Sigh.

gistfile1.txt

     c90:	ba 04 00 00 00       	mov    edx,0x4
     c95:	48 29 ca             	sub    rdx,rcx
     c98:	41 8a 14 13          	mov    dl,BYTE PTR [r11+rdx*1]
     c9c:	41 88 53 04          	mov    BYTE PTR [r11+0x4],dl
     ca0:	ba 05 00 00 00       	mov    edx,0x5
     ca5:	48 29 ca             	sub    rdx,rcx
     ca8:	41 8a 14 13          	mov    dl,BYTE PTR [r11+rdx*1]
     cac:	41 88 53 05          	mov    BYTE PTR [r11+0x5],dl
     cb0:	ba 06 00 00 00       	mov    edx,0x6
     cb5:	48 29 ca             	sub    rdx,rcx
     cb8:	41 8a 14 13          	mov    dl,BYTE PTR [r11+rdx*1]
     cbc:	41 88 53 06          	mov    BYTE PTR [r11+0x6],dl
     cc0:	ba 07 00 00 00       	mov    edx,0x7
     cc5:	48 29 ca             	sub    rdx,rcx
     cc8:	41 8a 14 13          	mov    dl,BYTE PTR [r11+rdx*1]
     ccc:	41 88 53 07          	mov    BYTE PTR [r11+0x7],dl

this is compiled from code

	to[4] = from[4];
	to[5] = from[5];
	to[6] = from[6];
	to[7] = from[7];

What we need to improve this is clearly... to add a comment! Well, not a
normal comment. An inline-assembly comment. Yes, that will do nicely.

+	__asm__("#%0" : "+r"(from)); // FG TEST: are we doing this again?

Look, it's *magic*!

     b08:	4c 89 d9             	mov    rcx,r11
     b0b:	48 29 c1             	sub    rcx,rax
     [..]
     b5e:	8a 51 04             	mov    dl,BYTE PTR [rcx+0x4]
     b61:	41 88 53 04          	mov    BYTE PTR [r11+0x4],dl
     b65:	8a 51 05             	mov    dl,BYTE PTR [rcx+0x5]
     b68:	41 88 53 05          	mov    BYTE PTR [r11+0x5],dl
     b6c:	8a 51 06             	mov    dl,BYTE PTR [rcx+0x6]
     b6f:	41 88 53 06          	mov    BYTE PTR [r11+0x6],dl
     b73:	8a 51 07             	mov    dl,BYTE PTR [rcx+0x7]
     b76:	41 88 53 07          	mov    BYTE PTR [r11+0x7],dl

Can you explain what that comment does?

Someone forgot to enable optimizations. Everyone can go home now.