Logo Questions Linux Laravel Mysql Ubuntu Git Menu
 

error: invalid suffix "b11111111111111111111111111111111" on integer constant

Tags:

c++

g++

I am using g++ version 4.1.2 on a RHEL 5.7 x86_64 box. This builds just fine with g++ version 4.4.5 which comes with RHEL 6.0 x86_64. What does this compiler error mean and how do you fix it?

[mehoggan@hoggant35002 C]$ g++ -Wall -o binary ./binary.cpp 
./binary.cpp:2:5: error: invalid suffix "b11111111111111111111111111111111" on integer constant
./binary.cpp:3:5: error: invalid suffix "b11111111111111111111111111111110" on integer constant
./binary.cpp:4:5: error: invalid suffix "b11111111111111111111111111111100" on integer constant
./binary.cpp:5:5: error: invalid suffix "b11111111111111111111111111111000" on integer constant
./binary.cpp:6:5: error: invalid suffix "b11111111111111111111111111110000" on integer constant
./binary.cpp:7:5: error: invalid suffix "b11111111111111111111111111100000" on integer constant
./binary.cpp:8:5: error: invalid suffix "b11111111111111111111111111000000" on integer constant
./binary.cpp:9:5: error: invalid suffix "b11111111111111111111111110000000" on integer constant
./binary.cpp:10:5: error: invalid suffix "b11111111111111111111111100000000" on integer constant
./binary.cpp:11:5: error: invalid suffix "b11111111111111111111111000000000" on integer constant
./binary.cpp:12:5: error: invalid suffix "b11111111111111111111110000000000" on integer constant
./binary.cpp:13:5: error: invalid suffix "b11111111111111111111100000000000" on integer constant
./binary.cpp:14:5: error: invalid suffix "b11111111111111111111000000000000" on integer constant
./binary.cpp:15:5: error: invalid suffix "b11111111111111111110000000000000" on integer constant
./binary.cpp:16:5: error: invalid suffix "b11111111111111111100000000000000" on integer constant
./binary.cpp:17:5: error: invalid suffix "b11111111111111111000000000000000" on integer constant
./binary.cpp:18:5: error: invalid suffix "b11111111111111110000000000000000" on integer constant
./binary.cpp:19:5: error: invalid suffix "b11111111111111100000000000000000" on integer constant
./binary.cpp:20:5: error: invalid suffix "b11111111111111000000000000000000" on integer constant
./binary.cpp:21:5: error: invalid suffix "b11111111111110000000000000000000" on integer constant
./binary.cpp:22:5: error: invalid suffix "b11111111111100000000000000000000" on integer constant
./binary.cpp:23:5: error: invalid suffix "b11111111111000000000000000000000" on integer constant
./binary.cpp:24:5: error: invalid suffix "b11111111110000000000000000000000" on integer constant
./binary.cpp:25:5: error: invalid suffix "b11111111100000000000000000000000" on integer constant
./binary.cpp:26:5: error: invalid suffix "b11111111000000000000000000000000" on integer constant
./binary.cpp:27:5: error: invalid suffix "b11111110000000000000000000000000" on integer constant
./binary.cpp:28:5: error: invalid suffix "b11111100000000000000000000000000" on integer constant
./binary.cpp:29:5: error: invalid suffix "b11111000000000000000000000000000" on integer constant
./binary.cpp:30:5: error: invalid suffix "b11110000000000000000000000000000" on integer constant
./binary.cpp:31:5: error: invalid suffix "b11100000000000000000000000000000" on integer constant
./binary.cpp:32:5: error: invalid suffix "b11000000000000000000000000000000" on integer constant
./binary.cpp:33:5: error: invalid suffix "b10000000000000000000000000000000" on integer constant

The code:

static int s_bitCountMask[32] = {
    0b11111111111111111111111111111111,
    0b11111111111111111111111111111110,
    0b11111111111111111111111111111100,
    0b11111111111111111111111111111000,
    0b11111111111111111111111111110000,
    0b11111111111111111111111111100000,
    0b11111111111111111111111111000000,
    0b11111111111111111111111110000000,
    0b11111111111111111111111100000000,
    0b11111111111111111111111000000000,
    0b11111111111111111111110000000000,
    0b11111111111111111111100000000000,
    0b11111111111111111111000000000000,
    0b11111111111111111110000000000000,
    0b11111111111111111100000000000000,
    0b11111111111111111000000000000000,
    0b11111111111111110000000000000000,
    0b11111111111111100000000000000000,
    0b11111111111111000000000000000000,
    0b11111111111110000000000000000000,
    0b11111111111100000000000000000000,
    0b11111111111000000000000000000000,
    0b11111111110000000000000000000000,
    0b11111111100000000000000000000000,
    0b11111111000000000000000000000000,
    0b11111110000000000000000000000000,
    0b11111100000000000000000000000000,
    0b11111000000000000000000000000000,
    0b11110000000000000000000000000000,
    0b11100000000000000000000000000000,
    0b11000000000000000000000000000000,
    0b10000000000000000000000000000000,
};

#include <stdio.h>

int main(int argc, char *argv[])
{
    for (int i = 0; i < 32; i++) {
        printf("%d\n",s_bitCountMask[i]);
    }
}
like image 465
Matthew Hoggan Avatar asked Nov 21 '11 19:11

Matthew Hoggan


2 Answers

It means the older verison of g++ won't accept the 0b000 binary constant syntax. You can rewrite the numbers as hex instead - binary easily translates into hex four bits at a time:

0xffffffff, 0xfffffffe, 0xfffffffc, 0xfffffff8,
0xfffffff0, 0xffffffe0, 0xffffffc0, 0xffffff80,
0xffffff00, 0xfffffe00, 0xfffffc00, 0xfffff800,
0xfffff000, 0xffffe000, 0xffffc000, 0xffff8000,
0xffff0000, 0xfffe0000, 0xfffc0000, 0xfff80000,
0xfff00000, 0xffe00000, 0xffc00000, 0xff800000,
0xff000000, 0xfe000000, 0xfc000000, 0xf8000000,
0xf0000000, 0xe0000000, 0xc0000000, 0x80000000
like image 82
Rup Avatar answered Nov 05 '22 12:11

Rup


Rup has the correct answer to this. But I'll add this:

Depending on how you're accessing that table, it might be more appropriate to just generate the value on the spot with this:

unsigned mask = ~((1 << index) - 1);
unsigned mask = -1 << index;          //  Better solution (by Rup)

Alternatively, you could also declare the table like this:

static int s_bitCountMask[32] = {
    -1 <<  0,
    -1 <<  1,
    -1 <<  2,
    -1 <<  3,

    ...

    -1 << 31,
};

*Assuming that -1 is all one-bits (which it is on nearly every machine now).

like image 24
Mysticial Avatar answered Nov 05 '22 12:11

Mysticial