[igt-dev] [PATCH] lib/igt_chamelium: Extend check of valid ports to port=0

[Alexandru M Stan]

On Fri, Jul 15, 2022 at 3:22 PM Mark Yacoub <markyacoub at chromium.org> wrote:
>
> [Why]
> Chamelium V3 has a valid port at 0 so 0 should not be rejected.

Can confirm, nowhere in the API specifications does it say port_id==0
is invalid and I have started using 0 on v3 since they're just indexes
in my internal chameleond ports array, so they start at 0.

>
> [How]
> Check for port values between 0 and CHAMELIUM_MAX_PORTS which should
> also include the V2 ports range.
>
> Test: ./kms_chamelium --run-subtest dp-hp
> Tested on: Volteer & cv3
>
> Signed-off-by: Mark Yacoub <markyacoub at chromium.org>
> ---
>  lib/igt_chamelium.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/lib/igt_chamelium.c b/lib/igt_chamelium.c
> index 55cbfa12..5d68372b 100644
> --- a/lib/igt_chamelium.c
> +++ b/lib/igt_chamelium.c
> @@ -2311,7 +2311,7 @@ static bool chamelium_read_port_mappings(struct chamelium *chamelium,
>                 port->id = g_key_file_get_integer(igt_key_file, group,
>                                                   "ChameliumPortID",
>                                                   &error);
> -               if (!port->id) {
> +               if (port->id < 0 || port->id > CHAMELIUM_MAX_PORTS) {

CHAMELIUM_MAX_PORTS seems to be #defined to 4 in another file. That
looks a little wrong.

I would do an RPC call for .GetSupportedPorts() to get the exact valid
set of port_ids instead of relying on a range. You might have some
invalid ones in there if you do it this way (no plans to do such gaps,
but again the API does not really specify).

>                         igt_warn("Failed to read chamelium port ID for %s: %s\n",
>                                  map_name, error->message);
>                         ret = false;
> --
> 2.37.0.170.g444d1eabd0-goog
>

Alexandru Stan (amstan)