rpc : turn GGML_RPC_DEBUG into a verbosity level and add logs

GGML_RPC_DEBUG is now parsed as a number: 0/unset disables debug logs,
1-3 emit increasingly detailed output (events, per-command trace,
transport detail). Non-numeric values fall back to 1.

The duplicated env/macro blocks in ggml-rpc.cpp and transport.cpp are
replaced by a shared log.h, which becomes the single choke point for
all logging of the RPC backend: LOG_ERROR/LOG_WARN/LOG_INFO for
unconditional severity logs and LOG_DBG/LOG_DBG2/LOG_DBG3 for the
verbosity-gated ones. The transport files no longer need ggml-impl.h,
and the server banner now also goes through the ggml logger (stderr).

Missing logs are added on both the client (handshake, buffer ops,
tensor transfers, graph computes, cache decisions) and the server
(per-command dispatch, graph nodes), including the negotiated
transport via the new socket_t::transport_name().

Assisted-by: pi:llama.cpp/MiMo-V2.6-Flash-RL

tools/rpc/README.md

@@ -113,8 +113,18 @@ $ GGML_RPC_NO_RDMA=1 bin/ggml-rpc-server
 
 ### Troubleshooting
 
-Use the `GGML_RPC_DEBUG` environment variable to enable debug messages from `ggml-rpc-server`:
+The `GGML_RPC_DEBUG` environment variable controls the verbosity of the logs emitted by the RPC backend.
+It can be set on the server, on the client (e.g. `llama-cli`), or both. Larger values produce more detailed output:
+
+- unset / `0` - disabled (only warnings and errors are printed)
+- `1` - high-level events: connections, handshake, buffer operations, tensor transfers, graph computes
+- `2` - per-command trace: every RPC message sent/received, cache hits/misses, queue events
+- `3` - transport detail: byte counts, per-command timings, graph node details
+
+Non-numeric values are treated as `1`.
+
 ```bash
 $ GGML_RPC_DEBUG=1 bin/ggml-rpc-server
+$ GGML_RPC_DEBUG=2 bin/llama-cli -hf ggml-org/gemma-3-1b-it-GGUF -ngl 99 --rpc 192.168.88.10:50052
 ```