);
 
-	if (has_regulator && !on) {
-		int err;
+	return 0;
 
-		err = regulator_disable(domain->regulator);
-		if (err)
-			dev_err(domain->dev,
-				"failed to disable regulator: %d\n", err);
-		/* Preserve earlier error code */
-		ret = ret ?: err;
-	}
+out_clk_disable:
+	clk_bulk_disable_unprepare(domain->num_clks, domain->clks);
+out_regulator_disable:
+	if (!IS_ERR(domain->regulator))
+		regulator_disable(domain->regulator);
 
 	return ret;
 }
 
-static int imx_gpc_pu_pgc_sw_pup_req(struct generic_pm_domain *genpd)
+static int imx_pgc_power_down(struct generic_pm_domain *genpd)
 {
-	return imx_gpc_pu_pgc_sw_pxx_req(genpd, true);
-}
+	struct imx_pgc_domain *domain = to_imx_pgc_domain(genpd);
+	u32 reg_val;
+	int ret;
 
-static int imx_gpc_pu_pgc_sw_pdn_req(struct generic_pm_domain *genpd)
-{
-	return imx_gpc_pu_pgc_sw_pxx_req(genpd, false);
+	/* Enable reset clocks for all devices in the domain */
+	ret = clk_bulk_prepare_enable(domain->num_clks, domain->clks);
+	if (ret) {
+		dev_err(domain->dev, "failed to enable reset clocks\n");
+		return ret;
+	}
+
+	/* request the ADB400 to power down */
+	if (domain->bits.hsk)
+		regmap_clear_bits(domain->regmap, GPC_PU_PWRHSK,
+				  domain->bits.hsk);
+
+	/* enable power control */
+	regmap_update_bits(domain->regmap, GPC_PGC_CTRL(domain->pgc),
+			   GPC_PGC_CTRL_PCR, GPC_PGC_CTRL_PCR);
+
+	/* request the domain to power down */
+	regmap_update_bits(domain->regmap, GPC_PU_PGC_SW_PDN_REQ,
+			   domain->bits.pxx, domain->bits.pxx);
+	/*
+	 * As per "5.5.9.4 Example Code 4" in IMX7DRM.pdf wait
+	 * for PUP_REQ/PDN_REQ bit to be cleared
+	 */
+	ret = regmap_read_poll_timeout(domain->regmap, GPC_PU_PGC_SW_PDN_REQ,
+				       reg_val, !(reg_val & domain->bits.pxx),
+				       0, USEC_PER_MSEC);
+	if (ret) {
+		dev_err(domain->dev, "failed to command PGC\n");
+		goto out_clk_disable;
+	}
+
+	/* Disable reset clocks for all devices in the domain */
+	clk_bulk_disable_unprepare(domain->num_clks, domain->clks);
+
+	if (!IS_ERR(domain->regulator)) {
+		ret = regulator_disable(domain->regulator);
+		if (ret) {
+			dev_err(domain->dev, "failed to disable regulator\n");
+			return ret;
+		}
+	}
+
+	return 0;
+
+out_clk_disable:
+	clk_bulk_disable_unprepare(domain->num_clks, domain->clks);
+
+	return ret;
 }
 
 static const struct imx_pgc_domain imx7_pgc_domains[] = {
@@ -590,8 +621,8 @@ static int imx_gpcv2_probe(struct platform_device *pdev)
 
 		domain = pd_pdev->dev.platform_data;
 		domain->regmap = regmap;
-		domain->genpd.power_on  = imx_gpc_pu_pgc_sw_pup_req;
-		domain->genpd.power_off = imx_gpc_pu_pgc_sw_pdn_req;
+		domain->genpd.power_on  = imx_pgc_power_up;
+		domain->genpd.power_off = imx_pgc_power_down;
 
 		pd_pdev->dev.parent = dev;
 		pd_pdev->dev.of_node = np;
-- 
2.30.0


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH V2 04/13] soc: imx: gpcv2: split power up and power down sequence control

[Frieder Schrempf]

On 06.05.21 03:04, Peng Fan (OSS) wrote:
> From: Lucas Stach <l.stach@pengutronix.de>
> 
> The current mixed function to control both power up and power down
> sequences is very hard to follow and already contains some sequence
> errors like triggering the ADB400 handshake at the wrong time due to
> this. Split the function into two, which results in slightly more
> code, but is way easier to get right.
> 
> Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
> Signed-off-by: Peng Fan <peng.fan@nxp.com>

Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>

> ---
>  drivers/soc/imx/gpcv2.c | 141 ++++++++++++++++++++++++----------------
>  1 file changed, 86 insertions(+), 55 deletions(-)
> 
> diff --git a/drivers/soc/imx/gpcv2.c b/drivers/soc/imx/gpcv2.c
> index 4222b6e87e7c..bcf1f338b0bf 100644
> --- a/drivers/soc/imx/gpcv2.c
> +++ b/drivers/soc/imx/gpcv2.c
> @@ -125,20 +125,19 @@ struct imx_pgc_domain_data {
>  	const struct regmap_access_table *reg_access_table;
>  };
>  
> -static int imx_gpc_pu_pgc_sw_pxx_req(struct generic_pm_domain *genpd,
> -				      bool on)
> +static inline struct imx_pgc_domain *
> +to_imx_pgc_domain(struct generic_pm_domain *genpd)
>  {
> -	struct imx_pgc_domain *domain = container_of(genpd,
> -						      struct imx_pgc_domain,
> -						      genpd);
> -	unsigned int offset = on ?
> -		GPC_PU_PGC_SW_PUP_REQ : GPC_PU_PGC_SW_PDN_REQ;
> -	const bool enable_power_control = !on;
> -	const bool has_regulator = !IS_ERR(domain->regulator);
> -	int i, ret = 0;
> -	u32 pxx_req;
> -
> -	if (has_regulator && on) {
> +	return container_of(genpd, struct imx_pgc_domain, genpd);
> +}
> +
> +static int imx_pgc_power_up(struct generic_pm_domain *genpd)
> +{
> +	struct imx_pgc_domain *domain = to_imx_pgc_domain(genpd);
> +	u32 reg_val;
> +	int ret;
> +
> +	if (!IS_ERR(domain->regulator)) {
>  		ret = regulator_enable(domain->regulator);
>  		if (ret) {
>  			dev_err(domain->dev, "failed to enable regulator\n");
> @@ -150,69 +149,101 @@ static int imx_gpc_pu_pgc_sw_pxx_req(struct generic_pm_domain *genpd,
>  	ret = clk_bulk_prepare_enable(domain->num_clks, domain->clks);
>  	if (ret) {
>  		dev_err(domain->dev, "failed to enable reset clocks\n");
> -		regulator_disable(domain->regulator);
> -		return ret;
> +		goto out_regulator_disable;
>  	}
>  
> -	if (enable_power_control)
> -		regmap_update_bits(domain->regmap, GPC_PGC_CTRL(domain->pgc),
> -				   GPC_PGC_CTRL_PCR, GPC_PGC_CTRL_PCR);
> -
> -	if (domain->bits.hsk)
> -		regmap_update_bits(domain->regmap, GPC_PU_PWRHSK,
> -				   domain->bits.hsk, on ? domain->bits.hsk : 0);
> -
> -	regmap_update_bits(domain->regmap, offset,
> +	/* request the domain to power up */
> +	regmap_update_bits(domain->regmap, GPC_PU_PGC_SW_PUP_REQ,
>  			   domain->bits.pxx, domain->bits.pxx);
> -
>  	/*
>  	 * As per "5.5.9.4 Example Code 4" in IMX7DRM.pdf wait
>  	 * for PUP_REQ/PDN_REQ bit to be cleared
>  	 */
> -	ret = regmap_read_poll_timeout(domain->regmap, offset, pxx_req,
> -				       !(pxx_req & domain->bits.pxx),
> +	ret = regmap_read_poll_timeout(domain->regmap, GPC_PU_PGC_SW_PUP_REQ,
> +				       reg_val, !(reg_val & domain->bits.pxx),
>  				       0, USEC_PER_MSEC);
>  	if (ret) {
>  		dev_err(domain->dev, "failed to command PGC\n");
> -		/*
> -		 * If we were in a process of enabling a
> -		 * domain and failed we might as well disable
> -		 * the regulator we just enabled. And if it
> -		 * was the opposite situation and we failed to
> -		 * power down -- keep the regulator on
> -		 */
> -		on = !on;
> +		goto out_clk_disable;
>  	}
>  
> -	if (enable_power_control)
> -		regmap_update_bits(domain->regmap, GPC_PGC_CTRL(domain->pgc),
> -				   GPC_PGC_CTRL_PCR, 0);
> +	/* disable power control */
> +	regmap_clear_bits(domain->regmap, GPC_PGC_CTRL(domain->pgc),
> +			  GPC_PGC_CTRL_PCR);
> +
> +	/* request the ADB400 to power up */
> +	if (domain->bits.hsk)
> +		regmap_update_bits(domain->regmap, GPC_PU_PWRHSK,
> +				   domain->bits.hsk, domain->bits.hsk);
>  
>  	/* Disable reset clocks for all devices in the domain */
>  	clk_bulk_disable_unprepare(domain->num_clks, domain->clks);
>  
> -	if (has_regulator && !on) {
> -		int err;
> +	return 0;
>  
> -		err = regulator_disable(domain->regulator);
> -		if (err)
> -			dev_err(domain->dev,
> -				"failed to disable regulator: %d\n", err);
> -		/* Preserve earlier error code */
> -		ret = ret ?: err;
> -	}
> +out_clk_disable:
> +	clk_bulk_disable_unprepare(domain->num_clks, domain->clks);
> +out_regulator_disable:
> +	if (!IS_ERR(domain->regulator))
> +		regulator_disable(domain->regulator);
>  
>  	return ret;
>  }
>  
> -static int imx_gpc_pu_pgc_sw_pup_req(struct generic_pm_domain *genpd)
> +static int imx_pgc_power_down(struct generic_pm_domain *genpd)
>  {
> -	return imx_gpc_pu_pgc_sw_pxx_req(genpd, true);
> -}
> +	struct imx_pgc_domain *domain = to_imx_pgc_domain(genpd);
> +	u32 reg_val;
> +	int ret;
>  
> -static int imx_gpc_pu_pgc_sw_pdn_req(struct generic_pm_domain *genpd)
> -{
> -	return imx_gpc_pu_pgc_sw_pxx_req(genpd, false);
> +	/* Enable reset clocks for all devices in the domain */
> +	ret = clk_bulk_prepare_enable(domain->num_clks, domain->clks);
> +	if (ret) {
> +		dev_err(domain->dev, "failed to enable reset clocks\n");
> +		return ret;
> +	}
> +
> +	/* request the ADB400 to power down */
> +	if (domain->bits.hsk)
> +		regmap_clear_bits(domain->regmap, GPC_PU_PWRHSK,
> +				  domain->bits.hsk);
> +
> +	/* enable power control */
> +	regmap_update_bits(domain->regmap, GPC_PGC_CTRL(domain->pgc),
> +			   GPC_PGC_CTRL_PCR, GPC_PGC_CTRL_PCR);
> +
> +	/* request the domain to power down */
> +	regmap_update_bits(domain->regmap, GPC_PU_PGC_SW_PDN_REQ,
> +			   domain->bits.pxx, domain->bits.pxx);
> +	/*
> +	 * As per "5.5.9.4 Example Code 4" in IMX7DRM.pdf wait
> +	 * for PUP_REQ/PDN_REQ bit to be cleared
> +	 */
> +	ret = regmap_read_poll_timeout(domain->regmap, GPC_PU_PGC_SW_PDN_REQ,
> +				       reg_val, !(reg_val & domain->bits.pxx),
> +				       0, USEC_PER_MSEC);
> +	if (ret) {
> +		dev_err(domain->dev, "failed to command PGC\n");
> +		goto out_clk_disable;
> +	}
> +
> +	/* Disable reset clocks for all devices in the domain */
> +	clk_bulk_disable_unprepare(domain->num_clks, domain->clks);
> +
> +	if (!IS_ERR(domain->regulator)) {
> +		ret = regulator_disable(domain->regulator);
> +		if (ret) {
> +			dev_err(domain->dev, "failed to disable regulator\n");
> +			return ret;
> +		}
> +	}
> +
> +	return 0;
> +
> +out_clk_disable:
> +	clk_bulk_disable_unprepare(domain->num_clks, domain->clks);
> +
> +	return ret;
>  }
>  
>  static const struct imx_pgc_domain imx7_pgc_domains[] = {
> @@ -590,8 +621,8 @@ static int imx_gpcv2_probe(struct platform_device *pdev)
>  
>  		domain = pd_pdev->dev.platform_data;
>  		domain->regmap = regmap;
> -		domain->genpd.power_on  = imx_gpc_pu_pgc_sw_pup_req;
> -		domain->genpd.power_off = imx_gpc_pu_pgc_sw_pdn_req;
> +		domain->genpd.power_on  = imx_pgc_power_up;
> +		domain->genpd.power_off = imx_pgc_power_down;
>  
>  		pd_pdev->dev.parent = dev;
>  		pd_pdev->dev.of_node = np;
> 

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH V2 05/13] soc: imx: gpcv2: wait for ADB400 handshake

[Peng Fan (OSS)]

From: Lucas Stach <l.stach@pengutronix.de>

New reference manuals show that there is actually a status bit for
the ADB400 handshake. Add a poll loop to wait for the ADB400 to
acknowledge our request.

[Peng Fan: i.MX8MM has blk ctl module, the handshake can only finish
 after setting blk ctl. The blk ctl driver will set the bus clk bit and
 the handshake will finish there. we just add a delay and suppose the
 handshake will finish after that.]

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
---
 drivers/soc/imx/gpcv2.c | 47 ++++++++++++++++++++++++++++++++++-------
 1 file changed, 39 insertions(+), 8 deletions(-)

diff --git a/drivers/soc/imx/gpcv2.c b/drivers/soc/imx/gpcv2.c
index bcf1f338b0bf..558b8b8af9af 100644
--- a/drivers/soc/imx/gpcv2.c
+++ b/drivers/soc/imx/gpcv2.c
@@ -69,6 +69,9 @@
 
 #define GPC_PU_PWRHSK			0x1fc
 
+#define IMX8M_GPU_HSK_PWRDNACKN			BIT(26)
+#define IMX8M_VPU_HSK_PWRDNACKN			BIT(25)
+#define IMX8M_DISP_HSK_PWRDNACKN		BIT(24)
 #define IMX8M_GPU_HSK_PWRDNREQN			BIT(6)
 #define IMX8M_VPU_HSK_PWRDNREQN			BIT(5)
 #define IMX8M_DISP_HSK_PWRDNREQN		BIT(4)
@@ -112,7 +115,8 @@ struct imx_pgc_domain {
 	const struct {
 		u32 pxx;
 		u32 map;
-		u32 hsk;
+		u32 hskreq;
+		u32 hskack;
 	} bits;
 
 	const int voltage;
@@ -172,9 +176,23 @@ static int imx_pgc_power_up(struct generic_pm_domain *genpd)
 			  GPC_PGC_CTRL_PCR);
 
 	/* request the ADB400 to power up */
-	if (domain->bits.hsk)
+	if (domain->bits.hskreq) {
 		regmap_update_bits(domain->regmap, GPC_PU_PWRHSK,
-				   domain->bits.hsk, domain->bits.hsk);
+				   domain->bits.hskreq, domain->bits.hskreq);
+
+		/*
+		 * ret = regmap_read_poll_timeout(domain->regmap, GPC_PU_PWRHSK, reg_val,
+		 *				  (reg_val & domain->bits.hskack), 0,
+		 *				  USEC_PER_MSEC);
+		 * Technically we need the commented code to wait handshake. But that needs
+		 * the BLK-CTL module BUS clk-en bit being set.
+		 *
+		 * There is a separate BLK-CTL module and we will have such a driver for it,
+		 * that driver will set the BUS clk-en bit and handshake will be triggered
+		 * automatically there. Just add a delay and suppose the handshake finish
+		 * after that.
+		 */
+	}
 
 	/* Disable reset clocks for all devices in the domain */
 	clk_bulk_disable_unprepare(domain->num_clks, domain->clks);
@@ -204,9 +222,19 @@ static int imx_pgc_power_down(struct generic_pm_domain *genpd)
 	}
 
 	/* request the ADB400 to power down */
-	if (domain->bits.hsk)
+	if (domain->bits.hskreq) {
 		regmap_clear_bits(domain->regmap, GPC_PU_PWRHSK,
-				  domain->bits.hsk);
+				  domain->bits.hskreq);
+
+		ret = regmap_read_poll_timeout(domain->regmap, GPC_PU_PWRHSK,
+					       reg_val,
+					       !(reg_val & domain->bits.hskack),
+					       0, USEC_PER_MSEC);
+		if (ret) {
+			dev_err(domain->dev, "failed to power down ADB400\n");
+			goto out_clk_disable;
+		}
+	}
 
 	/* enable power control */
 	regmap_update_bits(domain->regmap, GPC_PGC_CTRL(domain->pgc),
@@ -369,7 +397,8 @@ static const struct imx_pgc_domain imx8m_pgc_domains[] = {
 		.bits  = {
 			.pxx = IMX8M_GPU_SW_Pxx_REQ,
 			.map = IMX8M_GPU_A53_DOMAIN,
-			.hsk = IMX8M_GPU_HSK_PWRDNREQN,
+			.hskreq = IMX8M_GPU_HSK_PWRDNREQN,
+			.hskack = IMX8M_GPU_HSK_PWRDNACKN,
 		},
 		.pgc   = IMX8M_PGC_GPU,
 	},
@@ -381,7 +410,8 @@ static const struct imx_pgc_domain imx8m_pgc_domains[] = {
 		.bits  = {
 			.pxx = IMX8M_VPU_SW_Pxx_REQ,
 			.map = IMX8M_VPU_A53_DOMAIN,
-			.hsk = IMX8M_VPU_HSK_PWRDNREQN,
+			.hskreq = IMX8M_VPU_HSK_PWRDNREQN,
+			.hskack = IMX8M_VPU_HSK_PWRDNACKN,
 		},
 		.pgc   = IMX8M_PGC_VPU,
 	},
@@ -393,7 +423,8 @@ static const struct imx_pgc_domain imx8m_pgc_domains[] = {
 		.bits  = {
 			.pxx = IMX8M_DISP_SW_Pxx_REQ,
 			.map = IMX8M_DISP_A53_DOMAIN,
-			.hsk = IMX8M_DISP_HSK_PWRDNREQN,
+			.hskreq = IMX8M_DISP_HSK_PWRDNREQN,
+			.hskack = IMX8M_DISP_HSK_PWRDNACKN,
 		},
 		.pgc   = IMX8M_PGC_DISP,
 	},
-- 
2.30.0


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH V2 06/13] soc: imx: gpcv2: add runtime PM support for power-domains

[Peng Fan (OSS)]

From: Lucas Stach <l.stach@pengutronix.de>

This allows to nest domains into other power domains and have the
parent domain powered up/down as required by the child domains.

Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
---
 drivers/soc/imx/gpcv2.c | 18 +++++++++++++++++-
 1 file changed, 17 insertions(+), 1 deletion(-)

diff --git a/drivers/soc/imx/gpcv2.c b/drivers/soc/imx/gpcv2.c
index 558b8b8af9af..bfea8560cb64 100644
--- a/drivers/soc/imx/gpcv2.c
+++ b/drivers/soc/imx/gpcv2.c
@@ -12,6 +12,7 @@
 #include <linux/of_device.h>
 #include <linux/platform_device.h>
 #include <linux/pm_domain.h>
+#include <linux/pm_runtime.h>
 #include <linux/regmap.h>
 #include <linux/regulator/consumer.h>
 #include <linux/sizes.h>
@@ -141,11 +142,17 @@ static int imx_pgc_power_up(struct generic_pm_domain *genpd)
 	u32 reg_val;
 	int ret;
 
+	ret = pm_runtime_get_sync(domain->dev);
+	if (ret < 0) {
+		pm_runtime_put_noidle(domain->dev);
+		return ret;
+	}
+
 	if (!IS_ERR(domain->regulator)) {
 		ret = regulator_enable(domain->regulator);
 		if (ret) {
 			dev_err(domain->dev, "failed to enable regulator\n");
-			return ret;
+			goto out_put_pm;
 		}
 	}
 
@@ -204,6 +211,8 @@ static int imx_pgc_power_up(struct generic_pm_domain *genpd)
 out_regulator_disable:
 	if (!IS_ERR(domain->regulator))
 		regulator_disable(domain->regulator);
+out_put_pm:
+	pm_runtime_put(domain->dev);
 
 	return ret;
 }
@@ -266,6 +275,8 @@ static int imx_pgc_power_down(struct generic_pm_domain *genpd)
 		}
 	}
 
+	pm_runtime_put(domain->dev);
+
 	return 0;
 
 out_clk_disable:
@@ -523,6 +534,8 @@ static int imx_pgc_domain_probe(struct platform_device *pdev)
 		return dev_err_probe(domain->dev, domain->num_clks,
 				     "Failed to get domain's clocks\n");
 
+	pm_runtime_enable(domain->dev);
+
 	regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
 			   domain->bits.map, domain->bits.map);
 
@@ -546,6 +559,7 @@ static int imx_pgc_domain_probe(struct platform_device *pdev)
 out_domain_unmap:
 	regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
 			   domain->bits.map, 0);
+	pm_runtime_disable(domain->dev);
 
 	return ret;
 }
@@ -560,6 +574,8 @@ static int imx_pgc_domain_remove(struct platform_device *pdev)
 	regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
 			   domain->bits.map, 0);
 
+	pm_runtime_disable(domain->dev);
+
 	return 0;
 }
 
-- 
2.30.0


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH V2 07/13] soc: imx: gpcv2: allow domains without power-sequence control

[Peng Fan (OSS)]

From: Lucas Stach <l.stach@pengutronix.de>

Some of the PGC domains only control the handshake with the ADB400
and don't have any power sequence controls. Make such domains work
by allowing the pxx and map bits to be empty and skip all actions
using those controls.

Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
---
 drivers/soc/imx/gpcv2.c | 89 +++++++++++++++++++++++------------------
 1 file changed, 49 insertions(+), 40 deletions(-)

diff --git a/drivers/soc/imx/gpcv2.c b/drivers/soc/imx/gpcv2.c
index bfea8560cb64..0428f0eddee0 100644
--- a/drivers/soc/imx/gpcv2.c
+++ b/drivers/soc/imx/gpcv2.c
@@ -163,24 +163,27 @@ static int imx_pgc_power_up(struct generic_pm_domain *genpd)
 		goto out_regulator_disable;
 	}
 
-	/* request the domain to power up */
-	regmap_update_bits(domain->regmap, GPC_PU_PGC_SW_PUP_REQ,
-			   domain->bits.pxx, domain->bits.pxx);
-	/*
-	 * As per "5.5.9.4 Example Code 4" in IMX7DRM.pdf wait
-	 * for PUP_REQ/PDN_REQ bit to be cleared
-	 */
-	ret = regmap_read_poll_timeout(domain->regmap, GPC_PU_PGC_SW_PUP_REQ,
-				       reg_val, !(reg_val & domain->bits.pxx),
-				       0, USEC_PER_MSEC);
-	if (ret) {
-		dev_err(domain->dev, "failed to command PGC\n");
-		goto out_clk_disable;
-	}
+	if (domain->bits.pxx) {
+		/* request the domain to power up */
+		regmap_update_bits(domain->regmap, GPC_PU_PGC_SW_PUP_REQ,
+				   domain->bits.pxx, domain->bits.pxx);
+		/*
+		 * As per "5.5.9.4 Example Code 4" in IMX7DRM.pdf wait
+		 * for PUP_REQ/PDN_REQ bit to be cleared
+		 */
+		ret = regmap_read_poll_timeout(domain->regmap,
+					       GPC_PU_PGC_SW_PUP_REQ, reg_val,
+					       !(reg_val & domain->bits.pxx),
+					       0, USEC_PER_MSEC);
+		if (ret) {
+			dev_err(domain->dev, "failed to command PGC\n");
+			goto out_clk_disable;
+		}
 
-	/* disable power control */
-	regmap_clear_bits(domain->regmap, GPC_PGC_CTRL(domain->pgc),
-			  GPC_PGC_CTRL_PCR);
+		/* disable power control */
+		regmap_clear_bits(domain->regmap, GPC_PGC_CTRL(domain->pgc),
+				  GPC_PGC_CTRL_PCR);
+	}
 
 	/* request the ADB400 to power up */
 	if (domain->bits.hskreq) {
@@ -245,23 +248,26 @@ static int imx_pgc_power_down(struct generic_pm_domain *genpd)
 		}
 	}
 
-	/* enable power control */
-	regmap_update_bits(domain->regmap, GPC_PGC_CTRL(domain->pgc),
-			   GPC_PGC_CTRL_PCR, GPC_PGC_CTRL_PCR);
-
-	/* request the domain to power down */
-	regmap_update_bits(domain->regmap, GPC_PU_PGC_SW_PDN_REQ,
-			   domain->bits.pxx, domain->bits.pxx);
-	/*
-	 * As per "5.5.9.4 Example Code 4" in IMX7DRM.pdf wait
-	 * for PUP_REQ/PDN_REQ bit to be cleared
-	 */
-	ret = regmap_read_poll_timeout(domain->regmap, GPC_PU_PGC_SW_PDN_REQ,
-				       reg_val, !(reg_val & domain->bits.pxx),
-				       0, USEC_PER_MSEC);
-	if (ret) {
-		dev_err(domain->dev, "failed to command PGC\n");
-		goto out_clk_disable;
+	if (domain->bits.pxx) {
+		/* enable power control */
+		regmap_update_bits(domain->regmap, GPC_PGC_CTRL(domain->pgc),
+				   GPC_PGC_CTRL_PCR, GPC_PGC_CTRL_PCR);
+
+		/* request the domain to power down */
+		regmap_update_bits(domain->regmap, GPC_PU_PGC_SW_PDN_REQ,
+				   domain->bits.pxx, domain->bits.pxx);
+		/*
+		 * As per "5.5.9.4 Example Code 4" in IMX7DRM.pdf wait
+		 * for PUP_REQ/PDN_REQ bit to be cleared
+		 */
+		ret = regmap_read_poll_timeout(domain->regmap,
+					       GPC_PU_PGC_SW_PDN_REQ, reg_val,
+					       !(reg_val & domain->bits.pxx),
+					       0, USEC_PER_MSEC);
+		if (ret) {
+			dev_err(domain->dev, "failed to command PGC\n");
+			goto out_clk_disable;
+		}
 	}
 
 	/* Disable reset clocks for all devices in the domain */
@@ -536,8 +542,9 @@ static int imx_pgc_domain_probe(struct platform_device *pdev)
 
 	pm_runtime_enable(domain->dev);
 
-	regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
-			   domain->bits.map, domain->bits.map);
+	if (domain->bits.map)
+		regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
+				   domain->bits.map, domain->bits.map);
 
 	ret = pm_genpd_init(&domain->genpd, NULL, true);
 	if (ret) {
@@ -557,8 +564,9 @@ static int imx_pgc_domain_probe(struct platform_device *pdev)
 out_genpd_remove:
 	pm_genpd_remove(&domain->genpd);
 out_domain_unmap:
-	regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
-			   domain->bits.map, 0);
+	if (domain->bits.map)
+		regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
+				   domain->bits.map, 0);
 	pm_runtime_disable(domain->dev);
 
 	return ret;
@@ -571,8 +579,9 @@ static int imx_pgc_domain_remove(struct platform_device *pdev)
 	of_genpd_del_provider(domain->dev->of_node);
 	pm_genpd_remove(&domain->genpd);
 
-	regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
-			   domain->bits.map, 0);
+	if (domain->bits.map)
+		regmap_update_bits(domain->regmap, GPC_PGC_CPU_MAPPING,
+				   domain->bits.map, 0);
 
 	pm_runtime_disable(domain->dev);
 
-- 
2.30.0


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH V2 08/13] dt-bindings: imx: gpcv2: add support for optional resets

[Peng Fan (OSS)]

From: Lucas Stach <l.stach@pengutronix.de>

For some domains the resets of the devices in the domain are not
automatically triggered. Add an optional resets property to allow
the GPC driver to trigger those resets explicitly.

The resets belong to devices located inside the power domain,
which need to be held in reset across the power-up sequence. So we
have no means to specify what each reset is in a generic power-domain
binding. Same situation as with the clocks in this binding actually.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
---
 Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml b/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml
index a96e6dbf1858..4330c73a2c30 100644
--- a/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml
+++ b/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml
@@ -66,6 +66,13 @@ properties:
 
           power-supply: true
 
+          resets:
+            description: |
+              A number of phandles to resets that need to be asserted during
+              power-up sequencing of the domain.
+            minItems: 1
+            maxItems: 4
+
         required:
           - '#power-domain-cells'
           - reg
-- 
2.30.0


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH V2 08/13] dt-bindings: imx: gpcv2: add support for optional resets

[Frieder Schrempf]

On 06.05.21 03:04, Peng Fan (OSS) wrote:
> From: Lucas Stach <l.stach@pengutronix.de>
> 
> For some domains the resets of the devices in the domain are not
> automatically triggered. Add an optional resets property to allow
> the GPC driver to trigger those resets explicitly.
> 
> The resets belong to devices located inside the power domain,
> which need to be held in reset across the power-up sequence. So we
> have no means to specify what each reset is in a generic power-domain
> binding. Same situation as with the clocks in this binding actually.

My understanding was that Rob wanted this explanation to be contained in the binding docs itself and not only in the commit message, but I might be wrong.

> 
> Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
> Signed-off-by: Peng Fan <peng.fan@nxp.com>
> ---
>  Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml | 7 +++++++
>  1 file changed, 7 insertions(+)
> 
> diff --git a/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml b/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml
> index a96e6dbf1858..4330c73a2c30 100644
> --- a/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml
> +++ b/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml
> @@ -66,6 +66,13 @@ properties:
>  
>            power-supply: true
>  
> +          resets:
> +            description: |
> +              A number of phandles to resets that need to be asserted during
> +              power-up sequencing of the domain.
> +            minItems: 1
> +            maxItems: 4
> +
>          required:
>            - '#power-domain-cells'
>            - reg
> 

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH V2 08/13] dt-bindings: imx: gpcv2: add support for optional resets

[Rob Herring]

On Thu, May 06, 2021 at 08:43:17AM +0200, Frieder Schrempf wrote:
> On 06.05.21 03:04, Peng Fan (OSS) wrote:
> > From: Lucas Stach <l.stach@pengutronix.de>
> > 
> > For some domains the resets of the devices in the domain are not
> > automatically triggered. Add an optional resets property to allow
> > the GPC driver to trigger those resets explicitly.
> > 
> > The resets belong to devices located inside the power domain,
> > which need to be held in reset across the power-up sequence. So we
> > have no means to specify what each reset is in a generic power-domain
> > binding. Same situation as with the clocks in this binding actually.
> 
> My understanding was that Rob wanted this explanation to be contained in the binding docs itself and not only in the commit message, but I might be wrong.

Yes, that would be better.

> 
> > 
> > Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
> > Signed-off-by: Peng Fan <peng.fan@nxp.com>
> > ---
> >  Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml | 7 +++++++
> >  1 file changed, 7 insertions(+)
> > 
> > diff --git a/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml b/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml
> > index a96e6dbf1858..4330c73a2c30 100644
> > --- a/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml
> > +++ b/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml
> > @@ -66,6 +66,13 @@ properties:
> >  
> >            power-supply: true
> >  
> > +          resets:
> > +            description: |
> > +              A number of phandles to resets that need to be asserted during
> > +              power-up sequencing of the domain.
> > +            minItems: 1
> > +            maxItems: 4
> > +
> >          required:
> >            - '#power-domain-cells'
> >            - reg
> > 

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

RE: [PATCH V2 08/13] dt-bindings: imx: gpcv2: add support for optional resets

[Peng Fan]

> Subject: Re: [PATCH V2 08/13] dt-bindings: imx: gpcv2: add support for
> optional resets
> 
> On Thu, May 06, 2021 at 08:43:17AM +0200, Frieder Schrempf wrote:
> > On 06.05.21 03:04, Peng Fan (OSS) wrote:
> > > From: Lucas Stach <l.stach@pengutronix.de>
> > >
> > > For some domains the resets of the devices in the domain are not
> > > automatically triggered. Add an optional resets property to allow
> > > the GPC driver to trigger those resets explicitly.
> > >
> > > The resets belong to devices located inside the power domain, which
> > > need to be held in reset across the power-up sequence. So we have no
> > > means to specify what each reset is in a generic power-domain
> > > binding. Same situation as with the clocks in this binding actually.
> >
> > My understanding was that Rob wanted this explanation to be contained in
> the binding docs itself and not only in the commit message, but I might be
> wrong.
> 
> Yes, that would be better.

Sure, I will include that in V3.

Thanks,
Peng.

> 
> >
> > >
> > > Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
> > > Signed-off-by: Peng Fan <peng.fan@nxp.com>
> > > ---
> > >  Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml | 7
> > > +++++++
> > >  1 file changed, 7 insertions(+)
> > >
> > > diff --git
> > > a/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml
> > > b/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml
> > > index a96e6dbf1858..4330c73a2c30 100644
> > > --- a/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml
> > > +++ b/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml
> > > @@ -66,6 +66,13 @@ properties:
> > >
> > >            power-supply: true
> > >
> > > +          resets:
> > > +            description: |
> > > +              A number of phandles to resets that need to be
> asserted during
> > > +              power-up sequencing of the domain.
> > > +            minItems: 1
> > > +            maxItems: 4
> > > +
> > >          required:
> > >            - '#power-domain-cells'
> > >            - reg
> > >

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH V2 09/13] soc: imx: gpcv2: add support for optional resets

[Peng Fan (OSS)]

From: Lucas Stach <l.stach@pengutronix.de>

Normally the reset for the devices inside the power domain is
triggered automatically from the PGC in the power-up sequencing,
however on i.MX8MM this doesn't work for the GPU power domains.

Add support for triggering the reset explicitly during the power
up sequencing.

Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
---
 drivers/soc/imx/gpcv2.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/drivers/soc/imx/gpcv2.c b/drivers/soc/imx/gpcv2.c
index 0428f0eddee0..341c2352ca08 100644
--- a/drivers/soc/imx/gpcv2.c
+++ b/drivers/soc/imx/gpcv2.c
@@ -15,6 +15,7 @@
 #include <linux/pm_runtime.h>
 #include <linux/regmap.h>
 #include <linux/regulator/consumer.h>
+#include <linux/reset.h>
 #include <linux/sizes.h>
 #include <dt-bindings/power/imx7-power.h>
 #include <dt-bindings/power/imx8mq-power.h>
@@ -108,6 +109,7 @@ struct imx_pgc_domain {
 	struct generic_pm_domain genpd;
 	struct regmap *regmap;
 	struct regulator *regulator;
+	struct reset_control *reset;
 	struct clk_bulk_data *clks;
 	int num_clks;
 
@@ -163,6 +165,8 @@ static int imx_pgc_power_up(struct generic_pm_domain *genpd)
 		goto out_regulator_disable;
 	}
 
+	reset_control_assert(domain->reset);
+
 	if (domain->bits.pxx) {
 		/* request the domain to power up */
 		regmap_update_bits(domain->regmap, GPC_PU_PGC_SW_PUP_REQ,
@@ -185,6 +189,11 @@ static int imx_pgc_power_up(struct generic_pm_domain *genpd)
 				  GPC_PGC_CTRL_PCR);
 	}
 
+	/* delay for reset to propagate */
+	udelay(5);
+
+	reset_control_deassert(domain->reset);
+
 	/* request the ADB400 to power up */
 	if (domain->bits.hskreq) {
 		regmap_update_bits(domain->regmap, GPC_PU_PWRHSK,
@@ -540,6 +549,11 @@ static int imx_pgc_domain_probe(struct platform_device *pdev)
 		return dev_err_probe(domain->dev, domain->num_clks,
 				     "Failed to get domain's clocks\n");
 
+	domain->reset = devm_reset_control_array_get_optional_exclusive(domain->dev);
+	if (IS_ERR(domain->reset))
+		return dev_err_probe(domain->dev, PTR_ERR(domain->reset),
+				     "Failed to get domain's resets\n");
+
 	pm_runtime_enable(domain->dev);
 
 	if (domain->bits.map)
-- 
2.30.0


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH V2 10/13] dt-bindings: power: add defines for i.MX8MM power domains

[Peng Fan (OSS)]

From: Lucas Stach <l.stach@pengutronix.de>

Adding defines for i.MX8MM GPC power domains.

Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
---
 .../bindings/power/fsl,imx-gpcv2.yaml         |  2 ++
 include/dt-bindings/power/imx8mm-power.h      | 22 +++++++++++++++++++
 2 files changed, 24 insertions(+)
 create mode 100644 include/dt-bindings/power/imx8mm-power.h

diff --git a/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml b/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml
index 4330c73a2c30..d3539569d45f 100644
--- a/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml
+++ b/Documentation/devicetree/bindings/power/fsl,imx-gpcv2.yaml
@@ -26,6 +26,7 @@ properties:
     enum:
       - fsl,imx7d-gpc
       - fsl,imx8mq-gpc
+      - fsl,imx8mm-gpc
 
   reg:
     maxItems: 1
@@ -54,6 +55,7 @@ properties:
               Power domain index. Valid values are defined in
               include/dt-bindings/power/imx7-power.h for fsl,imx7d-gpc and
               include/dt-bindings/power/imx8m-power.h for fsl,imx8mq-gpc
+              include/dt-bindings/power/imx8mm-power.h for fsl,imx8mm-gpc
             maxItems: 1
 
           clocks:
diff --git a/include/dt-bindings/power/imx8mm-power.h b/include/dt-bindings/power/imx8mm-power.h
new file mode 100644
index 000000000000..fc9c2e16aadc
--- /dev/null
+++ b/include/dt-bindings/power/imx8mm-power.h
@@ -0,0 +1,22 @@
+/* SPDX-License-Identifier: (GPL-2.0 OR MIT) */
+/*
+ *  Copyright (C) 2020 Pengutronix, Lucas Stach <kernel@pengutronix.de>
+ */
+
+#ifndef __DT_BINDINGS_IMX8MM_POWER_H__
+#define __DT_BINDINGS_IMX8MM_POWER_H__
+
+#define IMX8MM_POWER_DOMAIN_HSIOMIX	0
+#define IMX8MM_POWER_DOMAIN_PCIE	1
+#define IMX8MM_POWER_DOMAIN_OTG1	2
+#define IMX8MM_POWER_DOMAIN_OTG2	3
+#define IMX8MM_POWER_DOMAIN_GPUMIX	4
+#define IMX8MM_POWER_DOMAIN_GPU		5
+#define IMX8MM_POWER_DOMAIN_VPUMIX	6
+#define IMX8MM_POWER_DOMAIN_VPUG1	7
+#define IMX8MM_POWER_DOMAIN_VPUG2	8
+#define IMX8MM_POWER_DOMAIN_VPUH1	9
+#define IMX8MM_POWER_DOMAIN_DISPMIX	10
+#define IMX8MM_POWER_DOMAIN_MIPI	11
+
+#endif
-- 
2.30.0


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH V2 11/13] soc: imx: gpcv2: add support for i.MX8MM power domains

[Peng Fan (OSS)]

From: Lucas Stach <l.stach@pengutronix.de>

This adds support for the power domains found on i.MX8MM. The 2D and 3D
GPU domains are abstracted as a single domain in the driver, as they can't
be powered up/down individually due to a shared reset.

Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
---
 drivers/soc/imx/gpcv2.c | 168 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 168 insertions(+)

diff --git a/drivers/soc/imx/gpcv2.c b/drivers/soc/imx/gpcv2.c
index 341c2352ca08..b9437d6d82a6 100644
--- a/drivers/soc/imx/gpcv2.c
+++ b/drivers/soc/imx/gpcv2.c
@@ -19,6 +19,7 @@
 #include <linux/sizes.h>
 #include <dt-bindings/power/imx7-power.h>
 #include <dt-bindings/power/imx8mq-power.h>
+#include <dt-bindings/power/imx8mm-power.h>
 
 #define GPC_LPCR_A_CORE_BSC			0x000
 
@@ -44,6 +45,19 @@
 #define IMX8M_PCIE1_A53_DOMAIN			BIT(3)
 #define IMX8M_MIPI_A53_DOMAIN			BIT(2)
 
+#define IMX8MM_VPUH1_A53_DOMAIN			BIT(15)
+#define IMX8MM_VPUG2_A53_DOMAIN			BIT(14)
+#define IMX8MM_VPUG1_A53_DOMAIN			BIT(13)
+#define IMX8MM_DISPMIX_A53_DOMAIN		BIT(12)
+#define IMX8MM_VPUMIX_A53_DOMAIN		BIT(10)
+#define IMX8MM_GPUMIX_A53_DOMAIN		BIT(9)
+#define IMX8MM_GPU_A53_DOMAIN			(BIT(8) | BIT(11))
+#define IMX8MM_DDR1_A53_DOMAIN			BIT(7)
+#define IMX8MM_OTG2_A53_DOMAIN			BIT(5)
+#define IMX8MM_OTG1_A53_DOMAIN			BIT(4)
+#define IMX8MM_PCIE_A53_DOMAIN			BIT(3)
+#define IMX8MM_MIPI_A53_DOMAIN			BIT(2)
+
 #define GPC_PU_PGC_SW_PUP_REQ		0x0f8
 #define GPC_PU_PGC_SW_PDN_REQ		0x104
 
@@ -67,6 +81,19 @@
 #define IMX8M_PCIE1_SW_Pxx_REQ			BIT(1)
 #define IMX8M_MIPI_SW_Pxx_REQ			BIT(0)
 
+#define IMX8MM_VPUH1_SW_Pxx_REQ			BIT(13)
+#define IMX8MM_VPUG2_SW_Pxx_REQ			BIT(12)
+#define IMX8MM_VPUG1_SW_Pxx_REQ			BIT(11)
+#define IMX8MM_DISPMIX_SW_Pxx_REQ		BIT(10)
+#define IMX8MM_VPUMIX_SW_Pxx_REQ		BIT(8)
+#define IMX8MM_GPUMIX_SW_Pxx_REQ		BIT(7)
+#define IMX8MM_GPU_SW_Pxx_REQ			(BIT(6) | BIT(9))
+#define IMX8MM_DDR1_SW_Pxx_REQ			BIT(5)
+#define IMX8MM_OTG2_SW_Pxx_REQ			BIT(3)
+#define IMX8MM_OTG1_SW_Pxx_REQ			BIT(2)
+#define IMX8MM_PCIE_SW_Pxx_REQ			BIT(1)
+#define IMX8MM_MIPI_SW_Pxx_REQ			BIT(0)
+
 #define GPC_M4_PU_PDN_FLG		0x1bc
 
 #define GPC_PU_PWRHSK			0x1fc
@@ -78,6 +105,17 @@
 #define IMX8M_VPU_HSK_PWRDNREQN			BIT(5)
 #define IMX8M_DISP_HSK_PWRDNREQN		BIT(4)
 
+
+#define IMX8MM_GPUMIX_HSK_PWRDNACKN		BIT(29)
+#define IMX8MM_GPU_HSK_PWRDNACKN		(BIT(27) | BIT(28))
+#define IMX8MM_VPUMIX_HSK_PWRDNACKN		BIT(26)
+#define IMX8MM_DISPMIX_HSK_PWRDNACKN		BIT(25)
+#define IMX8MM_HSIO_HSK_PWRDNACKN		(BIT(23) | BIT(24))
+#define IMX8MM_GPUMIX_HSK_PWRDNREQN		BIT(11)
+#define IMX8MM_GPU_HSK_PWRDNREQN		(BIT(9) | BIT(10))
+#define IMX8MM_VPUMIX_HSK_PWRDNREQN		BIT(8)
+#define IMX8MM_DISPMIX_HSK_PWRDNREQN		BIT(7)
+#define IMX8MM_HSIO_HSK_PWRDNREQN		(BIT(5) | BIT(6))
 /*
  * The PGC offset values in Reference Manual
  * (Rev. 1, 01/2018 and the older ones) GPC chapter's
@@ -100,6 +138,20 @@
 #define IMX8M_PGC_MIPI_CSI2		28
 #define IMX8M_PGC_PCIE2			29
 
+#define IMX8MM_PGC_MIPI			16
+#define IMX8MM_PGC_PCIE			17
+#define IMX8MM_PGC_OTG1			18
+#define IMX8MM_PGC_OTG2			19
+#define IMX8MM_PGC_DDR1			21
+#define IMX8MM_PGC_GPU2D		22
+#define IMX8MM_PGC_GPUMIX		23
+#define IMX8MM_PGC_VPUMIX		24
+#define IMX8MM_PGC_GPU3D		25
+#define IMX8MM_PGC_DISPMIX		26
+#define IMX8MM_PGC_VPUG1		27
+#define IMX8MM_PGC_VPUG2		28
+#define IMX8MM_PGC_VPUH1		29
+
 #define GPC_PGC_CTRL(n)			(0x800 + (n) * 0x40)
 #define GPC_PGC_SR(n)			(GPC_PGC_CTRL(n) + 0xc)
 
@@ -527,6 +579,121 @@ static const struct imx_pgc_domain_data imx8m_pgc_domain_data = {
 	.reg_access_table = &imx8m_access_table,
 };
 
+static const struct imx_pgc_domain imx8mm_pgc_domains[] = {
+	[IMX8MM_POWER_DOMAIN_HSIOMIX] = {
+		.genpd = {
+			.name = "hsiomix",
+		},
+		.bits  = {
+			.pxx = 0, /* no power sequence control */
+			.map = 0, /* no power sequence control */
+			.hskreq = IMX8MM_HSIO_HSK_PWRDNREQN,
+			.hskack = IMX8MM_HSIO_HSK_PWRDNACKN,
+		},
+	},
+
+	[IMX8MM_POWER_DOMAIN_PCIE] = {
+		.genpd = {
+			.name = "pcie",
+		},
+		.bits  = {
+			.pxx = IMX8MM_PCIE_SW_Pxx_REQ,
+			.map = IMX8MM_PCIE_A53_DOMAIN,
+		},
+		.pgc   = IMX8MM_PGC_PCIE,
+	},
+
+	[IMX8MM_POWER_DOMAIN_OTG1] = {
+		.genpd = {
+			.name = "usb-otg1",
+		},
+		.bits  = {
+			.pxx = IMX8MM_OTG1_SW_Pxx_REQ,
+			.map = IMX8MM_OTG1_A53_DOMAIN,
+		},
+		.pgc   = IMX8MM_PGC_OTG1,
+	},
+
+	[IMX8MM_POWER_DOMAIN_OTG2] = {
+		.genpd = {
+			.name = "usb-otg2",
+		},
+		.bits  = {
+			.pxx = IMX8MM_OTG2_SW_Pxx_REQ,
+			.map = IMX8MM_OTG2_A53_DOMAIN,
+		},
+		.pgc   = IMX8MM_PGC_OTG2,
+	},
+
+	[IMX8MM_POWER_DOMAIN_GPUMIX] = {
+		.genpd = {
+			.name = "gpumix",
+		},
+		.bits  = {
+			.pxx = IMX8MM_GPUMIX_SW_Pxx_REQ,
+			.map = IMX8MM_GPUMIX_A53_DOMAIN,
+			.hskreq = IMX8MM_GPUMIX_HSK_PWRDNREQN,
+			.hskack = IMX8MM_GPUMIX_HSK_PWRDNACKN,
+		},
+		.pgc   = IMX8MM_PGC_GPUMIX,
+	},
+
+	[IMX8MM_POWER_DOMAIN_GPU] = {
+		.genpd = {
+			.name = "gpu",
+		},
+		.bits  = {
+			.pxx = IMX8MM_GPU_SW_Pxx_REQ,
+			.map = IMX8MM_GPU_A53_DOMAIN,
+			.hskreq = IMX8MM_GPU_HSK_PWRDNREQN,
+			.hskack = IMX8MM_GPU_HSK_PWRDNACKN,
+		},
+		.pgc   = IMX8MM_PGC_GPU2D,
+	},
+};
+
+static const struct regmap_range imx8mm_yes_ranges[] = {
+		regmap_reg_range(GPC_LPCR_A_CORE_BSC,
+				 GPC_PU_PWRHSK),
+		regmap_reg_range(GPC_PGC_CTRL(IMX8MM_PGC_MIPI),
+				 GPC_PGC_SR(IMX8MM_PGC_MIPI)),
+		regmap_reg_range(GPC_PGC_CTRL(IMX8MM_PGC_PCIE),
+				 GPC_PGC_SR(IMX8MM_PGC_PCIE)),
+		regmap_reg_range(GPC_PGC_CTRL(IMX8MM_PGC_OTG1),
+				 GPC_PGC_SR(IMX8MM_PGC_OTG1)),
+		regmap_reg_range(GPC_PGC_CTRL(IMX8MM_PGC_OTG2),
+				 GPC_PGC_SR(IMX8MM_PGC_OTG2)),
+		regmap_reg_range(GPC_PGC_CTRL(IMX8MM_PGC_DDR1),
+				 GPC_PGC_SR(IMX8MM_PGC_DDR1)),
+		regmap_reg_range(GPC_PGC_CTRL(IMX8MM_PGC_GPU2D),
+				 GPC_PGC_SR(IMX8MM_PGC_GPU2D)),
+		regmap_reg_range(GPC_PGC_CTRL(IMX8MM_PGC_GPUMIX),
+				 GPC_PGC_SR(IMX8MM_PGC_GPUMIX)),
+		regmap_reg_range(GPC_PGC_CTRL(IMX8MM_PGC_VPUMIX),
+				 GPC_PGC_SR(IMX8MM_PGC_VPUMIX)),
+		regmap_reg_range(GPC_PGC_CTRL(IMX8MM_PGC_GPU3D),
+				 GPC_PGC_SR(IMX8MM_PGC_GPU3D)),
+		regmap_reg_range(GPC_PGC_CTRL(IMX8MM_PGC_DISPMIX),
+				 GPC_PGC_SR(IMX8MM_PGC_DISPMIX)),
+		regmap_reg_range(GPC_PGC_CTRL(IMX8MM_PGC_VPUG1),
+				 GPC_PGC_SR(IMX8MM_PGC_VPUG1)),
+		regmap_reg_range(GPC_PGC_CTRL(IMX8MM_PGC_VPUG2),
+				 GPC_PGC_SR(IMX8MM_PGC_VPUG2)),
+		regmap_reg_range(GPC_PGC_CTRL(IMX8MM_PGC_VPUH1),
+				 GPC_PGC_SR(IMX8MM_PGC_VPUH1)),
+};
+
+static const struct regmap_access_table imx8mm_access_table = {
+	.yes_ranges	= imx8mm_yes_ranges,
+	.n_yes_ranges	= ARRAY_SIZE(imx8mm_yes_ranges),
+};
+
+static const struct imx_pgc_domain_data imx8mm_pgc_domain_data = {
+	.domains = imx8mm_pgc_domains,
+	.domains_num = ARRAY_SIZE(imx8mm_pgc_domains),
+	.reg_access_table = &imx8mm_access_table,
+};
+
 static int imx_pgc_domain_probe(struct platform_device *pdev)
 {
 	struct imx_pgc_domain *domain = pdev->dev.platform_data;
@@ -710,6 +877,7 @@ static int imx_gpcv2_probe(struct platform_device *pdev)
 
 static const struct of_device_id imx_gpcv2_dt_ids[] = {
 	{ .compatible = "fsl,imx7d-gpc", .data = &imx7_pgc_domain_data, },
+	{ .compatible = "fsl,imx8mm-gpc", .data = &imx8mm_pgc_domain_data, },
 	{ .compatible = "fsl,imx8mq-gpc", .data = &imx8m_pgc_domain_data, },
 	{ }
 };
-- 
2.30.0


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH V2 12/13] soc: imx: gpcv2: Add support for missing i.MX8MM VPU/DISPMIX power domains

[Peng Fan (OSS)]

From: Lucas Stach <l.stach@pengutronix.de>

With the BLK-CTL driver now in place, let's add the missing domains.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Signed-off-by: Peng Fan <peng.fan@nxp.com>
---
 drivers/soc/imx/gpcv2.c | 70 +++++++++++++++++++++++++++++++++++++++++
 1 file changed, 70 insertions(+)

diff --git a/drivers/soc/imx/gpcv2.c b/drivers/soc/imx/gpcv2.c
index b9437d6d82a6..65b562cbcc6d 100644
--- a/drivers/soc/imx/gpcv2.c
+++ b/drivers/soc/imx/gpcv2.c
@@ -650,6 +650,76 @@ static const struct imx_pgc_domain imx8mm_pgc_domains[] = {
 		},
 		.pgc   = IMX8MM_PGC_GPU2D,
 	},
+
+	[IMX8MM_POWER_DOMAIN_VPUMIX] = {
+		.genpd = {
+			.name = "vpumix",
+		},
+		.bits  = {
+			.pxx = IMX8MM_VPUMIX_SW_Pxx_REQ,
+			.map = IMX8MM_VPUMIX_A53_DOMAIN,
+			.hskreq = IMX8MM_VPUMIX_HSK_PWRDNREQN,
+			.hskack = IMX8MM_VPUMIX_HSK_PWRDNACKN,
+		},
+		.pgc   = IMX8MM_PGC_VPUMIX,
+	},
+
+	[IMX8MM_POWER_DOMAIN_VPUG1] = {
+		.genpd = {
+			.name = "vpu-g1",
+		},
+		.bits  = {
+			.pxx = IMX8MM_VPUG1_SW_Pxx_REQ,
+			.map = IMX8MM_VPUG1_A53_DOMAIN,
+		},
+		.pgc   = IMX8MM_PGC_VPUG1,
+	},
+
+	[IMX8MM_POWER_DOMAIN_VPUG2] = {
+		.genpd = {
+			.name = "vpu-g2",
+		},
+		.bits  = {
+			.pxx = IMX8MM_VPUG2_SW_Pxx_REQ,
+			.map = IMX8MM_VPUG2_A53_DOMAIN,
+		},
+		.pgc   = IMX8MM_PGC_VPUG2,
+	},
+
+	[IMX8MM_POWER_DOMAIN_VPUH1] = {
+		.genpd = {
+			.name = "vpu-h1",
+		},
+		.bits  = {
+			.pxx = IMX8MM_VPUH1_SW_Pxx_REQ,
+			.map = IMX8MM_VPUH1_A53_DOMAIN,
+		},
+		.pgc   = IMX8MM_PGC_VPUH1,
+	},
+
+	[IMX8MM_POWER_DOMAIN_DISPMIX] = {
+		.genpd = {
+			.name = "dispmix",
+		},
+		.bits  = {
+			.pxx = IMX8MM_DISPMIX_SW_Pxx_REQ,
+			.map = IMX8MM_DISPMIX_A53_DOMAIN,
+			.hskreq = IMX8MM_DISPMIX_HSK_PWRDNREQN,
+			.hskack = IMX8MM_DISPMIX_HSK_PWRDNACKN,
+		},
+		.pgc   = IMX8MM_PGC_DISPMIX,
+	},
+
+	[IMX8MM_POWER_DOMAIN_MIPI] = {
+		.genpd = {
+			.name = "mipi",
+		},
+		.bits  = {
+			.pxx = IMX8MM_MIPI_SW_Pxx_REQ,
+			.map = IMX8MM_MIPI_A53_DOMAIN,
+		},
+		.pgc   = IMX8MM_PGC_MIPI,
+	},
 };
 
 static const struct regmap_range imx8mm_yes_ranges[] = {
-- 
2.30.0


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH V2 13/13] soc: imx: gpcv2: move reset assert after requesting domain power up

[Peng Fan (OSS)]

From: Peng Fan <peng.fan@nxp.com>

The i.MX8MM VPU power up sequence is a bit special, it must follow:
1. request power up
2. reset assert
3. reset deassert

This change in this patch will not affect other domains, because
the power domain default is in asserted state, unless bootloader
deassert the reset. It also applies to GPU power domain.

Signed-off-by: Peng Fan <peng.fan@nxp.com>
---
 drivers/soc/imx/gpcv2.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/soc/imx/gpcv2.c b/drivers/soc/imx/gpcv2.c
index 65b562cbcc6d..325a34833ffa 100644
--- a/drivers/soc/imx/gpcv2.c
+++ b/drivers/soc/imx/gpcv2.c
@@ -217,8 +217,6 @@ static int imx_pgc_power_up(struct generic_pm_domain *genpd)
 		goto out_regulator_disable;
 	}
 
-	reset_control_assert(domain->reset);
-
 	if (domain->bits.pxx) {
 		/* request the domain to power up */
 		regmap_update_bits(domain->regmap, GPC_PU_PGC_SW_PUP_REQ,
@@ -241,6 +239,8 @@ static int imx_pgc_power_up(struct generic_pm_domain *genpd)
 				  GPC_PGC_CTRL_PCR);
 	}
 
+	reset_control_assert(domain->reset);
+
 	/* delay for reset to propagate */
 	udelay(5);
 
-- 
2.30.0


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH V2 13/13] soc: imx: gpcv2: move reset assert after requesting domain power up

[Frieder Schrempf]

On 06.05.21 03:04, Peng Fan (OSS) wrote:
> From: Peng Fan <peng.fan@nxp.com>
> 
> The i.MX8MM VPU power up sequence is a bit special, it must follow:
> 1. request power up
> 2. reset assert
> 3. reset deassert
> 
> This change in this patch will not affect other domains, because
> the power domain default is in asserted state, unless bootloader
> deassert the reset. It also applies to GPU power domain.
> 
> Signed-off-by: Peng Fan <peng.fan@nxp.com>

I don't really like that there is a dependency on TF-A/bootloader as we never really know what they will do, but from my point of view the approach is ok and it seems to work properly as far as I have tested it (only with GPU).

Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Tested-by: Frieder Schrempf <frieder.schrempf@kontron.de>

> ---
>  drivers/soc/imx/gpcv2.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/soc/imx/gpcv2.c b/drivers/soc/imx/gpcv2.c
> index 65b562cbcc6d..325a34833ffa 100644
> --- a/drivers/soc/imx/gpcv2.c
> +++ b/drivers/soc/imx/gpcv2.c
> @@ -217,8 +217,6 @@ static int imx_pgc_power_up(struct generic_pm_domain *genpd)
>  		goto out_regulator_disable;
>  	}
>  
> -	reset_control_assert(domain->reset);
> -
>  	if (domain->bits.pxx) {
>  		/* request the domain to power up */
>  		regmap_update_bits(domain->regmap, GPC_PU_PGC_SW_PUP_REQ,
> @@ -241,6 +239,8 @@ static int imx_pgc_power_up(struct generic_pm_domain *genpd)
>  				  GPC_PGC_CTRL_PCR);
>  	}
>  
> +	reset_control_assert(domain->reset);
> +
>  	/* delay for reset to propagate */
>  	udelay(5);
>  
> 

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH V2 00/13] soc: imx: gpcv2: support i.MX8MM

[Frieder Schrempf]

On 06.05.21 03:04, Peng Fan (OSS) wrote:
> From: Peng Fan <peng.fan@nxp.com>
> 
> 
> V2:
>  - Add R-b/A-b tag
>  - Merge V1 patch 13 to V2 patch 6
>  - Drop V1 patch 15
>  - Merge V1 patch 16 to V2 patch 5 and add comments in patch 5 to explain
>  details
>  - Add explaination in patch 8 for "why the resets are not defined"
> 
> This patchset is a pick up Lucas's gpcv2 work for i.MX8MM and several
> minor changes from me to make it could work with i.MX BLK-CTL driver.
> 
> Thanks for Lucas's work and suggestion, Frieder Schrempf for collecting
> all the patches, Jacky Bai on help debug issues.

I tested this series together with the BLK CTL patches by using the GPU and the display stack. Everything looks good to me.

Tested-by: Frieder Schrempf <frieder.schrempf@kontron.de> 

> 
> Lucas Stach (12):
>   soc: imx: gpcv2: move to more ideomatic error handling in probe
>   soc: imx: gpcv2: move domain mapping to domain driver probe
>   soc: imx: gpcv2: switch to clk_bulk_* API
>   soc: imx: gpcv2: split power up and power down sequence control
>   soc: imx: gpcv2: wait for ADB400 handshake
>   soc: imx: gpcv2: add runtime PM support for power-domains
>   soc: imx: gpcv2: allow domains without power-sequence control
>   dt-bindings: imx: gpcv2: add support for optional resets
>   soc: imx: gpcv2: add support for optional resets
>   dt-bindings: power: add defines for i.MX8MM power domains
>   soc: imx: gpcv2: add support for i.MX8MM power domains
>   soc: imx: gpcv2: Add support for missing i.MX8MM VPU/DISPMIX power
>     domains
> 
> Peng Fan (1):
>   soc: imx: gpcv2: move reset assert after requesting domain power up
> 
>  .../bindings/power/fsl,imx-gpcv2.yaml         |   9 +
>  drivers/soc/imx/gpcv2.c                       | 542 ++++++++++++++----
>  include/dt-bindings/power/imx8mm-power.h      |  22 +
>  3 files changed, 458 insertions(+), 115 deletions(-)
>  create mode 100644 include/dt-bindings/power/imx8mm-power.h
> 

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH V2 00/13] soc: imx: gpcv2: support i.MX8MM

[Frieder Schrempf]

On 06.05.21 10:32, Frieder Schrempf wrote:
> On 06.05.21 03:04, Peng Fan (OSS) wrote:
>> From: Peng Fan <peng.fan@nxp.com>
>>
>>
>> V2:
>>  - Add R-b/A-b tag
>>  - Merge V1 patch 13 to V2 patch 6
>>  - Drop V1 patch 15
>>  - Merge V1 patch 16 to V2 patch 5 and add comments in patch 5 to explain
>>  details
>>  - Add explaination in patch 8 for "why the resets are not defined"
>>
>> This patchset is a pick up Lucas's gpcv2 work for i.MX8MM and several
>> minor changes from me to make it could work with i.MX BLK-CTL driver.
>>
>> Thanks for Lucas's work and suggestion, Frieder Schrempf for collecting
>> all the patches, Jacky Bai on help debug issues.
> 
> I tested this series together with the BLK CTL patches by using the GPU and the display stack. Everything looks good to me.
> 
> Tested-by: Frieder Schrempf <frieder.schrempf@kontron.de>

So after some more testing on different hardware I stumbled upon the problem that USB autosuspend doesn't work properly anymore.

I have an onboard LTE module that is connected to OTG2 on the i.MX8MM. When using the mainline TF-A (that enables USB power-domains by default) and removing the power-domain control from the kernel, the device comes up after a few seconds and is enumerated on the bus.

Now, when I let the kernel control the power-domains, the device comes up at boot, but isn't enumerated on the USB bus. As soon as I disable autosuspend for the port, it comes up.

Is this something that needs to be fixed on the USB driver side or is something to be considered for the GPCv2 driver?


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH V2 00/13] soc: imx: gpcv2: support i.MX8MM

[Frieder Schrempf]

On 19.05.21 18:09, Frieder Schrempf wrote:
> On 06.05.21 10:32, Frieder Schrempf wrote:
>> On 06.05.21 03:04, Peng Fan (OSS) wrote:
>>> From: Peng Fan <peng.fan@nxp.com>
>>>
>>>
>>> V2:
>>>  - Add R-b/A-b tag
>>>  - Merge V1 patch 13 to V2 patch 6
>>>  - Drop V1 patch 15
>>>  - Merge V1 patch 16 to V2 patch 5 and add comments in patch 5 to explain
>>>  details
>>>  - Add explaination in patch 8 for "why the resets are not defined"
>>>
>>> This patchset is a pick up Lucas's gpcv2 work for i.MX8MM and several
>>> minor changes from me to make it could work with i.MX BLK-CTL driver.
>>>
>>> Thanks for Lucas's work and suggestion, Frieder Schrempf for collecting
>>> all the patches, Jacky Bai on help debug issues.
>>
>> I tested this series together with the BLK CTL patches by using the GPU and the display stack. Everything looks good to me.
>>
>> Tested-by: Frieder Schrempf <frieder.schrempf@kontron.de>
> 
> So after some more testing on different hardware I stumbled upon the problem that USB autosuspend doesn't work properly anymore.
> 
> I have an onboard LTE module that is connected to OTG2 on the i.MX8MM. When using the mainline TF-A (that enables USB power-domains by default) and removing the power-domain control from the kernel, the device comes up after a few seconds and is enumerated on the bus.
> 
> Now, when I let the kernel control the power-domains, the device comes up at boot, but isn't enumerated on the USB bus. As soon as I disable autosuspend for the port, it comes up.
> 
> Is this something that needs to be fixed on the USB driver side or is something to be considered for the GPCv2 driver?

So I think this is something that needs to be covered on the USB driver side. I would expect that a device appearing on the bus should resume the autosuspended bus, but I don't really know much about USB so there might be other things I miss. For now I will disable autosuspend in this case.

A different, probably more severe problem is that I was still able to reliably run into lockups with suspend/resume and the GPU. I thought I had tested this before as it was one of the things that already failed with the previous implementation, but I must have missed something as it still fails with kernel v5.12.1 + v2 of the GPC patches.

This is how I run into the lockup:

echo mem > /sys/power/state  # Sleep
                             # Wake up again
glmark2-es2-drm              # Use the GPU
                             # Device locks up

Peng, is this something you can reproduce?

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH V2 00/13] soc: imx: gpcv2: support i.MX8MM

[Lucas Stach]

Hi Frieder,

Am Donnerstag, dem 20.05.2021 um 17:16 +0200 schrieb Frieder Schrempf:
> On 19.05.21 18:09, Frieder Schrempf wrote:
> > On 06.05.21 10:32, Frieder Schrempf wrote:
> > > On 06.05.21 03:04, Peng Fan (OSS) wrote:
> > > > From: Peng Fan <peng.fan@nxp.com>
> > > > 
> > > > 
> > > > V2:
> > > >  - Add R-b/A-b tag
> > > >  - Merge V1 patch 13 to V2 patch 6
> > > >  - Drop V1 patch 15
> > > >  - Merge V1 patch 16 to V2 patch 5 and add comments in patch 5
> > > > to explain
> > > >  details
> > > >  - Add explaination in patch 8 for "why the resets are not
> > > > defined"
> > > > 
> > > > This patchset is a pick up Lucas's gpcv2 work for i.MX8MM and
> > > > several
> > > > minor changes from me to make it could work with i.MX BLK-CTL
> > > > driver.
> > > > 
> > > > Thanks for Lucas's work and suggestion, Frieder Schrempf for
> > > > collecting
> > > > all the patches, Jacky Bai on help debug issues.
> > > 
> > > I tested this series together with the BLK CTL patches by using
> > > the GPU and the display stack. Everything looks good to me.
> > > 
> > > Tested-by: Frieder Schrempf <frieder.schrempf@kontron.de>
> > 
> > So after some more testing on different hardware I stumbled upon
> > the problem that USB autosuspend doesn't work properly anymore.
> > 
> > I have an onboard LTE module that is connected to OTG2 on the
> > i.MX8MM. When using the mainline TF-A (that enables USB power-
> > domains by default) and removing the power-domain control from the
> > kernel, the device comes up after a few seconds and is enumerated
> > on the bus.
> > 
> > Now, when I let the kernel control the power-domains, the device
> > comes up at boot, but isn't enumerated on the USB bus. As soon as I
> > disable autosuspend for the port, it comes up.
> > 
> > Is this something that needs to be fixed on the USB driver side or
> > is something to be considered for the GPCv2 driver?
> 
> So I think this is something that needs to be covered on the USB
> driver side. I would expect that a device appearing on the bus should
> resume the autosuspended bus, but I don't really know much about USB
> so there might be other things I miss. For now I will disable
> autosuspend in this case.
> 
> A different, probably more severe problem is that I was still able to
> reliably run into lockups with suspend/resume and the GPU. I thought
> I had tested this before as it was one of the things that already
> failed with the previous implementation, but I must have missed
> something as it still fails with kernel v5.12.1 + v2 of the GPC
> patches.
> 
> This is how I run into the lockup:
> 
> echo mem > /sys/power/state  # Sleep
>                              # Wake up again
> glmark2-es2-drm              # Use the GPU
>                              # Device locks up
> 
> Peng, is this something you can reproduce?

I could reproduce this issue on my last GPC+BLK_CTRL series. This was
caused by a bad interaction between our slightly unusual way to control
the nested power domains via runtime PM and the system suspend/resume
code, which lead to some of the power domains not properly coming up
again in the resume path. v2 of my series fixes this issue and the
above sequence works as expected.

Regards,
Lucas



_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH V2 00/13] soc: imx: gpcv2: support i.MX8MM

[Frieder Schrempf]

Hi Lucas,

On 21.07.21 22:51, Lucas Stach wrote:
> Hi Frieder,
> 
> Am Donnerstag, dem 20.05.2021 um 17:16 +0200 schrieb Frieder Schrempf:
>> On 19.05.21 18:09, Frieder Schrempf wrote:
>>> On 06.05.21 10:32, Frieder Schrempf wrote:
>>>> On 06.05.21 03:04, Peng Fan (OSS) wrote:
>>>>> From: Peng Fan <peng.fan@nxp.com>
>>>>>
>>>>>
>>>>> V2:
>>>>>  - Add R-b/A-b tag
>>>>>  - Merge V1 patch 13 to V2 patch 6
>>>>>  - Drop V1 patch 15
>>>>>  - Merge V1 patch 16 to V2 patch 5 and add comments in patch 5
>>>>> to explain
>>>>>  details
>>>>>  - Add explaination in patch 8 for "why the resets are not
>>>>> defined"
>>>>>
>>>>> This patchset is a pick up Lucas's gpcv2 work for i.MX8MM and
>>>>> several
>>>>> minor changes from me to make it could work with i.MX BLK-CTL
>>>>> driver.
>>>>>
>>>>> Thanks for Lucas's work and suggestion, Frieder Schrempf for
>>>>> collecting
>>>>> all the patches, Jacky Bai on help debug issues.
>>>>
>>>> I tested this series together with the BLK CTL patches by using
>>>> the GPU and the display stack. Everything looks good to me.
>>>>
>>>> Tested-by: Frieder Schrempf <frieder.schrempf@kontron.de>
>>>
>>> So after some more testing on different hardware I stumbled upon
>>> the problem that USB autosuspend doesn't work properly anymore.
>>>
>>> I have an onboard LTE module that is connected to OTG2 on the
>>> i.MX8MM. When using the mainline TF-A (that enables USB power-
>>> domains by default) and removing the power-domain control from the
>>> kernel, the device comes up after a few seconds and is enumerated
>>> on the bus.
>>>
>>> Now, when I let the kernel control the power-domains, the device
>>> comes up at boot, but isn't enumerated on the USB bus. As soon as I
>>> disable autosuspend for the port, it comes up.
>>>
>>> Is this something that needs to be fixed on the USB driver side or
>>> is something to be considered for the GPCv2 driver?
>>
>> So I think this is something that needs to be covered on the USB
>> driver side. I would expect that a device appearing on the bus should
>> resume the autosuspended bus, but I don't really know much about USB
>> so there might be other things I miss. For now I will disable
>> autosuspend in this case.
>>
>> A different, probably more severe problem is that I was still able to
>> reliably run into lockups with suspend/resume and the GPU. I thought
>> I had tested this before as it was one of the things that already
>> failed with the previous implementation, but I must have missed
>> something as it still fails with kernel v5.12.1 + v2 of the GPC
>> patches.
>>
>> This is how I run into the lockup:
>>
>> echo mem > /sys/power/state  # Sleep
>>                              # Wake up again
>> glmark2-es2-drm              # Use the GPU
>>                              # Device locks up
>>
>> Peng, is this something you can reproduce?
> 
> I could reproduce this issue on my last GPC+BLK_CTRL series. This was
> caused by a bad interaction between our slightly unusual way to control
> the nested power domains via runtime PM and the system suspend/resume
> code, which lead to some of the power domains not properly coming up
> again in the resume path. v2 of my series fixes this issue and the
> above sequence works as expected.

Glad you could reproduce and fix this issue. Thanks for the effort. I will try to do some tests myself soon.

Thanks
Frieder

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH V2 00/13] soc: imx: gpcv2: support i.MX8MM

[Ezequiel Garcia]

Hi Peng, Lucas,

On Wed, 5 May 2021 at 21:32, Peng Fan (OSS) <peng.fan@oss.nxp.com> wrote:
>
> From: Peng Fan <peng.fan@nxp.com>
>
>
> V2:
>  - Add R-b/A-b tag
>  - Merge V1 patch 13 to V2 patch 6
>  - Drop V1 patch 15
>  - Merge V1 patch 16 to V2 patch 5 and add comments in patch 5 to explain
>  details
>  - Add explaination in patch 8 for "why the resets are not defined"
>
> This patchset is a pick up Lucas's gpcv2 work for i.MX8MM and several
> minor changes from me to make it could work with i.MX BLK-CTL driver.
>
> Thanks for Lucas's work and suggestion, Frieder Schrempf for collecting
> all the patches, Jacky Bai on help debug issues.
>
> Lucas Stach (12):
>   soc: imx: gpcv2: move to more ideomatic error handling in probe
>   soc: imx: gpcv2: move domain mapping to domain driver probe
>   soc: imx: gpcv2: switch to clk_bulk_* API
>   soc: imx: gpcv2: split power up and power down sequence control
>   soc: imx: gpcv2: wait for ADB400 handshake
>   soc: imx: gpcv2: add runtime PM support for power-domains
>   soc: imx: gpcv2: allow domains without power-sequence control
>   dt-bindings: imx: gpcv2: add support for optional resets
>   soc: imx: gpcv2: add support for optional resets
>   dt-bindings: power: add defines for i.MX8MM power domains
>   soc: imx: gpcv2: add support for i.MX8MM power domains
>   soc: imx: gpcv2: Add support for missing i.MX8MM VPU/DISPMIX power
>     domains
>

It's nice to see this finally moving forward!

As you know, Hantro G2 support for i.MX8MQ (and i.MX8MP, i.MX8MM) is currently
blocked, as you have requested:

https://lore.kernel.org/driverdev-devel/5aa5700b862234895a7a6eb251ca3c80fdc1a6d3.camel@collabora.com/

So, I think we really need to include i.MX8MP and i.MX8MQ on this series.
It's been quite a while and we really need to have that. To be honest,
I fear that
if we merge this series as-is, MX8MP and MX8MP support will fall in
the oblivion,
and Hantro G2 VPU will remain unusable.

We are planning to submit Hantro G2 VP9 support soon, and we've been testing
on various platforms, but it will also be blocked by lack of power-domains.

In the future, please Cc the linux-media mailing list, as well as
Benjamin, Andrzej and myself, so we can follow this.

Thanks a lot for the hard work!
Ezequiel

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH V2 00/13] soc: imx: gpcv2: support i.MX8MM

[Lucas Stach]

Hi Ezequiel,

Am Mittwoch, dem 04.08.2021 um 11:30 -0300 schrieb Ezequiel Garcia:
> Hi Peng, Lucas,
> 
> On Wed, 5 May 2021 at 21:32, Peng Fan (OSS) <peng.fan@oss.nxp.com> wrote:
> > 
> > From: Peng Fan <peng.fan@nxp.com>
> > 
> > 
> > V2:
> >  - Add R-b/A-b tag
> >  - Merge V1 patch 13 to V2 patch 6
> >  - Drop V1 patch 15
> >  - Merge V1 patch 16 to V2 patch 5 and add comments in patch 5 to explain
> >  details
> >  - Add explaination in patch 8 for "why the resets are not defined"
> > 
> > This patchset is a pick up Lucas's gpcv2 work for i.MX8MM and several
> > minor changes from me to make it could work with i.MX BLK-CTL driver.
> > 
> > Thanks for Lucas's work and suggestion, Frieder Schrempf for collecting
> > all the patches, Jacky Bai on help debug issues.
> > 
> > Lucas Stach (12):
> >   soc: imx: gpcv2: move to more ideomatic error handling in probe
> >   soc: imx: gpcv2: move domain mapping to domain driver probe
> >   soc: imx: gpcv2: switch to clk_bulk_* API
> >   soc: imx: gpcv2: split power up and power down sequence control
> >   soc: imx: gpcv2: wait for ADB400 handshake
> >   soc: imx: gpcv2: add runtime PM support for power-domains
> >   soc: imx: gpcv2: allow domains without power-sequence control
> >   dt-bindings: imx: gpcv2: add support for optional resets
> >   soc: imx: gpcv2: add support for optional resets
> >   dt-bindings: power: add defines for i.MX8MM power domains
> >   soc: imx: gpcv2: add support for i.MX8MM power domains
> >   soc: imx: gpcv2: Add support for missing i.MX8MM VPU/DISPMIX power
> >     domains
> > 
> 
> It's nice to see this finally moving forward!
> 
> As you know, Hantro G2 support for i.MX8MQ (and i.MX8MP, i.MX8MM) is currently
> blocked, as you have requested:
> 
> https://lore.kernel.org/driverdev-devel/5aa5700b862234895a7a6eb251ca3c80fdc1a6d3.camel@collabora.com/
> 
> So, I think we really need to include i.MX8MP and i.MX8MQ on this series.
> It's been quite a while and we really need to have that. To be honest,
> I fear that
> if we merge this series as-is, MX8MP and MX8MP support will fall in
> the oblivion,
> and Hantro G2 VPU will remain unusable.
> 
> We are planning to submit Hantro G2 VP9 support soon, and we've been testing
> on various platforms, but it will also be blocked by lack of power-domains.
> 
> In the future, please Cc the linux-media mailing list, as well as
> Benjamin, Andrzej and myself, so we can follow this.

Please take a look at [1], which is the current state of this work. I
intend to add both i.MX8MQ and i.MX8MP support to the series now, as it
seems that there have been no big objections to my approach over the
last 2 weeks, where I was on vacation. ;)

Regards,
Lucas

[1]
https://lore.kernel.org/linux-arm-kernel/20210716232916.3572966-14-l.stach@pengutronix.de/T/#m43cbf6b8615b2a37ff2abb0346e7e7f6594976d1



_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH V2 00/13] soc: imx: gpcv2: support i.MX8MM

[Benjamin Gaignard]

Le 09/08/2021 à 10:15, Lucas Stach a écrit :
> Hi Ezequiel,
>
> Am Mittwoch, dem 04.08.2021 um 11:30 -0300 schrieb Ezequiel Garcia:
>> Hi Peng, Lucas,
>>
>> On Wed, 5 May 2021 at 21:32, Peng Fan (OSS) <peng.fan@oss.nxp.com> wrote:
>>> From: Peng Fan <peng.fan@nxp.com>
>>>
>>>
>>> V2:
>>>   - Add R-b/A-b tag
>>>   - Merge V1 patch 13 to V2 patch 6
>>>   - Drop V1 patch 15
>>>   - Merge V1 patch 16 to V2 patch 5 and add comments in patch 5 to explain
>>>   details
>>>   - Add explaination in patch 8 for "why the resets are not defined"
>>>
>>> This patchset is a pick up Lucas's gpcv2 work for i.MX8MM and several
>>> minor changes from me to make it could work with i.MX BLK-CTL driver.
>>>
>>> Thanks for Lucas's work and suggestion, Frieder Schrempf for collecting
>>> all the patches, Jacky Bai on help debug issues.
>>>
>>> Lucas Stach (12):
>>>    soc: imx: gpcv2: move to more ideomatic error handling in probe
>>>    soc: imx: gpcv2: move domain mapping to domain driver probe
>>>    soc: imx: gpcv2: switch to clk_bulk_* API
>>>    soc: imx: gpcv2: split power up and power down sequence control
>>>    soc: imx: gpcv2: wait for ADB400 handshake
>>>    soc: imx: gpcv2: add runtime PM support for power-domains
>>>    soc: imx: gpcv2: allow domains without power-sequence control
>>>    dt-bindings: imx: gpcv2: add support for optional resets
>>>    soc: imx: gpcv2: add support for optional resets
>>>    dt-bindings: power: add defines for i.MX8MM power domains
>>>    soc: imx: gpcv2: add support for i.MX8MM power domains
>>>    soc: imx: gpcv2: Add support for missing i.MX8MM VPU/DISPMIX power
>>>      domains
>>>
>> It's nice to see this finally moving forward!
>>
>> As you know, Hantro G2 support for i.MX8MQ (and i.MX8MP, i.MX8MM) is currently
>> blocked, as you have requested:
>>
>> https://lore.kernel.org/driverdev-devel/5aa5700b862234895a7a6eb251ca3c80fdc1a6d3.camel@collabora.com/
>>
>> So, I think we really need to include i.MX8MP and i.MX8MQ on this series.
>> It's been quite a while and we really need to have that. To be honest,
>> I fear that
>> if we merge this series as-is, MX8MP and MX8MP support will fall in
>> the oblivion,
>> and Hantro G2 VPU will remain unusable.
>>
>> We are planning to submit Hantro G2 VP9 support soon, and we've been testing
>> on various platforms, but it will also be blocked by lack of power-domains.
>>
>> In the future, please Cc the linux-media mailing list, as well as
>> Benjamin, Andrzej and myself, so we can follow this.
> Please take a look at [1], which is the current state of this work. I
> intend to add both i.MX8MQ and i.MX8MP support to the series now, as it
> seems that there have been no big objections to my approach over the
> last 2 weeks, where I was on vacation. ;)

Hi Lucas,

I have tried to implement the block control driver for imx8mq.
I didn't manage to get it working.
My implementation is here:
https://gitlab.collabora.com/benjamin.gaignard/for-upstream/-/tree/IMX8MQ_BLK_CTRL

While you have the same in IMX8MM do you have also made changes in Hantro driver ?
If it is that can you share these changes ? I have include mine in the above branch.

Regards,
Benjamin

>
> Regards,
> Lucas
>
> [1]
> https://lore.kernel.org/linux-arm-kernel/20210716232916.3572966-14-l.stach@pengutronix.de/T/#m43cbf6b8615b2a37ff2abb0346e7e7f6594976d1
>
>

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel