diff options
Diffstat (limited to 'drivers/accel/ivpu/ivpu_pm.c')
| -rw-r--r-- | drivers/accel/ivpu/ivpu_pm.c | 15 |
1 files changed, 12 insertions, 3 deletions
diff --git a/drivers/accel/ivpu/ivpu_pm.c b/drivers/accel/ivpu/ivpu_pm.c index ea30db181cd7..eacda1dbe840 100644 --- a/drivers/accel/ivpu/ivpu_pm.c +++ b/drivers/accel/ivpu/ivpu_pm.c @@ -33,8 +33,11 @@ static unsigned long ivpu_tdr_timeout_ms; module_param_named(tdr_timeout_ms, ivpu_tdr_timeout_ms, ulong, 0644); MODULE_PARM_DESC(tdr_timeout_ms, "Timeout for device hang detection, in milliseconds, 0 - default"); +static unsigned long ivpu_inference_timeout_ms; +module_param_named(inference_timeout_ms, ivpu_inference_timeout_ms, ulong, 0644); +MODULE_PARM_DESC(inference_timeout_ms, "Inference maximum duration, in milliseconds, 0 - default"); + #define PM_RESCHEDULE_LIMIT 5 -#define PM_TDR_HEARTBEAT_LIMIT 30 static void ivpu_pm_prepare_cold_boot(struct ivpu_device *vdev) { @@ -191,6 +194,10 @@ static void ivpu_job_timeout_work(struct work_struct *work) { struct ivpu_pm_info *pm = container_of(work, struct ivpu_pm_info, job_timeout_work.work); struct ivpu_device *vdev = pm->vdev; + unsigned long timeout_ms = ivpu_tdr_timeout_ms ? ivpu_tdr_timeout_ms : vdev->timeout.tdr; + unsigned long inference_timeout_ms = ivpu_inference_timeout_ms ? ivpu_inference_timeout_ms : + vdev->timeout.inference; + u64 inference_max_retries; u64 heartbeat; if (ivpu_jsm_get_heartbeat(vdev, 0, &heartbeat) || heartbeat <= vdev->fw->last_heartbeat) { @@ -198,8 +205,10 @@ static void ivpu_job_timeout_work(struct work_struct *work) goto recovery; } - if (atomic_fetch_inc(&vdev->job_timeout_counter) > PM_TDR_HEARTBEAT_LIMIT) { - ivpu_err(vdev, "Job timeout detected, heartbeat limit exceeded\n"); + inference_max_retries = DIV_ROUND_UP(inference_timeout_ms, timeout_ms); + if (atomic_fetch_inc(&vdev->job_timeout_counter) >= inference_max_retries) { + ivpu_err(vdev, "Job timeout detected, heartbeat limit (%lld) exceeded\n", + inference_max_retries); goto recovery; } |