archmage 0.2.0

Safely invoke your intrinsic power, using the tokens granted to you by the CPU. Cast primitive magics faster than any mage alive.
Documentation 
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

# Archmage build and test commands

# Default: run all tests
default: test

# Run all tests (excludes sleef which requires nightly)
test:
    cargo test --features "std macros bytemuck wide __composite avx512"

# Run tests with all features (requires nightly for sleef)
test-nightly:
    cargo +nightly test --all-features

# Run clippy (excludes sleef which requires nightly)
lint:
    cargo clippy --features "std macros bytemuck wide __composite avx512" -- -D warnings

# Format code
fmt:
    cargo fmt

# Check formatting
fmt-check:
    cargo fmt -- --check

# Run Miri tests (token logic only, no SIMD)
miri:
    rustup run nightly cargo miri test --test miri_safe --all-features

# Regenerate safe_unaligned_simd wrappers
generate:
    cargo run -p xtask -- generate

# ============================================================================
# Intel SDE testing (requires Intel SDE to be installed)
# Download from: https://www.intel.com/content/www/us/en/download/684897/intel-software-development-emulator.html
# ============================================================================

# Test as Pentium 4 (SSE2 only, no SSE3/SSSE3/SSE4)
test-p4:
    sde64 -p4 -- cargo test --all-features

# Test as Nehalem (SSE4.2, no AVX)
test-nehalem:
    sde64 -nhm -- cargo test --all-features

# Test as Haswell (AVX2 + FMA, no AVX-512)
test-haswell:
    sde64 -hsw -- cargo test --all-features

# Test as Skylake-X (AVX-512)
test-skylake:
    sde64 -skx -- cargo test --all-features

# Test as Ice Lake (AVX-512 + VBMI2)
test-icelake:
    sde64 -icl -- cargo test --all-features

# Run all SDE tests (requires Intel SDE)
test-all-cpus: test-p4 test-nehalem test-haswell test-skylake test-icelake

# ============================================================================
# Cross-compilation testing (requires cargo-cross)
# Install: cargo install cross --git https://github.com/cross-rs/cross
# ============================================================================

# Test on 32-bit x86 (via QEMU)
test-i686:
    cross test --all-features --target i686-unknown-linux-gnu

# Test on aarch64 (via QEMU) - lib and cross-platform tests only
test-aarch64:
    cross test --lib --target aarch64-unknown-linux-gnu
    cross test --test miri_safe --test feature_consistency --target aarch64-unknown-linux-gnu

# Test on armv7 (via QEMU) - lib and cross-platform tests only
test-armv7:
    cross test --lib --target armv7-unknown-linux-gnueabihf
    cross test --test miri_safe --test feature_consistency --target armv7-unknown-linux-gnueabihf

# Build for all cross targets (faster than running tests)
build-cross:
    cross build --all-features --target i686-unknown-linux-gnu
    cross build --all-features --target aarch64-unknown-linux-gnu
    cross build --all-features --target armv7-unknown-linux-gnueabihf

# Run tests on all cross targets
test-cross: test-i686 test-aarch64 test-armv7
    @echo "All cross-compilation tests passed!"

# Clippy for x86_64
clippy-x86_64:
    cargo clippy --all-features --target x86_64-unknown-linux-gnu -- -D warnings

# Clippy for aarch64
clippy-aarch64:
    cargo clippy --all-features --target aarch64-unknown-linux-gnu -- -D warnings

# Clippy for i686
clippy-i686:
    cargo clippy --all-features --target i686-unknown-linux-gnu -- -D warnings

# Clippy for all targets
clippy-all: clippy-x86_64 clippy-aarch64 clippy-i686
    @echo "All clippy checks passed!"

# ============================================================================
# CI-style comprehensive test
# ============================================================================

# Full CI check (no SDE)
ci: fmt-check lint test miri
    @echo "All CI checks passed!"

# Full validation with SDE (for local development)
validate: ci test-all-cpus
    @echo "Full validation complete!"

# Full validation including cross-compilation
validate-all: ci test-cross clippy-all
    @echo "Full cross-platform validation complete!"

# ============================================================================
# Benchmarking (requires -C target-cpu=native for accurate results)
# ============================================================================

# Run all benchmarks with native CPU optimizations
bench:
    RUSTFLAGS="-C target-cpu=native" cargo bench --features wide

# Run wide comparison benchmark (quick mode)
bench-wide:
    RUSTFLAGS="-C target-cpu=native" cargo bench --bench wide_comparison --features wide -- --quick

# Run wide comparison benchmark (full)
bench-wide-full:
    RUSTFLAGS="-C target-cpu=native" cargo bench --bench wide_comparison --features wide

# Run transcendental benchmarks
bench-transcendental:
    RUSTFLAGS="-C target-cpu=native" cargo bench --bench transcendental_accuracy

# Run edge case benchmarks
bench-edge-cases:
    RUSTFLAGS="-C target-cpu=native" cargo bench --bench edge_case_perf

# IMPORTANT: Without -C target-cpu=native, intrinsics won't inline properly
# and benchmarks will show archmage being 4-5x slower than wide.
# With native CPU targeting, archmage is 1.2-1.4x faster than wide.