Re: A patch for linux 2.1.127

Steven Roberts (strobert@ata-sd.com)
Wed, 11 Nov 1998 16:39:50 -0800

"Peter T. Breuer" wrote:
> "A month of sundays ago Steven Roberts wrote:"
> > Isn't that what the "inline" keyword is for???
>
> Sort of. It's for when you know beyond a shadow of a doubt that something
> should work best inline. But I generate code, I don't write it. Maybe if
> I read the output code I'd be able to make sensible decisions about what to
> inline and what not to, but I certainly don't and the people who use my code
> generator don't. I do emit a few "inlines" but I generally would have
> to duplicate gcc's calculations in order to tell it to inline when it
> should anyway have inlined just by virtue of being asked to optimize.
> I'm not going to do that.
>
I think the main point is, it really doesn't matter if egcs tries to do
auto-magic inlining, it did it wrong in the case that started this
thread, and if it has troubles doing it, it should not try.

I also agree with a bunch of folks on here who feel that having the
compiler try to figure out what should be inlined is probably not a goog
idea. situations like:

static void slow_foo(int stuff)
{
// code that shouldn't be bothered with inline
}

static inline void foo(int stuff)
{
if (foo_needed)
slow_foo(stuff);
}

Now I can easily see where a compiler doing auto-inlining could try to
do both inline. I guess I consider that level of optimization best left
to the programmer. You can't try to have a compiler do everything, good
programers have to have good smart programmers working on them.

Regards,
Steve Roberts

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