llama-cpp: object file has no dynamic section

[sozercan]

LocalAI version:
v2.19.2 and v2.19.3 (v2.19.1 is fine)

Environment, CPU architecture, OS, and Version:
Ubuntu 22.04, amd64

Describe the bug
Error below:

DBG GRPC(Meta-Llama-3-8B-Instruct.Q4_K_M.gguf-127.0.0.1:44439): stderr /tmp/localai/backend_data/backend-assets/grpc/llama-cpp-avx2: error while loading shared libraries: /tmp/localai/backend_data/backend-assets/grpc/llama-cpp-avx2: object file has no dynamic section

To Reproduce
./local-ai --debug --config-file test.yaml
curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{ "model": "llama-3-8b-instruct", "messages": [{"role": "user", "content": "hello"}] }'

- name: llama-3-8b-instruct
  backend: llama
  parameters:
    model: Meta-Llama-3-8B-Instruct.Q4_K_M.gguf
  context_size: 8192
  # cut for brevity

Expected behavior
Should work as expected, like in v2.19.1

Logs
see above

Additional context

[mudler]

mh. that sounds weird. Is that the binary only? Might be glibc - or either probably caused by: #2947

[mudler]

probably related: https://bugzilla.redhat.com/show_bug.cgi?id=1381315