Ubuntu - Illegal instruction (core dumped)

[carlberg74]

Trying to run som FaceEncodings on linux, processing some images.
but it crashes on line FaceRecognition.Create below.

My code:
Console.WriteLine("- Create Model Parameters!");
var modelParameter = new ModelParameter
{
PosePredictor68FaceLandmarksModel = File.ReadAllBytes("models/shape_predictor_68_face_landmarks.dat"),
PosePredictor5FaceLandmarksModel = File.ReadAllBytes("models/shape_predictor_5_face_landmarks.dat"),
FaceRecognitionModel = File.ReadAllBytes("models/dlib_face_recognition_resnet_model_v1.dat"),
CnnFaceDetectorModel = File.ReadAllBytes("models/mmod_human_face_detector.dat")
};

            Console.WriteLine("- FaceRecognition.Create");
            _fr = FaceRecognition.Create(modelParameter);

Added some gdb dump output:

/publish$ gdb ./ConsoleApp2
GNU gdb (Ubuntu 9.2-0ubuntu120.04.1) 9.2
Copyright (C) 2020 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later http://gnu.org/licenses/gpl.html
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.
Type "show copying" and "show warranty" for details.
This GDB was configured as "x86_64-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
http://www.gnu.org/software/gdb/bugs/.
Find the GDB manual and other documentation resources online at:
http://www.gnu.org/software/gdb/documentation/.

For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from ./ConsoleApp2...
(No debugging symbols found in ./ConsoleApp2)
(gdb) run
Starting program: /home/user/publish/ConsoleApp2
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
[New Thread 0x7ffff7072700 (LWP 14050)]
[New Thread 0x7ffff6871700 (LWP 14051)]
[New Thread 0x7ffff6050700 (LWP 14052)]
[New Thread 0x7ffff584f700 (LWP 14053)]
[New Thread 0x7ffff4a66700 (LWP 14054)]
[New Thread 0x7ffff3f28700 (LWP 14055)]
[New Thread 0x7ffff1d0a700 (LWP 14056)]
[New Thread 0x7ffff1505700 (LWP 14057)]
[New Thread 0x7ffff14c0700 (LWP 14058)]
Hello Face Recognition!
Init

• Create Model Parameters!
  [New Thread 0x7ffff0cbf700 (LWP 14060)]
• FaceRecognition.Create
  [Thread 0x7ffff1505700 (LWP 14057) exited]
  [New Thread 0x7ffff00f8700 (LWP 14071)]
  [New Thread 0x7fffef8f7700 (LWP 14072)]
  [New Thread 0x7fff7d48c700 (LWP 14073)]

Thread 1 "ConsoleApp2" received signal SIGILL, Illegal instruction.
0x00007fff282e0cbc in get_frontal_face_detector () from** /home/user/publish/libDlibDotNetNative.so**
(gdb)

I have tested:
Paths to models; if those are incorrect I get an clear error that models are not found.
file access: all models files are directory is accessable..
root: ran as root, same error.
.net versions: .net 5 and 6, same error.
windows: code works on windows.
lib: .so files are exists in directory
~/publish$ ls *.so
libDlibDotNetNativeDnnAgeClassification.so
libDlibDotNetNativeDnn.so
libDlibDotNetNativeDnnEmotionClassification.so
libDlibDotNetNative.so
libDlibDotNetNativeDnnGenderClassification.so

Any sugestions what I can try?

Cheers
-Anders

[takuya-takeuchi]

@carlberg74
I've never tried to run on Ubuntu 20.04.
So there is any issue about dependencies.

Please tell me more information about environementals.

[takuya-takeuchi]

BTW, did you try https://github.com/takuya-takeuchi/FaceRecognitionDotNet/wiki/Quickstart#ubuntudebian ?

[carlberg74]

Yes, I have followed the quickguide.

Env:
Ubuntu 20.04
Linux snafu 5.4.0-117-generic #132-Ubuntu SMP Thu Jun 2 00:39:06 UTC 2022 x86_64 x86_64 x86_64 GNU/Linux

Compiling on Windows with runtime target: linux-x64, target framework .net 5.0 (and 6.0) and copied to linux.
What linux envrioment have you tested on?

Tested with python3, just to see if I was missing some dependencies.
pip3 install dlib

Script:
import dlib
detector = dlib.get_frontal_face_detector()
facerec = dlib.face_recognition_model_v1("/home/user/publish/models/dlib_face_recognition_resnet_model_v1.dat")
img = dlib.load_rgb_image("/home/user/publish/Persons/test.jpg")
dets = detector(img, 1)
print(dets)

Runs without error and outputs cordinates for a rectagle.

[carlberg74]

Any idea what else I can try?
What linux environment has this been tested on?

[takuya-takeuchi]

Sorry, I am very busy so I have no time to go through issue untill this weekend.

[takuya-takeuchi]

@carlberg74

I created Ubuntu 20.04 Desktop on Hyper-V and try to wrok FRDN program.

1. Install .NET SDK

$ wget https://packages.microsoft.com/config/ubuntu/20.04/packages-microsoft-prod.deb -O packages-microsoft-prod.deb
$ sudo dpkg -i packages-microsoft-prod.deb
$ rm packages-microsoft-prod.deb
$ sudo apt-get update; \
  sudo apt-get install -y apt-transport-https && \
  sudo apt-get update && \
  sudo apt-get install -y dotnet-sdk-2.1

2. Install git and clone repo

$ sudo aot-get install git
$ git clone https://github.com/takuya-takeuchi/FaceRecognitionDotNet

3. Install dependencies

$ sudo apt-get install libx11-6 libopenblas-dev

4. Try Example

$ cd examples/Benchmark
$ dotnet remove reference ../../src/FaceRecognitionDotNet/FaceRecognitionDotNet.csproj
$ dotnet add package FaceRecognitionDotNet
$ git clone https://github.com/ageitgey/face_recognition_models
$ dotnet run -c Release -- -m=face_recognition_models/face_recognition_models/models
Benchmarks

Timings at 240p:
 - Face locations: 0.0396s (25.25 fps)
 - Face landmarks: 0.0012s (833.33 fps)
 - Encode face (inc. landmarks): 0.0172s (58.14 fps)
 - End-to-end: 0.0570s (17.54 fps)

Timings at 480p:
 - Face locations: 0.1516s (6.60 fps)
 - Face landmarks: 0.0012s (833.33 fps)
 - Encode face (inc. landmarks): 0.0172s (58.14 fps)
 - End-to-end: 0.1688s (5.92 fps)

Timings at 720p:
 - Face locations: 0.3480s (2.87 fps)
 - Face landmarks: 0.0012s (833.33 fps)
 - End-to-end: 0.3674s (2.72 fps)0.0174s (57.47 fps)

Timings at 1080p:
 - Face locations: 0.7806s (1.28 fps)
 - Face landmarks: 0.0012s (833.33 fps)
 - Encode face (inc. landmarks): 0.0174s (57.47 fps)
 - End-to-end: 0.7978s (1.25 fps

It works fine so wiki is not wrong.

[takuya-takeuchi]

@carlberg74
Just in case, could you show your machine cpu?
Legacy cpu occur issue.

[carlberg74]

$ lscpu
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Byte Order: Little Endian
Address sizes: 36 bits physical, 48 bits virtual
CPU(s): 4
On-line CPU(s) list: 0-3
Thread(s) per core: 1
Core(s) per socket: 4
Socket(s): 1
NUMA node(s): 1
Vendor ID: GenuineIntel
CPU family: 6
Model: 76
Model name: Intel(R) Celeron(R) CPU N3150 @ 1.60GHz
Stepping: 3
CPU MHz: 939.320
CPU max MHz: 2080.0000
CPU min MHz: 480.0000
BogoMIPS: 3200.00
Virtualization: VT-x
L1d cache: 96 KiB
L1i cache: 128 KiB
L2 cache: 2 MiB
NUMA node0 CPU(s): 0-3
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Mitigation; Clear CPU buffers; SMT disabled
Vulnerability Meltdown: Mitigation; PTI
Vulnerability Spec store bypass: Not affected
Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2: Mitigation; Retpolines, IBPB conditional, IBRS_FW, STIBP disabled, RSB filling
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Not affected
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx rdtscp lm constant_tsc arch_perfmon
pebs bts rep_good nopl xtopology tsc_reliable nonstop_tsc cpuid aperfmperf tsc_known_freq pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm sse4_1
sse4_2 movbe popcnt tsc_deadline_timer aes rdrand lahf_lm 3dnowprefetch epb pti ibrs ibpb stibp tpr_shadow vnmi flexpriority ept vpid tsc_adjust smep erms dtherm ida
arat md_clear

[takuya-takeuchi]

@carlberg74
Celeron N3150 does not support AVX.
Plesae refer this issue.
takuya-takeuchi/DlibDotNet#153 (comment)

[takuya-takeuchi]

No response and feedback