fix: unpin onnxruntime, fix HUM-6564 at the source

Containerfile

@@ -19,16 +19,30 @@ ARG SOURCE_VERSION=2026-08-30h
 # Install native libs needed by onnxruntime (libgomp) and numpy (libstdc++)
 USER 0
 RUN dnf install -y --setopt=install_weak_deps=False libgomp libstdc++ tar gzip && dnf clean all
+
+# onnxruntime vendors cpp_client_telemetry, which reads /etc/machine-id for a
+# device ID. When the value is unusable it falls back to popen("blkid"), and
+# popen() returns NULL with no /bin/sh. The SDK does not check that, so the
+# following pclose(NULL) segfaults the interpreter on `import onnxruntime`.
+# That is HUM-6564 / onnxruntime#32173.
+#
+# The file must be POPULATED — measured against 1.29.0 with telemetry on:
+#   absent -> SIGSEGV(139) | empty -> SIGSEGV(139) | populated -> clean
+# The empty "first boot" state is NOT enough. Needs root, so it lives here.
+RUN tr -d - < /proc/sys/kernel/random/uuid > /etc/machine-id.seed
+
 USER 65532
 
 # Download and extract upstream source
 RUN curl -sL https://github.com/fatherlinux/mcp-memory-service/archive/refs/heads/main.tar.gz \
     | tar xz --strip-components=1 -C /app
 
-# Install CPU-only PyTorch then the package with ONNX embedding support
-# Pin onnxruntime<1.20 — 1.29 segfaults in Hummingbird distroless (PyInit crash)
+# Install CPU-only PyTorch then the package with ONNX embedding support.
+# onnxruntime is no longer pinned: the <1.20 pin was a workaround for the
+# distroless segfault, and it held the embedding runtime nine minor versions
+# back. The populated /etc/machine-id above and ORT_DISABLE_TELEMETRY below
+# each independently prevent that crash, verified against 1.29.0.
 RUN pip3.12 install --no-cache-dir torch --index-url https://download.pytorch.org/whl/cpu && \
-    pip3.12 install --no-cache-dir "onnxruntime<1.20" && \
     pip3.12 install --no-cache-dir -e ".[sqlite]"
 
 RUN mkdir -p /app/sqlite_db /app/backups
@@ -54,14 +68,24 @@ COPY --from=builder /usr/lib64/libstdc++.so* /usr/lib64/
 COPY --from=builder /tmp/.local /tmp/.local
 COPY --from=builder /app /app
 
+# Keeps onnxruntime's telemetry off its shell-based fallback — see stage 1
+COPY --from=builder /etc/machine-id.seed /etc/machine-id
+
 ENV PYTHONUNBUFFERED=1 \
     PYTHONPATH=/app/src \
     PATH="/tmp/.local/bin:${PATH}" \
     MCP_MEMORY_SQLITE_PATH=/app/sqlite_db/memory.db \
     MCP_MEMORY_BACKUPS_PATH=/app/backups \
     MCP_SSE_HOST=0.0.0.0 \
-    MCP_SSE_PORT=8765 \
-    HF_HUB_DISABLE_TELEMETRY=1
+    MCP_SSE_PORT=8765
+
+# Second, independent guard against the HUM-6564 crash, and appropriate on its
+# own merits: left enabled, onnxruntime's init reads /etc/machine-id and
+# /etc/os-release, writes /tmp/mat-debug-1.log and creates a session file at
+# /tmp/.ses — not wanted in the image holding the whole memory corpus.
+ENV HF_HUB_DISABLE_TELEMETRY=1 \
+    ORT_DISABLE_TELEMETRY=1 \
+    DO_NOT_TRACK=1
 
 VOLUME ["/app/sqlite_db", "/app/backups"]