| Issue |
118936
|
| Summary |
[clang++][precision] Cannot use float sqrt with cmath
|
| Labels |
clang
|
| Assignees |
|
| Reporter |
whokeke
|
hey guys, I found a confusing case, can anyone explain it?
Case 1 with camth :
```c++
#include <cmath>
float test(float a)
{
return a + sqrt(a);
}
```
With the command `clang++ -O3 -ffast-math -S`, we get
```asm
test(float):
fcvt d0, s0
fsqrt d1, d0
fadd d0, d1, d0
fcvt s0, d0
ret
```
But Case 2 with math.h
```c++
#include <math.h>
float test(float a)
{
return a + sqrt(a);
}
```
With the same command `clang++ -O3 -ffast-math -S`, we get
```asm
test(float):
fsqrt s1, s0
fadd s0, s1, s0
ret
```
I'm not familiar with the difference between math.h and cmath. Can anyone figure me out?
And can we just use float sqrt even with cmath ?
![]()
_______________________________________________
llvm-bugs mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
