Clients/services running on a remoteproc that booted up might need to be
aware of the state of already running remoteprocs. When a remoteproc boots
up (fresh or after recovery) it is not aware of the remoteprocs that booted
before it, i.e., the system state is incomplete. So to keep track of it we
send sysmon on behalf of all 'ONLINE' remoteprocs.
Signed-off-by: Siddharth Gupta <[email protected]>
---
drivers/remoteproc/qcom_sysmon.c | 25 +++++++++++++++++++++++++
1 file changed, 25 insertions(+)
diff --git a/drivers/remoteproc/qcom_sysmon.c b/drivers/remoteproc/qcom_sysmon.c
index 851664e..8d8996d 100644
--- a/drivers/remoteproc/qcom_sysmon.c
+++ b/drivers/remoteproc/qcom_sysmon.c
@@ -453,10 +453,20 @@ static int sysmon_prepare(struct rproc_subdev *subdev)
return 0;
}
+/**
+ * sysmon_start() - start callback for the sysmon remoteproc subdevice
+ * @subdev: instance of the sysmon subdevice
+ *
+ * Inform all the listners of sysmon notifications that the rproc associated
+ * to @subdev has booted up. The rproc that booted up also needs to know
+ * which rprocs are already up and running, so send start notifications
+ * on behalf of all the online rprocs.
+ */
static int sysmon_start(struct rproc_subdev *subdev)
{
struct qcom_sysmon *sysmon = container_of(subdev, struct qcom_sysmon,
subdev);
+ struct qcom_sysmon *target;
struct sysmon_event event = {
.subsys_name = sysmon->name,
.ssr_event = SSCTL_SSR_EVENT_AFTER_POWERUP
@@ -464,6 +474,21 @@ static int sysmon_start(struct rproc_subdev *subdev)
blocking_notifier_call_chain(&sysmon_notifiers, 0, (void *)&event);
+ mutex_lock(&sysmon_lock);
+ list_for_each_entry(target, &sysmon_list, node) {
+ if (target == sysmon ||
+ target->rproc->state != RPROC_RUNNING)
+ continue;
+
+ event.subsys_name = target->name;
+
+ if (sysmon->ssctl_version == 2)
+ ssctl_send_event(sysmon, &event);
+ else if (sysmon->ept)
+ sysmon_send_event(sysmon, &event);
+ }
+ mutex_unlock(&sysmon_lock);
+
return 0;
}
--
Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project
On Wed, Apr 08, 2020 at 04:36:40PM -0700, Siddharth Gupta wrote:
> Clients/services running on a remoteproc that booted up might need to be
> aware of the state of already running remoteprocs. When a remoteproc boots
> up (fresh or after recovery) it is not aware of the remoteprocs that booted
> before it, i.e., the system state is incomplete. So to keep track of it we
> send sysmon on behalf of all 'ONLINE' remoteprocs.
>
> Signed-off-by: Siddharth Gupta <[email protected]>
> ---
> drivers/remoteproc/qcom_sysmon.c | 25 +++++++++++++++++++++++++
> 1 file changed, 25 insertions(+)
>
> diff --git a/drivers/remoteproc/qcom_sysmon.c b/drivers/remoteproc/qcom_sysmon.c
> index 851664e..8d8996d 100644
> --- a/drivers/remoteproc/qcom_sysmon.c
> +++ b/drivers/remoteproc/qcom_sysmon.c
> @@ -453,10 +453,20 @@ static int sysmon_prepare(struct rproc_subdev *subdev)
> return 0;
> }
>
> +/**
> + * sysmon_start() - start callback for the sysmon remoteproc subdevice
> + * @subdev: instance of the sysmon subdevice
> + *
> + * Inform all the listners of sysmon notifications that the rproc associated
> + * to @subdev has booted up. The rproc that booted up also needs to know
> + * which rprocs are already up and running, so send start notifications
> + * on behalf of all the online rprocs.
> + */
> static int sysmon_start(struct rproc_subdev *subdev)
> {
> struct qcom_sysmon *sysmon = container_of(subdev, struct qcom_sysmon,
> subdev);
> + struct qcom_sysmon *target;
> struct sysmon_event event = {
> .subsys_name = sysmon->name,
> .ssr_event = SSCTL_SSR_EVENT_AFTER_POWERUP
> @@ -464,6 +474,21 @@ static int sysmon_start(struct rproc_subdev *subdev)
>
> blocking_notifier_call_chain(&sysmon_notifiers, 0, (void *)&event);
>
> + mutex_lock(&sysmon_lock);
> + list_for_each_entry(target, &sysmon_list, node) {
> + if (target == sysmon ||
> + target->rproc->state != RPROC_RUNNING)
> + continue;
> +
> + event.subsys_name = target->name;
> +
> + if (sysmon->ssctl_version == 2)
> + ssctl_send_event(sysmon, &event);
> + else if (sysmon->ept)
> + sysmon_send_event(sysmon, &event);
> + }
> + mutex_unlock(&sysmon_lock);
> +
Acked-by: Mathieu Poirier <[email protected]>
> return 0;
> }
>
> --
> Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
> a Linux Foundation Collaborative Project
On Wed 08 Apr 16:36 PDT 2020, Siddharth Gupta wrote:
> Clients/services running on a remoteproc that booted up might need to be
> aware of the state of already running remoteprocs. When a remoteproc boots
> up (fresh or after recovery) it is not aware of the remoteprocs that booted
> before it, i.e., the system state is incomplete. So to keep track of it we
> send sysmon on behalf of all 'ONLINE' remoteprocs.
>
> Signed-off-by: Siddharth Gupta <[email protected]>
Reviewed-by: Bjorn Andersson <[email protected]>
Regards,
Bjorn
> ---
> drivers/remoteproc/qcom_sysmon.c | 25 +++++++++++++++++++++++++
> 1 file changed, 25 insertions(+)
>
> diff --git a/drivers/remoteproc/qcom_sysmon.c b/drivers/remoteproc/qcom_sysmon.c
> index 851664e..8d8996d 100644
> --- a/drivers/remoteproc/qcom_sysmon.c
> +++ b/drivers/remoteproc/qcom_sysmon.c
> @@ -453,10 +453,20 @@ static int sysmon_prepare(struct rproc_subdev *subdev)
> return 0;
> }
>
> +/**
> + * sysmon_start() - start callback for the sysmon remoteproc subdevice
> + * @subdev: instance of the sysmon subdevice
> + *
> + * Inform all the listners of sysmon notifications that the rproc associated
> + * to @subdev has booted up. The rproc that booted up also needs to know
> + * which rprocs are already up and running, so send start notifications
> + * on behalf of all the online rprocs.
> + */
> static int sysmon_start(struct rproc_subdev *subdev)
> {
> struct qcom_sysmon *sysmon = container_of(subdev, struct qcom_sysmon,
> subdev);
> + struct qcom_sysmon *target;
> struct sysmon_event event = {
> .subsys_name = sysmon->name,
> .ssr_event = SSCTL_SSR_EVENT_AFTER_POWERUP
> @@ -464,6 +474,21 @@ static int sysmon_start(struct rproc_subdev *subdev)
>
> blocking_notifier_call_chain(&sysmon_notifiers, 0, (void *)&event);
>
> + mutex_lock(&sysmon_lock);
> + list_for_each_entry(target, &sysmon_list, node) {
> + if (target == sysmon ||
> + target->rproc->state != RPROC_RUNNING)
> + continue;
> +
> + event.subsys_name = target->name;
> +
> + if (sysmon->ssctl_version == 2)
> + ssctl_send_event(sysmon, &event);
> + else if (sysmon->ept)
> + sysmon_send_event(sysmon, &event);
> + }
> + mutex_unlock(&sysmon_lock);
> +
> return 0;
> }
>
> --
> Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
> a Linux Foundation Collaborative Project