Move the files from host/include/i386 to host/include/x86_64, replacing the stub headers that redirected to i386.
Remove linux-user/include/host/i386. Remove common-user/host/i386. Reviewed-by: Thomas Huth <[email protected]> Reviewed-by: Pierrick Bouvier <[email protected]> Signed-off-by: Richard Henderson <[email protected]> --- host/include/i386/host/cpuinfo.h | 41 ------ host/include/i386/host/crypto/aes-round.h | 152 ------------------- host/include/i386/host/crypto/clmul.h | 29 ---- host/include/x86_64/host/cpuinfo.h | 42 +++++- host/include/x86_64/host/crypto/aes-round.h | 153 +++++++++++++++++++- host/include/x86_64/host/crypto/clmul.h | 30 +++- linux-user/include/host/i386/host-signal.h | 38 ----- common-user/host/i386/safe-syscall.inc.S | 127 ---------------- host/include/i386/host/bufferiszero.c.inc | 125 ---------------- host/include/x86_64/host/bufferiszero.c.inc | 126 +++++++++++++++- 10 files changed, 347 insertions(+), 516 deletions(-) delete mode 100644 host/include/i386/host/cpuinfo.h delete mode 100644 host/include/i386/host/crypto/aes-round.h delete mode 100644 host/include/i386/host/crypto/clmul.h delete mode 100644 linux-user/include/host/i386/host-signal.h delete mode 100644 common-user/host/i386/safe-syscall.inc.S delete mode 100644 host/include/i386/host/bufferiszero.c.inc diff --git a/host/include/i386/host/cpuinfo.h b/host/include/i386/host/cpuinfo.h deleted file mode 100644 index 93d029d499..0000000000 --- a/host/include/i386/host/cpuinfo.h +++ /dev/null @@ -1,41 +0,0 @@ -/* - * SPDX-License-Identifier: GPL-2.0-or-later - * Host specific cpu identification for x86. - */ - -#ifndef HOST_CPUINFO_H -#define HOST_CPUINFO_H - -/* Digested version of <cpuid.h> */ - -#define CPUINFO_ALWAYS (1u << 0) /* so cpuinfo is nonzero */ -#define CPUINFO_OSXSAVE (1u << 1) -#define CPUINFO_MOVBE (1u << 2) -#define CPUINFO_LZCNT (1u << 3) -#define CPUINFO_POPCNT (1u << 4) -#define CPUINFO_BMI1 (1u << 5) -#define CPUINFO_BMI2 (1u << 6) -#define CPUINFO_SSE2 (1u << 7) -#define CPUINFO_AVX1 (1u << 9) -#define CPUINFO_AVX2 (1u << 10) -#define CPUINFO_AVX512F (1u << 11) -#define CPUINFO_AVX512VL (1u << 12) -#define CPUINFO_AVX512BW (1u << 13) -#define CPUINFO_AVX512DQ (1u << 14) -#define CPUINFO_AVX512VBMI2 (1u << 15) -#define CPUINFO_ATOMIC_VMOVDQA (1u << 16) -#define CPUINFO_ATOMIC_VMOVDQU (1u << 17) -#define CPUINFO_AES (1u << 18) -#define CPUINFO_PCLMUL (1u << 19) -#define CPUINFO_GFNI (1u << 20) - -/* Initialized with a constructor. */ -extern unsigned cpuinfo; - -/* - * We cannot rely on constructor ordering, so other constructors must - * use the function interface rather than the variable above. - */ -unsigned cpuinfo_init(void); - -#endif /* HOST_CPUINFO_H */ diff --git a/host/include/i386/host/crypto/aes-round.h b/host/include/i386/host/crypto/aes-round.h deleted file mode 100644 index 59a64130f7..0000000000 --- a/host/include/i386/host/crypto/aes-round.h +++ /dev/null @@ -1,152 +0,0 @@ -/* - * x86 specific aes acceleration. - * SPDX-License-Identifier: GPL-2.0-or-later - */ - -#ifndef X86_HOST_CRYPTO_AES_ROUND_H -#define X86_HOST_CRYPTO_AES_ROUND_H - -#include "host/cpuinfo.h" -#include <immintrin.h> - -#if defined(__AES__) && defined(__SSSE3__) -# define HAVE_AES_ACCEL true -# define ATTR_AES_ACCEL -#else -# define HAVE_AES_ACCEL likely(cpuinfo & CPUINFO_AES) -# define ATTR_AES_ACCEL __attribute__((target("aes,ssse3"))) -#endif - -static inline __m128i ATTR_AES_ACCEL -aes_accel_bswap(__m128i x) -{ - return _mm_shuffle_epi8(x, _mm_set_epi8(0, 1, 2, 3, 4, 5, 6, 7, 8, - 9, 10, 11, 12, 13, 14, 15)); -} - -static inline void ATTR_AES_ACCEL -aesenc_MC_accel(AESState *ret, const AESState *st, bool be) -{ - __m128i t = (__m128i)st->v; - __m128i z = _mm_setzero_si128(); - - if (be) { - t = aes_accel_bswap(t); - t = _mm_aesdeclast_si128(t, z); - t = _mm_aesenc_si128(t, z); - t = aes_accel_bswap(t); - } else { - t = _mm_aesdeclast_si128(t, z); - t = _mm_aesenc_si128(t, z); - } - ret->v = (AESStateVec)t; -} - -static inline void ATTR_AES_ACCEL -aesenc_SB_SR_AK_accel(AESState *ret, const AESState *st, - const AESState *rk, bool be) -{ - __m128i t = (__m128i)st->v; - __m128i k = (__m128i)rk->v; - - if (be) { - t = aes_accel_bswap(t); - k = aes_accel_bswap(k); - t = _mm_aesenclast_si128(t, k); - t = aes_accel_bswap(t); - } else { - t = _mm_aesenclast_si128(t, k); - } - ret->v = (AESStateVec)t; -} - -static inline void ATTR_AES_ACCEL -aesenc_SB_SR_MC_AK_accel(AESState *ret, const AESState *st, - const AESState *rk, bool be) -{ - __m128i t = (__m128i)st->v; - __m128i k = (__m128i)rk->v; - - if (be) { - t = aes_accel_bswap(t); - k = aes_accel_bswap(k); - t = _mm_aesenc_si128(t, k); - t = aes_accel_bswap(t); - } else { - t = _mm_aesenc_si128(t, k); - } - ret->v = (AESStateVec)t; -} - -static inline void ATTR_AES_ACCEL -aesdec_IMC_accel(AESState *ret, const AESState *st, bool be) -{ - __m128i t = (__m128i)st->v; - - if (be) { - t = aes_accel_bswap(t); - t = _mm_aesimc_si128(t); - t = aes_accel_bswap(t); - } else { - t = _mm_aesimc_si128(t); - } - ret->v = (AESStateVec)t; -} - -static inline void ATTR_AES_ACCEL -aesdec_ISB_ISR_AK_accel(AESState *ret, const AESState *st, - const AESState *rk, bool be) -{ - __m128i t = (__m128i)st->v; - __m128i k = (__m128i)rk->v; - - if (be) { - t = aes_accel_bswap(t); - k = aes_accel_bswap(k); - t = _mm_aesdeclast_si128(t, k); - t = aes_accel_bswap(t); - } else { - t = _mm_aesdeclast_si128(t, k); - } - ret->v = (AESStateVec)t; -} - -static inline void ATTR_AES_ACCEL -aesdec_ISB_ISR_AK_IMC_accel(AESState *ret, const AESState *st, - const AESState *rk, bool be) -{ - __m128i t = (__m128i)st->v; - __m128i k = (__m128i)rk->v; - - if (be) { - t = aes_accel_bswap(t); - k = aes_accel_bswap(k); - t = _mm_aesdeclast_si128(t, k); - t = _mm_aesimc_si128(t); - t = aes_accel_bswap(t); - } else { - t = _mm_aesdeclast_si128(t, k); - t = _mm_aesimc_si128(t); - } - ret->v = (AESStateVec)t; -} - -static inline void ATTR_AES_ACCEL -aesdec_ISB_ISR_IMC_AK_accel(AESState *ret, const AESState *st, - const AESState *rk, bool be) -{ - __m128i t = (__m128i)st->v; - __m128i k = (__m128i)rk->v; - - if (be) { - t = aes_accel_bswap(t); - k = aes_accel_bswap(k); - t = _mm_aesdec_si128(t, k); - t = aes_accel_bswap(t); - } else { - t = _mm_aesdec_si128(t, k); - } - ret->v = (AESStateVec)t; -} - -#endif /* X86_HOST_CRYPTO_AES_ROUND_H */ diff --git a/host/include/i386/host/crypto/clmul.h b/host/include/i386/host/crypto/clmul.h deleted file mode 100644 index dc3c814797..0000000000 --- a/host/include/i386/host/crypto/clmul.h +++ /dev/null @@ -1,29 +0,0 @@ -/* - * x86 specific clmul acceleration. - * SPDX-License-Identifier: GPL-2.0-or-later - */ - -#ifndef X86_HOST_CRYPTO_CLMUL_H -#define X86_HOST_CRYPTO_CLMUL_H - -#include "host/cpuinfo.h" -#include <immintrin.h> - -#if defined(__PCLMUL__) -# define HAVE_CLMUL_ACCEL true -# define ATTR_CLMUL_ACCEL -#else -# define HAVE_CLMUL_ACCEL likely(cpuinfo & CPUINFO_PCLMUL) -# define ATTR_CLMUL_ACCEL __attribute__((target("pclmul"))) -#endif - -static inline Int128 ATTR_CLMUL_ACCEL -clmul_64_accel(uint64_t n, uint64_t m) -{ - union { __m128i v; Int128 s; } u; - - u.v = _mm_clmulepi64_si128(_mm_set_epi64x(0, n), _mm_set_epi64x(0, m), 0); - return u.s; -} - -#endif /* X86_HOST_CRYPTO_CLMUL_H */ diff --git a/host/include/x86_64/host/cpuinfo.h b/host/include/x86_64/host/cpuinfo.h index 67debab9a0..93d029d499 100644 --- a/host/include/x86_64/host/cpuinfo.h +++ b/host/include/x86_64/host/cpuinfo.h @@ -1 +1,41 @@ -#include "host/include/i386/host/cpuinfo.h" +/* + * SPDX-License-Identifier: GPL-2.0-or-later + * Host specific cpu identification for x86. + */ + +#ifndef HOST_CPUINFO_H +#define HOST_CPUINFO_H + +/* Digested version of <cpuid.h> */ + +#define CPUINFO_ALWAYS (1u << 0) /* so cpuinfo is nonzero */ +#define CPUINFO_OSXSAVE (1u << 1) +#define CPUINFO_MOVBE (1u << 2) +#define CPUINFO_LZCNT (1u << 3) +#define CPUINFO_POPCNT (1u << 4) +#define CPUINFO_BMI1 (1u << 5) +#define CPUINFO_BMI2 (1u << 6) +#define CPUINFO_SSE2 (1u << 7) +#define CPUINFO_AVX1 (1u << 9) +#define CPUINFO_AVX2 (1u << 10) +#define CPUINFO_AVX512F (1u << 11) +#define CPUINFO_AVX512VL (1u << 12) +#define CPUINFO_AVX512BW (1u << 13) +#define CPUINFO_AVX512DQ (1u << 14) +#define CPUINFO_AVX512VBMI2 (1u << 15) +#define CPUINFO_ATOMIC_VMOVDQA (1u << 16) +#define CPUINFO_ATOMIC_VMOVDQU (1u << 17) +#define CPUINFO_AES (1u << 18) +#define CPUINFO_PCLMUL (1u << 19) +#define CPUINFO_GFNI (1u << 20) + +/* Initialized with a constructor. */ +extern unsigned cpuinfo; + +/* + * We cannot rely on constructor ordering, so other constructors must + * use the function interface rather than the variable above. + */ +unsigned cpuinfo_init(void); + +#endif /* HOST_CPUINFO_H */ diff --git a/host/include/x86_64/host/crypto/aes-round.h b/host/include/x86_64/host/crypto/aes-round.h index 2773cc9f10..59a64130f7 100644 --- a/host/include/x86_64/host/crypto/aes-round.h +++ b/host/include/x86_64/host/crypto/aes-round.h @@ -1 +1,152 @@ -#include "host/include/i386/host/crypto/aes-round.h" +/* + * x86 specific aes acceleration. + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef X86_HOST_CRYPTO_AES_ROUND_H +#define X86_HOST_CRYPTO_AES_ROUND_H + +#include "host/cpuinfo.h" +#include <immintrin.h> + +#if defined(__AES__) && defined(__SSSE3__) +# define HAVE_AES_ACCEL true +# define ATTR_AES_ACCEL +#else +# define HAVE_AES_ACCEL likely(cpuinfo & CPUINFO_AES) +# define ATTR_AES_ACCEL __attribute__((target("aes,ssse3"))) +#endif + +static inline __m128i ATTR_AES_ACCEL +aes_accel_bswap(__m128i x) +{ + return _mm_shuffle_epi8(x, _mm_set_epi8(0, 1, 2, 3, 4, 5, 6, 7, 8, + 9, 10, 11, 12, 13, 14, 15)); +} + +static inline void ATTR_AES_ACCEL +aesenc_MC_accel(AESState *ret, const AESState *st, bool be) +{ + __m128i t = (__m128i)st->v; + __m128i z = _mm_setzero_si128(); + + if (be) { + t = aes_accel_bswap(t); + t = _mm_aesdeclast_si128(t, z); + t = _mm_aesenc_si128(t, z); + t = aes_accel_bswap(t); + } else { + t = _mm_aesdeclast_si128(t, z); + t = _mm_aesenc_si128(t, z); + } + ret->v = (AESStateVec)t; +} + +static inline void ATTR_AES_ACCEL +aesenc_SB_SR_AK_accel(AESState *ret, const AESState *st, + const AESState *rk, bool be) +{ + __m128i t = (__m128i)st->v; + __m128i k = (__m128i)rk->v; + + if (be) { + t = aes_accel_bswap(t); + k = aes_accel_bswap(k); + t = _mm_aesenclast_si128(t, k); + t = aes_accel_bswap(t); + } else { + t = _mm_aesenclast_si128(t, k); + } + ret->v = (AESStateVec)t; +} + +static inline void ATTR_AES_ACCEL +aesenc_SB_SR_MC_AK_accel(AESState *ret, const AESState *st, + const AESState *rk, bool be) +{ + __m128i t = (__m128i)st->v; + __m128i k = (__m128i)rk->v; + + if (be) { + t = aes_accel_bswap(t); + k = aes_accel_bswap(k); + t = _mm_aesenc_si128(t, k); + t = aes_accel_bswap(t); + } else { + t = _mm_aesenc_si128(t, k); + } + ret->v = (AESStateVec)t; +} + +static inline void ATTR_AES_ACCEL +aesdec_IMC_accel(AESState *ret, const AESState *st, bool be) +{ + __m128i t = (__m128i)st->v; + + if (be) { + t = aes_accel_bswap(t); + t = _mm_aesimc_si128(t); + t = aes_accel_bswap(t); + } else { + t = _mm_aesimc_si128(t); + } + ret->v = (AESStateVec)t; +} + +static inline void ATTR_AES_ACCEL +aesdec_ISB_ISR_AK_accel(AESState *ret, const AESState *st, + const AESState *rk, bool be) +{ + __m128i t = (__m128i)st->v; + __m128i k = (__m128i)rk->v; + + if (be) { + t = aes_accel_bswap(t); + k = aes_accel_bswap(k); + t = _mm_aesdeclast_si128(t, k); + t = aes_accel_bswap(t); + } else { + t = _mm_aesdeclast_si128(t, k); + } + ret->v = (AESStateVec)t; +} + +static inline void ATTR_AES_ACCEL +aesdec_ISB_ISR_AK_IMC_accel(AESState *ret, const AESState *st, + const AESState *rk, bool be) +{ + __m128i t = (__m128i)st->v; + __m128i k = (__m128i)rk->v; + + if (be) { + t = aes_accel_bswap(t); + k = aes_accel_bswap(k); + t = _mm_aesdeclast_si128(t, k); + t = _mm_aesimc_si128(t); + t = aes_accel_bswap(t); + } else { + t = _mm_aesdeclast_si128(t, k); + t = _mm_aesimc_si128(t); + } + ret->v = (AESStateVec)t; +} + +static inline void ATTR_AES_ACCEL +aesdec_ISB_ISR_IMC_AK_accel(AESState *ret, const AESState *st, + const AESState *rk, bool be) +{ + __m128i t = (__m128i)st->v; + __m128i k = (__m128i)rk->v; + + if (be) { + t = aes_accel_bswap(t); + k = aes_accel_bswap(k); + t = _mm_aesdec_si128(t, k); + t = aes_accel_bswap(t); + } else { + t = _mm_aesdec_si128(t, k); + } + ret->v = (AESStateVec)t; +} + +#endif /* X86_HOST_CRYPTO_AES_ROUND_H */ diff --git a/host/include/x86_64/host/crypto/clmul.h b/host/include/x86_64/host/crypto/clmul.h index f25eced416..dc3c814797 100644 --- a/host/include/x86_64/host/crypto/clmul.h +++ b/host/include/x86_64/host/crypto/clmul.h @@ -1 +1,29 @@ -#include "host/include/i386/host/crypto/clmul.h" +/* + * x86 specific clmul acceleration. + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef X86_HOST_CRYPTO_CLMUL_H +#define X86_HOST_CRYPTO_CLMUL_H + +#include "host/cpuinfo.h" +#include <immintrin.h> + +#if defined(__PCLMUL__) +# define HAVE_CLMUL_ACCEL true +# define ATTR_CLMUL_ACCEL +#else +# define HAVE_CLMUL_ACCEL likely(cpuinfo & CPUINFO_PCLMUL) +# define ATTR_CLMUL_ACCEL __attribute__((target("pclmul"))) +#endif + +static inline Int128 ATTR_CLMUL_ACCEL +clmul_64_accel(uint64_t n, uint64_t m) +{ + union { __m128i v; Int128 s; } u; + + u.v = _mm_clmulepi64_si128(_mm_set_epi64x(0, n), _mm_set_epi64x(0, m), 0); + return u.s; +} + +#endif /* X86_HOST_CRYPTO_CLMUL_H */ diff --git a/linux-user/include/host/i386/host-signal.h b/linux-user/include/host/i386/host-signal.h deleted file mode 100644 index e2b64f077f..0000000000 --- a/linux-user/include/host/i386/host-signal.h +++ /dev/null @@ -1,38 +0,0 @@ -/* - * host-signal.h: signal info dependent on the host architecture - * - * Copyright (c) 2003-2005 Fabrice Bellard - * Copyright (c) 2021 Linaro Limited - * - * This work is licensed under the terms of the GNU LGPL, version 2.1 or later. - * See the COPYING file in the top-level directory. - */ - -#ifndef I386_HOST_SIGNAL_H -#define I386_HOST_SIGNAL_H - -/* The third argument to a SA_SIGINFO handler is ucontext_t. */ -typedef ucontext_t host_sigcontext; - -static inline uintptr_t host_signal_pc(host_sigcontext *uc) -{ - return uc->uc_mcontext.gregs[REG_EIP]; -} - -static inline void host_signal_set_pc(host_sigcontext *uc, uintptr_t pc) -{ - uc->uc_mcontext.gregs[REG_EIP] = pc; -} - -static inline void *host_signal_mask(host_sigcontext *uc) -{ - return &uc->uc_sigmask; -} - -static inline bool host_signal_write(siginfo_t *info, host_sigcontext *uc) -{ - return uc->uc_mcontext.gregs[REG_TRAPNO] == 0xe - && (uc->uc_mcontext.gregs[REG_ERR] & 0x2); -} - -#endif diff --git a/common-user/host/i386/safe-syscall.inc.S b/common-user/host/i386/safe-syscall.inc.S deleted file mode 100644 index db2ed09839..0000000000 --- a/common-user/host/i386/safe-syscall.inc.S +++ /dev/null @@ -1,127 +0,0 @@ -/* - * safe-syscall.inc.S : host-specific assembly fragment - * to handle signals occurring at the same time as system calls. - * This is intended to be included by common-user/safe-syscall.S - * - * Written by Richard Henderson <[email protected]> - * Copyright (C) 2016 Red Hat, Inc. - * - * This work is licensed under the terms of the GNU GPL, version 2 or later. - * See the COPYING file in the top-level directory. - */ - - .global safe_syscall_base - .global safe_syscall_start - .global safe_syscall_end - .type safe_syscall_base, @function - - /* This is the entry point for making a system call. The calling - * convention here is that of a C varargs function with the - * first argument an 'int *' to the signal_pending flag, the - * second one the system call number (as a 'long'), and all further - * arguments being syscall arguments (also 'long'). - */ -safe_syscall_base: - .cfi_startproc - push %ebp - .cfi_adjust_cfa_offset 4 - .cfi_rel_offset ebp, 0 - push %esi - .cfi_adjust_cfa_offset 4 - .cfi_rel_offset esi, 0 - push %edi - .cfi_adjust_cfa_offset 4 - .cfi_rel_offset edi, 0 - push %ebx - .cfi_adjust_cfa_offset 4 - .cfi_rel_offset ebx, 0 - - /* The syscall calling convention isn't the same as the C one: - * we enter with 0(%esp) == return address - * 4(%esp) == &signal_pending - * 8(%esp) == syscall number - * 12(%esp) ... 32(%esp) == syscall arguments - * and return the result in eax - * and the syscall instruction needs - * eax == syscall number - * ebx, ecx, edx, esi, edi, ebp == syscall arguments - * and returns the result in eax - * Shuffle everything around appropriately. - * Note the 16 bytes that we pushed to save registers. - */ - mov 12+16(%esp), %ebx /* the syscall arguments */ - mov 16+16(%esp), %ecx - mov 20+16(%esp), %edx - mov 24+16(%esp), %esi - mov 28+16(%esp), %edi - mov 32+16(%esp), %ebp - - /* This next sequence of code works in conjunction with the - * rewind_if_safe_syscall_function(). If a signal is taken - * and the interrupted PC is anywhere between 'safe_syscall_start' - * and 'safe_syscall_end' then we rewind it to 'safe_syscall_start'. - * The code sequence must therefore be able to cope with this, and - * the syscall instruction must be the final one in the sequence. - */ -safe_syscall_start: - /* if signal_pending is non-zero, don't do the call */ - mov 4+16(%esp), %eax /* signal_pending */ - cmpl $0, (%eax) - jnz 2f - mov 8+16(%esp), %eax /* syscall number */ - int $0x80 -safe_syscall_end: - - /* code path for having successfully executed the syscall */ -#if defined(__linux__) - /* Linux kernel returns (small) negative errno. */ - cmp $-4095, %eax - jae 0f -#elif defined(__FreeBSD__) - /* FreeBSD kernel returns positive errno and C bit set. */ - jc 1f -#else -#error "unsupported os" -#endif - pop %ebx - .cfi_remember_state - .cfi_adjust_cfa_offset -4 - .cfi_restore ebx - pop %edi - .cfi_adjust_cfa_offset -4 - .cfi_restore edi - pop %esi - .cfi_adjust_cfa_offset -4 - .cfi_restore esi - pop %ebp - .cfi_adjust_cfa_offset -4 - .cfi_restore ebp - ret - .cfi_restore_state - -#if defined(__linux__) -0: neg %eax - jmp 1f -#endif - - /* code path when we didn't execute the syscall */ -2: mov $QEMU_ERESTARTSYS, %eax - - /* code path setting errno */ -1: pop %ebx - .cfi_adjust_cfa_offset -4 - .cfi_restore ebx - pop %edi - .cfi_adjust_cfa_offset -4 - .cfi_restore edi - pop %esi - .cfi_adjust_cfa_offset -4 - .cfi_restore esi - pop %ebp - .cfi_adjust_cfa_offset -4 - .cfi_restore ebp - mov %eax, 4(%esp) - jmp safe_syscall_set_errno_tail - - .cfi_endproc - .size safe_syscall_base, .-safe_syscall_base diff --git a/host/include/i386/host/bufferiszero.c.inc b/host/include/i386/host/bufferiszero.c.inc deleted file mode 100644 index 74ae98580f..0000000000 --- a/host/include/i386/host/bufferiszero.c.inc +++ /dev/null @@ -1,125 +0,0 @@ -/* - * SPDX-License-Identifier: GPL-2.0-or-later - * buffer_is_zero acceleration, x86 version. - */ - -#if defined(CONFIG_AVX2_OPT) || defined(__SSE2__) -#include <immintrin.h> - -/* Helper for preventing the compiler from reassociating - chains of binary vector operations. */ -#define SSE_REASSOC_BARRIER(vec0, vec1) asm("" : "+x"(vec0), "+x"(vec1)) - -/* Note that these vectorized functions may assume len >= 256. */ - -static bool __attribute__((target("sse2"))) -buffer_zero_sse2(const void *buf, size_t len) -{ - /* Unaligned loads at head/tail. */ - __m128i v = *(__m128i_u *)(buf); - __m128i w = *(__m128i_u *)(buf + len - 16); - /* Align head/tail to 16-byte boundaries. */ - const __m128i *p = QEMU_ALIGN_PTR_DOWN(buf + 16, 16); - const __m128i *e = QEMU_ALIGN_PTR_DOWN(buf + len - 1, 16); - __m128i zero = { 0 }; - - /* Collect a partial block at tail end. */ - v |= e[-1]; w |= e[-2]; - SSE_REASSOC_BARRIER(v, w); - v |= e[-3]; w |= e[-4]; - SSE_REASSOC_BARRIER(v, w); - v |= e[-5]; w |= e[-6]; - SSE_REASSOC_BARRIER(v, w); - v |= e[-7]; v |= w; - - /* - * Loop over complete 128-byte blocks. - * With the head and tail removed, e - p >= 14, so the loop - * must iterate at least once. - */ - do { - v = _mm_cmpeq_epi8(v, zero); - if (unlikely(_mm_movemask_epi8(v) != 0xFFFF)) { - return false; - } - v = p[0]; w = p[1]; - SSE_REASSOC_BARRIER(v, w); - v |= p[2]; w |= p[3]; - SSE_REASSOC_BARRIER(v, w); - v |= p[4]; w |= p[5]; - SSE_REASSOC_BARRIER(v, w); - v |= p[6]; w |= p[7]; - SSE_REASSOC_BARRIER(v, w); - v |= w; - p += 8; - } while (p < e - 7); - - return _mm_movemask_epi8(_mm_cmpeq_epi8(v, zero)) == 0xFFFF; -} - -#ifdef CONFIG_AVX2_OPT -static bool __attribute__((target("avx2"))) -buffer_zero_avx2(const void *buf, size_t len) -{ - /* Unaligned loads at head/tail. */ - __m256i v = *(__m256i_u *)(buf); - __m256i w = *(__m256i_u *)(buf + len - 32); - /* Align head/tail to 32-byte boundaries. */ - const __m256i *p = QEMU_ALIGN_PTR_DOWN(buf + 32, 32); - const __m256i *e = QEMU_ALIGN_PTR_DOWN(buf + len - 1, 32); - __m256i zero = { 0 }; - - /* Collect a partial block at tail end. */ - v |= e[-1]; w |= e[-2]; - SSE_REASSOC_BARRIER(v, w); - v |= e[-3]; w |= e[-4]; - SSE_REASSOC_BARRIER(v, w); - v |= e[-5]; w |= e[-6]; - SSE_REASSOC_BARRIER(v, w); - v |= e[-7]; v |= w; - - /* Loop over complete 256-byte blocks. */ - for (; p < e - 7; p += 8) { - /* PTEST is not profitable here. */ - v = _mm256_cmpeq_epi8(v, zero); - if (unlikely(_mm256_movemask_epi8(v) != 0xFFFFFFFF)) { - return false; - } - v = p[0]; w = p[1]; - SSE_REASSOC_BARRIER(v, w); - v |= p[2]; w |= p[3]; - SSE_REASSOC_BARRIER(v, w); - v |= p[4]; w |= p[5]; - SSE_REASSOC_BARRIER(v, w); - v |= p[6]; w |= p[7]; - SSE_REASSOC_BARRIER(v, w); - v |= w; - } - - return _mm256_movemask_epi8(_mm256_cmpeq_epi8(v, zero)) == 0xFFFFFFFF; -} -#endif /* CONFIG_AVX2_OPT */ - -static biz_accel_fn const accel_table[] = { - buffer_is_zero_int_ge256, - buffer_zero_sse2, -#ifdef CONFIG_AVX2_OPT - buffer_zero_avx2, -#endif -}; - -static unsigned best_accel(void) -{ - unsigned info = cpuinfo_init(); - -#ifdef CONFIG_AVX2_OPT - if (info & CPUINFO_AVX2) { - return 2; - } -#endif - return info & CPUINFO_SSE2 ? 1 : 0; -} - -#else -# include "host/include/generic/host/bufferiszero.c.inc" -#endif diff --git a/host/include/x86_64/host/bufferiszero.c.inc b/host/include/x86_64/host/bufferiszero.c.inc index 1d3f1fd6f5..74ae98580f 100644 --- a/host/include/x86_64/host/bufferiszero.c.inc +++ b/host/include/x86_64/host/bufferiszero.c.inc @@ -1 +1,125 @@ -#include "host/include/i386/host/bufferiszero.c.inc" +/* + * SPDX-License-Identifier: GPL-2.0-or-later + * buffer_is_zero acceleration, x86 version. + */ + +#if defined(CONFIG_AVX2_OPT) || defined(__SSE2__) +#include <immintrin.h> + +/* Helper for preventing the compiler from reassociating + chains of binary vector operations. */ +#define SSE_REASSOC_BARRIER(vec0, vec1) asm("" : "+x"(vec0), "+x"(vec1)) + +/* Note that these vectorized functions may assume len >= 256. */ + +static bool __attribute__((target("sse2"))) +buffer_zero_sse2(const void *buf, size_t len) +{ + /* Unaligned loads at head/tail. */ + __m128i v = *(__m128i_u *)(buf); + __m128i w = *(__m128i_u *)(buf + len - 16); + /* Align head/tail to 16-byte boundaries. */ + const __m128i *p = QEMU_ALIGN_PTR_DOWN(buf + 16, 16); + const __m128i *e = QEMU_ALIGN_PTR_DOWN(buf + len - 1, 16); + __m128i zero = { 0 }; + + /* Collect a partial block at tail end. */ + v |= e[-1]; w |= e[-2]; + SSE_REASSOC_BARRIER(v, w); + v |= e[-3]; w |= e[-4]; + SSE_REASSOC_BARRIER(v, w); + v |= e[-5]; w |= e[-6]; + SSE_REASSOC_BARRIER(v, w); + v |= e[-7]; v |= w; + + /* + * Loop over complete 128-byte blocks. + * With the head and tail removed, e - p >= 14, so the loop + * must iterate at least once. + */ + do { + v = _mm_cmpeq_epi8(v, zero); + if (unlikely(_mm_movemask_epi8(v) != 0xFFFF)) { + return false; + } + v = p[0]; w = p[1]; + SSE_REASSOC_BARRIER(v, w); + v |= p[2]; w |= p[3]; + SSE_REASSOC_BARRIER(v, w); + v |= p[4]; w |= p[5]; + SSE_REASSOC_BARRIER(v, w); + v |= p[6]; w |= p[7]; + SSE_REASSOC_BARRIER(v, w); + v |= w; + p += 8; + } while (p < e - 7); + + return _mm_movemask_epi8(_mm_cmpeq_epi8(v, zero)) == 0xFFFF; +} + +#ifdef CONFIG_AVX2_OPT +static bool __attribute__((target("avx2"))) +buffer_zero_avx2(const void *buf, size_t len) +{ + /* Unaligned loads at head/tail. */ + __m256i v = *(__m256i_u *)(buf); + __m256i w = *(__m256i_u *)(buf + len - 32); + /* Align head/tail to 32-byte boundaries. */ + const __m256i *p = QEMU_ALIGN_PTR_DOWN(buf + 32, 32); + const __m256i *e = QEMU_ALIGN_PTR_DOWN(buf + len - 1, 32); + __m256i zero = { 0 }; + + /* Collect a partial block at tail end. */ + v |= e[-1]; w |= e[-2]; + SSE_REASSOC_BARRIER(v, w); + v |= e[-3]; w |= e[-4]; + SSE_REASSOC_BARRIER(v, w); + v |= e[-5]; w |= e[-6]; + SSE_REASSOC_BARRIER(v, w); + v |= e[-7]; v |= w; + + /* Loop over complete 256-byte blocks. */ + for (; p < e - 7; p += 8) { + /* PTEST is not profitable here. */ + v = _mm256_cmpeq_epi8(v, zero); + if (unlikely(_mm256_movemask_epi8(v) != 0xFFFFFFFF)) { + return false; + } + v = p[0]; w = p[1]; + SSE_REASSOC_BARRIER(v, w); + v |= p[2]; w |= p[3]; + SSE_REASSOC_BARRIER(v, w); + v |= p[4]; w |= p[5]; + SSE_REASSOC_BARRIER(v, w); + v |= p[6]; w |= p[7]; + SSE_REASSOC_BARRIER(v, w); + v |= w; + } + + return _mm256_movemask_epi8(_mm256_cmpeq_epi8(v, zero)) == 0xFFFFFFFF; +} +#endif /* CONFIG_AVX2_OPT */ + +static biz_accel_fn const accel_table[] = { + buffer_is_zero_int_ge256, + buffer_zero_sse2, +#ifdef CONFIG_AVX2_OPT + buffer_zero_avx2, +#endif +}; + +static unsigned best_accel(void) +{ + unsigned info = cpuinfo_init(); + +#ifdef CONFIG_AVX2_OPT + if (info & CPUINFO_AVX2) { + return 2; + } +#endif + return info & CPUINFO_SSE2 ? 1 : 0; +} + +#else +# include "host/include/generic/host/bufferiszero.c.inc" +#endif -- 2.43.0
