hal_intrinsics operators breakage in 4.11 #27267
Description
System Information
System: Ubuntu x64 20.04 but not really matter
gcc 9.3
Detailed description
4.11 breaks backward compatibility for hal/intrin.hpp
Steps to reproduce
Sample code
#include <opencv2/core/hal/intrin.hpp>
#include <iostream>
int main()
{
cv::v_float32x4 x=cv::v_setall_f32(1.0f);
cv::v_float32x4 y=cv::v_setall_f32(2.0f);
cv::v_float32x4 z=x+y;
std::cout << z.get0() << std::endl;
}
Using OpenCV 4.10 and below:
$ g++ -I /path/to/opencv4.10/opencv4 -O2 -g -Wall test.cpp
$ ./a.out
3
Using OpenCV 4.11
$ g++ -I /path/to/opencv4.11/opencv4 -O2 -g -Wall test.cpp
test.cpp: In function 'int main()':
test.cpp:7:21: error: no match for 'operator+' (operand types are 'cv::hal_baseline::v_float32x4' and 'cv::hal_baseline::v_float32x4')
7 | cv::v_float32x4 z=x+y;
| ~^~
| | |
| | cv::hal_baseline::v_float32x4
| cv::hal_baseline::v_float32x4
Issue submission checklist
- I report the issue, it's not a question
- I checked the problem with documentation, FAQ, open issues, forum.opencv.org, Stack Overflow, etc and have not found any solution
- I updated to the latest OpenCV version and the issue is still there
- There is reproducer code and related data files (videos, images, onnx, etc)