-
-
Notifications
You must be signed in to change notification settings - Fork 56.5k
Description
System Information
OpenCV: 4.11.0
Operating System / Platform Fedora rawhide (42)
Compiler & compiler version: gcc: 15.0.0-0.4.fc42
here is the scratch build (only available for a few days) : https://koji.fedoraproject.org/koji/taskinfo?taskID=127846000
"
See https://gcc.gnu.org/r15-1919 and
https://gcc.gnu.org/pipermail/gcc-patches/2024-April/649747.html
In short, for intrinsics on the various architectures, the supported API
are the intrinsics declared in the header, not the underlying builtin
functions, which can be added/removed as needed, those are just
implementation details how to implement those intrinsics.
opencv is clearly using a builtin directly, don't.
That said, I don't know much about the VSX intrinsics, for vector
conversions there is also __builtin_convertvector (both in GCC and LLVM)
which allows to say convert a vector of floats to vector of doubles or vice
versa.
In any case, I think raise this upstream and if they don't know what to do,
get them talk to the author of those changes.
Jakub
"
Let me know if you need more information
Thank you,
Detailed description
[ 1%] Building CXX object CMakeFiles/ade.dir/3rdparty/ade/ade-0.1.2e/sources/ade/source/memory_accessor.cpp.o
/usr/bin/g++ -DVK_NO_PROTOTYPES -I/builddir/build/BUILD/opencv-4.11.0-build/opencv-4.11.0/redhat-linux-build -I/usr/include/vulkan -I/builddir/build/BUILD/opencv-4.11.0-build/opencv-4.11.0/redhat-linux-build/3rdparty/ade/ade-0.1.2e/sources/ade/include -isystem /usr/include/gdal -isystem /usr/include/coin -isystem /usr/include/openblas -O2 -fexceptions -g -grecord-gcc-switches -pipe -Wall -Werror=format-security -Wp,-U_FORTIFY_SOURCE,-D_FORTIFY_SOURCE=3 -Wp,-D_GLIBCXX_ASSERTIONS -specs=/usr/lib/rpm/redhat/redhat-hardened-cc1 -fstack-protector-strong -specs=/usr/lib/rpm/redhat/redhat-annobin-cc1 -m64 -mcpu=power8 -mtune=power8 -fasynchronous-unwind-tables -fstack-clash-protection -fsigned-char -W -Wall -Wreturn-type -Wnon-virtual-dtor -Waddress -Wsequence-point -Wformat -Wformat-security -Wmissing-declarations -Wundef -Winit-self -Wpointer-arith -Wshadow -Wsign-promo -Wuninitialized -Wsuggest-override -Wno-delete-non-virtual-dtor -Wno-comment -Wimplicit-fallthrough=3 -Wno-strict-overflow -fdiagnostics-show-option -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -fvisibility=hidden -fvisibility-inlines-hidden -fopenmp -DNDEBUG -DNDEBUG -std=c++17 -fPIC -MD -MT CMakeFiles/ade.dir/3rdparty/ade/ade-0.1.2e/sources/ade/source/memory_accessor.cpp.o -MF CMakeFiles/ade.dir/3rdparty/ade/ade-0.1.2e/sources/ade/source/memory_accessor.cpp.o.d -o CMakeFiles/ade.dir/3rdparty/ade/ade-0.1.2e/sources/ade/source/memory_accessor.cpp.o -c /builddir/build/BUILD/opencv-4.11.0-build/opencv-4.11.0/redhat-linux-build/3rdparty/ade/ade-0.1.2e/sources/ade/source/memory_accessor.cpp
In file included from /builddir/build/BUILD/opencv-4.11.0-build/opencv-4.11.0/modules/core/include/opencv2/core/base.hpp:679,
from /builddir/build/BUILD/opencv-4.11.0-build/opencv-4.11.0/modules/core/include/opencv2/core.hpp:53,
from /builddir/build/BUILD/opencv-4.11.0-build/opencv-4.11.0/modules/core/include/opencv2/core/utility.hpp:56,
from /builddir/build/BUILD/opencv-4.11.0-build/opencv-4.11.0/modules/core/src/precomp.hpp:53,
from /builddir/build/BUILD/opencv-4.11.0-build/opencv-4.11.0/modules/core/src/alloc.cpp:43:
/builddir/build/BUILD/opencv-4.11.0-build/opencv-4.11.0/modules/core/include/opencv2/core/vsx_utils.hpp: In function ‘vec_float4 vec_cvfo(const vec_double2&)’:
/builddir/build/BUILD/opencv-4.11.0-build/opencv-4.11.0/modules/core/include/opencv2/core/vsx_utils.hpp:260:54: error: ‘__builtin_vsx_xvcvdpsp’ was not declared in this scope; did you mean ‘__builtin_vsx_xscvdpsp’?
260 | VSX_REDIRECT_1RG(vec_float4, vec_double2, vec_cvfo, __builtin_vsx_xvcvdpsp)
| ^~~~~~~~~~~~~~~~~~~~~~
/builddir/build/BUILD/opencv-4.11.0-build/opencv-4.11.0/modules/core/include/opencv2/core/vsx_utils.hpp:102:43: note: in definition of macro ‘VSX_REDIRECT_1RG’
102 | VSX_FINLINE(rt) fnm(const rg& a) { return fn2(a); }
| ^~~
/builddir/build/BUILD/opencv-4.11.0-build/opencv-4.11.0/modules/core/include/opencv2/core/vsx_utils.hpp: In function ‘vec_double2 vec_cvfo(const vec_float4&)’:
/builddir/build/BUILD/opencv-4.11.0-build/opencv-4.11.0/modules/core/include/opencv2/core/vsx_utils.hpp:261:54: error: ‘__builtin_vsx_xvcvspdp’ was not declared in this scope; did you mean ‘__builtin_vsx_xscvspdp’?
261 | VSX_REDIRECT_1RG(vec_double2, vec_float4, vec_cvfo, __builtin_vsx_xvcvspdp)
| ^~~~~~~~~~~~~~~~~~~~~~
/builddir/build/BUILD/opencv-4.11.0-build/opencv-4.11.0/modules/core/include/opencv2/core/vsx_utils.hpp:102:43: note: in definition of macro ‘VSX_REDIRECT_1RG’
102 | VSX_FINLINE(rt) fnm(const rg& a) { return fn2(a); }
| ^~~
Steps to reproduce
git clone https://src.fedoraproject.org/forks/packit/rpms/opencv.git
git checkout rawhide-update-pull_from_upstream
fedpkg --release rawhide scratch-build --target f42-build-side-103300 (this one one need to be a fedora packager ) you can build locally with koji side-tag f42-build-side-1103300 repo, let me know if you need that instructions
Thank you
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)