Skip to contentSkip to conversation

remoteproc: core: Add core functionality to the remoteproc framework

PATCH In review Siddharth Gupta 3 participants 6 messages Apr 8, 2020 – Apr 16, 2020

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

v2 cover remoteproc: core: Add core functionality to the remoteproc framework

This patch series adds core functionality to the remoteproc framework.

Patch 1 adds a new API to the framework which allows kernel clients to update
the firmware name for the specified remoteproc.
Patch 2 intends to improve the user experience by preventing the system from
going to sleep while the remoteproc is recovering from a crash.

Changes since v1:
- The API in patch 1 has been modified based on comments by Mathieu [1]. Since this patch cannot be merged in without a user this is more of a RFC for merging in later once a user is upstreamed.
- In patch 2 updated the device being used to the parent device of the rproc, based on comments from Mathieu [2] and Bjorn [3].

[1]: https://lore.kernel.org/linux-arm-msm/[email protected]/T/#mc36cf66598238a67d1a3e77ab8362b90bd56882e
[2]: https://lore.kernel.org/linux-arm-msm/[email protected]/T/#mbafae326ff7e0e5dd1ba5370355affe042a4e21f
[3]: https://lore.kernel.org/linux-arm-msm/[email protected]/T/#m4694a60c2c6c720ff6eb3c83008373dfe18ea06b

Siddharth Gupta (2):
remoteproc: core: Add an API for changing firmware name
remoteproc: core: Prevent sleep when rproc crashes

 drivers/remoteproc/qcom_q6v5_pas.c   |  1 +
 drivers/remoteproc/remoteproc_core.c | 47 ++++++++++++++++++++++++++++++++++++
 include/linux/remoteproc.h           |  1 +
 3 files changed, 49 insertions(+)
Signature · 2 lines
Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project
#1 lore raw

PATCH v2 1/2 remoteproc: core: Add an API for changing firmware name

Add an API which allows to change the name of the firmware to be booted on the specified rproc. This change gives us the flixibility to change the firmware at run-time depending on the usecase. Some remoteprocs might use a different firmware for testing, production and development purposes, which may be selected based on the fuse settings during bootup.

---
 drivers/remoteproc/remoteproc_core.c | 43 ++++++++++++++++++++++++++++++++++++
 include/linux/remoteproc.h           |  1 +
 2 files changed, 44 insertions(+)
drivers/remoteproc/remoteproc_core.c +43 -0
index fb9c813..9f99fe2 100644
@@ -1796,6 +1796,49 @@ int rproc_boot(struct rproc *rproc)
 EXPORT_SYMBOL(rproc_boot);
 
 /**
+ * rproc_set_firmware_name() - change the firmware name for specified remoteproc
+ * @rproc: handle of a remote processor
+ * @firmware: name of the firmware to boot with
+ *
+ * Change the name of the firmware to be loaded to @firmware in the rproc
+ * structure. We should ensure that the remoteproc is not running.
+ *
+ * Returns 0 on success, and an appropriate error value otherwise.
+ */
+int rproc_set_firmware_name(struct rproc *rproc, const char *firmware)
+{
+	int len, ret = 0;
+	char *p;
+
+	if (!rproc || !firmware)
+		return -EINVAL;
+
+	len = strcspn(firmware, "\n");
+	if (!len)
+		return -EINVAL;
+
+	mutex_lock(&rproc->lock);
+
+	if (rproc->state != RPROC_OFFLINE) {
+		ret = -EBUSY;
+		goto out;
+	}
+
+	p = kstrndup(firmware, len, GFP_KERNEL);
+	if (!p) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	kfree(rproc->firmware);
+	rproc->firmware = p;
+out:
+	mutex_unlock(&rproc->lock);
+	return ret;
+}
+EXPORT_SYMBOL(rproc_set_firmware_name);
+
+/**
  * rproc_shutdown() - power off the remote processor
  * @rproc: the remote processor
  *
include/linux/remoteproc.h +1 -0
index 9c07d79..c5d36e6 100644
@@ -613,6 +613,7 @@ rproc_of_resm_mem_entry_init(struct device *dev, u32 of_resm_idx, size_t len,
 			     u32 da, const char *name, ...);
 
 int rproc_boot(struct rproc *rproc);
+int rproc_set_firmware_name(struct rproc *rproc, const char *firmware);
 void rproc_shutdown(struct rproc *rproc);
 void rproc_report_crash(struct rproc *rproc, enum rproc_crash_type type);
 int rproc_coredump_add_segment(struct rproc *rproc, dma_addr_t da, size_t size);
Signature · 2 lines
Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project
#3 lore raw

PATCH v2 2/2 remoteproc: core: Prevent sleep when rproc crashes

In reply to Siddharth Gupta

Remoteproc recovery should be fast and any delay will have an impact on the user-experience. Add a wakeup source to remoteproc which ensures that the system does not go into idle state while a remoteproc is recovering, thus prevent any delays that might occur during system resume.

---
 drivers/remoteproc/qcom_q6v5_pas.c   | 1 +
 drivers/remoteproc/remoteproc_core.c | 4 ++++
 2 files changed, 5 insertions(+)
drivers/remoteproc/qcom_q6v5_pas.c +1 -0
index 7a63efb..6bb2c7d 100644
@@ -401,6 +401,7 @@ static int adsp_probe(struct platform_device *pdev)
 
 	adsp = (struct qcom_adsp *)rproc->priv;
 	adsp->dev = &pdev->dev;
+	device_wakeup_enable(adsp->dev);
 	adsp->rproc = rproc;
 	adsp->pas_id = desc->pas_id;
 	adsp->has_aggre2_clk = desc->has_aggre2_clk;
drivers/remoteproc/remoteproc_core.c +4 -0
index 9f99fe2..19a360d 100644
@@ -1729,6 +1729,8 @@ static void rproc_crash_handler_work(struct work_struct *work)
 
 	if (!rproc->recovery_disabled)
 		rproc_trigger_recovery(rproc);
+
+	pm_relax(rproc->dev.parent);
 }
 
 /**
@@ -2273,6 +2275,8 @@ void rproc_report_crash(struct rproc *rproc, enum rproc_crash_type type)
 		return;
 	}
 
+	pm_stay_awake(rproc->dev.parent);
+
 	dev_err(&rproc->dev, "crash detected in %s: type %s\n",
 		rproc->name, rproc_crash_to_string(type));
 
Signature · 2 lines
Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project
#4 lore raw

Re: [PATCH v2 2/2] remoteproc: core: Prevent sleep when rproc crashes

Quoted 5 lines · Siddharth Gupta

Remoteproc recovery should be fast and any delay will have an impact on the user-experience. Add a wakeup source to remoteproc which ensures that the system does not go into idle state while a remoteproc is recovering, thus prevent any delays that might occur during system resume.

This is better, but I think it can be reworded further to show that it's not a matter of making it "fast", it's a matter of preventing suspend from interrupting the recovery of a remoteproc.

Quoted 11 lines
---
 drivers/remoteproc/qcom_q6v5_pas.c   | 1 +
 drivers/remoteproc/remoteproc_core.c | 4 ++++
 2 files changed, 5 insertions(+)
drivers/remoteproc/qcom_q6v5_pas.c +0 -0
index 7a63efb..6bb2c7d 100644

drivers/remoteproc/qcom_q6v5_pas.c static int adsp_probe(struct platform_device *pdev) · line 404

Quoted 4 lines
drivers/remoteproc/qcom_q6v5_pas.c +1 -0
@@ -401,6 +401,7 @@ static int adsp_probe(struct platform_device *pdev)
 
 	adsp = (struct qcom_adsp *)rproc->priv;
 	adsp->dev = &pdev->dev;
+	device_wakeup_enable(adsp->dev);

Move this 5 lines down and give it an empty line before and after.

Quoted 17 lines
drivers/remoteproc/qcom_q6v5_pas.c +0 -0
 	adsp->rproc = rproc;
 	adsp->pas_id = desc->pas_id;
 	adsp->has_aggre2_clk = desc->has_aggre2_clk;
drivers/remoteproc/remoteproc_core.c +2 -0
index 9f99fe2..19a360d 100644
@@ -1729,6 +1729,8 @@ static void rproc_crash_handler_work(struct work_struct *work)
 
 	if (!rproc->recovery_disabled)
 		rproc_trigger_recovery(rproc);
+
+	pm_relax(rproc->dev.parent);
 }
 
 /**

drivers/remoteproc/remoteproc_core.c void rproc_report_crash(struct rproc *rproc, enum rproc_crash_type type) · line 2277

Quoted 3 lines
drivers/remoteproc/remoteproc_core.c +0 -0
@@ -2273,6 +2275,8 @@ void rproc_report_crash(struct rproc *rproc, enum rproc_crash_type type)
 		return;
 	}
 
	/* Prevent suspend while the remoteproc is being recovered */

PS. This patch is unrelated to patch 1/2, so please resubmit it separately.

Regards,
Bjorn

Quoted 8 lines
drivers/remoteproc/remoteproc_core.c +2 -0
+	pm_stay_awake(rproc->dev.parent);
+
 	dev_err(&rproc->dev, "crash detected in %s: type %s\n",
 		rproc->name, rproc_crash_to_string(type));
 
Signature · 2 lines
Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project
maintainer
#2 lore raw

Re: [PATCH v2 1/2] remoteproc: core: Add an API for changing firmware name

In reply to Siddharth Gupta

Hi Siddharth,

Quoted 52 lines · Siddharth Gupta

Add an API which allows to change the name of the firmware to be booted on the specified rproc. This change gives us the flixibility to change the firmware at run-time depending on the usecase. Some remoteprocs might use a different firmware for testing, production and development purposes, which may be selected based on the fuse settings during bootup.

---
 drivers/remoteproc/remoteproc_core.c | 43 ++++++++++++++++++++++++++++++++++++
 include/linux/remoteproc.h           |  1 +
 2 files changed, 44 insertions(+)
drivers/remoteproc/remoteproc_core.c +32 -0
index fb9c813..9f99fe2 100644
@@ -1796,6 +1796,49 @@ int rproc_boot(struct rproc *rproc)
 EXPORT_SYMBOL(rproc_boot);
 
 /**
+ * rproc_set_firmware_name() - change the firmware name for specified remoteproc
+ * @rproc: handle of a remote processor
+ * @firmware: name of the firmware to boot with
+ *
+ * Change the name of the firmware to be loaded to @firmware in the rproc
+ * structure. We should ensure that the remoteproc is not running.
+ *
+ * Returns 0 on success, and an appropriate error value otherwise.
+ */
+int rproc_set_firmware_name(struct rproc *rproc, const char *firmware)
+{
+	int len, ret = 0;
+	char *p;
+
+	if (!rproc || !firmware)
+		return -EINVAL;
+
+	len = strcspn(firmware, "\n");
+	if (!len)
+		return -EINVAL;
+
+	mutex_lock(&rproc->lock);
+
+	if (rproc->state != RPROC_OFFLINE) {
+		ret = -EBUSY;
+		goto out;
+	}
+
+	p = kstrndup(firmware, len, GFP_KERNEL);
+	if (!p) {
+		ret = -ENOMEM;
+		goto out;

drivers/remoteproc/remoteproc_core.c int rproc_boot(struct rproc *rproc) · line 1840

Quoted 10 lines
drivers/remoteproc/remoteproc_core.c +10 -0
+	}
+
+	kfree(rproc->firmware);
+	rproc->firmware = p;
+out:
+	mutex_unlock(&rproc->lock);
+	return ret;
+}
+EXPORT_SYMBOL(rproc_set_firmware_name);
+

This is much better, thanks for cleaning things up. Keep in mind that when you do resend this rproc->firmware will likely have become a "const char *", requiring the use of kstrndup_const()... But that is for a later time.

Mathieu

Quoted 19 lines
drivers/remoteproc/remoteproc_core.c +1 -0
+/**
  * rproc_shutdown() - power off the remote processor
  * @rproc: the remote processor
  *
include/linux/remoteproc.h +1 -0
index 9c07d79..c5d36e6 100644
@@ -613,6 +613,7 @@ rproc_of_resm_mem_entry_init(struct device *dev, u32 of_resm_idx, size_t len,
 			     u32 da, const char *name, ...);
 
 int rproc_boot(struct rproc *rproc);
+int rproc_set_firmware_name(struct rproc *rproc, const char *firmware);
 void rproc_shutdown(struct rproc *rproc);
 void rproc_report_crash(struct rproc *rproc, enum rproc_crash_type type);
 int rproc_coredump_add_segment(struct rproc *rproc, dma_addr_t da, size_t size);
Signature · 2 lines
Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project
maintainer Acked-by on v2 2/2
#5 lore raw

Re: [PATCH v2 2/2] remoteproc: core: Prevent sleep when rproc crashes

In reply to Siddharth Gupta
Quoted 37 lines · Siddharth Gupta

Remoteproc recovery should be fast and any delay will have an impact on the user-experience. Add a wakeup source to remoteproc which ensures that the system does not go into idle state while a remoteproc is recovering, thus prevent any delays that might occur during system resume.

---
 drivers/remoteproc/qcom_q6v5_pas.c   | 1 +
 drivers/remoteproc/remoteproc_core.c | 4 ++++
 2 files changed, 5 insertions(+)
drivers/remoteproc/qcom_q6v5_pas.c +1 -0
index 7a63efb..6bb2c7d 100644
@@ -401,6 +401,7 @@ static int adsp_probe(struct platform_device *pdev)
 
 	adsp = (struct qcom_adsp *)rproc->priv;
 	adsp->dev = &pdev->dev;
+	device_wakeup_enable(adsp->dev);
 	adsp->rproc = rproc;
 	adsp->pas_id = desc->pas_id;
 	adsp->has_aggre2_clk = desc->has_aggre2_clk;
drivers/remoteproc/remoteproc_core.c +2 -0
index 9f99fe2..19a360d 100644
@@ -1729,6 +1729,8 @@ static void rproc_crash_handler_work(struct work_struct *work)
 
 	if (!rproc->recovery_disabled)
 		rproc_trigger_recovery(rproc);
+
+	pm_relax(rproc->dev.parent);
 }
 
 /**

drivers/remoteproc/remoteproc_core.c void rproc_report_crash(struct rproc *rproc, enum rproc_crash_type type) · line 2279

Quoted 5 lines
drivers/remoteproc/remoteproc_core.c +2 -0
@@ -2273,6 +2275,8 @@ void rproc_report_crash(struct rproc *rproc, enum rproc_crash_type type)
 		return;
 	}
 
+	pm_stay_awake(rproc->dev.parent);
+

Much better:

Quoted 6 lines
drivers/remoteproc/remoteproc_core.c +0 -0
 	dev_err(&rproc->dev, "crash detected in %s: type %s\n",
 		rproc->name, rproc_crash_to_string(type));
 
Signature · 2 lines
Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project