Quoting Mika Kuoppala (2018-03-22 13:24:13) > Chris Wilson <chris@xxxxxxxxxxxxxxxxxx> writes: > > > Both request_submit and request_unsubmit deal with transferring the > > request from the client's timeline onto the execution timeline and back > > again. As both functions deal with a pair of timeline's, using a > > shorthand for just one of them is slightly confusing, especially as the > > The amount confusion is correlated with readers previous knowledge about > timelines. Slightly is an understatement for the newcomers but > we can accept that as a good amount of confusion, on average. > > > different functions use the shorthand for the alternate timeline. > > Instead, use the full version of each timeline so it should be easier to > > keep track of the transfer between the request/client and the engine. > > > > I pondered if the comment can even be removed on this, much > easier to read, version. But as the comment gives a > proper names to timelines, it has value. > > Only one thing remains is that should the 'client's timeline' > be changed to 'per context timeline' on this commit msg, > to reflect the comment in the code. I use client/context quite interchangeably. Every context is a client, but a client may have more than one context. I should probably avoid "client" altogether since we only have fd and contexts, and a hint of process (which is more or less limited to mmu, more or less meaningless when we look at the GPU structs). -Chris _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
