This question already has answers here:
In C bits, multiply by 3 and divide by 16
(5 answers)
Closed 8 years ago.
I need to multiply a number by 3/16, rounding to zero using only bitwise operations such as ! ~ & ^ | + << >>. So far I have the following, the only problem is it doesn't work when the number is negative, it always rounds down rather than to zero. I know there should be bitwise if statement that if x is negative then add 15. But I dont know how to implement it, any help is appreciated.
int ezThreeSixteenths(int x) {
int times_two = x << 1;
int times_three = times_two + x;
int divide_eight = times_three >> 4;
int a = 0b11111111;
int a1 = a << 8;
int a2 = a << 16;
int a3 = 0b11111 << 24;
int mask = a | a1 | a2 | a3;
int final = divide_eight & mask;
return final;
}
If you have a function that you are satisfied works when it's positive, test the MSB to detect a negative bit, if so take the two's complement (you don't say whether you can use - as well as + but you can use ^ and +), run your function, then take the two's complement again.
Use twos complement to convert a negative number to a positive number. Then when you're done convert the positive number back to a negative one?
Related
I am doing CSAPP's datalab, the isGreater function.
Here's the description
isGreater - if x > y then return 1, else return 0
Example: isGreater(4,5) = 0, isGreater(5,4) = 1
Legal ops: ! ~ & ^ | + << >>
Max ops: 24
Rating: 3
x and y are both int type.
So i consider to simulate the jg instruction to implement it.Here's my code
int isGreater(int x, int y)
{
int yComplement = ~y + 1;
int minusResult = x + yComplement; // 0xffffffff
int SF = (minusResult >> 31) & 0x1; // 1
int ZF = !minusResult; // 0
int xSign = (x >> 31) & 0x1; // 0
int ySign = (yComplement >> 31) & 0x1; // 1
int OF = !(xSign ^ ySign) & (xSign ^ SF); // 0
return !(OF ^ SF) & !ZF;
}
The jg instruction need SF == OF and ZF == 0.
But it can't pass a special case, that is, x = 0x7fffffff(INT_MAX), y = 0x80000000(INT_MIN).
I deduce it like this:
x + yComplement = 0xffffffff, so SF = 1, ZF = 0, since xSign != ySign, the OF is set to 0.
So, what's wrong with my code, is my OF setting operation wrong?
You're detecting overflow in the addition x + yComplement, rather than in the overall subtraction
-INT_MIN itself overflows in 2's complement; INT_MIN == -INT_MIN. This is the 2's complement anomaly1.
You should be getting fast-positive overflow detection for any negative number (other than INT_MIN) minus INT_MIN. The resulting addition will have signed overflow. e.g. -10 + INT_MIN overflows.
http://teaching.idallen.com/dat2343/10f/notes/040_overflow.txt has a table of input/output signs for add and subtraction. The cases that overflow are where the inputs signs are opposite but the result sign matches y.
SUBTRACTION SIGN BITS (for num1 - num2 = sum)
num1sign num2sign sumsign
---------------------------
0 0 0
0 0 1
0 1 0
*OVER* 0 1 1 (subtracting a negative is the same as adding a positive)
*OVER* 1 0 0 (subtracting a positive is the same as adding a negative)
1 0 1
1 1 0
1 1 1
You could use this directly with the original x and y, and only use yComplement as part of getting the minusResult. Adjust your logic to match this truth table.
Or you could use int ySign = (~y) >> 31; and leave the rest of your code unmodified. (Use a tmp to hold ~y so you only do the operation once, for this and yComplement). The one's complement inverse (~) does not suffer from the 2's complement anomaly.
Footnote 1: sign/magnitude and one's complement have two redundant ways to represent 0, instead of an value with no inverse.
Fun fact: if you make an integer absolute-value function, you should consider the result unsigned to avoid this problem. int can't represent the absolute value of INT_MIN.
Efficiency improvements:
If you use unsigned int, you don't need & 1 after a shift because logical shifts don't sign-extend. (And as a bonus, it would avoid C signed-overflow undefined behaviour in +: http://blog.llvm.org/2011/05/what-every-c-programmer-should-know.html).
Then (if you used uint32_t, or sizeof(unsigned) * CHAR_BIT instead of 31) you'd have a safe and portable implementation of 2's complement comparison. (signed shift semantics for negative numbers are implementation-defined in C.) I think you're using C as a sort of pseudo-code for bit operations, and aren't interested in actually writing a portable implementation, and that's fine. The way you're doing things will work on normal compilers on normal CPUs.
Or you can use & 0x80000000 to leave the high bits in place (but then you'd have to left shift your ! result).
It's just the lab's restriction, you can't use unsigned or any constant larger than 0xff(255)
Ok, so you don't have access to logical right shift. Still, you need at most one &1. It's ok to work with numbers where all you care about is the low bit, but where the rest hold garbage.
You eventually do & !ZF, which is either &0 or &1. Thus, any high garbage in OF` is wiped away.
You can also delay the >> 31 until after XORing together two numbers.
This is a fun problem that I want to optimize myself:
// untested, 13 operations
int isGreater_optimized(int x, int y)
{
int not_y = ~y;
int minus_y = not_y + 1;
int sum = x + minus_y;
int x_vs_y = x ^ y; // high bit = 1 if they were opposite signs: OF is possible
int x_vs_sum = x ^ sum; // high bit = 1 if they were opposite signs: OF is possible
int OF = (x_vs_y & x_vs_sum) >> 31; // high bits hold garbage
int SF = sum >> 31;
int non_zero = !!sum; // 0 or 1
return (~(OF ^ SF)) & non_zero; // high garbage is nuked by `& 1`
}
Note the use of ~ instead of ! to invert a value that has high garbage.
It looks like there's still some redundancy in calculating OF separately from SF, but actually the XORing of sum twice doesn't cancel out. x ^ sum is an input for &, and we XOR with sum after that.
We can delay the shifts even later, though, and I found some more optimizations by avoiding an extra inversion. This is 11 operations
// replace 31 with sizeof(int) * CHAR_BIT if you want. #include <limit.h>
// or use int32_t
int isGreater_optimized2(int x, int y)
{
int not_y = ~y;
int minus_y = not_y + 1;
int sum = x + minus_y;
int SF = sum; // value in the high bit, rest are garbage
int x_vs_y = x ^ y; // high bit = 1 if they were opposite signs: OF is possible
int x_vs_sum = x ^ sum; // high bit = 1 if they were opposite signs: OF is possible
int OF = x_vs_y & x_vs_sum; // low bits hold garbage
int less = (OF ^ SF);
int ZF = !sum; // 0 or 1
int le = (less >> 31) & ZF; // clears high garbage
return !le; // jg == jnle
}
I wondered if any compilers might see through this manual compare and optimize it into cmp edi, esi/ setg al, but no such luck :/ I guess that's not a pattern that they look for, because code that could have been written as x > y tends to be written that way :P
But anyway, here's the x86 asm output from gcc and clang on the Godbolt compiler explorer.
Assuming two's complement, INT_MIN's absolute value isn't representable as an int. So, yComplement == y (ie. still negative), and ySign is 1 instead of the desired 0.
You could instead calculate the sign of y like this (changing as little as possible in your code) :
int ySign = !((y >> 31) & 0x1);
For a more detailed analysis, and a more optimal alternative, check Peter Cordes' answer.
I'm working on a way to divide a signed integer by a power of 2 using only binary operators (<< >> + ^ ~ & | !), and the result has to be round toward 0. I came across this question also on Stackoverflow on the problem, however, I cannot understand why it works. Here's the solution:
int divideByPowerOf2(int x, int n)
{
return (x + ((x >> 31) & ((1 << n) + ~0))) >> n;
}
I understand the x >> 31 part (only add the next part if x is negative, because if it's positive x will be automatically round toward 0). But what's bothering me is the (1 << n) + ~0 part. How can it work?
Assuming 2-complement, just bit-shifting the dividend is equivalent to a certain kind of division: not the conventional division where we round the dividend to next multiple of divisor toward zero. But another kind where we round the dividend toward negative infinity. I rediscovered that in Smalltalk, see http://smallissimo.blogspot.fr/2015/03/is-bitshift-equivalent-to-division-in.html.
For example, let's divide -126 by 8. traditionally, we would write
-126 = -15 * 8 - 6
But if we round toward infinity, we get a positive remainder and write it:
-126 = -16 * 8 + 2
The bit-shifting is performing the second operation, in term of bit patterns (assuming 8 bits long int for the sake of being short):
1000|0010 >> 3 = 1111|0000
1000|0010 = 1111|0000 * 0000|1000 + 0000|0010
So what if we want the traditional division with quotient rounded toward zero and remainder of same sign as dividend? Simple, we just have to add 1 to the quotient - if and only if the dividend is negative and the division is inexact.
You saw that x>>31 corresponds to first condition, dividend is negative, assuming int has 32 bits.
The second term corresponds to the second condition, if division is inexact.
See how are encoded -1, -2, -4, ... in two complement: 1111|1111 , 1111|1110 , 1111|1100. So the negation of nth power of two has n trailing zeros.
When the dividend has n trailing zeros and we divide by 2^n, then no need to add 1 to final quotient. In any other case, we need to add 1.
What ((1 << n) + ~0) is doing is creating a mask with n trailing ones.
The n last bits don't really matter, because we are going to shift to the right and just throw them away. So, if the division is exact, the n trailing bits of dividend are zero, and we just add n 1s that will be skipped. On the contrary, if the division is inexact, then one or more of the n trailing bits of the dividend is 1, and we are sure to cause a carry to the n+1 bit position: that's how we add 1 to the quotient (we add 2^n to the dividend). Does that explain it a bit more?
This is "write-only code": instead of trying to understand the code, try to create it by yourself.
For example, let's divide a number by 8 (shift right by 3).
If the number is negative, the normal right-shift rounds in the wrong direction. Let's "fix" it by adding a number:
int divideBy8(int x)
{
if (x >= 0)
return x >> 3;
else
return (x + whatever) >> 3;
}
Here you can come up with a mathematical formula for whatever, or do some trial and error. Anyway, here whatever = 7:
int divideBy8(int x)
{
if (x >= 0)
return x >> 3;
else
return (x + 7) >> 3;
}
How to unify the two cases? You need to make an expression that looks like this:
(x + stuff) >> 3
where stuff is 7 for negative x, and 0 for positive x. The trick here is using x >> 31, which is a 32-bit number whose bits are equal to the sign-bit of x: all 0 or all 1. So stuff is
(x >> 31) & 7
Combining all these, and replacing 8 and 7 by the more general power of 2, you get the code you asked about.
Note: in the description above, I assume that int represents a 32-bit hardware register, and hardware uses two's complement representation to do right shift.
OP's reference is of a C# code and so many subtle differences that cause it to be bad code with C, as this post is tagged.
int is not necessarily 32-bits so using a magic number of 32 does not make for a robust solution.
In particular (1 << n) + ~0 results in implementation defined behavior when n causes a bit to be shifted into the sign place. Not good coding.
Restricting code to only using "binary" operators << >> + ^ ~ & | ! encourages a coder to assume things about int which is not portable nor compliant with the C spec. So OP's posted code does not "work" in general, although may work in many common implementations.
OP code fails when int is not 2's complement, not uses the range [-2147483648 .. 2147483647] or when 1 << n uses implementation behavior that is not as expected.
// weak code
int divideByPowerOf2(int x, int n) {
return (x + ((x >> 31) & ((1 << n) + ~0))) >> n;
}
A simple alternative, assuming long long exceeds the range of int follows. I doubt this meets some corner of OP's goals, but OP's given goals encourages non-robust coding.
int divideByPowerOf2(int x, int n) {
long long ill = x;
if (x < 0) ill = -ill;
while (n--) ill >>= 1;
if (x < 0) ill = -ill;
return (int) ill;
}
I've been working on this puzzle for awhile. I'm trying to figure out how to rotate 4 bits in a number (x) around to the left (with wrapping) by n where 0 <= n <= 31.. The code will look like:
moveNib(int x, int n){
//... some code here
}
The trick is that I can only use these operators:
~ & ^ | + << >>
and of them only a combination of 25. I also can not use If statements, loops, function calls. And I may only use type int.
An example would be moveNib(0x87654321,1) = 0x76543218.
My attempt: I have figured out how to use a mask to store the the bits and all but I can't figure out how to move by an arbitrary number. Any help would be appreciated thank you!
How about:
uint32_t moveNib(uint32_t x, int n) { return x<<(n<<2) | x>>((8-n)<<2); }
It uses <<2 to convert from nibbles to bits, and then shifts the bits by that much. To handle wraparound, we OR by a copy of the number which has been shifted by the opposite amount in the opposite direciton. For example, with x=0x87654321 and n=1, the left part is shifted 4 bits to the left and becomes 0x76543210, and the right part is shifted 28 bits to the right and becomes 0x00000008, and when ORed together, the result is 0x76543218, as requested.
Edit: If - really isn't allowed, then this will get the same result (assuming an architecture with two's complement integers) without using it:
uint32_t moveNib(uint32_t x, int n) { return x<<(n<<2) | x>>((9+~n)<<2); }
Edit2: OK. Since you aren't allowed to use anything but int, how about this, then?
int moveNib(int x, int n) { return (x&0xffffffff)<<(n<<2) | (x&0xffffffff)>>((9+~n)<<2); }
The logic is the same as before, but we force the calculation to use unsigned integers by ANDing with 0xffffffff. All this assumes 32 bit integers, though. Is there anything else I have missed now?
Edit3: Here's one more version, which should be a bit more portable:
int moveNib(int x, int n) { return ((x|0u)<<((n&7)<<2) | (x|0u)>>((9+~(n&7))<<2))&0xffffffff; }
It caps n as suggested by chux, and uses |0u to convert to unsigned in order to avoid the sign bit duplication you get with signed integers. This works because (from the standard):
Otherwise, if the operand that has unsigned integer type has rank greater or equal to the rank of the type of the other operand, then the operand with signed integer type is converted to the type of the operand with unsigned integer type.
Since int and 0u have the same rank, but 0u is unsigned, then the result is unsigned, even though ORing with 0 otherwise would be a null operation.
It then truncates the result to the range of a 32-bit int so that the function will still work if ints have more bits than this (though the rotation will still be performed on the lowest 32 bits in that case. A 64-bit version would replace 7 by 15, 9 by 17 and truncate using 0xffffffffffffffff).
This solution uses 12 operators (11 if you skip the truncation, 10 if you store n&7 in a variable).
To see what happens in detail here, let's go through it for the example you gave: x=0x87654321, n=1. x|0u results in a the unsigned number 0x87654321u. (n&7)<<2=4, so we will shift 4 bits to the left, while ((9+~(n&7))<<2=28, so we will shift 28 bits to the right. So putting this together, we will compute 0x87654321u<<4 | 0x87654321u >> 28. For 32-bit integers, this is 0x76543210|0x8=0x76543218. But for 64-bit integers it is 0x876543210|0x8=0x876543218, so in that case we need to truncate to 32 bits, which is what the final &0xffffffff does. If the integers are shorter than 32 bits, then this won't work, but your example in the question had 32 bits, so I assume the integer types are at least that long.
As a small side-note: If you allow one operator which is not on the list, the sizeof operator, then we can make a version that works with all the bits of a longer int automatically. Inspired by Aki, we get (using 16 operators (remember, sizeof is an operator in C)):
int moveNib(int x, int n) {
int nbit = (n&((sizeof(int)<<1)+~0u))<<2;
return (x|0u)<<nbit | (x|0u)>>((sizeof(int)<<3)+1u+~nbit);
}
Without the additional restrictions, the typical rotate_left operation (by 0 < n < 32) is trivial.
uint32_t X = (x << 4*n) | (x >> 4*(8-n));
Since we are talking about rotations, n < 0 is not a problem. Rotation right by 1 is the same as rotation left by 7 units. Ie. nn=n & 7; and we are through.
int nn = (n & 7) << 2; // Remove the multiplication
uint32_t X = (x << nn) | (x >> (32-nn));
When nn == 0, x would be shifted by 32, which is undefined. This can be replaced simply with x >> 0, i.e. no rotation at all. (x << 0) | (x >> 0) == x.
Replacing the subtraction with addition: a - b = a + (~b+1) and simplifying:
int nn = (n & 7) << 2;
int mm = (33 + ~nn) & 31;
uint32_t X = (x << nn) | (x >> mm); // when nn=0, also mm=0
Now the only problem is in shifting a signed int x right, which would duplicate the sign bit. That should be cured by a mask: (x << nn) - 1
int nn = (n & 7) << 2;
int mm = (33 + ~nn) & 31;
int result = (x << nn) | ((x >> mm) & ((1 << nn) + ~0));
At this point we have used just 12 of the allowed operations -- next we can start to dig into the problem of sizeof(int)...
int nn = (n & (sizeof(int)-1)) << 2; // etc.
I need to convert from two's complement to sign-magnitude in C using only the operators
! ~ & ^ | + << >>
My approach is to find sign:
int sign = !(!(a>>31));
basically, if sign == 1 . I want to flip the number and add 1 else just want to display the number.
The thing is I can't use any loops, if statements etc.
This is what I'm working on:
int s_M = ((((a+1)>>31)^sign)+1)&sign;
any suggestions?
From http://graphics.stanford.edu/~seander/bithacks.html#IntegerAbs
int const mask = v >> 31;
unsigned int r = (v + mask) ^ mask;
Gives the absolute value (magnitude). If you wish to add the sign bit back simply mask and or the 32nd bit:
unsigned int s_M = r | (v & 0x80000000);
Or if you're looking for a one liner:
unsigned int s_M = ((v + (v >> 31)) ^ (v >> 31)) | (v & 0x80000000);
When you're converting from 2 complement, you should subtract 1, not add.
I'm not entirely sure what the output should be, but to obtain the magnitude you can do something like this:
int m = (a^(a>>31)) + sign;
Basically, shifting a negative number 31 bits to the right will make it all 1's, or 0xffffffff, which you can then use to xor the input number and make it positive. As you correctly noted sign needs to be added then for the correct result in that case.
If the input number was positive to begin with, the shift results in a zero and so the xor does nothing. Adding sign in that case also doesn't do anything, so it results in the input number.
To get the last bit you could use mask operation
int last_bit = 32 bit integer & 0x80000000
o/p may be 0 or 0x80000000
if it is 0 just display the given number else you have to perform the following operations to represent in signed magnitude
1) Subtract 1 from the number
2) perform 1s complement on the resultant ( that is negation ~)
3) Set the last bit of the resultant number
I mean ( ~ (num -`1) ) | 0x7fffffff
since your restricted not to use - operator. Perform the 2's complement on -1 and add it to the num.
To put it simple in one line
num & 0x80000000 ? printf("%d",(~(num+((~1)+1))) | 0x7fffffff) : printf("%d",num) ;
I was wondering how to get C to not extend my binary number when I bitshift to the left
int main ()
{
unsigned int binary_temp = 0b0100;
binary_temp = binary_temp << 2;
printf("%d", binary_temp);
return 0;
}
When I run that I want a return value of 0 since it has extended past the 4 digits I have, but right now it returns 16 (10000). How would I get C not to extend my number?
Edit: I would like to be able to work with the number in binary form so I need to have only 4 digits, and not just outputting the right number.
It does not extend your number but saves it as unsigned int type which is 4 bytes (32 bits) in size. You only fill the last 4 bits. To treat it as only 4 bits, use Bitwise AND with a Mask value. Here's example code:
int main()
{
unsigned int binary_temp = 0b0100;
binary_temp = (binary_temp << 2) & 0b1111;
printf("%u", binary_temp);
return 0;
}
You can bitwise AND the result with a 4 bit mask value:
binary_temp = (binary_temp << 2) & 0xF;
There is no 0b in standard C. You could use 4.
unsigned int /* prepare for wtf identifier: */
binary_temp = 4;
Left shifting by 2 is multiplying by 4. Why not?
binary_temp *= 4;
... and then reduce modulo 16?
binary_temp %= 16;
What sense is there to using binary operators, in this case? I see none.
The %d directive corresponds to an int argument, but the argument you're giving printf is an unsigned int. That's undefined behaviour.
printf("%u", binary_temp);
I'm sure whichever book you're reading will tell you about the %u directive.