Re: [RFC PATCH 08/12] media: iris: Avoid updating frame size to firmware during reconfig

[Bryan O'Donoghue]

On 05/03/2025 10:43, Dikshita Agarwal wrote:
> During the reconfig, firmware sends the resolution aligned by 8 byte,
> if driver set the same resoluton to firmware, it will be aligned to 16
> byte causing another sequence change which would be incorrect.

During reconfig the firmware sends the resolution aligned to 8 bytes. If 
the driver sends the same resolution back to the firmware the resolution 
will be aligned to 16 bytes not 8.

The alignment mismatch would then subsequently cause the firmware to 
send another redundant sequence change.

> Fix this by not setting the updated resolution to firmware during
> reconfig.

Fix this by not setting the resolution property during reconfig.
> Signed-off-by: Dikshita Agarwal <quic_dikshita@xxxxxxxxxxx>
> ---
>   .../platform/qcom/iris/iris_hfi_gen1_command.c    | 15 ++++++++-------
>   .../platform/qcom/iris/iris_hfi_gen1_response.c   |  1 +
>   drivers/media/platform/qcom/iris/iris_instance.h  |  2 ++
>   drivers/media/platform/qcom/iris/iris_vdec.c      |  4 ++++
>   4 files changed, 15 insertions(+), 7 deletions(-)
>
> diff --git a/drivers/media/platform/qcom/iris/iris_hfi_gen1_command.c b/drivers/media/platform/qcom/iris/iris_hfi_gen1_command.c
> index a160ae915886..d5e81049d37e 100644
> --- a/drivers/media/platform/qcom/iris/iris_hfi_gen1_command.c
> +++ b/drivers/media/platform/qcom/iris/iris_hfi_gen1_command.c
> @@ -562,14 +562,15 @@ static int iris_hfi_gen1_set_resolution(struct iris_inst *inst)
>   	struct hfi_framesize fs;
>   	int ret;
>   
> -	fs.buffer_type = HFI_BUFFER_INPUT;
> -	fs.width = inst->fmt_src->fmt.pix_mp.width;
> -	fs.height = inst->fmt_src->fmt.pix_mp.height;
> -
> -	ret = hfi_gen1_set_property(inst, ptype, &fs, sizeof(fs));
> -	if (ret)
> -		return ret;
> +	if (!inst->in_reconfig) {
> +		fs.buffer_type = HFI_BUFFER_INPUT;
> +		fs.width = inst->fmt_src->fmt.pix_mp.width;
> +		fs.height = inst->fmt_src->fmt.pix_mp.height;
>   
> +		ret = hfi_gen1_set_property(inst, ptype, &fs, sizeof(fs));
> +		if (ret)
> +			return ret;
> +	}
>   	fs.buffer_type = HFI_BUFFER_OUTPUT2;
>   	fs.width = inst->fmt_dst->fmt.pix_mp.width;
>   	fs.height = inst->fmt_dst->fmt.pix_mp.height;
> diff --git a/drivers/media/platform/qcom/iris/iris_hfi_gen1_response.c b/drivers/media/platform/qcom/iris/iris_hfi_gen1_response.c
> index 91d95eed68aa..6576496fdbdf 100644
> --- a/drivers/media/platform/qcom/iris/iris_hfi_gen1_response.c
> +++ b/drivers/media/platform/qcom/iris/iris_hfi_gen1_response.c
> @@ -155,6 +155,7 @@ static void iris_hfi_gen1_read_changed_params(struct iris_inst *inst,
>   		inst->crop.height = event.height;
>   	}
>   
> +	inst->in_reconfig = true;

This flag can be changed by iris_hfi_isr_handler() down the chain.


> @@ -453,6 +453,8 @@ static int iris_vdec_process_streamon_input(struct iris_inst *inst)
>   	if (ret)
>   		return ret;
>   
> +	inst->in_reconfig = false;
> +
>   	return iris_inst_change_sub_state(inst, 0, set_sub_state);
>   }
>   
> @@ -544,6 +546,8 @@ static int iris_vdec_process_streamon_output(struct iris_inst *inst)
>   	if (ret)
>   		return ret;
>   
> +	inst->in_reconfig = false;
> +

Are these usages of the in_reconfig flag then thread-safe ?

i.e. are both iris_vdec_process_streamon_input() and 
iris_vdec_process_streamon_output() guaranteed not to run @ the same time ?

I don't see any obvious locking here.

---
bod