139 lines
4.6 KiB
C
139 lines
4.6 KiB
C
/*
|
|
* VaptVupt — Cross-platform portability macros
|
|
*
|
|
* Provides unified abstractions for compiler intrinsics used throughout
|
|
* the codebase. Supports GCC, Clang, MSVC, and Intel compilers.
|
|
*
|
|
* SPDX-License-Identifier: GPL-3.0-or-later
|
|
*/
|
|
|
|
#ifndef VV_PLATFORM_H
|
|
#define VV_PLATFORM_H
|
|
|
|
#include <string.h>
|
|
#include <stdint.h>
|
|
|
|
/* ─── Branch prediction hints ─── */
|
|
#if defined(__GNUC__) || defined(__clang__)
|
|
#define VV_LIKELY(x) __builtin_expect(!!(x), 1)
|
|
#define VV_UNLIKELY(x) __builtin_expect(!!(x), 0)
|
|
#else
|
|
#define VV_LIKELY(x) (x)
|
|
#define VV_UNLIKELY(x) (x)
|
|
#endif
|
|
|
|
/* ─── Prefetch hint ─── */
|
|
#if defined(__GNUC__) || defined(__clang__)
|
|
#define VV_PREFETCH(p) __builtin_prefetch((p), 0, 1)
|
|
#define VV_PREFETCH_RW(p) __builtin_prefetch((p), 1, 1)
|
|
#elif defined(_MSC_VER) && (defined(_M_X64) || defined(_M_IX86))
|
|
#include <emmintrin.h>
|
|
#define VV_PREFETCH(p) _mm_prefetch((const char*)(p), _MM_HINT_T1)
|
|
#define VV_PREFETCH_RW(p) _mm_prefetch((const char*)(p), _MM_HINT_T1)
|
|
#else
|
|
#define VV_PREFETCH(p) ((void)0)
|
|
#define VV_PREFETCH_RW(p) ((void)0)
|
|
#endif
|
|
|
|
/* ─── Always-inline / never-inline ─── */
|
|
#if defined(__GNUC__) || defined(__clang__)
|
|
#define VV_ALWAYS_INLINE static inline __attribute__((always_inline))
|
|
#define VV_NOINLINE __attribute__((noinline))
|
|
#elif defined(_MSC_VER)
|
|
#define VV_ALWAYS_INLINE static __forceinline
|
|
#define VV_NOINLINE __declspec(noinline)
|
|
#else
|
|
#define VV_ALWAYS_INLINE static inline
|
|
#define VV_NOINLINE
|
|
#endif
|
|
|
|
/* ─── Unused parameter suppression ─── */
|
|
#if defined(__GNUC__) || defined(__clang__)
|
|
#define VV_UNUSED __attribute__((unused))
|
|
#else
|
|
#define VV_UNUSED
|
|
#endif
|
|
|
|
/* ─── Portable unaligned load/store via memcpy (compiler optimizes to single instr) ─── */
|
|
static inline uint16_t vv_load16(const void *p) {
|
|
uint16_t v; memcpy(&v, p, 2); return v;
|
|
}
|
|
static inline uint32_t vv_load32(const void *p) {
|
|
uint32_t v; memcpy(&v, p, 4); return v;
|
|
}
|
|
static inline uint64_t vv_load64(const void *p) {
|
|
uint64_t v; memcpy(&v, p, 8); return v;
|
|
}
|
|
static inline void vv_store16(void *p, uint16_t v) { memcpy(p, &v, 2); }
|
|
static inline void vv_store32(void *p, uint32_t v) { memcpy(p, &v, 4); }
|
|
static inline void vv_store64(void *p, uint64_t v) { memcpy(p, &v, 8); }
|
|
|
|
/* ─── Count trailing zeros (for hash/match optimization) ─── */
|
|
#if defined(__GNUC__) || defined(__clang__)
|
|
static inline int vv_ctz32(uint32_t x) { return __builtin_ctz(x); }
|
|
static inline int vv_ctz64(uint64_t x) { return __builtin_ctzll(x); }
|
|
#elif defined(_MSC_VER)
|
|
#include <intrin.h>
|
|
static inline int vv_ctz32(uint32_t x) {
|
|
unsigned long idx; _BitScanForward(&idx, x); return (int)idx;
|
|
}
|
|
static inline int vv_ctz64(uint64_t x) {
|
|
#if defined(_M_X64) || defined(_M_ARM64)
|
|
unsigned long idx; _BitScanForward64(&idx, x); return (int)idx;
|
|
#else
|
|
uint32_t lo = (uint32_t)x;
|
|
if (lo) return vv_ctz32(lo);
|
|
return 32 + vv_ctz32((uint32_t)(x >> 32));
|
|
#endif
|
|
}
|
|
#else
|
|
static inline int vv_ctz32(uint32_t x) {
|
|
int n = 0; while (!(x & 1)) { x >>= 1; n++; } return n;
|
|
}
|
|
static inline int vv_ctz64(uint64_t x) {
|
|
int n = 0; while (!(x & 1)) { x >>= 1; n++; } return n;
|
|
}
|
|
#endif
|
|
|
|
/* ─── SIMD capability detection macros ─── */
|
|
#if defined(__AVX2__)
|
|
#define VV_HAS_AVX2 1
|
|
#else
|
|
#define VV_HAS_AVX2 0
|
|
#endif
|
|
|
|
#if defined(__SSE2__) || defined(_M_X64) || (defined(_M_IX86_FP) && _M_IX86_FP >= 2)
|
|
#define VV_HAS_SSE2 1
|
|
#else
|
|
#define VV_HAS_SSE2 0
|
|
#endif
|
|
|
|
#if defined(__aarch64__) && defined(__ARM_NEON)
|
|
#define VV_HAS_NEON 1
|
|
#else
|
|
#define VV_HAS_NEON 0
|
|
#endif
|
|
|
|
/* Sprint 117: explicit no_sanitize annotation for hardened builds.
|
|
*
|
|
* Several hot paths use intentional unsigned modular arithmetic:
|
|
* - Knuth multiplicative hashes in the LZ matcher
|
|
* - xxh64 round mixers (multiplication, left-shift)
|
|
* - Post-decrement loop guards (uint32_t depth-- > 0)
|
|
*
|
|
* C11 §6.2.5p9 defines unsigned overflow as wraparound, so these are
|
|
* NOT undefined behavior — but `-fsanitize=integer` and the related
|
|
* `-fsanitize=shift-base` flags warn anyway, breaking hardened-build
|
|
* deployments. Apply this attribute to the affected functions to
|
|
* silence the false positives without disabling the checks globally.
|
|
*
|
|
* The annotation is clang-only (gcc has no equivalent and does not
|
|
* accept -fsanitize=integer in the first place). */
|
|
#if defined(__clang__) && (__clang_major__ >= 4)
|
|
# define VV_NO_SANITIZE_INTEGER \
|
|
__attribute__((no_sanitize("unsigned-integer-overflow", "shift", "shift-base", "shift-exponent")))
|
|
#else
|
|
# define VV_NO_SANITIZE_INTEGER
|
|
#endif
|
|
|
|
#endif /* VV_PLATFORM_H */
|