UPSTREAM PR #17903: docs: use port 8080 in Docker examples

[loci-dev]

Mirrored from ggml-org/llama.cpp#17903

Fixes #10262.

This PR fixes the documentation bug described in #10262, which was closed automatically by the stale bot.

The current Docker example:

docker run -v /path/to/models:/models \
  -p 8000:8000 \
  ghcr.io/ggml-org/llama.cpp:server \
  -m /models/7B/ggml-model-q4_0.gguf \
  --port 8000 \
  --host 0.0.0.0 \
  -n 512

causes the container health check to fail and the container status to become unhealthy.

The reason is that the example configures the server to listen on port 8000 (--port 8000), while the container image’s HEALTHCHECK is hard-coded to probe port 8080. This PR updates the documentation to use port 8080 so that the health check works correctly.

[loci-review]

Explore the complete analysis inside the Version Insights

Performance Analysis Summary: PR #509

This PR modifies only documentation (docs/docker.md), correcting Docker port examples from 8000 to 8080. No source code, binaries, or performance-critical functions were changed. Analysis confirms zero performance impact across all metrics.