runbound 0.4.14

A DNS server. Just for fun.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179

# Runbound build system
# Requires: cargo, llvm-profdata (from clang/llvm-tools-preview)
#
# Install llvm tools:
#   rustup component add llvm-tools-preview
#
# Usage:
#   make              — debug build
#   make release      — optimised release build
#   make pgo          — profile-guided optimised build (+10–20 % QPS)
#   make bench        — quick dnsperf benchmark (requires dnsperf + running server)
#   make install      — install as systemd service (requires root)

BINARY    := target/release/runbound
PGO_DIR   := /tmp/runbound-pgo
MERGED    := $(PGO_DIR)/merged.profdata
QUERIES   := bench/queries.txt
BENCH_DUR := 30
BENCH_QPS := 150000

.PHONY: all build release pgo pgo-instrument pgo-merge pgo-optimise \
        bench install clean \
        audit deny sbom audit-full

# ── Default ───────────────────────────────────────────────────────────────────
all: build

build:
	cargo build

release:
	cargo build --release

# ── PGO — three-step pipeline ─────────────────────────────────────────────────
#
# How PGO works:
#   1. Instrumented build  — compiler inserts counters in every branch
#   2. Profiling run       — real traffic writes which branches are hottest
#   3. Optimised build     — compiler reorders code so the hot path lands in
#                            the CPU instruction-cache's first cache line,
#                            eliminating branch-predictor misses on most queries
#
# Result: +10–20 % raw QPS without changing a single line of application code.

pgo: pgo-instrument pgo-profile pgo-merge pgo-optimise
	@echo ""
	@echo "━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━"
	@echo " PGO build complete: $(BINARY)"
	@echo " Run 'make bench' to measure the improvement."
	@echo "━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━"

pgo-instrument:
	@echo "▶ Step 1/3 — Instrumented build (profiling counters injected)..."
	@mkdir -p $(PGO_DIR)
	RUSTFLAGS="-Cprofile-generate=$(PGO_DIR)" \
	    cargo build --release
	@echo "  Done: $(BINARY) (instrumented — ~20% slower, for profiling only)"

# This step is interactive: it starts the server, runs dnsperf, then stops it.
pgo-profile: $(BINARY)
	@echo ""
	@echo "▶ Step 2/3 — Collecting profile data..."
	@echo "  Starting Runbound in background (instrumented build)..."
	@$(BINARY) /etc/runbound/unbound.conf &
	@sleep 1
	@if ! command -v dnsperf > /dev/null 2>&1; then \
	    echo "  [WARN] dnsperf not found — generating synthetic queries instead"; \
	    $(MAKE) _pgo-synthetic-profile; \
	else \
	    echo "  Running dnsperf for $(BENCH_DUR)s at $(BENCH_QPS) QPS..."; \
	    dnsperf -s 127.0.0.1 -d $(QUERIES) -Q $(BENCH_QPS) -l $(BENCH_DUR) || true; \
	fi
	@echo "  Stopping instrumented server..."
	@pkill -f "runbound.*unbound.conf" 2>/dev/null || true
	@sleep 1
	@echo "  Profile data written to $(PGO_DIR)/"

# Fallback when dnsperf is unavailable: use dig in a tight loop
_pgo-synthetic-profile:
	@echo "  Synthetic profiling: 50 000 queries via dig..."
	@for i in $$(seq 1 10000); do \
	    dig +short @127.0.0.1 google.com A    > /dev/null 2>&1; \
	    dig +short @127.0.0.1 cloudflare.com AAAA > /dev/null 2>&1; \
	    dig +short @127.0.0.1 github.com A    > /dev/null 2>&1; \
	    dig +short @127.0.0.1 example.com MX  > /dev/null 2>&1; \
	    dig +short @127.0.0.1 rust-lang.org A > /dev/null 2>&1; \
	done

pgo-merge:
	@echo ""
	@echo "▶ Step 3a/3 — Merging profile data..."
	@# llvm-profdata may be in the rustup toolchain or system llvm
	@LLVM_PROFDATA=$$(find $$(rustup toolchain list -v | grep default | awk '{print $$2}') \
	    -name "llvm-profdata" 2>/dev/null | head -1); \
	if [ -z "$$LLVM_PROFDATA" ]; then \
	    LLVM_PROFDATA=$$(command -v llvm-profdata 2>/dev/null); \
	fi; \
	if [ -z "$$LLVM_PROFDATA" ]; then \
	    echo "  [ERROR] llvm-profdata not found."; \
	    echo "  Install with: rustup component add llvm-tools-preview"; \
	    exit 1; \
	fi; \
	echo "  Using: $$LLVM_PROFDATA"; \
	$$LLVM_PROFDATA merge -output=$(MERGED) $(PGO_DIR)/*.profraw
	@echo "  Merged: $(MERGED)"

pgo-optimise:
	@echo ""
	@echo "▶ Step 3b/3 — Final optimised build with profile data..."
	RUSTFLAGS="-Cprofile-use=$(MERGED) -Cllvm-args=-pgo-warn-missing-function" \
	    cargo build --release
	@echo "  Done: $(BINARY) (PGO-optimised)"

# ── Benchmark ─────────────────────────────────────────────────────────────────

bench:
	@if ! command -v dnsperf > /dev/null 2>&1; then \
	    echo "[ERROR] dnsperf not installed. On Debian/Ubuntu:"; \
	    echo "        apt install dnsperf"; \
	    exit 1; \
	fi
	@if [ ! -f $(QUERIES) ]; then \
	    echo "[WARN] No query file at $(QUERIES) — using built-in queries"; \
	    mkdir -p bench; \
	    printf "google.com A\ncloudflare.com AAAA\ngithub.com A\nexample.com MX\nrust-lang.org A\n" | \
	        awk 'BEGIN{for(i=0;i<2000;i++) {getline line < "/dev/stdin"; print line}}' \
	        > $(QUERIES) 2>/dev/null || \
	    for i in $$(seq 10000); do \
	        echo "google.com A"; echo "cloudflare.com A"; \
	    done > $(QUERIES); \
	fi
	@echo "━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━"
	@echo " Benchmarking Runbound at 127.0.0.1:53 for $(BENCH_DUR)s"
	@echo " Target: $(BENCH_QPS) QPS"
	@echo "━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━"
	dnsperf -s 127.0.0.1 -d $(QUERIES) -Q $(BENCH_QPS) -l $(BENCH_DUR)

# ── Install ───────────────────────────────────────────────────────────────────

install: release
	sudo ./install.sh

# ── Security audit ────────────────────────────────────────────────────────────
#
# Tools required (install once):
#   cargo install cargo-audit cargo-deny cargo-cyclonedx cargo-outdated
#
# See docs/audit.md for the full audit process and release procedure.

audit:
	@command -v cargo-audit >/dev/null 2>&1 || \
	    { echo "[ERROR] cargo-audit not installed. Run: cargo install cargo-audit"; exit 1; }
	cargo audit --deny warnings

deny:
	@command -v cargo-deny >/dev/null 2>&1 || \
	    { echo "[ERROR] cargo-deny not installed. Run: cargo install cargo-deny"; exit 1; }
	cargo deny check

sbom:
	@command -v cargo-cyclonedx >/dev/null 2>&1 || \
	    { echo "[ERROR] cargo-cyclonedx not installed. Run: cargo install cargo-cyclonedx"; exit 1; }
	cargo cyclonedx --format json --output sbom.cdx.json
	@echo "SBOM written to sbom.cdx.json"

audit-full: audit deny sbom
	@command -v cargo-outdated >/dev/null 2>&1 || \
	    { echo "[WARN] cargo-outdated not installed. Run: cargo install cargo-outdated"; }
	@command -v cargo-outdated >/dev/null 2>&1 && cargo outdated || true
	@echo ""
	@echo "━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━"
	@echo " Full audit complete. Attach sbom.cdx.json to the GitHub release."
	@echo "━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━"

# ── Clean ─────────────────────────────────────────────────────────────────────

clean:
	cargo clean
	rm -rf $(PGO_DIR)