[Intel-gfx] [PATCH v3 08/14] drm/i915/scheduler: Execute requests in order of priorities

[Tvrtko Ursulin]

On 14/11/2016 11:41, Chris Wilson wrote:
> On Mon, Nov 14, 2016 at 11:15:52AM +0000, Tvrtko Ursulin wrote:
>> On 14/11/2016 08:56, Chris Wilson wrote:
>>> +static void execlists_schedule(struct drm_i915_gem_request *request, int prio)
>>> +{
>>> +	struct intel_engine_cs *engine = NULL;
>>> +	struct i915_dependency *dep, *p;
>>> +	struct i915_dependency stack;
>>> +	LIST_HEAD(dfs);
>>> +
>>> +	if (prio <= READ_ONCE(request->priotree.priority))
>>> +		return;
>>> +
>>> +	/* Need BKL in order to use the temporary link inside i915_dependency */
>>> +	lockdep_assert_held(&request->i915->drm.struct_mutex);
>>> +
>>> +	stack.signaler = &request->priotree;
>>> +	list_add(&stack.dfs_link, &dfs);
>>> +
>>> +	/* Recursively bump all dependent priorities to match the new request */
>>
>> Missed last time round that the comment needs updating.
>
> It still is a recursive design though, just flat. That one word was
> saving a paragraph :|
>
> I think the easiest way to describe what the code is doing here is to
> show the recursive version in the comment and then hope for inspiration
> in describing how that maps onto the search list.

I can see that angle yes. Maybe the just add a second sentence saying 
something like "To avoid having recursive code to do this recursive 
update we build a flat list of dependencies in a depth first search 
manner."?

Regards,

Tvrtko