Rewrite Universal Intrinsic code: ImgProc (CV_SIMD_WIDTH related Part) #24166

hanliutong · 2023-08-16T14:55:08Z

Related PR: #24058, #24132. The goal of this series of PRs is to modify the SIMD code blocks in the opencv/modules/imgproc folder by using the new Universal Intrinsic API.

The modification of this PR mainly focuses on the code that uses the CV_SIMD_WIDTH macro. This macro is sometimes used for loop tail processing, such as box_filter.simd.hpp and morph.simd.hpp.

#if CV_SIMD
int i = 0;
for (i < n - v_uint16::nlanes; i += v_uint16::nlanes) {
// some universal intrinsic code
// e.g. v_uint16...
}
#if CV_SIMD_WIDTH > 16
for (i < n - v_uint16x8::nlanes; i += v_uint16x8::nlanes) {
// handle loop tail by 128 bit SIMD
// e.g. v_uint16x8
}
#endif //CV_SIMD_WIDTH 
#endif// CV_SIMD

The main contradiction is that the variable-length Universal Intrinsic backend cannot use 128bit fixed-length data structures. Therefore, this PR uses the scalar loop to handle the loop tail.

This PR is marked as draft because the modification of the box_filter.simd.hpp file caused a compilation error. The cause of the error is initially believed to be due to an internal error in the GCC compiler.

box_filter.simd.hpp:1162:5: internal compiler error: Segmentation fault
 1162 |     }
      |     ^
0xe03883 crash_signal
        /wafer/share/gcc/gcc/toplev.cc:314
0x7ff261c4251f ???
        ./signal/../sysdeps/unix/sysv/linux/x86_64/libc_sigaction.c:0
0x6bde48 hash_set<rtl_ssa::set_info*, false, default_hash_traits<rtl_ssa::set_info*> >::iterator::operator*()
        /wafer/share/gcc/gcc/hash-set.h:125
0x6bde48 extract_single_source
        /wafer/share/gcc/gcc/config/riscv/riscv-vsetvl.cc:1184
0x6bde48 extract_single_source
        /wafer/share/gcc/gcc/config/riscv/riscv-vsetvl.cc:1174
0x119ad9e pass_vsetvl::propagate_avl() const
        /wafer/share/gcc/gcc/config/riscv/riscv-vsetvl.cc:4087
0x119ceaf pass_vsetvl::execute(function*)
        /wafer/share/gcc/gcc/config/riscv/riscv-vsetvl.cc:4344
0x119ceaf pass_vsetvl::execute(function*)
        /wafer/share/gcc/gcc/config/riscv/riscv-vsetvl.cc:4325
Please submit a full bug report, with preprocessed source (by using -freport-bug).
Please include the complete backtrace with any bug report.

This PR can be compiled with Clang 16, and opencv_test_imgproc is passed on QEMU.

Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

I agree to contribute to the project under Apache 2 License.
To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
The PR is proposed to the proper branch
There is a reference to the original bug report and related work
There is accuracy test, performance test and test data in opencv_extra repository, if applicable
Patch to opencv_extra has the same branch name.
The feature is well documented and sample code can be built with the project CMake

force_builders=Custom
Xbuild_image:Custom=riscv-clang-rvv
build_image:Custom=riscv-gcc-rvv
Xbuild_image:Custom=riscv-clang-rvv-128
Xbuild_image:Custom=riscv-gcc-rvv-07
test_modules:Custom=core,imgproc,dnn
buildworker:Custom=linux-1
test_timeout:Custom=600
build_contrib:Custom=OFF

mshabunin · 2023-08-18T09:05:30Z

I've reproduced the GCC issue with 13.2.0, I propose to partially revert modifications made in the box_filter: leave changes made by automated refactoring tool and only revert preprocessor guards.

hanliutong · 2023-08-18T14:44:11Z

I reverted the necessary code block (only one, line 1137), and then my riscv64-unknown-linux-gnu-g++ 13.2.0 (gc891d8dc23e) can successfully compile it now.

mshabunin · 2023-08-22T13:33:34Z

Apparently the GCC issue have been fixed already on branch releases/gcc-13 (https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111074).

This reverts commit 8ab6c22.

hanliutong · 2023-08-23T02:21:24Z

Compiled on my device with riscv64-unknown-linux-gnu-g++ (g6c54fc2560c) 13.2.1 20230822

mshabunin · 2023-09-13T10:35:27Z

modules/imgproc/src/shapedescr.cpp

        }
+    #else // #if (CV_SIMD_WIDTH > 16 && !CV_SIMD) equal to "#elif CV_SIMD_SCALABLE"
+
+        if( i < npoints )


I think it would be better to extract a function or two here to avoid duplication of code.

mshabunin

This PR causes several test failures for me:

[  FAILED  ] Imgproc_Erode.accuracy
[  FAILED  ] Imgproc_Dilate.accuracy
[  FAILED  ] Imgproc_MorphologyEx.accuracy
[  FAILED  ] Imgproc_Blur.accuracy
[  FAILED  ] Imgproc_Morphology.iterated

I used GCC from releases/gcc-13 branch and qemu 7.1.0

hanliutong · 2023-09-13T14:01:28Z

[  FAILED  ] Imgproc_Erode.accuracy
[  FAILED  ] Imgproc_Dilate.accuracy
[  FAILED  ] Imgproc_MorphologyEx.accuracy
[  FAILED  ] Imgproc_Blur.accuracy
[  FAILED  ] Imgproc_Morphology.iterated

I can reproduce this problem using GCC 13, but not on Clang 16 (all tests pass). 🤔This is weird, I'm working on it.

mshabunin · 2023-09-13T16:15:46Z

Perhaps we can just disable some of the blocks for now in scalable mode while keeping refactored code. It could be GCC-specific issue which might be resolved later.

opencv-alalek · 2023-09-13T19:39:34Z

Perhaps we can just disable some of the blocks for now in scalable mode while keeping refactored code. It could be GCC-specific issue which might be resolved later.

I can confirm too, that Clang 16 tests are fine: https://pullrequest.opencv.org/buildbot/builders/precommit_custom_linux/builds/100347

I believe it is OK to keep code enabled on experimental platforms due to compiler issue (until we have compiler update fix or robust local fix with minimal impact).

If there are no other issues, then we could merge it.

BTW, GCC 13 (from upstream-202304) fails with ICE (internal compiler error):
https://pullrequest.opencv.org/buildbot/builders/precommit_custom_linux/builds/100348
(need to upgrade compiler)

hanliutong · 2023-09-14T03:36:26Z

Updated, I disabled the necessary code block for GCC.

mshabunin · 2023-09-14T07:18:29Z

modules/imgproc/src/morph.simd.hpp

-#if CV_SIMD
+// Only for clang temporarily. PR #24166
+// TODO: Remove __clang__ macro when GCC is available.
+#if (CV_SIMD || CV_SIMD_SCALABLE) && defined(__clang__)


Shouldn't it be:

Suggested change

#if (CV_SIMD || CV_SIMD_SCALABLE) && defined(__clang__)

#if (CV_SIMD || (CV_SIMD_SCALABLE && defined(__clang__))

Otherwise regular CV_SIMD will not work with GCC and MSVC on other platforms.

Both variants don't look correct.

It is not a problem of SCALABLE.
It is related to GCC for RISC-V platform.

BTW, I still recommend to avoid trying to disable code for experimental platforms on mainline branches. Compiler could be fixed, but disabled code is in repository forever.

I got your point about the macro that we need to ensure that this code block is only disable on GCC on the RISC-V platform, then we can use #if CV_SIMD || (CV_SIMD_SCALABLE && !(defined(__gnu_linux__) && defined(__riscv))) to disable the code block on riscv gcc.

I also got the point that we can keep #if (CV_SIMD || CV_SIMD_SCALABLE) and leave the broken part on riscv GCC and wait for GCC upstream to fix it. It also make sence.

Which one do we prefer?

Let's disable problematic blocks in SCALABLE mode to avoid breaking more tests, it will help us to establish CI testing. And then we will gradually resolve these issues (whether they are in GCC or qemu or our implementation) and re-enable these blocks. Our main goal is to get rid of obsolete constructions while keeping builds green (gcc-13 and clang-16).

It is not necessary to distinguish between clang and gcc, just leave #if CV_SIMD and add a TODO: comment.

Done

#if CV_SIMD // TODO: enable for CV_SIMD_SCALABLE, GCC 13 related

mshabunin · 2023-09-14T10:35:57Z

modules/imgproc/src/shapedescr.cpp

        return Rect();

-#if CV_SIMD
+#if (CV_SIMD || CV_SIMD_SCALABLE)


Let's also disable this block for SCALABLE mode and revert all unnecessary modifications leaving only automatically replaced intrinsics constructs, otherwise it becomes too complex. This function should be refactored separately: first for wide intrinsics, then for scalable.

Done

#if CV_SIMD // TODO: enable for CV_SIMD_SCALABLE, loop tail related.

asmorkalov

👍

Rewrite Universal Intrinsic code: float related part #24325 The goal of this series of PRs is to modify the SIMD code blocks guarded by CV_SIMD macro: rewrite them by using the new Universal Intrinsic API. The series of PRs is listed below: #23885 First patch, an example #23980 Core module #24058 ImgProc module, part 1 #24132 ImgProc module, part 2 #24166 ImgProc module, part 3 #24301 Features2d and calib3d module #24324 Gapi module This patch (hopefully) is the last one in the series. This patch mainly involves 3 parts 1. Add some modifications related to float (CV_SIMD_64F) 2. Use `#if (CV_SIMD || CV_SIMD_SCALABLE)` instead of `#if CV_SIMD || CV_SIMD_SCALABLE`, then we can get the `CV_SIMD` module that is not enabled for `CV_SIMD_SCALABLE` by looking for `if CV_SIMD` 3. Summary of `CV_SIMD` blocks that remains unmodified: Updated comments - Some blocks will cause test fail when enable for RVV, marked as `TODO: enable for CV_SIMD_SCALABLE, ....` - Some blocks can not be rewrited directly. (Not commented in the source code, just listed here) - ./modules/core/src/mathfuncs_core.simd.hpp (Vector type wrapped in class/struct) - ./modules/imgproc/src/color_lab.cpp (Array of vector type) - ./modules/imgproc/src/color_rgb.simd.hpp (Array of vector type) - ./modules/imgproc/src/sumpixels.simd.hpp (fixed length algorithm, strongly ralated with `CV_SIMD_WIDTH`) These algorithms will need to be redesigned to accommodate scalable backends. ### Pull Request Readiness Checklist See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request - [ ] I agree to contribute to the project under Apache 2 License. - [ ] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV - [ ] The PR is proposed to the proper branch - [ ] There is a reference to the original bug report and related work - [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable Patch to opencv_extra has the same branch name. - [ ] The feature is well documented and sample code can be built with the project CMake

Rewrite Universal Intrinsic code: float related part opencv#24325 The goal of this series of PRs is to modify the SIMD code blocks guarded by CV_SIMD macro: rewrite them by using the new Universal Intrinsic API. The series of PRs is listed below: opencv#23885 First patch, an example opencv#23980 Core module opencv#24058 ImgProc module, part 1 opencv#24132 ImgProc module, part 2 opencv#24166 ImgProc module, part 3 opencv#24301 Features2d and calib3d module opencv#24324 Gapi module This patch (hopefully) is the last one in the series. This patch mainly involves 3 parts 1. Add some modifications related to float (CV_SIMD_64F) 2. Use `#if (CV_SIMD || CV_SIMD_SCALABLE)` instead of `#if CV_SIMD || CV_SIMD_SCALABLE`, then we can get the `CV_SIMD` module that is not enabled for `CV_SIMD_SCALABLE` by looking for `if CV_SIMD` 3. Summary of `CV_SIMD` blocks that remains unmodified: Updated comments - Some blocks will cause test fail when enable for RVV, marked as `TODO: enable for CV_SIMD_SCALABLE, ....` - Some blocks can not be rewrited directly. (Not commented in the source code, just listed here) - ./modules/core/src/mathfuncs_core.simd.hpp (Vector type wrapped in class/struct) - ./modules/imgproc/src/color_lab.cpp (Array of vector type) - ./modules/imgproc/src/color_rgb.simd.hpp (Array of vector type) - ./modules/imgproc/src/sumpixels.simd.hpp (fixed length algorithm, strongly ralated with `CV_SIMD_WIDTH`) These algorithms will need to be redesigned to accommodate scalable backends. ### Pull Request Readiness Checklist See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request - [ ] I agree to contribute to the project under Apache 2 License. - [ ] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV - [ ] The PR is proposed to the proper branch - [ ] There is a reference to the original bug report and related work - [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable Patch to opencv_extra has the same branch name. - [ ] The feature is well documented and sample code can be built with the project CMake

Rewrite Universal Intrinsic code: ImgProc (CV_SIMD_WIDTH related Part) opencv#24166 Related PR: opencv#24058, opencv#24132. The goal of this series of PRs is to modify the SIMD code blocks in the opencv/modules/imgproc folder by using the new Universal Intrinsic API. The modification of this PR mainly focuses on the code that uses the `CV_SIMD_WIDTH` macro. This macro is sometimes used for loop tail processing, such as `box_filter.simd.hpp` and `morph.simd.hpp`. ```cpp #if CV_SIMD int i = 0; for (i < n - v_uint16::nlanes; i += v_uint16::nlanes) { // some universal intrinsic code // e.g. v_uint16... } #if CV_SIMD_WIDTH > 16 for (i < n - v_uint16x8::nlanes; i += v_uint16x8::nlanes) { // handle loop tail by 128 bit SIMD // e.g. v_uint16x8 } #endif //CV_SIMD_WIDTH #endif// CV_SIMD ``` The main contradiction is that the variable-length Universal Intrinsic backend cannot use 128bit fixed-length data structures. Therefore, this PR uses the scalar loop to handle the loop tail. This PR is marked as draft because the modification of the `box_filter.simd.hpp` file caused a compilation error. The cause of the error is initially believed to be due to an internal error in the GCC compiler. ```bash box_filter.simd.hpp:1162:5: internal compiler error: Segmentation fault 1162 | } | ^ 0xe03883 crash_signal /wafer/share/gcc/gcc/toplev.cc:314 0x7ff261c4251f ??? ./signal/../sysdeps/unix/sysv/linux/x86_64/libc_sigaction.c:0 0x6bde48 hash_set<rtl_ssa::set_info*, false, default_hash_traits<rtl_ssa::set_info*> >::iterator::operator*() /wafer/share/gcc/gcc/hash-set.h:125 0x6bde48 extract_single_source /wafer/share/gcc/gcc/config/riscv/riscv-vsetvl.cc:1184 0x6bde48 extract_single_source /wafer/share/gcc/gcc/config/riscv/riscv-vsetvl.cc:1174 0x119ad9e pass_vsetvl::propagate_avl() const /wafer/share/gcc/gcc/config/riscv/riscv-vsetvl.cc:4087 0x119ceaf pass_vsetvl::execute(function*) /wafer/share/gcc/gcc/config/riscv/riscv-vsetvl.cc:4344 0x119ceaf pass_vsetvl::execute(function*) /wafer/share/gcc/gcc/config/riscv/riscv-vsetvl.cc:4325 Please submit a full bug report, with preprocessed source (by using -freport-bug). Please include the complete backtrace with any bug report. ``` This PR can be compiled with Clang 16, and `opencv_test_imgproc` is passed on QEMU. ### Pull Request Readiness Checklist See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request - [ ] I agree to contribute to the project under Apache 2 License. - [ ] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV - [ ] The PR is proposed to the proper branch - [ ] There is a reference to the original bug report and related work - [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable Patch to opencv_extra has the same branch name. - [ ] The feature is well documented and sample code can be built with the project CMake

Rewrite Universal Intrinsic code: float related part opencv#24325 The goal of this series of PRs is to modify the SIMD code blocks guarded by CV_SIMD macro: rewrite them by using the new Universal Intrinsic API. The series of PRs is listed below: opencv#23885 First patch, an example opencv#23980 Core module opencv#24058 ImgProc module, part 1 opencv#24132 ImgProc module, part 2 opencv#24166 ImgProc module, part 3 opencv#24301 Features2d and calib3d module opencv#24324 Gapi module This patch (hopefully) is the last one in the series. This patch mainly involves 3 parts 1. Add some modifications related to float (CV_SIMD_64F) 2. Use `#if (CV_SIMD || CV_SIMD_SCALABLE)` instead of `#if CV_SIMD || CV_SIMD_SCALABLE`, then we can get the `CV_SIMD` module that is not enabled for `CV_SIMD_SCALABLE` by looking for `if CV_SIMD` 3. Summary of `CV_SIMD` blocks that remains unmodified: Updated comments - Some blocks will cause test fail when enable for RVV, marked as `TODO: enable for CV_SIMD_SCALABLE, ....` - Some blocks can not be rewrited directly. (Not commented in the source code, just listed here) - ./modules/core/src/mathfuncs_core.simd.hpp (Vector type wrapped in class/struct) - ./modules/imgproc/src/color_lab.cpp (Array of vector type) - ./modules/imgproc/src/color_rgb.simd.hpp (Array of vector type) - ./modules/imgproc/src/sumpixels.simd.hpp (fixed length algorithm, strongly ralated with `CV_SIMD_WIDTH`) These algorithms will need to be redesigned to accommodate scalable backends. ### Pull Request Readiness Checklist See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request - [ ] I agree to contribute to the project under Apache 2 License. - [ ] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV - [ ] The PR is proposed to the proper branch - [ ] There is a reference to the original bug report and related work - [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable Patch to opencv_extra has the same branch name. - [ ] The feature is well documented and sample code can be built with the project CMake

Rewrite Universal Intrinsic code: ImgProc (CV_SIMD_WIDTH related Part) opencv#24166 Related PR: opencv#24058, opencv#24132. The goal of this series of PRs is to modify the SIMD code blocks in the opencv/modules/imgproc folder by using the new Universal Intrinsic API. The modification of this PR mainly focuses on the code that uses the `CV_SIMD_WIDTH` macro. This macro is sometimes used for loop tail processing, such as `box_filter.simd.hpp` and `morph.simd.hpp`. ```cpp #if CV_SIMD int i = 0; for (i < n - v_uint16::nlanes; i += v_uint16::nlanes) { // some universal intrinsic code // e.g. v_uint16... } #if CV_SIMD_WIDTH > 16 for (i < n - v_uint16x8::nlanes; i += v_uint16x8::nlanes) { // handle loop tail by 128 bit SIMD // e.g. v_uint16x8 } #endif //CV_SIMD_WIDTH #endif// CV_SIMD ``` The main contradiction is that the variable-length Universal Intrinsic backend cannot use 128bit fixed-length data structures. Therefore, this PR uses the scalar loop to handle the loop tail. This PR is marked as draft because the modification of the `box_filter.simd.hpp` file caused a compilation error. The cause of the error is initially believed to be due to an internal error in the GCC compiler. ```bash box_filter.simd.hpp:1162:5: internal compiler error: Segmentation fault 1162 | } | ^ 0xe03883 crash_signal /wafer/share/gcc/gcc/toplev.cc:314 0x7ff261c4251f ??? ./signal/../sysdeps/unix/sysv/linux/x86_64/libc_sigaction.c:0 0x6bde48 hash_set<rtl_ssa::set_info*, false, default_hash_traits<rtl_ssa::set_info*> >::iterator::operator*() /wafer/share/gcc/gcc/hash-set.h:125 0x6bde48 extract_single_source /wafer/share/gcc/gcc/config/riscv/riscv-vsetvl.cc:1184 0x6bde48 extract_single_source /wafer/share/gcc/gcc/config/riscv/riscv-vsetvl.cc:1174 0x119ad9e pass_vsetvl::propagate_avl() const /wafer/share/gcc/gcc/config/riscv/riscv-vsetvl.cc:4087 0x119ceaf pass_vsetvl::execute(function*) /wafer/share/gcc/gcc/config/riscv/riscv-vsetvl.cc:4344 0x119ceaf pass_vsetvl::execute(function*) /wafer/share/gcc/gcc/config/riscv/riscv-vsetvl.cc:4325 Please submit a full bug report, with preprocessed source (by using -freport-bug). Please include the complete backtrace with any bug report. ``` This PR can be compiled with Clang 16, and `opencv_test_imgproc` is passed on QEMU. ### Pull Request Readiness Checklist See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request - [ ] I agree to contribute to the project under Apache 2 License. - [ ] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV - [ ] The PR is proposed to the proper branch - [ ] There is a reference to the original bug report and related work - [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable Patch to opencv_extra has the same branch name. - [ ] The feature is well documented and sample code can be built with the project CMake

Rewrite Universal Intrinsic code: float related part opencv#24325 The goal of this series of PRs is to modify the SIMD code blocks guarded by CV_SIMD macro: rewrite them by using the new Universal Intrinsic API. The series of PRs is listed below: opencv#23885 First patch, an example opencv#23980 Core module opencv#24058 ImgProc module, part 1 opencv#24132 ImgProc module, part 2 opencv#24166 ImgProc module, part 3 opencv#24301 Features2d and calib3d module opencv#24324 Gapi module This patch (hopefully) is the last one in the series. This patch mainly involves 3 parts 1. Add some modifications related to float (CV_SIMD_64F) 2. Use `#if (CV_SIMD || CV_SIMD_SCALABLE)` instead of `#if CV_SIMD || CV_SIMD_SCALABLE`, then we can get the `CV_SIMD` module that is not enabled for `CV_SIMD_SCALABLE` by looking for `if CV_SIMD` 3. Summary of `CV_SIMD` blocks that remains unmodified: Updated comments - Some blocks will cause test fail when enable for RVV, marked as `TODO: enable for CV_SIMD_SCALABLE, ....` - Some blocks can not be rewrited directly. (Not commented in the source code, just listed here) - ./modules/core/src/mathfuncs_core.simd.hpp (Vector type wrapped in class/struct) - ./modules/imgproc/src/color_lab.cpp (Array of vector type) - ./modules/imgproc/src/color_rgb.simd.hpp (Array of vector type) - ./modules/imgproc/src/sumpixels.simd.hpp (fixed length algorithm, strongly ralated with `CV_SIMD_WIDTH`) These algorithms will need to be redesigned to accommodate scalable backends. ### Pull Request Readiness Checklist See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request - [ ] I agree to contribute to the project under Apache 2 License. - [ ] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV - [ ] The PR is proposed to the proper branch - [ ] There is a reference to the original bug report and related work - [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable Patch to opencv_extra has the same branch name. - [ ] The feature is well documented and sample code can be built with the project CMake

Add remaining (CV_SIMD_WIDTH related) modifications.

895e1a3

mshabunin self-assigned this Aug 17, 2023

Temporarily revert a code block.

8ab6c22

hanliutong marked this pull request as ready for review August 18, 2023 14:44

Revert "Temporarily revert a code block."

4909c60

This reverts commit 8ab6c22.

mshabunin reviewed Sep 13, 2023

View reviewed changes

Disable RVV blocks for GCC temporarily.

c9d5a64

Extract functions to avoid duplication of code.

0e77302

hanliutong force-pushed the rewrite-remaining branch from 4a048c5 to 0e77302 Compare September 14, 2023 06:47

mshabunin reviewed Sep 14, 2023

View reviewed changes

Revert code.

d81ceac

mshabunin approved these changes Sep 14, 2023

View reviewed changes

asmorkalov added category: imgproc platform: riscv labels Sep 15, 2023

asmorkalov added this to the 4.9.0 milestone Sep 15, 2023

asmorkalov approved these changes Sep 19, 2023

View reviewed changes

asmorkalov merged commit 320c0bf into opencv:4.x Sep 19, 2023

asmorkalov added the GSoC label Sep 19, 2023

hanliutong mentioned this pull request Sep 27, 2023

Rewrite Universal Intrinsic code: float related part #24325

Merged

6 tasks

asmorkalov mentioned this pull request Sep 28, 2023

(5.x) Merge 4.x #24338

Merged

asmorkalov mentioned this pull request Oct 17, 2023

Speed up line merging in INTER_AREA #24412

Merged

6 tasks

asmorkalov mentioned this pull request Oct 31, 2023

About the performance of opencv of the sizeless instruction(Riscv vector, SVE .etc) #21780

Closed

4 tasks

	#if (CV_SIMD \|\| CV_SIMD_SCALABLE) && defined(__clang__)
	#if (CV_SIMD \|\| (CV_SIMD_SCALABLE && defined(__clang__))

Uh oh!

Rewrite Universal Intrinsic code: ImgProc (CV_SIMD_WIDTH related Part) #24166

Rewrite Universal Intrinsic code: ImgProc (CV_SIMD_WIDTH related Part) #24166

Uh oh!

Conversation

hanliutong commented Aug 16, 2023 • edited by opencv-alalek Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Pull Request Readiness Checklist

Uh oh!

mshabunin commented Aug 18, 2023

Uh oh!

hanliutong commented Aug 18, 2023

Uh oh!

mshabunin commented Aug 22, 2023

Uh oh!

hanliutong commented Aug 23, 2023

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mshabunin left a comment

Choose a reason for hiding this comment

Uh oh!

hanliutong commented Sep 13, 2023

Uh oh!

mshabunin commented Sep 13, 2023

Uh oh!

opencv-alalek commented Sep 13, 2023

Uh oh!

hanliutong commented Sep 14, 2023

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

asmorkalov left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

hanliutong commented Aug 16, 2023 •

edited by opencv-alalek

Loading