Re: [PATCH] acpi/iort: check output reference for the real used mapping

[Hanjun Guo <guohanjun@huawei.com>]

On 2020/2/25 17:01, Heyi Guo wrote:
> The function iort_node_map_id() does the sanity check against the
> first mapping in the node, but not the one which we really use.
> 
> Logically we need check the mapping we use, or check every mapping in
> the node. Choose the first fix for we are not firmware tester.
> 
> Signed-off-by: Heyi Guo <guoheyi@huawei.com>
> 
> ---
> Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
> Cc: Hanjun Guo <guohanjun@huawei.com>
> Cc: Sudeep Holla <sudeep.holla@arm.com>
> Cc: "Rafael J. Wysocki" <rjw@rjwysocki.net>
> Cc: Len Brown <lenb@kernel.org>
> Cc: linux-acpi@vger.kernel.org
> Cc: linux-arm-kernel@lists.infradead.org
> Cc: linux-kernel@vger.kernel.org
> ---
>  drivers/acpi/arm64/iort.c | 14 +++++++-------
>  1 file changed, 7 insertions(+), 7 deletions(-)
> 
> diff --git a/drivers/acpi/arm64/iort.c b/drivers/acpi/arm64/iort.c
> index ed3d2d1a7ae9..d0fe8d673240 100644
> --- a/drivers/acpi/arm64/iort.c
> +++ b/drivers/acpi/arm64/iort.c
> @@ -470,13 +470,6 @@ static struct acpi_iort_node *iort_node_map_id(struct acpi_iort_node *node,
>  		map = ACPI_ADD_PTR(struct acpi_iort_id_mapping, node,
>  				   node->mapping_offset);
>  
> -		/* Firmware bug! */
> -		if (!map->output_reference) {
> -			pr_err(FW_BUG "[node %p type %d] ID map has NULL parent reference\n",
> -			       node, node->type);
> -			goto fail_map;
> -		}
> -
>  		/*
>  		 * Get the special ID mapping index (if any) and skip its
>  		 * associated ID map to prevent erroneous multi-stage
> @@ -497,6 +490,13 @@ static struct acpi_iort_node *iort_node_map_id(struct acpi_iort_node *node,
>  		if (i == node->mapping_count)
>  			goto fail_map;
>  
> +		/* Firmware bug! */
> +		if (!map->output_reference) {
> +			pr_err(FW_BUG "[node %p type %d] ID map has NULL parent reference\n",
> +			       node, node->type);
> +			goto fail_map;
> +		}

I think we can warning on the NULL parent reference when
scanning the mappings in the node, but don't bail out for
the mappings we are not using.

Thanks
Hanjun