Re: [RFC][PATCH] Faster generic_fls

From: Linus Torvalds (torvalds@transmeta.com)
Date: Wed Apr 30 2003 - 20:07:13 EST


On 1 May 2003, Falk Hueffner wrote:
>
> There appears to be hardware support for fls, too. This is what gcc
> generates for

Yeah, except if you want best code generation you should probably use

        static inline int fls(int x)
        {
                int bit;
                /* Only well-defined for non-zero */
                asm("bsrl %1,%0":"=r" (bit):"rm" (x));
                return x ? bit : 32;
        }

which should generate (assuming you give "-march=xxxx" to tell gcc to use
cmov):

                movl $32, %eax
                bsrl %edx,%ecx
                testl %edx, %edx
                cmovne %ecx, %eax

which looks pretty optimal.

                Linus

-
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/



This archive was generated by hypermail 2b29 : Wed Apr 30 2003 - 22:00:37 EST