[igt-dev] [RFC PATCH i-g-t] lib/i915: Generate engine names at runtime

[Andi Shyti]

Hi Petri,

> @@ -120,12 +121,22 @@ static void init_engine(struct intel_execution_engine2 *e2,
>  		if (__e2->class == class && __e2->instance == instance)
>  			break;
>  
> -	if (__e2->name) {
> -		e2->name = __e2->name;
> +	if (__e2->name[0]) {
> +		strncpy(e2->name, __e2->name, sizeof(e2->name));
>  	} else {
> -		igt_warn("found unknown engine (%d, %d)\n", class, instance);
> -		e2->name = unknown_name;
> -		e2->flags = -1;
> +		for (cls = intel_execution_engine_class_map; cls->name; cls++) {
> +			if (cls->class == class) {
> +				snprintf(e2->name, sizeof(e2->name), "%s%d", cls->name, instance);
> +				igt_info("unknown but supported engine %s found\n", e2->name);
> +				break;
> +			}
> +		}

We need to make a decision here and refactor this library
accordingly. We either:

1. decide to get rid of all the static table and keep only
"intel_execution_engine_class" (and maybe we can spare that one
as well).

or

2. add the new classes in the static table as it's meant to be
used.

I honestly do not like any hybrid solution and when this library
was developed in a first place we had to accept lots of
compromises because we needed both.

Now that we have the dynamic_libraries, can we finally remove
all the static tables and finally go for the solution 1?

> +
> +		if (!e2->name[0]) {
> +			igt_info("found unknown engine (%d, %d)\n", class, instance);
> +			strncpy(e2->name, unknown_name, sizeof(e2->name));

nit: just strcpy would do, we know how big is unkown_name and it
won't change.

[...]

> +const struct intel_execution_engine_class intel_execution_engine_class_map[] = {
> +	{ "rcs", I915_ENGINE_CLASS_RENDER },
> +	{ "bcs", I915_ENGINE_CLASS_COPY },
> +	{ "vcs", I915_ENGINE_CLASS_VIDEO },
> +	{ "vecs", I915_ENGINE_CLASS_VIDEO_ENHANCE },
> +	{ }
> +};

another one? it would be the third const struct that is used to
get the engine names/instances. I thought we wanted to eliminate
them in the long run :'(

Andi