[PATCH V4 0/4] ufs: ufs-qcom: Add support firmware managed platforms

[PATCH V4 0/4] ufs: ufs-qcom: Add support firmware managed platforms

[Ram Kumar Dwivedi]

On Qualcomm automotive SoC SA8255P, platform resource like clocks,
interconnect, resets, regulators and PHY are configured remotely by
firmware.

Logical power domain is used to abstract these resources in firmware
and SCMI power protocol is used to request resource operations by using
runtime PM framework APIs such as pm_runtime_get/put_sync to invoke
power_on/_off calls from kernel respectively.

Changes from V1
1. Updated "dma-coherent" property type from boolean to just true.
2. Switched from QUIC mail ID to OSS mail ID. 
3. Added Acknowledged by tag from Dmitry for patch 1/3 
4. Added Reviewed-by tag from Manivannan for patch 1/3 

Changes from V2
1. Added new patch "scsi: ufs: core Enforce minimum pm level for sysfs
   configuration" to prevent users from selecting unsupported power
   levels via sysfs.
2. Set minimum power level (UFS_PM_LVL_5) for SA8255P platforms.
3. Changed DT example in qcom,sa8255p-ufshc.yaml to use single-cell
   addressing instead of dual-cell addressing.

Changes from V3
1. Removed address-cells and size-cells from example DT in
   qcom,sa8255p-ufshc.yaml.

Ram Kumar Dwivedi (4):
  MAINTAINERS: broaden UFS Qualcomm binding file pattern
  dt-bindings: ufs: Document bindings for SA8255P UFS Host Controller
  scsi: ufs: core Enforce minimum pm level for sysfs configuration
  ufs: ufs-qcom: Add support for firmware-managed resource abstraction

 .../bindings/ufs/qcom,sa8255p-ufshc.yaml      |  56 ++++++
 MAINTAINERS                                   |   2 +-
 drivers/ufs/core/ufs-sysfs.c                  |   2 +-
 drivers/ufs/host/ufs-qcom.c                   | 164 +++++++++++++++++-
 drivers/ufs/host/ufs-qcom.h                   |   1 +
 include/ufs/ufshcd.h                          |   1 +
 6 files changed, 223 insertions(+), 3 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/ufs/qcom,sa8255p-ufshc.yaml

-- 
2.34.1

[PATCH V4 1/4] MAINTAINERS: broaden UFS Qualcomm binding file pattern

[Ram Kumar Dwivedi]

Update the file pattern for UFS Qualcomm devicetree bindings to match
all files under `Documentation/devicetree/bindings/ufs/qcom*` instead
of only `qcom,ufs.yaml`. This ensures maintainers are correctly
notified for any related binding changes.

Acked-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Reviewed-by: Manivannan Sadhasivam <mani@kernel.org>
Signed-off-by: Ram Kumar Dwivedi <ram.dwivedi@oss.qualcomm.com>
---
 MAINTAINERS | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index 0dbf349fc1ed..70c43fc74401 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -26842,7 +26842,7 @@ M:	Manivannan Sadhasivam <mani@kernel.org>
 L:	linux-arm-msm@vger.kernel.org
 L:	linux-scsi@vger.kernel.org
 S:	Maintained
-F:	Documentation/devicetree/bindings/ufs/qcom,ufs.yaml
+F:	Documentation/devicetree/bindings/ufs/qcom*
 F:	drivers/ufs/host/ufs-qcom*
 
 UNIVERSAL FLASH STORAGE HOST CONTROLLER DRIVER RENESAS HOOKS
-- 
2.34.1

[PATCH V4 2/4] dt-bindings: ufs: Document bindings for SA8255P UFS Host Controller

[Ram Kumar Dwivedi]

Document the device tree bindings for UFS host controller on
Qualcomm SA8255P platform which integrates firmware-managed
resources.

The platform firmware implements the SCMI server and manages
resources such as the PHY, clocks, regulators and resets via the
SCMI power protocol. As a result, the OS-visible DT only describes
the controller’s MMIO, interrupt, IOMMU and power-domain interfaces.

The generic "qcom,ufshc" and "jedec,ufs-2.0" compatible strings are
removed from the binding, since this firmware managed design won't
be compatible with the drivers doing full resource management.

Co-developed-by: Anjana Hari <anjana.hari@oss.qualcomm.com>
Signed-off-by: Anjana Hari <anjana.hari@oss.qualcomm.com>
Signed-off-by: Ram Kumar Dwivedi <ram.dwivedi@oss.qualcomm.com>
---
 .../bindings/ufs/qcom,sa8255p-ufshc.yaml      | 56 +++++++++++++++++++
 1 file changed, 56 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ufs/qcom,sa8255p-ufshc.yaml

diff --git a/Documentation/devicetree/bindings/ufs/qcom,sa8255p-ufshc.yaml b/Documentation/devicetree/bindings/ufs/qcom,sa8255p-ufshc.yaml
new file mode 100644
index 000000000000..ea2f746117e5
--- /dev/null
+++ b/Documentation/devicetree/bindings/ufs/qcom,sa8255p-ufshc.yaml
@@ -0,0 +1,56 @@
+# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/ufs/qcom,sa8255p-ufshc.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Qualcomm SA8255P UFS Host Controller
+
+maintainers:
+  - Ram Kumar Dwivedi <ram.dwivedi@oss.qualcomm.com>
+
+properties:
+  compatible:
+    const: qcom,sa8255p-ufshc
+
+  reg:
+    maxItems: 1
+
+  interrupts:
+    maxItems: 1
+
+  iommus:
+    maxItems: 1
+
+  dma-coherent: true
+
+  power-domains:
+    maxItems: 1
+
+required:
+  - compatible
+  - reg
+  - interrupts
+  - power-domains
+  - iommus
+  - dma-coherent
+
+allOf:
+  - $ref: ufs-common.yaml
+
+unevaluatedProperties: false
+
+examples:
+  - |
+    #include <dt-bindings/interrupt-controller/arm-gic.h>
+
+        ufshc@1d84000 {
+            compatible = "qcom,sa8255p-ufshc";
+            reg = <0x01d84000 0x3000>;
+            interrupts = <GIC_SPI 265 IRQ_TYPE_LEVEL_HIGH>;
+            lanes-per-direction = <2>;
+
+            iommus = <&apps_smmu 0x100 0x0>;
+            power-domains = <&scmi3_pd 0>;
+            dma-coherent;
+        };
-- 
2.34.1

[PATCH V4 3/4] scsi: ufs: core Enforce minimum pm level for sysfs configuration

[Ram Kumar Dwivedi]

Some UFS platforms only support a limited subset of power levels.
Currently, the sysfs interface allows users to set any pm level
without validating the minimum supported value. If an unsupported
level is selected, suspend may fail.

Introduce an pm_lvl_min field in the ufs_hba structure and use it
to clamp the pm level requested via sysfs so that only supported
levels are accepted. Platforms that require a minimum pm level
can set this field during probe.

Signed-off-by: Ram Kumar Dwivedi <ram.dwivedi@oss.qualcomm.com>
---
 drivers/ufs/core/ufs-sysfs.c | 2 +-
 include/ufs/ufshcd.h         | 1 +
 2 files changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/ufs/core/ufs-sysfs.c b/drivers/ufs/core/ufs-sysfs.c
index b33f8656edb5..02e5468ad49d 100644
--- a/drivers/ufs/core/ufs-sysfs.c
+++ b/drivers/ufs/core/ufs-sysfs.c
@@ -141,7 +141,7 @@ static inline ssize_t ufs_sysfs_pm_lvl_store(struct device *dev,
 	if (kstrtoul(buf, 0, &value))
 		return -EINVAL;
 
-	if (value >= UFS_PM_LVL_MAX)
+	if (value >= UFS_PM_LVL_MAX || value < hba->pm_lvl_min)
 		return -EINVAL;
 
 	if (ufs_pm_lvl_states[value].dev_state == UFS_DEEPSLEEP_PWR_MODE &&
diff --git a/include/ufs/ufshcd.h b/include/ufs/ufshcd.h
index 19154228780b..ac8697a7355b 100644
--- a/include/ufs/ufshcd.h
+++ b/include/ufs/ufshcd.h
@@ -972,6 +972,7 @@ struct ufs_hba {
 	enum ufs_pm_level rpm_lvl;
 	/* Desired UFS power management level during system PM */
 	enum ufs_pm_level spm_lvl;
+	enum ufs_pm_level pm_lvl_min;
 	int pm_op_in_progress;
 
 	/* Auto-Hibernate Idle Timer register value */
-- 
2.34.1

[PATCH V4 4/4] ufs: ufs-qcom: Add support for firmware-managed resource abstraction

[Ram Kumar Dwivedi]

Add a compatible string for SA8255p platforms where resources such as
PHY, clocks, regulators, and resets are managed by firmware through an
SCMI server. Use the SCMI power protocol to abstract these resources and
invoke power operations via runtime PM APIs (pm_runtime_get/put_sync).

Introduce vendor operations (vops) for SA8255p targets to enable SCMI-
based resource control. In this model, capabilities like clock scaling
and gating are not yet supported; these will be added incrementally.

Co-developed-by: Anjana Hari <anjana.hari@oss.qualcomm.com>
Signed-off-by: Anjana Hari <anjana.hari@oss.qualcomm.com>
Co-developed-by: Shazad Hussain <shazad.hussain@oss.qualcomm.com>
Signed-off-by: Shazad Hussain <shazad.hussain@oss.qualcomm.com>
Signed-off-by: Ram Kumar Dwivedi <ram.dwivedi@oss.qualcomm.com>
---
 drivers/ufs/host/ufs-qcom.c | 164 +++++++++++++++++++++++++++++++++++-
 drivers/ufs/host/ufs-qcom.h |   1 +
 2 files changed, 164 insertions(+), 1 deletion(-)

diff --git a/drivers/ufs/host/ufs-qcom.c b/drivers/ufs/host/ufs-qcom.c
index 8ebee0cc5313..ddca7b344642 100644
--- a/drivers/ufs/host/ufs-qcom.c
+++ b/drivers/ufs/host/ufs-qcom.c
@@ -14,6 +14,7 @@
 #include <linux/of.h>
 #include <linux/phy/phy.h>
 #include <linux/platform_device.h>
+#include <linux/pm_domain.h>
 #include <linux/reset-controller.h>
 #include <linux/time.h>
 #include <linux/unaligned.h>
@@ -619,6 +620,27 @@ static int ufs_qcom_hce_enable_notify(struct ufs_hba *hba,
 	return err;
 }
 
+static int ufs_qcom_fw_managed_hce_enable_notify(struct ufs_hba *hba,
+						 enum ufs_notify_change_status status)
+{
+	struct ufs_qcom_host *host = ufshcd_get_variant(hba);
+
+	switch (status) {
+	case PRE_CHANGE:
+		ufs_qcom_select_unipro_mode(host);
+		break;
+	case POST_CHANGE:
+		ufs_qcom_enable_hw_clk_gating(hba);
+		ufs_qcom_ice_enable(host);
+		break;
+	default:
+		dev_err(hba->dev, "Invalid status %d\n", status);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
 /**
  * ufs_qcom_cfg_timers - Configure ufs qcom cfg timers
  *
@@ -789,6 +811,38 @@ static int ufs_qcom_resume(struct ufs_hba *hba, enum ufs_pm_op pm_op)
 	return ufs_qcom_ice_resume(host);
 }
 
+static int ufs_qcom_fw_managed_suspend(struct ufs_hba *hba, enum ufs_pm_op pm_op,
+				       enum ufs_notify_change_status status)
+{
+	struct ufs_qcom_host *host = ufshcd_get_variant(hba);
+
+	if (status == PRE_CHANGE)
+		return 0;
+
+	if (hba->spm_lvl != UFS_PM_LVL_5) {
+		dev_err(hba->dev, "Unsupported spm level %d\n", hba->spm_lvl);
+		return -EINVAL;
+	}
+
+	pm_runtime_put_sync(hba->dev);
+
+	return ufs_qcom_ice_suspend(host);
+}
+
+static int ufs_qcom_fw_managed_resume(struct ufs_hba *hba, enum ufs_pm_op pm_op)
+{
+	struct ufs_qcom_host *host = ufshcd_get_variant(hba);
+	int err;
+
+	err = pm_runtime_resume_and_get(hba->dev);
+	if (err) {
+		dev_err(hba->dev, "PM runtime resume failed: %d\n", err);
+		return err;
+	}
+
+	return ufs_qcom_ice_resume(host);
+}
+
 static void ufs_qcom_dev_ref_clk_ctrl(struct ufs_qcom_host *host, bool enable)
 {
 	if (host->dev_ref_clk_ctrl_mmio &&
@@ -1421,6 +1475,55 @@ static void ufs_qcom_exit(struct ufs_hba *hba)
 	phy_exit(host->generic_phy);
 }
 
+static int ufs_qcom_fw_managed_init(struct ufs_hba *hba)
+{
+	struct device *dev = hba->dev;
+	struct ufs_qcom_host *host;
+	int err;
+
+	host = devm_kzalloc(dev, sizeof(*host), GFP_KERNEL);
+	if (!host)
+		return -ENOMEM;
+
+	host->hba = hba;
+	ufshcd_set_variant(hba, host);
+
+	ufs_qcom_get_controller_revision(hba, &host->hw_ver.major,
+					 &host->hw_ver.minor, &host->hw_ver.step);
+
+	err = ufs_qcom_ice_init(host);
+	if (err)
+		goto out_variant_clear;
+
+	ufs_qcom_get_default_testbus_cfg(host);
+	err = ufs_qcom_testbus_config(host);
+	if (err)
+		/* Failure is non-fatal */
+		dev_warn(dev, "Failed to configure the testbus %d\n", err);
+
+	hba->caps |= UFSHCD_CAP_WB_EN;
+
+	ufs_qcom_advertise_quirks(hba);
+	host->hba->quirks &= ~UFSHCD_QUIRK_REINIT_AFTER_MAX_GEAR_SWITCH;
+
+	hba->pm_lvl_min = UFS_PM_LVL_5;
+	hba->spm_lvl = hba->rpm_lvl = hba->pm_lvl_min;
+
+	ufs_qcom_set_host_params(hba);
+	ufs_qcom_parse_gear_limits(hba);
+
+	return 0;
+
+out_variant_clear:
+	ufshcd_set_variant(hba, NULL);
+	return err;
+}
+
+static void ufs_qcom_fw_managed_exit(struct ufs_hba *hba)
+{
+	pm_runtime_put_sync(hba->dev);
+}
+
 /**
  * ufs_qcom_set_clk_40ns_cycles - Configure 40ns clk cycles
  *
@@ -1950,6 +2053,39 @@ static int ufs_qcom_device_reset(struct ufs_hba *hba)
 	return 0;
 }
 
+/**
+ * ufs_qcom_fw_managed_device_reset - Reset UFS device under FW-managed design
+ * @hba: pointer to UFS host bus adapter
+ *
+ * In the firmware-managed reset model, cold boot power-on is handled
+ * automatically by the PM domain framework during SCMI protocol init,
+ * before ufshcd_device_reset() is reached. For subsequent resets
+ * (such as suspend/resume or recovery), the UFS driver must explicitly
+ * invoke PM runtime operations to reset the subsystem.
+ *
+ * Return: 0 on success or a negative error code on failure.
+ */
+static int ufs_qcom_fw_managed_device_reset(struct ufs_hba *hba)
+{
+	static bool is_boot = true;
+	int err;
+
+	/* Skip reset on cold boot; perform it on subsequent calls */
+	if (is_boot) {
+		is_boot = false;
+		return 0;
+	}
+
+	pm_runtime_put_sync(hba->dev);
+	err = pm_runtime_resume_and_get(hba->dev);
+	if (err < 0) {
+		dev_err(hba->dev, "PM runtime resume failed: %d\n", err);
+		return err;
+	}
+
+	return 0;
+}
+
 static void ufs_qcom_config_scaling_param(struct ufs_hba *hba,
 					struct devfreq_dev_profile *p,
 					struct devfreq_simple_ondemand_data *d)
@@ -2229,6 +2365,20 @@ static const struct ufs_hba_variant_ops ufs_hba_qcom_vops = {
 	.freq_to_gear_speed	= ufs_qcom_freq_to_gear_speed,
 };
 
+static const struct ufs_hba_variant_ops ufs_hba_qcom_sa8255p_vops = {
+	.name                   = "qcom-sa8255p",
+	.init                   = ufs_qcom_fw_managed_init,
+	.exit                   = ufs_qcom_fw_managed_exit,
+	.hce_enable_notify      = ufs_qcom_fw_managed_hce_enable_notify,
+	.pwr_change_notify      = ufs_qcom_pwr_change_notify,
+	.apply_dev_quirks       = ufs_qcom_apply_dev_quirks,
+	.fixup_dev_quirks       = ufs_qcom_fixup_dev_quirks,
+	.suspend                = ufs_qcom_fw_managed_suspend,
+	.resume                 = ufs_qcom_fw_managed_resume,
+	.dbg_register_dump      = ufs_qcom_dump_dbg_regs,
+	.device_reset           = ufs_qcom_fw_managed_device_reset,
+};
+
 /**
  * ufs_qcom_probe - probe routine of the driver
  * @pdev: pointer to Platform device handle
@@ -2239,9 +2389,16 @@ static int ufs_qcom_probe(struct platform_device *pdev)
 {
 	int err;
 	struct device *dev = &pdev->dev;
+	const struct ufs_hba_variant_ops *vops;
+	const struct ufs_qcom_drvdata *drvdata = device_get_match_data(dev);
+
+	if (drvdata && drvdata->vops)
+		vops = drvdata->vops;
+	else
+		vops = &ufs_hba_qcom_vops;
 
 	/* Perform generic probe */
-	err = ufshcd_pltfrm_init(pdev, &ufs_hba_qcom_vops);
+	err = ufshcd_pltfrm_init(pdev, vops);
 	if (err)
 		return dev_err_probe(dev, err, "ufshcd_pltfrm_init() failed\n");
 
@@ -2269,10 +2426,15 @@ static const struct ufs_qcom_drvdata ufs_qcom_sm8550_drvdata = {
 	.no_phy_retention = true,
 };
 
+static const struct ufs_qcom_drvdata ufs_qcom_sa8255p_drvdata = {
+	.vops = &ufs_hba_qcom_sa8255p_vops
+};
+
 static const struct of_device_id ufs_qcom_of_match[] __maybe_unused = {
 	{ .compatible = "qcom,ufshc" },
 	{ .compatible = "qcom,sm8550-ufshc", .data = &ufs_qcom_sm8550_drvdata },
 	{ .compatible = "qcom,sm8650-ufshc", .data = &ufs_qcom_sm8550_drvdata },
+	{ .compatible = "qcom,sa8255p-ufshc", .data = &ufs_qcom_sa8255p_drvdata },
 	{},
 };
 MODULE_DEVICE_TABLE(of, ufs_qcom_of_match);
diff --git a/drivers/ufs/host/ufs-qcom.h b/drivers/ufs/host/ufs-qcom.h
index 380d02333d38..1111ab34da01 100644
--- a/drivers/ufs/host/ufs-qcom.h
+++ b/drivers/ufs/host/ufs-qcom.h
@@ -313,6 +313,7 @@ struct ufs_qcom_host {
 struct ufs_qcom_drvdata {
 	enum ufshcd_quirks quirks;
 	bool no_phy_retention;
+	const struct ufs_hba_variant_ops *vops;
 };
 
 static inline u32
-- 
2.34.1

Re: [PATCH V4 3/4] scsi: ufs: core Enforce minimum pm level for sysfs configuration

[Bart Van Assche]

On 1/6/26 5:40 AM, Ram Kumar Dwivedi wrote:
> diff --git a/include/ufs/ufshcd.h b/include/ufs/ufshcd.h
> index 19154228780b..ac8697a7355b 100644
> --- a/include/ufs/ufshcd.h
> +++ b/include/ufs/ufshcd.h
> @@ -972,6 +972,7 @@ struct ufs_hba {
>   	enum ufs_pm_level rpm_lvl;
>   	/* Desired UFS power management level during system PM */
>   	enum ufs_pm_level spm_lvl;
> +	enum ufs_pm_level pm_lvl_min;
>   	int pm_op_in_progress;
>   
>   	/* Auto-Hibernate Idle Timer register value */

Please do not introduce new kernel-doc warnings and update the 
documentation block above this data structure when adding new members.

Thanks,

Bart.

Re: [PATCH V4 3/4] scsi: ufs: core Enforce minimum pm level for sysfs configuration

[Shawn Lin]

在 2026/01/06 星期二 21:40, Ram Kumar Dwivedi 写道:
> Some UFS platforms only support a limited subset of power levels.
> Currently, the sysfs interface allows users to set any pm level
> without validating the minimum supported value. If an unsupported
> level is selected, suspend may fail.
> 
> Introduce an pm_lvl_min field in the ufs_hba structure and use it
> to clamp the pm level requested via sysfs so that only supported
> levels are accepted. Platforms that require a minimum pm level
> can set this field during probe.
> 
> Signed-off-by: Ram Kumar Dwivedi <ram.dwivedi@oss.qualcomm.com>
> ---
>   drivers/ufs/core/ufs-sysfs.c | 2 +-
>   include/ufs/ufshcd.h         | 1 +
>   2 files changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/ufs/core/ufs-sysfs.c b/drivers/ufs/core/ufs-sysfs.c
> index b33f8656edb5..02e5468ad49d 100644
> --- a/drivers/ufs/core/ufs-sysfs.c
> +++ b/drivers/ufs/core/ufs-sysfs.c
> @@ -141,7 +141,7 @@ static inline ssize_t ufs_sysfs_pm_lvl_store(struct device *dev,
>   	if (kstrtoul(buf, 0, &value))
>   		return -EINVAL;
>   
> -	if (value >= UFS_PM_LVL_MAX)
> +	if (value >= UFS_PM_LVL_MAX || value < hba->pm_lvl_min)

It makes sense that some platform support a limited subset of power
levels. But each level is in increasing order of power savings, and you
set it to UFS_PM_LVL_5. Don't you support UFS_PM_LVL_0 the full active
mode?

>   		return -EINVAL;
>   
>   	if (ufs_pm_lvl_states[value].dev_state == UFS_DEEPSLEEP_PWR_MODE &&
> diff --git a/include/ufs/ufshcd.h b/include/ufs/ufshcd.h
> index 19154228780b..ac8697a7355b 100644
> --- a/include/ufs/ufshcd.h
> +++ b/include/ufs/ufshcd.h
> @@ -972,6 +972,7 @@ struct ufs_hba {
>   	enum ufs_pm_level rpm_lvl;
>   	/* Desired UFS power management level during system PM */
>   	enum ufs_pm_level spm_lvl;
> +	enum ufs_pm_level pm_lvl_min;
>   	int pm_op_in_progress;
>   
>   	/* Auto-Hibernate Idle Timer register value */

Re: [PATCH V4 2/4] dt-bindings: ufs: Document bindings for SA8255P UFS Host Controller

[Krzysztof Kozlowski]

On Tue, Jan 06, 2026 at 07:10:06PM +0530, Ram Kumar Dwivedi wrote:
> +unevaluatedProperties: false
> +
> +examples:
> +  - |
> +    #include <dt-bindings/interrupt-controller/arm-gic.h>
> +
> +        ufshc@1d84000 {

Completely messed indentation.

Why did we ask to drop the unnecessary soc node? To make it simpler.

Even if you do not believe that code should be simpler, you should
immediately spot the odd indentation.

For such trivialities you cannot get this patch merged. It's third
revision which you send carelessly.

Best regards,
Krzysztof

Re: [PATCH V4 3/4] scsi: ufs: core Enforce minimum pm level for sysfs configuration

[Manivannan Sadhasivam]

On Tue, Jan 06, 2026 at 10:27:29PM +0800, Shawn Lin wrote:
> 在 2026/01/06 星期二 21:40, Ram Kumar Dwivedi 写道:
> > Some UFS platforms only support a limited subset of power levels.
> > Currently, the sysfs interface allows users to set any pm level
> > without validating the minimum supported value. If an unsupported
> > level is selected, suspend may fail.
> > 
> > Introduce an pm_lvl_min field in the ufs_hba structure and use it
> > to clamp the pm level requested via sysfs so that only supported
> > levels are accepted. Platforms that require a minimum pm level
> > can set this field during probe.
> > 
> > Signed-off-by: Ram Kumar Dwivedi <ram.dwivedi@oss.qualcomm.com>
> > ---
> >   drivers/ufs/core/ufs-sysfs.c | 2 +-
> >   include/ufs/ufshcd.h         | 1 +
> >   2 files changed, 2 insertions(+), 1 deletion(-)
> > 
> > diff --git a/drivers/ufs/core/ufs-sysfs.c b/drivers/ufs/core/ufs-sysfs.c
> > index b33f8656edb5..02e5468ad49d 100644
> > --- a/drivers/ufs/core/ufs-sysfs.c
> > +++ b/drivers/ufs/core/ufs-sysfs.c
> > @@ -141,7 +141,7 @@ static inline ssize_t ufs_sysfs_pm_lvl_store(struct device *dev,
> >   	if (kstrtoul(buf, 0, &value))
> >   		return -EINVAL;
> > -	if (value >= UFS_PM_LVL_MAX)
> > +	if (value >= UFS_PM_LVL_MAX || value < hba->pm_lvl_min)
> 
> It makes sense that some platform support a limited subset of power
> levels. But each level is in increasing order of power savings, and you
> set it to UFS_PM_LVL_5. Don't you support UFS_PM_LVL_0 the full active
> mode?
> 

These are the suspend levels, not runtime levels. So yes, our platform doesn't
support full power mode when it is in suspend state.

- Mani

> >   		return -EINVAL;
> >   	if (ufs_pm_lvl_states[value].dev_state == UFS_DEEPSLEEP_PWR_MODE &&
> > diff --git a/include/ufs/ufshcd.h b/include/ufs/ufshcd.h
> > index 19154228780b..ac8697a7355b 100644
> > --- a/include/ufs/ufshcd.h
> > +++ b/include/ufs/ufshcd.h
> > @@ -972,6 +972,7 @@ struct ufs_hba {
> >   	enum ufs_pm_level rpm_lvl;
> >   	/* Desired UFS power management level during system PM */
> >   	enum ufs_pm_level spm_lvl;
> > +	enum ufs_pm_level pm_lvl_min;
> >   	int pm_op_in_progress;
> >   	/* Auto-Hibernate Idle Timer register value */
> 

-- 
மணிவண்ணன் சதாசிவம்

Re: [PATCH V4 4/4] ufs: ufs-qcom: Add support for firmware-managed resource abstraction

[Manivannan Sadhasivam]

On Tue, Jan 06, 2026 at 07:10:08PM +0530, Ram Kumar Dwivedi wrote:
> Add a compatible string for SA8255p platforms where resources such as
> PHY, clocks, regulators, and resets are managed by firmware through an
> SCMI server. Use the SCMI power protocol to abstract these resources and
> invoke power operations via runtime PM APIs (pm_runtime_get/put_sync).
> 
> Introduce vendor operations (vops) for SA8255p targets to enable SCMI-
> based resource control. In this model, capabilities like clock scaling
> and gating are not yet supported; these will be added incrementally.
> 
> Co-developed-by: Anjana Hari <anjana.hari@oss.qualcomm.com>
> Signed-off-by: Anjana Hari <anjana.hari@oss.qualcomm.com>
> Co-developed-by: Shazad Hussain <shazad.hussain@oss.qualcomm.com>
> Signed-off-by: Shazad Hussain <shazad.hussain@oss.qualcomm.com>
> Signed-off-by: Ram Kumar Dwivedi <ram.dwivedi@oss.qualcomm.com>
> ---
>  drivers/ufs/host/ufs-qcom.c | 164 +++++++++++++++++++++++++++++++++++-
>  drivers/ufs/host/ufs-qcom.h |   1 +
>  2 files changed, 164 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/ufs/host/ufs-qcom.c b/drivers/ufs/host/ufs-qcom.c
> index 8ebee0cc5313..ddca7b344642 100644
> --- a/drivers/ufs/host/ufs-qcom.c
> +++ b/drivers/ufs/host/ufs-qcom.c
> @@ -14,6 +14,7 @@
>  #include <linux/of.h>
>  #include <linux/phy/phy.h>
>  #include <linux/platform_device.h>
> +#include <linux/pm_domain.h>
>  #include <linux/reset-controller.h>
>  #include <linux/time.h>
>  #include <linux/unaligned.h>
> @@ -619,6 +620,27 @@ static int ufs_qcom_hce_enable_notify(struct ufs_hba *hba,
>  	return err;
>  }
>  
> +static int ufs_qcom_fw_managed_hce_enable_notify(struct ufs_hba *hba,
> +						 enum ufs_notify_change_status status)
> +{
> +	struct ufs_qcom_host *host = ufshcd_get_variant(hba);
> +
> +	switch (status) {
> +	case PRE_CHANGE:
> +		ufs_qcom_select_unipro_mode(host);
> +		break;
> +	case POST_CHANGE:
> +		ufs_qcom_enable_hw_clk_gating(hba);
> +		ufs_qcom_ice_enable(host);
> +		break;
> +	default:
> +		dev_err(hba->dev, "Invalid status %d\n", status);
> +		return -EINVAL;
> +	}
> +
> +	return 0;
> +}
> +
>  /**
>   * ufs_qcom_cfg_timers - Configure ufs qcom cfg timers
>   *
> @@ -789,6 +811,38 @@ static int ufs_qcom_resume(struct ufs_hba *hba, enum ufs_pm_op pm_op)
>  	return ufs_qcom_ice_resume(host);
>  }
>  
> +static int ufs_qcom_fw_managed_suspend(struct ufs_hba *hba, enum ufs_pm_op pm_op,
> +				       enum ufs_notify_change_status status)
> +{
> +	struct ufs_qcom_host *host = ufshcd_get_variant(hba);
> +
> +	if (status == PRE_CHANGE)
> +		return 0;
> +
> +	if (hba->spm_lvl != UFS_PM_LVL_5) {
> +		dev_err(hba->dev, "Unsupported spm level %d\n", hba->spm_lvl);
> +		return -EINVAL;

After the sysfs change, do you still need this check?

> +	}
> +
> +	pm_runtime_put_sync(hba->dev);
> +
> +	return ufs_qcom_ice_suspend(host);
> +}
> +
> +static int ufs_qcom_fw_managed_resume(struct ufs_hba *hba, enum ufs_pm_op pm_op)
> +{
> +	struct ufs_qcom_host *host = ufshcd_get_variant(hba);
> +	int err;
> +
> +	err = pm_runtime_resume_and_get(hba->dev);
> +	if (err) {
> +		dev_err(hba->dev, "PM runtime resume failed: %d\n", err);
> +		return err;
> +	}
> +
> +	return ufs_qcom_ice_resume(host);
> +}
> +
>  static void ufs_qcom_dev_ref_clk_ctrl(struct ufs_qcom_host *host, bool enable)
>  {
>  	if (host->dev_ref_clk_ctrl_mmio &&
> @@ -1421,6 +1475,55 @@ static void ufs_qcom_exit(struct ufs_hba *hba)
>  	phy_exit(host->generic_phy);
>  }
>  
> +static int ufs_qcom_fw_managed_init(struct ufs_hba *hba)
> +{
> +	struct device *dev = hba->dev;
> +	struct ufs_qcom_host *host;
> +	int err;
> +
> +	host = devm_kzalloc(dev, sizeof(*host), GFP_KERNEL);
> +	if (!host)
> +		return -ENOMEM;
> +
> +	host->hba = hba;
> +	ufshcd_set_variant(hba, host);
> +
> +	ufs_qcom_get_controller_revision(hba, &host->hw_ver.major,
> +					 &host->hw_ver.minor, &host->hw_ver.step);
> +
> +	err = ufs_qcom_ice_init(host);
> +	if (err)
> +		goto out_variant_clear;
> +
> +	ufs_qcom_get_default_testbus_cfg(host);
> +	err = ufs_qcom_testbus_config(host);
> +	if (err)
> +		/* Failure is non-fatal */
> +		dev_warn(dev, "Failed to configure the testbus %d\n", err);
> +
> +	hba->caps |= UFSHCD_CAP_WB_EN;
> +
> +	ufs_qcom_advertise_quirks(hba);
> +	host->hba->quirks &= ~UFSHCD_QUIRK_REINIT_AFTER_MAX_GEAR_SWITCH;
> +
> +	hba->pm_lvl_min = UFS_PM_LVL_5;
> +	hba->spm_lvl = hba->rpm_lvl = hba->pm_lvl_min;

hba->spm_lvl = hba->rpm_lvl = hba->pm_lvl_min = UFS_PM_LVL_5; ?

> +
> +	ufs_qcom_set_host_params(hba);
> +	ufs_qcom_parse_gear_limits(hba);
> +
> +	return 0;
> +
> +out_variant_clear:
> +	ufshcd_set_variant(hba, NULL);
> +	return err;
> +}
> +
> +static void ufs_qcom_fw_managed_exit(struct ufs_hba *hba)
> +{
> +	pm_runtime_put_sync(hba->dev);
> +}
> +
>  /**
>   * ufs_qcom_set_clk_40ns_cycles - Configure 40ns clk cycles
>   *
> @@ -1950,6 +2053,39 @@ static int ufs_qcom_device_reset(struct ufs_hba *hba)
>  	return 0;
>  }
>  
> +/**
> + * ufs_qcom_fw_managed_device_reset - Reset UFS device under FW-managed design
> + * @hba: pointer to UFS host bus adapter
> + *
> + * In the firmware-managed reset model, cold boot power-on is handled
> + * automatically by the PM domain framework during SCMI protocol init,

No. genpd handles the power on for the power domain before UFS controller driver
probes. SCMI protocol init has nothing to do with it.

- Mani

-- 
மணிவண்ணன் சதாசிவம்

Re: [PATCH V4 4/4] ufs: ufs-qcom: Add support for firmware-managed resource abstraction

[Ram Kumar Dwivedi]

On 08-Jan-26 10:12 AM, Manivannan Sadhasivam wrote:
> On Tue, Jan 06, 2026 at 07:10:08PM +0530, Ram Kumar Dwivedi wrote:
>> Add a compatible string for SA8255p platforms where resources such as
>> PHY, clocks, regulators, and resets are managed by firmware through an
>> SCMI server. Use the SCMI power protocol to abstract these resources and
>> invoke power operations via runtime PM APIs (pm_runtime_get/put_sync).
>>
>> Introduce vendor operations (vops) for SA8255p targets to enable SCMI-
>> based resource control. In this model, capabilities like clock scaling
>> and gating are not yet supported; these will be added incrementally.
>>
>> Co-developed-by: Anjana Hari <anjana.hari@oss.qualcomm.com>
>> Signed-off-by: Anjana Hari <anjana.hari@oss.qualcomm.com>
>> Co-developed-by: Shazad Hussain <shazad.hussain@oss.qualcomm.com>
>> Signed-off-by: Shazad Hussain <shazad.hussain@oss.qualcomm.com>
>> Signed-off-by: Ram Kumar Dwivedi <ram.dwivedi@oss.qualcomm.com>
>> ---
>>  drivers/ufs/host/ufs-qcom.c | 164 +++++++++++++++++++++++++++++++++++-
>>  drivers/ufs/host/ufs-qcom.h |   1 +
>>  2 files changed, 164 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/ufs/host/ufs-qcom.c b/drivers/ufs/host/ufs-qcom.c
>> index 8ebee0cc5313..ddca7b344642 100644
>> --- a/drivers/ufs/host/ufs-qcom.c
>> +++ b/drivers/ufs/host/ufs-qcom.c
>> @@ -14,6 +14,7 @@
>>  #include <linux/of.h>
>>  #include <linux/phy/phy.h>
>>  #include <linux/platform_device.h>
>> +#include <linux/pm_domain.h>
>>  #include <linux/reset-controller.h>
>>  #include <linux/time.h>
>>  #include <linux/unaligned.h>
>> @@ -619,6 +620,27 @@ static int ufs_qcom_hce_enable_notify(struct ufs_hba *hba,
>>  	return err;
>>  }
>>  
>> +static int ufs_qcom_fw_managed_hce_enable_notify(struct ufs_hba *hba,
>> +						 enum ufs_notify_change_status status)
>> +{
>> +	struct ufs_qcom_host *host = ufshcd_get_variant(hba);
>> +
>> +	switch (status) {
>> +	case PRE_CHANGE:
>> +		ufs_qcom_select_unipro_mode(host);
>> +		break;
>> +	case POST_CHANGE:
>> +		ufs_qcom_enable_hw_clk_gating(hba);
>> +		ufs_qcom_ice_enable(host);
>> +		break;
>> +	default:
>> +		dev_err(hba->dev, "Invalid status %d\n", status);
>> +		return -EINVAL;
>> +	}
>> +
>> +	return 0;
>> +}
>> +
>>  /**
>>   * ufs_qcom_cfg_timers - Configure ufs qcom cfg timers
>>   *
>> @@ -789,6 +811,38 @@ static int ufs_qcom_resume(struct ufs_hba *hba, enum ufs_pm_op pm_op)
>>  	return ufs_qcom_ice_resume(host);
>>  }
>>  
>> +static int ufs_qcom_fw_managed_suspend(struct ufs_hba *hba, enum ufs_pm_op pm_op,
>> +				       enum ufs_notify_change_status status)
>> +{
>> +	struct ufs_qcom_host *host = ufshcd_get_variant(hba);
>> +
>> +	if (status == PRE_CHANGE)
>> +		return 0;
>> +
>> +	if (hba->spm_lvl != UFS_PM_LVL_5) {
>> +		dev_err(hba->dev, "Unsupported spm level %d\n", hba->spm_lvl);
>> +		return -EINVAL;
> 
> After the sysfs change, do you still need this check?

Hi Mani,

I will remove this check in the next version.

Thanks,
Ram.


> 
>> +	}
>> +
>> +	pm_runtime_put_sync(hba->dev);
>> +
>> +	return ufs_qcom_ice_suspend(host);
>> +}
>> +
>> +static int ufs_qcom_fw_managed_resume(struct ufs_hba *hba, enum ufs_pm_op pm_op)
>> +{
>> +	struct ufs_qcom_host *host = ufshcd_get_variant(hba);
>> +	int err;
>> +
>> +	err = pm_runtime_resume_and_get(hba->dev);
>> +	if (err) {
>> +		dev_err(hba->dev, "PM runtime resume failed: %d\n", err);
>> +		return err;
>> +	}
>> +
>> +	return ufs_qcom_ice_resume(host);
>> +}
>> +
>>  static void ufs_qcom_dev_ref_clk_ctrl(struct ufs_qcom_host *host, bool enable)
>>  {
>>  	if (host->dev_ref_clk_ctrl_mmio &&
>> @@ -1421,6 +1475,55 @@ static void ufs_qcom_exit(struct ufs_hba *hba)
>>  	phy_exit(host->generic_phy);
>>  }
>>  
>> +static int ufs_qcom_fw_managed_init(struct ufs_hba *hba)
>> +{
>> +	struct device *dev = hba->dev;
>> +	struct ufs_qcom_host *host;
>> +	int err;
>> +
>> +	host = devm_kzalloc(dev, sizeof(*host), GFP_KERNEL);
>> +	if (!host)
>> +		return -ENOMEM;
>> +
>> +	host->hba = hba;
>> +	ufshcd_set_variant(hba, host);
>> +
>> +	ufs_qcom_get_controller_revision(hba, &host->hw_ver.major,
>> +					 &host->hw_ver.minor, &host->hw_ver.step);
>> +
>> +	err = ufs_qcom_ice_init(host);
>> +	if (err)
>> +		goto out_variant_clear;
>> +
>> +	ufs_qcom_get_default_testbus_cfg(host);
>> +	err = ufs_qcom_testbus_config(host);
>> +	if (err)
>> +		/* Failure is non-fatal */
>> +		dev_warn(dev, "Failed to configure the testbus %d\n", err);
>> +
>> +	hba->caps |= UFSHCD_CAP_WB_EN;
>> +
>> +	ufs_qcom_advertise_quirks(hba);
>> +	host->hba->quirks &= ~UFSHCD_QUIRK_REINIT_AFTER_MAX_GEAR_SWITCH;
>> +
>> +	hba->pm_lvl_min = UFS_PM_LVL_5;
>> +	hba->spm_lvl = hba->rpm_lvl = hba->pm_lvl_min;
> 
> hba->spm_lvl = hba->rpm_lvl = hba->pm_lvl_min = UFS_PM_LVL_5; ?

Hi Mani,

I will update this to a single-line assignment as suggested in the next version.

Thanks,
Ram

> 
>> +
>> +	ufs_qcom_set_host_params(hba);
>> +	ufs_qcom_parse_gear_limits(hba);
>> +
>> +	return 0;
>> +
>> +out_variant_clear:
>> +	ufshcd_set_variant(hba, NULL);
>> +	return err;
>> +}
>> +
>> +static void ufs_qcom_fw_managed_exit(struct ufs_hba *hba)
>> +{
>> +	pm_runtime_put_sync(hba->dev);
>> +}
>> +
>>  /**
>>   * ufs_qcom_set_clk_40ns_cycles - Configure 40ns clk cycles
>>   *
>> @@ -1950,6 +2053,39 @@ static int ufs_qcom_device_reset(struct ufs_hba *hba)
>>  	return 0;
>>  }
>>  
>> +/**
>> + * ufs_qcom_fw_managed_device_reset - Reset UFS device under FW-managed design
>> + * @hba: pointer to UFS host bus adapter
>> + *
>> + * In the firmware-managed reset model, cold boot power-on is handled
>> + * automatically by the PM domain framework during SCMI protocol init,
> 
> No. genpd handles the power on for the power domain before UFS controller driver
> probes. SCMI protocol init has nothing to do with it.
Hi Mani,

I have updated the comment in the next version to clarify that 
genpd handles the power-on.

Thanks,
Ram.


> 
> - Mani
> 

Re: [PATCH V4 2/4] dt-bindings: ufs: Document bindings for SA8255P UFS Host Controller

[Ram Kumar Dwivedi]

On 07-Jan-26 1:08 PM, Krzysztof Kozlowski wrote:
> On Tue, Jan 06, 2026 at 07:10:06PM +0530, Ram Kumar Dwivedi wrote:
>> +unevaluatedProperties: false
>> +
>> +examples:
>> +  - |
>> +    #include <dt-bindings/interrupt-controller/arm-gic.h>
>> +
>> +        ufshc@1d84000 {
> 
> Completely messed indentation.
> 
> Why did we ask to drop the unnecessary soc node? To make it simpler.
> 
> Even if you do not believe that code should be simpler, you should
> immediately spot the odd indentation.
> 
> For such trivialities you cannot get this patch merged. It's third
> revision which you send carelessly.

Hi Krzysztof,

Apologies for the oversight. I missed correcting the indentation after removing the 'soc' node wrapper.
I will ensure the example is correctly indented in the next version.

Thanks,
Ram


> 
> Best regards,
> Krzysztof
> 

Re: [PATCH V4 3/4] scsi: ufs: core Enforce minimum pm level for sysfs configuration

[Ram Kumar Dwivedi]

On 06-Jan-26 7:26 PM, Bart Van Assche wrote:
> On 1/6/26 5:40 AM, Ram Kumar Dwivedi wrote:
>> diff --git a/include/ufs/ufshcd.h b/include/ufs/ufshcd.h
>> index 19154228780b..ac8697a7355b 100644
>> --- a/include/ufs/ufshcd.h
>> +++ b/include/ufs/ufshcd.h
>> @@ -972,6 +972,7 @@ struct ufs_hba {
>>       enum ufs_pm_level rpm_lvl;
>>       /* Desired UFS power management level during system PM */
>>       enum ufs_pm_level spm_lvl;
>> +    enum ufs_pm_level pm_lvl_min;
>>       int pm_op_in_progress;
>>         /* Auto-Hibernate Idle Timer register value */
> 
> Please do not introduce new kernel-doc warnings and update the documentation block above this data structure when adding new members.

Hi Bart,

I will add the kernel-doc description for pm_lvl_min in the next version.

Thanks, 
Ram


> 
> Thanks,
> 
> Bart.