Re: [PATCH] drm/v3d: Add DRM_IOCTL_V3D_PERFMON_SET_GLOBAL

[André Almeida]

Hi Christian!

Em 20/10/2024 17:41, Christian Gmeiner escreveu:
> From: Christian Gmeiner <cgmeiner@xxxxxxxxxx>
>
> This patch adds a new ioctl, DRM_IOCTL_V3D_PERFMON_SET_GLOBAL, which
> allows the configuration of a global performance monitor (perfmon).
> The global perfmon is used for all jobs, ensuring consistent performance
> tracking across submissions.

Usually we write in the imperative form:

Add a new ioctl, ...

> Signed-off-by: Christian Gmeiner <cgmeiner@xxxxxxxxxx>
> ---
>   drivers/gpu/drm/v3d/v3d_drv.c                 |  3 ++
>   drivers/gpu/drm/v3d/v3d_drv.h                 | 10 ++++
>   drivers/gpu/drm/v3d/v3d_perfmon.c             | 49 +++++++++++++++++++
>   .../gpu/drm/v3d/v3d_performance_counters.h    |  6 +++
>   drivers/gpu/drm/v3d/v3d_sched.c               | 10 +++-
>   include/uapi/drm/v3d_drm.h                    | 15 ++++++
>   6 files changed, 91 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpu/drm/v3d/v3d_drv.c b/drivers/gpu/drm/v3d/v3d_drv.c
> index d7ff1f5fa481..f1753ee2af25 100644
> --- a/drivers/gpu/drm/v3d/v3d_drv.c
> +++ b/drivers/gpu/drm/v3d/v3d_drv.c
> @@ -214,6 +214,7 @@ static const struct drm_ioctl_desc v3d_drm_ioctls[] = {
>   	DRM_IOCTL_DEF_DRV(V3D_PERFMON_GET_VALUES, v3d_perfmon_get_values_ioctl, DRM_RENDER_ALLOW),
>   	DRM_IOCTL_DEF_DRV(V3D_SUBMIT_CPU, v3d_submit_cpu_ioctl, DRM_RENDER_ALLOW | DRM_AUTH),
>   	DRM_IOCTL_DEF_DRV(V3D_PERFMON_GET_COUNTER, v3d_perfmon_get_counter_ioctl, DRM_RENDER_ALLOW),
> +	DRM_IOCTL_DEF_DRV(V3D_PERFMON_SET_GLOBAL, v3d_perfmon_set_global_ioctl, DRM_RENDER_ALLOW),
>   };
>   
>   static const struct drm_driver v3d_drm_driver = {
> @@ -350,6 +351,8 @@ static int v3d_platform_drm_probe(struct platform_device *pdev)
>   	if (ret)
>   		goto drm_unregister;
>   
> +	atomic_set(&v3d->num_perfmon, 0);
> +
>   	return 0;
>   
>   drm_unregister:
> diff --git a/drivers/gpu/drm/v3d/v3d_drv.h b/drivers/gpu/drm/v3d/v3d_drv.h
> index cf4b23369dc4..9491d730d99f 100644
> --- a/drivers/gpu/drm/v3d/v3d_drv.h
> +++ b/drivers/gpu/drm/v3d/v3d_drv.h
> @@ -61,6 +61,8 @@ struct v3d_queue_state {
>   	struct v3d_stats stats;
>   };
>   
> +struct v3d_dev;
> +

Forward declarations go in the beginning of the file, along with the 
other ones:

struct clk;
struct platform_device;
struct reset_control;
+struct v3d_dev;

>   /* Performance monitor object. The perform lifetime is controlled by userspace
>    * using perfmon related ioctls. A perfmon can be attached to a submit_cl
>    * request, and when this is the case, HW perf counters will be activated just
> @@ -68,6 +70,9 @@ struct v3d_queue_state {
>    * done. This way, only events related to a specific job will be counted.
>    */
>   struct v3d_perfmon {
> +	/* Pointer back to v3d instance this perfmon belongs. */
> +	struct v3d_dev *v3d;
> +
>   	/* Tracks the number of users of the perfmon, when this counter reaches
>   	 * zero the perfmon is destroyed.
>   	 */
> @@ -179,6 +184,9 @@ struct v3d_dev {
>   		u32 num_allocated;
>   		u32 pages_allocated;
>   	} bo_stats;
> +
> +	/* Keep track of current number of allocated perfmons. */
> +	atomic_t num_perfmon;
>   };
>   
>   static inline struct v3d_dev *
> @@ -584,6 +592,8 @@ int v3d_perfmon_get_values_ioctl(struct drm_device *dev, void *data,
>   				 struct drm_file *file_priv);
>   int v3d_perfmon_get_counter_ioctl(struct drm_device *dev, void *data,
>   				  struct drm_file *file_priv);
> +int v3d_perfmon_set_global_ioctl(struct drm_device *dev, void *data,
> +				 struct drm_file *file_priv);
>   
>   /* v3d_sysfs.c */
>   int v3d_sysfs_init(struct device *dev);

[...]

> diff --git a/include/uapi/drm/v3d_drm.h b/include/uapi/drm/v3d_drm.h
> index 87fc5bb0a61e..960d392d75a3 100644
> --- a/include/uapi/drm/v3d_drm.h
> +++ b/include/uapi/drm/v3d_drm.h
> @@ -43,6 +43,7 @@ extern "C" {
>   #define DRM_V3D_PERFMON_GET_VALUES                0x0a
>   #define DRM_V3D_SUBMIT_CPU                        0x0b
>   #define DRM_V3D_PERFMON_GET_COUNTER               0x0c
> +#define DRM_V3D_PERFMON_SET_GLOBAL                0x0d
>   
>   #define DRM_IOCTL_V3D_SUBMIT_CL           DRM_IOWR(DRM_COMMAND_BASE + DRM_V3D_SUBMIT_CL, struct drm_v3d_submit_cl)
>   #define DRM_IOCTL_V3D_WAIT_BO             DRM_IOWR(DRM_COMMAND_BASE + DRM_V3D_WAIT_BO, struct drm_v3d_wait_bo)
> @@ -61,6 +62,8 @@ extern "C" {
>   #define DRM_IOCTL_V3D_SUBMIT_CPU          DRM_IOW(DRM_COMMAND_BASE + DRM_V3D_SUBMIT_CPU, struct drm_v3d_submit_cpu)
>   #define DRM_IOCTL_V3D_PERFMON_GET_COUNTER DRM_IOWR(DRM_COMMAND_BASE + DRM_V3D_PERFMON_GET_COUNTER, \
>   						   struct drm_v3d_perfmon_get_counter)
> +#define DRM_IOCTL_V3D_PERFMON_SET_GLOBAL  DRM_IOW(DRM_COMMAND_BASE + DRM_V3D_PERFMON_SET_GLOBAL, \
> +						   struct drm_v3d_perfmon_set_global)
>   
>   #define DRM_V3D_SUBMIT_CL_FLUSH_CACHE             0x01
>   #define DRM_V3D_SUBMIT_EXTENSION		  0x02
> @@ -765,6 +768,18 @@ struct drm_v3d_perfmon_get_counter {
>   	__u8 reserved[7];
>   };
>   
> +/**

Using /** means that you are writting a kernel-doc comment [1], so make 
sure to describe each struct member, otherwise it's going to generate 
build warnings with W=1.

> + * struct drm_v3d_perfmon_set_global - ioctl to define a
> + * global performance counter that is used if a job has
> + * not assigned one on its own.
> + */
> +
> +#define DRM_V3D_PERFMON_CLEAR_GLOBAL    0x0001

I would keep this define above the struct comment.

> +struct drm_v3d_perfmon_set_global {
> +	__u32 flags;
> +	__u32 id;
> +};
> +
>   #if defined(__cplusplus)
>   }
>   #endif

[1] https://www.kernel.org/doc/html/latest/doc-guide/kernel-doc.html