On Fri, Dec 20, 2013 at 09:47:53PM -0500, Alan Stern wrote: > On Fri, 20 Dec 2013, Greg Kroah-Hartman wrote: > > > > > /* Check that the pipe's type matches the endpoint's type */ > > > > if (usb_pipetype(urb->pipe) != pipetypes[xfertype]) > > > > > > It looks as though it ought to be possible to make that check: > > > if (unlikely(xfertype != urb->pipe->valid_xfertype)) > > > > You should almost never use unlikely() on your own, in a driver, as the > > compiler should make better code without it. > > Is that really true? If it is, what point is there in having the > likely()/unlikely() macros in the first place? At this point in time, with modern processors and the huge pipelines that they have, almost nothing. Andi Kleen did a bunch of benchmarks a few years ago that show that removing them all makes things work better, and we, as programmers, usually get the "guess" wrong. Maybe, in the scheduler, or on a _very_ hot path in the kernel (like locks), that can be properly benchmarked and validated, can they make a difference. But in USB, with our line speeds, it's not an issue. greg k-h -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
