llm-proxy: propagate client cancellation to the backend (free the slot)

A pi-agent stop did not reach models served through the proxy: the forward loop
only noticed the client was gone when it next tried to WRITE, so during the
blocking non-streaming getresponse() (full generation) the backend ran to
completion holding the single-slot admission slot — a cancelled request kept
burning the NPU/DGX. Add a client-liveness watchdog: while a backend request is
in flight, select() on the client socket; on EOF close the backend conn so it
aborts and the slot frees. New _try_backend return "aborted" (no failover;
client is gone). _stream_response read loop hardened to stop cleanly when the
watchdog closes the conn mid-read. Live-verified: non-streaming request cut at
3s -> CLIENT-CANCELLED logged, slot freed, followup 200 in 4s.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01EWpfhDgYNA21tETDP9ueBE

local-backends.json

@@ -0,0 +1,4 @@
+[
+  {"name":"qwen3-4b","host":"boltzmann.fritz.box","port":8091,"slots":1},
+  {"name":"bosch-dspark","host":"bosch.fritz.box","port":8888,"slots":32}
+]

systemd/llama-server-qwen3-4b-npu.service

@@ -1,27 +0,0 @@
-[Unit]
-Description=rk-llama.cpp NPU backend: Qwen3 4B Q8_0 on RK3588 NPU (port 8091)
-After=network-online.target
-Wants=network-online.target
-
-[Service]
-Type=simple
-
-# Kernel gate: only run under an NPU-capable kernel (rkopnu present; matches *npu*).
-ExecCondition=/bin/sh -c 'case "$(uname -r)" in *npu*) exit 0;; *) exit 1;; esac'
-
-LimitNOFILE=65536
-
-WorkingDirectory=/home/mfritsche/npu
-ExecStart=/home/mfritsche/src/rk-llama.cpp/build/bin/llama-server\
-     -m /home/mfritsche/models/Qwen3-4B-Instruct-2507-Q8_0.gguf\
-     --alias qwen3-4b-2507-npu\
-     -c 65536 -t 4 --cache-type-k q8_0 --cache-type-v q8_0\
-     --host 0.0.0.0 --port 8091
-
-Restart=on-failure
-RestartSec=5
-
-TimeoutStartSec=180
-
-[Install]
-WantedBy=default.target