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(+)
--
Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project
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.
Signed-off-by: Siddharth Gupta <[email protected]>
---
drivers/remoteproc/qcom_q6v5_pas.c | 1 +
drivers/remoteproc/remoteproc_core.c | 4 ++++
2 files changed, 5 insertions(+)
diff --git a/drivers/remoteproc/qcom_q6v5_pas.c b/drivers/remoteproc/qcom_q6v5_pas.c
index 7a63efb..6bb2c7d 100644
--- a/drivers/remoteproc/qcom_q6v5_pas.c
+++ b/drivers/remoteproc/qcom_q6v5_pas.c
@@ -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;
diff --git a/drivers/remoteproc/remoteproc_core.c b/drivers/remoteproc/remoteproc_core.c
index 9f99fe2..19a360d 100644
--- a/drivers/remoteproc/remoteproc_core.c
+++ b/drivers/remoteproc/remoteproc_core.c
@@ -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));
--
Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project
On Wed 08 Apr 15:18 PDT 2020, Siddharth Gupta wrote:
> 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.
> Signed-off-by: Siddharth Gupta <[email protected]>
> ---
> drivers/remoteproc/qcom_q6v5_pas.c | 1 +
> drivers/remoteproc/remoteproc_core.c | 4 ++++
> 2 files changed, 5 insertions(+)
>
> diff --git a/drivers/remoteproc/qcom_q6v5_pas.c b/drivers/remoteproc/qcom_q6v5_pas.c
> index 7a63efb..6bb2c7d 100644
> --- a/drivers/remoteproc/qcom_q6v5_pas.c
> +++ b/drivers/remoteproc/qcom_q6v5_pas.c
> @@ -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.
> adsp->rproc = rproc;
> adsp->pas_id = desc->pas_id;
> adsp->has_aggre2_clk = desc->has_aggre2_clk;
> diff --git a/drivers/remoteproc/remoteproc_core.c b/drivers/remoteproc/remoteproc_core.c
> index 9f99fe2..19a360d 100644
> --- a/drivers/remoteproc/remoteproc_core.c
> +++ b/drivers/remoteproc/remoteproc_core.c
> @@ -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;
> }
>
/* Prevent suspend while the remoteproc is being recovered */
PS. This patch is unrelated to patch 1/2, so please resubmit it
separately.
Regards,
Bjorn
> + pm_stay_awake(rproc->dev.parent);
> +
> dev_err(&rproc->dev, "crash detected in %s: type %s\n",
> rproc->name, rproc_crash_to_string(type));
>
> --
> Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
> a Linux Foundation Collaborative Project
On Wed, Apr 08, 2020 at 03:18:25PM -0700, Siddharth Gupta wrote:
> 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.
>
> Signed-off-by: Siddharth Gupta <[email protected]>
> ---
> drivers/remoteproc/qcom_q6v5_pas.c | 1 +
> drivers/remoteproc/remoteproc_core.c | 4 ++++
> 2 files changed, 5 insertions(+)
>
> diff --git a/drivers/remoteproc/qcom_q6v5_pas.c b/drivers/remoteproc/qcom_q6v5_pas.c
> index 7a63efb..6bb2c7d 100644
> --- a/drivers/remoteproc/qcom_q6v5_pas.c
> +++ b/drivers/remoteproc/qcom_q6v5_pas.c
> @@ -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;
> diff --git a/drivers/remoteproc/remoteproc_core.c b/drivers/remoteproc/remoteproc_core.c
> index 9f99fe2..19a360d 100644
> --- a/drivers/remoteproc/remoteproc_core.c
> +++ b/drivers/remoteproc/remoteproc_core.c
> @@ -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);
> +
Much better:
Acked-by: Mathieu Poirier <[email protected]>
> dev_err(&rproc->dev, "crash detected in %s: type %s\n",
> rproc->name, rproc_crash_to_string(type));
>
> --
> Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
> a Linux Foundation Collaborative Project