Thanks Marc!
This certainly works for my original problem. But my question is why gcc is kind of changing its mind as it generates the code for the functions I sent in my original post -- it just seems illogical.
-Pavel
----- Original Message -----
From: Marc Glisse <marc.glisse@xxxxxxxx>
To: Pavel Tolkachev <paultolk@xxxxxxxxx>
Cc: GCC Users List <gcc-help@xxxxxxxxxxx>
Sent: Tuesday, August 16, 2011 1:40 AM
Subject: Re: an inconsistency in generated code?
On Mon, 15 Aug 2011, Pavel Tolkachev wrote:
> I needed to write portable and
> fast C/C++ function to bit-negate the most significant byte of a 32-bit
> integer. Currently, my target is i386. First, I wrote this function:
>
> uint32_t
> BitNegMsb_1(uint32_t v)
> {
> return (v & 0xFFFFFFu) | (~v &0xFF000000u);
> }
Did you try with a xor?
return v^0xFF000000u;
--
Marc Glisse
