Re: [PATCH 0/5] WorkStruct: Shrink work_struct by two thirds

[David Howells]

Andrew Morton <akpm@xxxxxxxx> wrote:

> waaaaaaaay too many rejects for me, sorry.  This is quite the worst time in
> the kernel cycle to be preparing patches like this.  Especially when they're
> against mainline when everyone has so much material pending.

Actually... there is a way to do this sort of incrementally, I think:

 (1) Turn this sort of thing:

	do_work(struct x *x)
	{
		...
	}

	queue_x(struct x *x)
	{
		INIT_WORK(&x->work, do_work, x);
		schedule_work(&x->work)
	}

    Into this sort of thing:

	#define DECLARE_IMMEDIATE_WORK(w, f) DECLARE_WORK((w), (f), (w))
	#define DECLARE_DELAYED_WORK(w, f) DECLARE_WORK((w), (f), (w))
	#define INIT_IMMEDIATE_WORK(w, f) INIT_WORK((w), (f), (w))
	#define INIT_DELAYED_WORK(w, f) INIT_WORK((w), (f), (w))

	do_work(struct work_struct *work)
	{
		struct x *x = container_of(work, struct x, work);
		...
	}

	queue_x(struct x *x)
	{
		INIT_IMMEDIATE_WORK(&x->work, do_work); //or
		INIT_DELAYED_WORK(&x->work, do_work);
		schedule_work(&x->work)
	}

 (2) Make delayed_work equivalent to work_struct:

	#define delayed_work work_struct

 (3) Then apply the rest of the patches such that they remove the #defines as
     appropriate.

Might that help?

David
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/