Re: [PATCH 7/8] remoteproc: qcom: Add support for memory sandbox

[Christophe JAILLET]

Hi,

the error handling below looks odd.

Le 03/08/2022 à 16:21, Srinivasa Rao Mandadapu a écrit :
> Add memory sandbox support for ADSP based platforms secure booting.
>
> Signed-off-by: Srinivasa Rao Mandadapu <quic_srivasam-jfJNa2p1gH1BDgjK7y7TUQ@xxxxxxxxxxxxxxxx>
> ---
>   drivers/remoteproc/qcom_q6v5_adsp.c | 101 +++++++++++++++++++++++++++++++++++-
>   1 file changed, 99 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/remoteproc/qcom_q6v5_adsp.c b/drivers/remoteproc/qcom_q6v5_adsp.c
> index 3dbd035..f81da47 100644
> --- a/drivers/remoteproc/qcom_q6v5_adsp.c
> +++ b/drivers/remoteproc/qcom_q6v5_adsp.c

>   static int adsp_start(struct rproc *rproc)
>   {
>   	struct qcom_adsp *adsp = (struct qcom_adsp *)rproc->priv;
> @@ -341,7 +429,13 @@ static int adsp_start(struct rproc *rproc)
>   	ret = qcom_q6v5_prepare(&adsp->q6v5);
>   	if (ret)
>   		return ret;
> -
> +	if (!adsp->is_wpss) {
> +		ret = adsp_map_smmu(adsp, rproc);
> +		if (ret) {
> +			dev_err(adsp->dev, "ADSP smmu mapping failed\n");
> +			goto adsp_smmu_unmap;
goto disable_irqs;?

> +		}
> +	}
>   	ret = clk_prepare_enable(adsp->xo);
>   	if (ret)
>   		goto disable_irqs;

goto adsp_smmu_unmap;?

> @@ -402,6 +496,9 @@ static int adsp_start(struct rproc *rproc)
>   	clk_disable_unprepare(adsp->xo);
>   disable_irqs:
>   	qcom_q6v5_unprepare(&adsp->q6v5);
> +adsp_smmu_unmap:
> +	iommu_unmap(adsp->iommu_dom, adsp->mem_phys, adsp->mem_size);
> +	iommu_domain_free(adsp->iommu_dom);

Should this new hunk be above disable_irqs?
And I think that it should be guardd with a "if (!adsp->is_wpss)".

CJ

>   
>   	return ret;
>   }