Skip to content

Fixed armv8 compilation warnings#24681

Merged
asmorkalov merged 2 commits intoopencv:4.xfrom
WanliZhong:err_armv8
Dec 12, 2023
Merged

Fixed armv8 compilation warnings#24681
asmorkalov merged 2 commits intoopencv:4.xfrom
WanliZhong:err_armv8

Conversation

@WanliZhong
Copy link
Copy Markdown
Member

@WanliZhong WanliZhong commented Dec 11, 2023 •
edited by asmorkalov
Loading

Fixes the following warning on armv8:

warning: dereferencing type-punned pointer will break strict-aliasing rules [-Wstrict-aliasing]

Buildbot: https://pullrequest.opencv.org/buildbot/builders/4_x_ARMv8-lin

@WanliZhong WanliZhong added this to the 4.9.0 milestone Dec 11, 2023
@WanliZhong WanliZhong requested a review from zihaomu December 11, 2023 07:19
@WanliZhong
Copy link
Copy Markdown
Member Author

WanliZhong commented Dec 11, 2023

Warnings are fixed. Errors still need to be checked.

@asmorkalov
Copy link
Copy Markdown
Contributor

asmorkalov commented Dec 11, 2023 

modules/dnn/src/layers/cpu_kernels/conv_block.simd.hpp:729: trailing whitespace.
+    }    
modules/dnn/src/layers/cpu_kernels/conv_winograd_f63.simd.hpp:921: trailing whitespace.
+    }

@asmorkalov asmorkalov added category: build/install cleanup Code cleanup (e.g, drop legacy C-API, legacy unmaintained code) category: dnn and removed cleanup Code cleanup (e.g, drop legacy C-API, legacy unmaintained code) labels Dec 11, 2023
@WanliZhong
Copy link
Copy Markdown
Member Author

WanliZhong commented Dec 12, 2023 •
edited
Loading

Part of the performance test results show the performance is not influenced by this modification.

fp16 is 2x faster than fp32 in this patch. I think it's ok to merge this fix.

MacOS M2, use Min (ms)

Name of Test 1
conv::Conv::(GFLOPS=6.641, K=[3 x 3], IN={1, 64, 150, 200}, OCN=192, PM=SAME, BIAS, OCV/CPU) 20.074
conv::Conv::(GFLOPS=6.641, K=[3 x 3], IN={1, 64, 150, 200}, OCN=192, PM=SAME, BIAS, OCV/CPU_FP16) 12.229
conv::Conv::(GFLOPS=6.641, K=[3 x 3], IN={1, 64, 300, 300}, OCN=64, P=[1 x 1], BIAS, OCV/CPU) 20.011
conv::Conv::(GFLOPS=6.641, K=[3 x 3], IN={1, 64, 300, 300}, OCN=64, P=[1 x 1], BIAS, OCV/CPU_FP16) 16.007
conv::Conv::(GFLOPS=6.814, K=[3 x 3], IN={1, 512, 38, 38}, OCN=512, P=[1 x 1], BIAS, OCV/CPU) 15.304
conv::Conv::(GFLOPS=6.814, K=[3 x 3], IN={1, 512, 38, 38}, OCN=512, P=[1 x 1], BIAS, OCV/CPU_FP16) 10.736
conv::Conv::(GFLOPS=8.025, K=[3 x 3], IN={1, 1024, 19, 19}, OCN=1206, P=[1 x 1], BIAS, OCV/CPU) 25.451
conv::Conv::(GFLOPS=8.025, K=[3 x 3], IN={1, 1024, 19, 19}, OCN=1206, P=[1 x 1], BIAS, OCV/CPU_FP16) 17.501
conv::Conv::(GFLOPS=9.986, K=[3 x 3], IN={1, 512, 46, 46}, OCN=512, P=[1 x 1], BIAS, OCV/CPU) 33.519
conv::Conv::(GFLOPS=9.986, K=[3 x 3], IN={1, 512, 46, 46}, OCN=512, P=[1 x 1], BIAS, OCV/CPU_FP16) 16.884
conv::Conv::(GFLOPS=9.987, K=[3 x 3], IN={1, 256, 92, 92}, OCN=256, P=[1 x 1], BIAS, OCV/CPU) 22.552
conv::Conv::(GFLOPS=9.987, K=[3 x 3], IN={1, 256, 92, 92}, OCN=256, P=[1 x 1], BIAS, OCV/CPU_FP16) 14.555
conv::Conv::(GFLOPS=9.989, K=[3 x 3], IN={1, 128, 184, 184}, OCN=128, P=[1 x 1], BIAS, OCV/CPU) 34.036
conv::Conv::(GFLOPS=9.989, K=[3 x 3], IN={1, 128, 184, 184}, OCN=128, P=[1 x 1], BIAS, OCV/CPU_FP16) 15.490
conv::Conv::(GFLOPS=9.993, K=[3 x 3], IN={1, 64, 368, 368}, OCN=64, P=[1 x 1], BIAS, OCV/CPU) 27.186
conv::Conv::(GFLOPS=9.993, K=[3 x 3], IN={1, 64, 368, 368}, OCN=64, P=[1 x 1], BIAS, OCV/CPU_FP16) 17.417
conv::Conv::(GFLOPS=10.087, K=[3 x 3], IN={1, 576, 38, 50}, OCN=512, PM=SAME, BIAS, OCV/CPU) 20.665
conv::Conv::(GFLOPS=10.087, K=[3 x 3], IN={1, 576, 38, 50}, OCN=512, PM=SAME, BIAS, OCV/CPU_FP16) 13.120
conv::Conv::(GFLOPS=10.701, K=[3 x 3], IN={1, 512, 38, 38}, OCN=804, P=[1 x 1], BIAS, OCV/CPU) 34.430
conv::Conv::(GFLOPS=10.701, K=[3 x 3], IN={1, 512, 38, 38}, OCN=804, P=[1 x 1], BIAS, OCV/CPU_FP16) 12.969
conv::Conv::(GFLOPS=11.797, K=[5 x 5], IN={1, 240, 64, 64}, OCN=240, PM=SAME, OCV/CPU) 67.966
conv::Conv::(GFLOPS=11.797, K=[5 x 5], IN={1, 240, 64, 64}, OCN=240, PM=SAME, OCV/CPU_FP16) 36.998
conv::Conv::(GFLOPS=11.797, K=[5 x 5], IN={1, 480, 32, 32}, OCN=480, PM=SAME, OCV/CPU) 72.575
conv::Conv::(GFLOPS=11.797, K=[5 x 5], IN={1, 480, 32, 32}, OCN=480, PM=SAME, OCV/CPU_FP16) 41.956
conv::Conv::(GFLOPS=16.987, K=[5 x 5], IN={1, 1152, 16, 16}, OCN=1152, PM=SAME, OCV/CPU) 64.109
conv::Conv::(GFLOPS=16.987, K=[5 x 5], IN={1, 1152, 16, 16}, OCN=1152, PM=SAME, OCV/CPU_FP16) 36.303
conv::Conv::(GFLOPS=23.122, K=[5 x 5], IN={1, 672, 32, 32}, OCN=672, PM=SAME, OCV/CPU) 158.706
conv::Conv::(GFLOPS=23.122, K=[5 x 5], IN={1, 672, 32, 32}, OCN=672, PM=SAME, OCV/CPU_FP16) 56.058

zihaomu
zihaomu approved these changes Dec 12, 2023
View reviewed changes
Copy link
Copy Markdown
Member

@zihaomu zihaomu left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

👍

@asmorkalov asmorkalov merged commit 9bbc890 into opencv:4.x Dec 12, 2023
@asmorkalov asmorkalov mentioned this pull request Jan 19, 2024
Merged
thewoz pushed a commit to thewoz/opencv that referenced this pull request May 29, 2024
@WanliZhong @thewoz
Merge pull request opencv#24681 from WanliZhong:err_armv8
a09fe06 
Fixed armv8 compilation warnings opencv#24681 

Fixes the following warning on  armv8:
```
warning: dereferencing type-punned pointer will break strict-aliasing rules [-Wstrict-aliasing]
```
Buildbot: https://pullrequest.opencv.org/buildbot/builders/4_x_ARMv8-lin
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@zihaomu zihaomu zihaomu approved these changes

Assignees

@zihaomu zihaomu

Labels

category: build/install category: dnn

Projects

None yet

Milestone

4.9.0

Development

Successfully merging this pull request may close these issues.

3 participants

@WanliZhong @asmorkalov @zihaomu