Re: [PATCH] Better __ashlDI3, __ashrDI3 and __lshrDI3 functions, plus fixed __bswapsi2 function

[Jakub Jelinek <jakub@redhat.com>]

On Wed, Nov 25, 2020 at 09:22:53PM +0100, Stefan Kanthak wrote:
> > As Jakub has already indicated, your change will result in infinite
> > recursion on avr. I happened to have a cr16 handy and it looks like
> > it'd generate infinite recursion there too.
> 
> JFTR: does GCC emit a warning then? If not: why not?

Why should it.  libgcc is something GCC has full control over and can assume
it is written in a way that it can't recurse infinitely, that is one of
libgcc design goals.

> Since I neither have an avr nor a cr16 here, and also no TR-440, no S/3x0,
> no Spectra-70, no PDP-11, no VAX, no SPARC, no MIPS, no PowerPC, no MC68k,
> no NSC16xxx and no NSC32xxx any more, GCC only gives me access to the x86
> code it generates.

You can always use cross-compilers.

> > On other targets the routines you're changing won't be used because they
> > either have 64 bit shifts or the compiler can synthesize them from other
> > primitives that are available.
> 
> These routines are documented in
> <https://gcc.gnu.org/onlinedocs/gccint/Integer-library-routines.html>
> and might be called by your users.

That is just theory, if people call them by hand instead of using normal C
arithmetics they will get worse code in any case (at least because of the
library call).
As has been said multiple times, trying to optimize routines that are never
called on x86 for x86 is just wasted energy, better invest your time in
functions that are actually ever called.  And even for those, care has to be
taken so that it doesn't break any other of the > 50 target architectures
GCC supports.

	Jakub