[PATCH] drm/amdgpu: add condition check for trace_amdgpu_cs()

[Christian König]

Am 17.08.20 um 11:04 schrieb Kevin Wang:
> add trace event enabled check to avoid nop loop when submit multi ibs
> in amdgpu_cs_ioctl() function.

Maybe we should change the trace point instead to trace all IBs with 
just a single call.

>
> Signed-off-by: Kevin Wang <kevin1.wang at amd.com>
> ---
>   drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 5 +++--
>   1 file changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
> index ffbcaf4bfb8b..409694f074ac 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
> @@ -1319,8 +1319,9 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
>   
>   	reserved_buffers = true;
>   
> -	for (i = 0; i < parser.job->num_ibs; i++)
> -		trace_amdgpu_cs(&parser, i);
> +	if (trace_amdgpu_cs_enabled())
> +		for (i = 0; i < parser.job->num_ibs; i++)
> +			trace_amdgpu_cs(&parser, i);
>   
>   	r = amdgpu_cs_vm_handling(&parser);
>   	if (r)