Updating fields of bits incorrectly

Question

I'm trying to solve a question. It says,

Initialize a new variable to the value 17512807u.

Assume we number the bits as usual from 0 as least significant (on the right) to 31 (most significant, on the left). Update bits 18 through 21 with the integer value 8 and bits 10 through 14 with value 17 (decimal). Print the resulting value as an eight digit hexadecimal number to show all of the digits.

Here's the code I came up with:

#include <stdio.h> int main(){ int value = 17512807u; int L = 21; // starting left position int R = 18; // starting right position int mask = (1 << (L - R + 1) - 1) << R; int newField = (8 << R) & mask; // integer value 8, shifting to right int newValue = value & (~mask); // remove range of bits value = newField | newValue; // update range of bits L = 14; R = 10; mask = (1 << (L - R + 1) - 1) << R; newField = (17 << R) & mask; newValue = value & (~mask); value = newField | newValue; printf("%08x\n", value); }

The answer I get is 012b7d67

However, I am told this is the incorrect answer. I do not know what the correct answer is.

The code pasted here didn't even compile for me. Two errors and two warnings. After fixing those, I got 01234567. — greedy52
– greedy52, Commented Feb 15, 2017 at 20:34
Your code can invoke undefined behaviour. int is not guaranteed to hold 32 bits. If you need a fixed bit-width, use fixed idth types. Oh, and it is in general better to use unsigned integer when shifting. signed integers can invoke undefined behaviour for certain values (not necessarily true for this example, just keep that in mind!) — too honest for this site
– too honest for this site, Commented Feb 15, 2017 at 20:49
@Olaf I had not noticed that int value = 17512807u; does not change int value to unsigned type. — Weather Vane
– Weather Vane, Commented Feb 15, 2017 at 20:57

greedy52 · Accepted Answer · 2017-02-15 20:36:43Z

1

int mask = ((1 << (L - R + 1)) - 1) << R;

answered Feb 15, 2017 at 20:36

greedy52

1,41510 silver badges8 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

chqrlie · Accepted Answer · 2017-02-15 21:59:44Z

Your expressions to compute mask are incorrect. You changed the parentheses in the second expression but both are incorrect, and do not even compile:

int mask = ((1 << (L - R + 1) - 1) << R; ... mask = ((1 << (L - R + 1) - 1 << R);

should be written:

mask = ((1UL << (L - R + 1)) - 1) << R;

YePhIcK · Accepted Answer · 2017-02-16 18:14:50Z

Looks like what you are asked to do is use the bit fields. For example, given the type

struct bits{ int a:5; unsigned short b:3; unsigned char c:2; bool d:1; };

The above struct will have 4 members, each of specific bit length.

If you union that struct with an int you get a "dual-view" of the bits. As a list of fields or as a single integer:

union U{ struct bits fields; int i; };

Now the code like

U u; u.i = 0; u.fields.b = true;

becomes valid and gives you access to either the whole number or individual bit fields.

Collectives™ on Stack Overflow

Updating fields of bits incorrectly

3 Answers 3

Comments

Comments

Comments

Linked

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

Comments

Comments

Comments

Linked

Related