qemu: uncaught target signal 4 (Illegal instruction) - core dumped

[cowhca]

I have docker running and run
docker run -it --rm \ -v "INPUT_DIR" \ -v "OUTPUT_DIR" \ dullage/eyelign:latest

It tells me this and hangs:
qemu: uncaught target signal 4 (Illegal instruction) - core dumped

I am running the command in a directory with INPUT_DIR and OUTPUT_DIR exist and INPUT_DIR has all the images as jpg files.

[cowhca]

It also gives me
WARNING: The requested image's platform (linux/amd64) does not match the detected host platform (linux/arm64/v8) and no specific platform was requested
but that goes away with --platform linux/amd64 flag

[dullage]

Hi @cowhca. Are you running this on a machine with an arm processor. Perhaps a newer Mac?

If so, that's the problem as the Docker image I published was only built for x86.

[cowhca]

Yes, I'm running it on an M1 MacBook Pro. Thanks for letting me know, unfortunate it won't work for me.

[dullage]

When I get a chance, I'll see if I can do an arm build. It may be that some of the dependencies don't support it but I'll give it a shot and let you know.

[dullage]

@cowhca - There is now a build for arm/v8 available on dockerhub. Hopefully, the example docker run command from the README page should work for you now.

[duceduc]

Hello. I am experiencing similar issue. Trying this on a rpi4.

ngx@ngx:/tmp $ docker pull dullage/eyelign:latest
latest: Pulling from dullage/eyelign
no matching manifest for linux/arm64/v8 in the manifest list entries

update: I have ran the below with the --platform tag mentioned above. It installed, but error when trying to run it.

ngx@ngx:/tmp $ docker pull --platform linux/amd64 dullage/eyelign:latest

Error:

ngx@ngx:/tmp $ docker run -it --rm \
  -v "/var/www/log/shots:/input" \
  -v "/var/www/log/aligned:/ouput" \
  dullage/eyelign:latest
WARNING: The requested image's platform (linux/amd64) does not match the detected host platform (linux/arm64/v8) and no specific platform was requested
exec /usr/local/bin/python3: exec format error

[dullage]

@duceduc - I haven't tried running it on a Pi4 but I did just try pulling the image on my Pi4 and it worked.

Can you run the following command and see what's in the output: docker manifest inspect dullage/eyelign:latest.

[duceduc]

Hi. This is the output.

ngx@ngx:/tmp $ docker manifest inspect dullage/eyelign:latest
{
   "schemaVersion": 2,
   "mediaType": "application/vnd.docker.distribution.manifest.list.v2+json",
   "manifests": [
      {
         "mediaType": "application/vnd.docker.distribution.manifest.v2+json",
         "size": 2843,
         "digest": "sha256:e85474190cb45016746d1fab2ba973c4f572fe9960e7454d23e02831de31a8ed",
         "platform": {
            "architecture": "amd64",
            "os": "linux"
         }
      },
      {
         "mediaType": "application/vnd.docker.distribution.manifest.v2+json",
         "size": 2843,
         "digest": "sha256:75368129361bb4a4bd268535efbd445fbc4e1f37e89c27c50906f1ef05a3da33",
         "platform": {
            "architecture": "arm",
            "os": "linux",
            "variant": "v8"
         }
      }
   ]
}

My rpi4

ngx@ngx:/tmp $ cat /sys/firmware/devicetree/base/model;echo
Raspberry Pi 4 Model B Rev 1.4

[dullage]

Thanks @duceduc. I've pushed a new build-up that uses slightly different architecture tags. Can you give a pull another try please.

[duceduc]

Hi. Thank you for for the update. It is working now. Now I just need to play with the options. Having trouble finding the eyes.