How can I get GCC to optimize this bit-shifting instruction into a move?

user_10795151 0
Viewed 182

I'm trying to use the following code to emulate a 16-bit half-float in software:

typedef struct half
{
    unsigned short mantissa:10;
    unsigned short exponent:5;
    unsigned short sign:1;
} half;

unsigned short from_half(half h)
{
    return h.mantissa | h.exponent << 10 | h.sign << 15;
}

half to_half(unsigned short s)
{
    half result = { s, s >> 10, s >> 15 };
    return result;
}

I set this up so that it could easily be optimized into a move instruction, but lo and behold, in from_half, GCC does the bit-shifting anyway (even at -O3):

from_half:
        mov     edx, edi
        mov     eax, edi
        and     di, 1023
        shr     dx, 15
        and     eax, 31744
        movzx   edx, dl
        sal     edx, 15
        or      eax, edx
        or      eax, edi
        ret

while to_half is optimized nicely:

to_half:
        mov     eax, edi
        ret

Godbolt

I've tried different optimization levels (-O1, -O2, -Os) but none optimize it into what I was hoping.

Clang does this how I would expect even at -O1:

from_half:                              # @from_half
        mov     eax, edi
        ret
to_half:                                # @to_half
        mov     eax, edi
        ret

Godbolt

How can I get GCC to optimize this into a move? Why isn't it optimized that way already?

compiler-optimization bit-fields gcc c
2 Answers

In addition to Booboo's answer, you can try the following which answers your question

How can I get GCC to optimize this into a move?

Just cast each shifted bit-field expression to unsigned short 

unsigned short from_half(half h)
{
    return (unsigned short)h.mantissa | (unsigned short)(h.exponent << 10) | (unsigned short)(h.sign << 15);
}

https://godbolt.org/z/CfZSgC

It results in:

from_half:
        mov     eax, edi
        ret

Why isn't it optimized that way already?

I am not sure I have a solid answer on this one. Apparently the intermediate promotion of the bit-fields to int confuses the optimizer... But this is just a guess.

user_52182770

It's been a while since I have coded in C, but it seems the use of a union should work:

#include <stdint.h>
#include <stdbool.h>
#include <stdio.h>

static bool useUnion;

__attribute__ ((__constructor__)) // supported by gcc compiler
static void initUseUnion()
{
    union {
       uint16_t i;
       char c[2];
    } n = { 0x0001 };
    useUnion = n.c[0]; // little endian
}

typedef struct half
{
    unsigned short mantissa:10;
    unsigned short exponent:5;
    unsigned short sign:1;
} half;

typedef union half_short
{
    half h;
    uint16_t s;
} half_short;

unsigned short from_half(half h)
{
    if (useUnion) {
        half_short hs;
        hs.h = h;
        return hs.s;
    }
    else {
        return h.mantissa | h.exponent << 10 | h.sign << 15;
    }
}

half to_half(unsigned short s)
{
    if (useUnion) {
        half_short hs;
        hs.s = s;
        return hs.h;
    }
    else {
        half result = { s, s >> 10, s >> 15 };
        return result;
    }
}

int main(int argc, char* argv[])
{
    printf("%d\n", useUnion);
    return 0;
}
user_28237190
Related