Messages in this thread | | | From | Frank Seidel <> | Subject | Re: [RFC 13/13] Char: nozomi, cleanup read and write | Date | Mon, 12 Nov 2007 10:43:41 +0100 |
| |
On Montag 12 November 2007 08:54:52, you (Adrian Bunk) wrote: > On Sat, Nov 10, 2007 at 11:04:41PM +0100, Jiri Slaby wrote: > > Why? Anyway I think this is the case. The body of the then branch is executed at > > most once, while the else branch each time but last. If you write/read 1002 > > bytes, it means 250:1. ...and it's invoked from interrupt too... > > AFAIK there is no well defined semantics whether likely/unlikely means > 10:1 or 10000:1 that is guaranteed to not change during the next > 10 years. > > Unless there's a measurable difference, it's best to write readable > C code and simply leave all optimizations to the compiler.
From my (totally) beginners point of view i would have guessed a chance of very well below one percent that this condition is true could be called "unlikely", but i have to admit this is most probably a much too naive way of thinking, especially in regards of compiler optimizations. And in this special case now there are already most calls (about 80 percent) caught by the switch-case-shortcuts anyway.
So, i'll revert it.
Thanks a lot for your feedback, Frank - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |