From 0a499a7e9819e7a0980408f18df68160a0b55f2e Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Sun, 1 Dec 2024 17:08:26 -0800 Subject: lib/crc32: drop leading underscores from __crc32c_le_base Remove the leading underscores from __crc32c_le_base(). This is in preparation for adding crc32c_le_arch() and eventually renaming __crc32c_le() to crc32c_le(). Reviewed-by: Ard Biesheuvel Link: https://lore.kernel.org/r/20241202010844.144356-2-ebiggers@kernel.org Signed-off-by: Eric Biggers --- include/linux/crc32.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/crc32.h b/include/linux/crc32.h index 87f788c0d607..5b07fc9081c4 100644 --- a/include/linux/crc32.h +++ b/include/linux/crc32.h @@ -39,7 +39,7 @@ static inline u32 crc32_le_combine(u32 crc1, u32 crc2, size_t len2) } u32 __pure __crc32c_le(u32 crc, unsigned char const *p, size_t len); -u32 __pure __crc32c_le_base(u32 crc, unsigned char const *p, size_t len); +u32 __pure crc32c_le_base(u32 crc, unsigned char const *p, size_t len); /** * __crc32c_le_combine - Combine two crc32c check values into one. For two -- cgit v1.2.3 From d36cebe03c3ae4ea1fde20cfc797fab8729c3ab5 Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Sun, 1 Dec 2024 17:08:27 -0800 Subject: lib/crc32: improve support for arch-specific overrides Currently the CRC32 library functions are defined as weak symbols, and the arm64 and riscv architectures override them. This method of arch-specific overrides has the limitation that it only works when both the base and arch code is built-in. Also, it makes the arch-specific code be silently not used if it is accidentally built with lib-y instead of obj-y; unfortunately the RISC-V code does this. This commit reorganizes the code to have explicit *_arch() functions that are called when they are enabled, similar to how some of the crypto library code works (e.g. chacha_crypt() calls chacha_crypt_arch()). Make the existing kconfig choice for the CRC32 implementation also control whether the arch-optimized implementation (if one is available) is enabled or not. Make it enabled by default if CRC32 is also enabled. The result is that arch-optimized CRC32 library functions will be included automatically when appropriate, but it is now possible to disable them. They can also now be built as a loadable module if the CRC32 library functions happen to be used only by loadable modules, in which case the arch and base CRC32 modules will be automatically loaded via direct symbol dependency when appropriate. Reviewed-by: Ard Biesheuvel Link: https://lore.kernel.org/r/20241202010844.144356-3-ebiggers@kernel.org Signed-off-by: Eric Biggers --- arch/arm64/Kconfig | 1 + arch/arm64/lib/Makefile | 3 +- arch/arm64/lib/crc32-glue.c | 13 +- arch/riscv/Kconfig | 1 + arch/riscv/lib/Makefile | 3 +- arch/riscv/lib/crc32-riscv.c | 301 +++++++++++++++++++++++++++++++++++++++++++ arch/riscv/lib/crc32.c | 294 ------------------------------------------ crypto/crc32_generic.c | 4 +- crypto/crc32c_generic.c | 4 +- include/linux/crc32.h | 35 ++++- lib/Kconfig | 70 +++++++--- lib/crc32.c | 22 +--- 12 files changed, 409 insertions(+), 342 deletions(-) create mode 100644 arch/riscv/lib/crc32-riscv.c delete mode 100644 arch/riscv/lib/crc32.c (limited to 'include/linux') diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index 100570a048c5..71f6310c8240 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -21,6 +21,7 @@ config ARM64 select ARCH_ENABLE_THP_MIGRATION if TRANSPARENT_HUGEPAGE select ARCH_HAS_CACHE_LINE_SIZE select ARCH_HAS_CC_PLATFORM + select ARCH_HAS_CRC32 select ARCH_HAS_CURRENT_STACK_POINTER select ARCH_HAS_DEBUG_VIRTUAL select ARCH_HAS_DEBUG_VM_PGTABLE diff --git a/arch/arm64/lib/Makefile b/arch/arm64/lib/Makefile index 8e882f479d98..5fbcf0d56665 100644 --- a/arch/arm64/lib/Makefile +++ b/arch/arm64/lib/Makefile @@ -13,7 +13,8 @@ endif lib-$(CONFIG_ARCH_HAS_UACCESS_FLUSHCACHE) += uaccess_flushcache.o -obj-$(CONFIG_CRC32) += crc32.o crc32-glue.o +obj-$(CONFIG_CRC32_ARCH) += crc32-arm64.o +crc32-arm64-y := crc32.o crc32-glue.o obj-$(CONFIG_FUNCTION_ERROR_INJECTION) += error-inject.o diff --git a/arch/arm64/lib/crc32-glue.c b/arch/arm64/lib/crc32-glue.c index ad015223d15d..d7f6e1cbf0d2 100644 --- a/arch/arm64/lib/crc32-glue.c +++ b/arch/arm64/lib/crc32-glue.c @@ -2,6 +2,7 @@ #include #include +#include #include #include @@ -21,7 +22,7 @@ asmlinkage u32 crc32_le_arm64_4way(u32 crc, unsigned char const *p, size_t len); asmlinkage u32 crc32c_le_arm64_4way(u32 crc, unsigned char const *p, size_t len); asmlinkage u32 crc32_be_arm64_4way(u32 crc, unsigned char const *p, size_t len); -u32 __pure crc32_le(u32 crc, unsigned char const *p, size_t len) +u32 __pure crc32_le_arch(u32 crc, const u8 *p, size_t len) { if (!alternative_has_cap_likely(ARM64_HAS_CRC32)) return crc32_le_base(crc, p, len); @@ -40,8 +41,9 @@ u32 __pure crc32_le(u32 crc, unsigned char const *p, size_t len) return crc32_le_arm64(crc, p, len); } +EXPORT_SYMBOL(crc32_le_arch); -u32 __pure __crc32c_le(u32 crc, unsigned char const *p, size_t len) +u32 __pure crc32c_le_arch(u32 crc, const u8 *p, size_t len) { if (!alternative_has_cap_likely(ARM64_HAS_CRC32)) return crc32c_le_base(crc, p, len); @@ -60,8 +62,9 @@ u32 __pure __crc32c_le(u32 crc, unsigned char const *p, size_t len) return crc32c_le_arm64(crc, p, len); } +EXPORT_SYMBOL(crc32c_le_arch); -u32 __pure crc32_be(u32 crc, unsigned char const *p, size_t len) +u32 __pure crc32_be_arch(u32 crc, const u8 *p, size_t len) { if (!alternative_has_cap_likely(ARM64_HAS_CRC32)) return crc32_be_base(crc, p, len); @@ -80,3 +83,7 @@ u32 __pure crc32_be(u32 crc, unsigned char const *p, size_t len) return crc32_be_arm64(crc, p, len); } +EXPORT_SYMBOL(crc32_be_arch); + +MODULE_LICENSE("GPL"); +MODULE_DESCRIPTION("arm64-optimized CRC32 functions"); diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig index d4a7ca0388c0..7d5718667e39 100644 --- a/arch/riscv/Kconfig +++ b/arch/riscv/Kconfig @@ -24,6 +24,7 @@ config RISCV select ARCH_ENABLE_SPLIT_PMD_PTLOCK if PGTABLE_LEVELS > 2 select ARCH_ENABLE_THP_MIGRATION if TRANSPARENT_HUGEPAGE select ARCH_HAS_BINFMT_FLAT + select ARCH_HAS_CRC32 if RISCV_ISA_ZBC select ARCH_HAS_CURRENT_STACK_POINTER select ARCH_HAS_DEBUG_VIRTUAL if MMU select ARCH_HAS_DEBUG_VM_PGTABLE diff --git a/arch/riscv/lib/Makefile b/arch/riscv/lib/Makefile index 8eec6b69a875..79368a895fee 100644 --- a/arch/riscv/lib/Makefile +++ b/arch/riscv/lib/Makefile @@ -15,8 +15,7 @@ endif lib-$(CONFIG_MMU) += uaccess.o lib-$(CONFIG_64BIT) += tishift.o lib-$(CONFIG_RISCV_ISA_ZICBOZ) += clear_page.o -lib-$(CONFIG_RISCV_ISA_ZBC) += crc32.o - +obj-$(CONFIG_CRC32_ARCH) += crc32-riscv.o obj-$(CONFIG_FUNCTION_ERROR_INJECTION) += error-inject.o lib-$(CONFIG_RISCV_ISA_V) += xor.o lib-$(CONFIG_RISCV_ISA_V) += riscv_v_helpers.o diff --git a/arch/riscv/lib/crc32-riscv.c b/arch/riscv/lib/crc32-riscv.c new file mode 100644 index 000000000000..a3ff7db2a1ce --- /dev/null +++ b/arch/riscv/lib/crc32-riscv.c @@ -0,0 +1,301 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Accelerated CRC32 implementation with Zbc extension. + * + * Copyright (C) 2024 Intel Corporation + */ + +#include +#include +#include + +#include +#include +#include +#include +#include +#include + +/* + * Refer to https://www.corsix.org/content/barrett-reduction-polynomials for + * better understanding of how this math works. + * + * let "+" denotes polynomial add (XOR) + * let "-" denotes polynomial sub (XOR) + * let "*" denotes polynomial multiplication + * let "/" denotes polynomial floor division + * let "S" denotes source data, XLEN bit wide + * let "P" denotes CRC32 polynomial + * let "T" denotes 2^(XLEN+32) + * let "QT" denotes quotient of T/P, with the bit for 2^XLEN being implicit + * + * crc32(S, P) + * => S * (2^32) - S * (2^32) / P * P + * => lowest 32 bits of: S * (2^32) / P * P + * => lowest 32 bits of: S * (2^32) * (T / P) / T * P + * => lowest 32 bits of: S * (2^32) * quotient / T * P + * => lowest 32 bits of: S * quotient / 2^XLEN * P + * => lowest 32 bits of: (clmul_high_part(S, QT) + S) * P + * => clmul_low_part(clmul_high_part(S, QT) + S, P) + * + * In terms of below implementations, the BE case is more intuitive, since the + * higher order bit sits at more significant position. + */ + +#if __riscv_xlen == 64 +/* Slide by XLEN bits per iteration */ +# define STEP_ORDER 3 + +/* Each below polynomial quotient has an implicit bit for 2^XLEN */ + +/* Polynomial quotient of (2^(XLEN+32))/CRC32_POLY, in LE format */ +# define CRC32_POLY_QT_LE 0x5a72d812fb808b20 + +/* Polynomial quotient of (2^(XLEN+32))/CRC32C_POLY, in LE format */ +# define CRC32C_POLY_QT_LE 0xa434f61c6f5389f8 + +/* Polynomial quotient of (2^(XLEN+32))/CRC32_POLY, in BE format, it should be + * the same as the bit-reversed version of CRC32_POLY_QT_LE + */ +# define CRC32_POLY_QT_BE 0x04d101df481b4e5a + +static inline u64 crc32_le_prep(u32 crc, unsigned long const *ptr) +{ + return (u64)crc ^ (__force u64)__cpu_to_le64(*ptr); +} + +static inline u32 crc32_le_zbc(unsigned long s, u32 poly, unsigned long poly_qt) +{ + u32 crc; + + /* We don't have a "clmulrh" insn, so use clmul + slli instead. */ + asm volatile (".option push\n" + ".option arch,+zbc\n" + "clmul %0, %1, %2\n" + "slli %0, %0, 1\n" + "xor %0, %0, %1\n" + "clmulr %0, %0, %3\n" + "srli %0, %0, 32\n" + ".option pop\n" + : "=&r" (crc) + : "r" (s), + "r" (poly_qt), + "r" ((u64)poly << 32) + :); + return crc; +} + +static inline u64 crc32_be_prep(u32 crc, unsigned long const *ptr) +{ + return ((u64)crc << 32) ^ (__force u64)__cpu_to_be64(*ptr); +} + +#elif __riscv_xlen == 32 +# define STEP_ORDER 2 +/* Each quotient should match the upper half of its analog in RV64 */ +# define CRC32_POLY_QT_LE 0xfb808b20 +# define CRC32C_POLY_QT_LE 0x6f5389f8 +# define CRC32_POLY_QT_BE 0x04d101df + +static inline u32 crc32_le_prep(u32 crc, unsigned long const *ptr) +{ + return crc ^ (__force u32)__cpu_to_le32(*ptr); +} + +static inline u32 crc32_le_zbc(unsigned long s, u32 poly, unsigned long poly_qt) +{ + u32 crc; + + /* We don't have a "clmulrh" insn, so use clmul + slli instead. */ + asm volatile (".option push\n" + ".option arch,+zbc\n" + "clmul %0, %1, %2\n" + "slli %0, %0, 1\n" + "xor %0, %0, %1\n" + "clmulr %0, %0, %3\n" + ".option pop\n" + : "=&r" (crc) + : "r" (s), + "r" (poly_qt), + "r" (poly) + :); + return crc; +} + +static inline u32 crc32_be_prep(u32 crc, unsigned long const *ptr) +{ + return crc ^ (__force u32)__cpu_to_be32(*ptr); +} + +#else +# error "Unexpected __riscv_xlen" +#endif + +static inline u32 crc32_be_zbc(unsigned long s) +{ + u32 crc; + + asm volatile (".option push\n" + ".option arch,+zbc\n" + "clmulh %0, %1, %2\n" + "xor %0, %0, %1\n" + "clmul %0, %0, %3\n" + ".option pop\n" + : "=&r" (crc) + : "r" (s), + "r" (CRC32_POLY_QT_BE), + "r" (CRC32_POLY_BE) + :); + return crc; +} + +#define STEP (1 << STEP_ORDER) +#define OFFSET_MASK (STEP - 1) + +typedef u32 (*fallback)(u32 crc, unsigned char const *p, size_t len); + +static inline u32 crc32_le_unaligned(u32 crc, unsigned char const *p, + size_t len, u32 poly, + unsigned long poly_qt) +{ + size_t bits = len * 8; + unsigned long s = 0; + u32 crc_low = 0; + + for (int i = 0; i < len; i++) + s = ((unsigned long)*p++ << (__riscv_xlen - 8)) | (s >> 8); + + s ^= (unsigned long)crc << (__riscv_xlen - bits); + if (__riscv_xlen == 32 || len < sizeof(u32)) + crc_low = crc >> bits; + + crc = crc32_le_zbc(s, poly, poly_qt); + crc ^= crc_low; + + return crc; +} + +static inline u32 __pure crc32_le_generic(u32 crc, unsigned char const *p, + size_t len, u32 poly, + unsigned long poly_qt, + fallback crc_fb) +{ + size_t offset, head_len, tail_len; + unsigned long const *p_ul; + unsigned long s; + + asm goto(ALTERNATIVE("j %l[legacy]", "nop", 0, + RISCV_ISA_EXT_ZBC, 1) + : : : : legacy); + + /* Handle the unaligned head. */ + offset = (unsigned long)p & OFFSET_MASK; + if (offset && len) { + head_len = min(STEP - offset, len); + crc = crc32_le_unaligned(crc, p, head_len, poly, poly_qt); + p += head_len; + len -= head_len; + } + + tail_len = len & OFFSET_MASK; + len = len >> STEP_ORDER; + p_ul = (unsigned long const *)p; + + for (int i = 0; i < len; i++) { + s = crc32_le_prep(crc, p_ul); + crc = crc32_le_zbc(s, poly, poly_qt); + p_ul++; + } + + /* Handle the tail bytes. */ + p = (unsigned char const *)p_ul; + if (tail_len) + crc = crc32_le_unaligned(crc, p, tail_len, poly, poly_qt); + + return crc; + +legacy: + return crc_fb(crc, p, len); +} + +u32 __pure crc32_le_arch(u32 crc, const u8 *p, size_t len) +{ + return crc32_le_generic(crc, p, len, CRC32_POLY_LE, CRC32_POLY_QT_LE, + crc32_le_base); +} +EXPORT_SYMBOL(crc32_le_arch); + +u32 __pure crc32c_le_arch(u32 crc, const u8 *p, size_t len) +{ + return crc32_le_generic(crc, p, len, CRC32C_POLY_LE, + CRC32C_POLY_QT_LE, crc32c_le_base); +} +EXPORT_SYMBOL(crc32c_le_arch); + +static inline u32 crc32_be_unaligned(u32 crc, unsigned char const *p, + size_t len) +{ + size_t bits = len * 8; + unsigned long s = 0; + u32 crc_low = 0; + + s = 0; + for (int i = 0; i < len; i++) + s = *p++ | (s << 8); + + if (__riscv_xlen == 32 || len < sizeof(u32)) { + s ^= crc >> (32 - bits); + crc_low = crc << bits; + } else { + s ^= (unsigned long)crc << (bits - 32); + } + + crc = crc32_be_zbc(s); + crc ^= crc_low; + + return crc; +} + +u32 __pure crc32_be_arch(u32 crc, const u8 *p, size_t len) +{ + size_t offset, head_len, tail_len; + unsigned long const *p_ul; + unsigned long s; + + asm goto(ALTERNATIVE("j %l[legacy]", "nop", 0, + RISCV_ISA_EXT_ZBC, 1) + : : : : legacy); + + /* Handle the unaligned head. */ + offset = (unsigned long)p & OFFSET_MASK; + if (offset && len) { + head_len = min(STEP - offset, len); + crc = crc32_be_unaligned(crc, p, head_len); + p += head_len; + len -= head_len; + } + + tail_len = len & OFFSET_MASK; + len = len >> STEP_ORDER; + p_ul = (unsigned long const *)p; + + for (int i = 0; i < len; i++) { + s = crc32_be_prep(crc, p_ul); + crc = crc32_be_zbc(s); + p_ul++; + } + + /* Handle the tail bytes. */ + p = (unsigned char const *)p_ul; + if (tail_len) + crc = crc32_be_unaligned(crc, p, tail_len); + + return crc; + +legacy: + return crc32_be_base(crc, p, len); +} +EXPORT_SYMBOL(crc32_be_arch); + +MODULE_LICENSE("GPL"); +MODULE_DESCRIPTION("Accelerated CRC32 implementation with Zbc extension"); diff --git a/arch/riscv/lib/crc32.c b/arch/riscv/lib/crc32.c deleted file mode 100644 index 333fb7af1192..000000000000 --- a/arch/riscv/lib/crc32.c +++ /dev/null @@ -1,294 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only -/* - * Accelerated CRC32 implementation with Zbc extension. - * - * Copyright (C) 2024 Intel Corporation - */ - -#include -#include -#include - -#include -#include -#include -#include -#include - -/* - * Refer to https://www.corsix.org/content/barrett-reduction-polynomials for - * better understanding of how this math works. - * - * let "+" denotes polynomial add (XOR) - * let "-" denotes polynomial sub (XOR) - * let "*" denotes polynomial multiplication - * let "/" denotes polynomial floor division - * let "S" denotes source data, XLEN bit wide - * let "P" denotes CRC32 polynomial - * let "T" denotes 2^(XLEN+32) - * let "QT" denotes quotient of T/P, with the bit for 2^XLEN being implicit - * - * crc32(S, P) - * => S * (2^32) - S * (2^32) / P * P - * => lowest 32 bits of: S * (2^32) / P * P - * => lowest 32 bits of: S * (2^32) * (T / P) / T * P - * => lowest 32 bits of: S * (2^32) * quotient / T * P - * => lowest 32 bits of: S * quotient / 2^XLEN * P - * => lowest 32 bits of: (clmul_high_part(S, QT) + S) * P - * => clmul_low_part(clmul_high_part(S, QT) + S, P) - * - * In terms of below implementations, the BE case is more intuitive, since the - * higher order bit sits at more significant position. - */ - -#if __riscv_xlen == 64 -/* Slide by XLEN bits per iteration */ -# define STEP_ORDER 3 - -/* Each below polynomial quotient has an implicit bit for 2^XLEN */ - -/* Polynomial quotient of (2^(XLEN+32))/CRC32_POLY, in LE format */ -# define CRC32_POLY_QT_LE 0x5a72d812fb808b20 - -/* Polynomial quotient of (2^(XLEN+32))/CRC32C_POLY, in LE format */ -# define CRC32C_POLY_QT_LE 0xa434f61c6f5389f8 - -/* Polynomial quotient of (2^(XLEN+32))/CRC32_POLY, in BE format, it should be - * the same as the bit-reversed version of CRC32_POLY_QT_LE - */ -# define CRC32_POLY_QT_BE 0x04d101df481b4e5a - -static inline u64 crc32_le_prep(u32 crc, unsigned long const *ptr) -{ - return (u64)crc ^ (__force u64)__cpu_to_le64(*ptr); -} - -static inline u32 crc32_le_zbc(unsigned long s, u32 poly, unsigned long poly_qt) -{ - u32 crc; - - /* We don't have a "clmulrh" insn, so use clmul + slli instead. */ - asm volatile (".option push\n" - ".option arch,+zbc\n" - "clmul %0, %1, %2\n" - "slli %0, %0, 1\n" - "xor %0, %0, %1\n" - "clmulr %0, %0, %3\n" - "srli %0, %0, 32\n" - ".option pop\n" - : "=&r" (crc) - : "r" (s), - "r" (poly_qt), - "r" ((u64)poly << 32) - :); - return crc; -} - -static inline u64 crc32_be_prep(u32 crc, unsigned long const *ptr) -{ - return ((u64)crc << 32) ^ (__force u64)__cpu_to_be64(*ptr); -} - -#elif __riscv_xlen == 32 -# define STEP_ORDER 2 -/* Each quotient should match the upper half of its analog in RV64 */ -# define CRC32_POLY_QT_LE 0xfb808b20 -# define CRC32C_POLY_QT_LE 0x6f5389f8 -# define CRC32_POLY_QT_BE 0x04d101df - -static inline u32 crc32_le_prep(u32 crc, unsigned long const *ptr) -{ - return crc ^ (__force u32)__cpu_to_le32(*ptr); -} - -static inline u32 crc32_le_zbc(unsigned long s, u32 poly, unsigned long poly_qt) -{ - u32 crc; - - /* We don't have a "clmulrh" insn, so use clmul + slli instead. */ - asm volatile (".option push\n" - ".option arch,+zbc\n" - "clmul %0, %1, %2\n" - "slli %0, %0, 1\n" - "xor %0, %0, %1\n" - "clmulr %0, %0, %3\n" - ".option pop\n" - : "=&r" (crc) - : "r" (s), - "r" (poly_qt), - "r" (poly) - :); - return crc; -} - -static inline u32 crc32_be_prep(u32 crc, unsigned long const *ptr) -{ - return crc ^ (__force u32)__cpu_to_be32(*ptr); -} - -#else -# error "Unexpected __riscv_xlen" -#endif - -static inline u32 crc32_be_zbc(unsigned long s) -{ - u32 crc; - - asm volatile (".option push\n" - ".option arch,+zbc\n" - "clmulh %0, %1, %2\n" - "xor %0, %0, %1\n" - "clmul %0, %0, %3\n" - ".option pop\n" - : "=&r" (crc) - : "r" (s), - "r" (CRC32_POLY_QT_BE), - "r" (CRC32_POLY_BE) - :); - return crc; -} - -#define STEP (1 << STEP_ORDER) -#define OFFSET_MASK (STEP - 1) - -typedef u32 (*fallback)(u32 crc, unsigned char const *p, size_t len); - -static inline u32 crc32_le_unaligned(u32 crc, unsigned char const *p, - size_t len, u32 poly, - unsigned long poly_qt) -{ - size_t bits = len * 8; - unsigned long s = 0; - u32 crc_low = 0; - - for (int i = 0; i < len; i++) - s = ((unsigned long)*p++ << (__riscv_xlen - 8)) | (s >> 8); - - s ^= (unsigned long)crc << (__riscv_xlen - bits); - if (__riscv_xlen == 32 || len < sizeof(u32)) - crc_low = crc >> bits; - - crc = crc32_le_zbc(s, poly, poly_qt); - crc ^= crc_low; - - return crc; -} - -static inline u32 __pure crc32_le_generic(u32 crc, unsigned char const *p, - size_t len, u32 poly, - unsigned long poly_qt, - fallback crc_fb) -{ - size_t offset, head_len, tail_len; - unsigned long const *p_ul; - unsigned long s; - - asm goto(ALTERNATIVE("j %l[legacy]", "nop", 0, - RISCV_ISA_EXT_ZBC, 1) - : : : : legacy); - - /* Handle the unaligned head. */ - offset = (unsigned long)p & OFFSET_MASK; - if (offset && len) { - head_len = min(STEP - offset, len); - crc = crc32_le_unaligned(crc, p, head_len, poly, poly_qt); - p += head_len; - len -= head_len; - } - - tail_len = len & OFFSET_MASK; - len = len >> STEP_ORDER; - p_ul = (unsigned long const *)p; - - for (int i = 0; i < len; i++) { - s = crc32_le_prep(crc, p_ul); - crc = crc32_le_zbc(s, poly, poly_qt); - p_ul++; - } - - /* Handle the tail bytes. */ - p = (unsigned char const *)p_ul; - if (tail_len) - crc = crc32_le_unaligned(crc, p, tail_len, poly, poly_qt); - - return crc; - -legacy: - return crc_fb(crc, p, len); -} - -u32 __pure crc32_le(u32 crc, unsigned char const *p, size_t len) -{ - return crc32_le_generic(crc, p, len, CRC32_POLY_LE, CRC32_POLY_QT_LE, - crc32_le_base); -} - -u32 __pure __crc32c_le(u32 crc, unsigned char const *p, size_t len) -{ - return crc32_le_generic(crc, p, len, CRC32C_POLY_LE, - CRC32C_POLY_QT_LE, crc32c_le_base); -} - -static inline u32 crc32_be_unaligned(u32 crc, unsigned char const *p, - size_t len) -{ - size_t bits = len * 8; - unsigned long s = 0; - u32 crc_low = 0; - - s = 0; - for (int i = 0; i < len; i++) - s = *p++ | (s << 8); - - if (__riscv_xlen == 32 || len < sizeof(u32)) { - s ^= crc >> (32 - bits); - crc_low = crc << bits; - } else { - s ^= (unsigned long)crc << (bits - 32); - } - - crc = crc32_be_zbc(s); - crc ^= crc_low; - - return crc; -} - -u32 __pure crc32_be(u32 crc, unsigned char const *p, size_t len) -{ - size_t offset, head_len, tail_len; - unsigned long const *p_ul; - unsigned long s; - - asm goto(ALTERNATIVE("j %l[legacy]", "nop", 0, - RISCV_ISA_EXT_ZBC, 1) - : : : : legacy); - - /* Handle the unaligned head. */ - offset = (unsigned long)p & OFFSET_MASK; - if (offset && len) { - head_len = min(STEP - offset, len); - crc = crc32_be_unaligned(crc, p, head_len); - p += head_len; - len -= head_len; - } - - tail_len = len & OFFSET_MASK; - len = len >> STEP_ORDER; - p_ul = (unsigned long const *)p; - - for (int i = 0; i < len; i++) { - s = crc32_be_prep(crc, p_ul); - crc = crc32_be_zbc(s); - p_ul++; - } - - /* Handle the tail bytes. */ - p = (unsigned char const *)p_ul; - if (tail_len) - crc = crc32_be_unaligned(crc, p, tail_len); - - return crc; - -legacy: - return crc32_be_base(crc, p, len); -} diff --git a/crypto/crc32_generic.c b/crypto/crc32_generic.c index 6a55d206fab3..cc064ea8240e 100644 --- a/crypto/crc32_generic.c +++ b/crypto/crc32_generic.c @@ -160,12 +160,12 @@ static struct shash_alg algs[] = {{ static int __init crc32_mod_init(void) { /* register the arch flavor only if it differs from the generic one */ - return crypto_register_shashes(algs, 1 + (&crc32_le != &crc32_le_base)); + return crypto_register_shashes(algs, 1 + IS_ENABLED(CONFIG_CRC32_ARCH)); } static void __exit crc32_mod_fini(void) { - crypto_unregister_shashes(algs, 1 + (&crc32_le != &crc32_le_base)); + crypto_unregister_shashes(algs, 1 + IS_ENABLED(CONFIG_CRC32_ARCH)); } subsys_initcall(crc32_mod_init); diff --git a/crypto/crc32c_generic.c b/crypto/crc32c_generic.c index 635599b255ec..04b03d825cf4 100644 --- a/crypto/crc32c_generic.c +++ b/crypto/crc32c_generic.c @@ -200,12 +200,12 @@ static struct shash_alg algs[] = {{ static int __init crc32c_mod_init(void) { /* register the arch flavor only if it differs from the generic one */ - return crypto_register_shashes(algs, 1 + (&__crc32c_le != &crc32c_le_base)); + return crypto_register_shashes(algs, 1 + IS_ENABLED(CONFIG_CRC32_ARCH)); } static void __exit crc32c_mod_fini(void) { - crypto_unregister_shashes(algs, 1 + (&__crc32c_le != &crc32c_le_base)); + crypto_unregister_shashes(algs, 1 + IS_ENABLED(CONFIG_CRC32_ARCH)); } subsys_initcall(crc32c_mod_init); diff --git a/include/linux/crc32.h b/include/linux/crc32.h index 5b07fc9081c4..58c632533b08 100644 --- a/include/linux/crc32.h +++ b/include/linux/crc32.h @@ -8,10 +8,34 @@ #include #include -u32 __pure crc32_le(u32 crc, unsigned char const *p, size_t len); -u32 __pure crc32_le_base(u32 crc, unsigned char const *p, size_t len); -u32 __pure crc32_be(u32 crc, unsigned char const *p, size_t len); -u32 __pure crc32_be_base(u32 crc, unsigned char const *p, size_t len); +u32 __pure crc32_le_arch(u32 crc, const u8 *p, size_t len); +u32 __pure crc32_le_base(u32 crc, const u8 *p, size_t len); +u32 __pure crc32_be_arch(u32 crc, const u8 *p, size_t len); +u32 __pure crc32_be_base(u32 crc, const u8 *p, size_t len); +u32 __pure crc32c_le_arch(u32 crc, const u8 *p, size_t len); +u32 __pure crc32c_le_base(u32 crc, const u8 *p, size_t len); + +static inline u32 __pure crc32_le(u32 crc, const u8 *p, size_t len) +{ + if (IS_ENABLED(CONFIG_CRC32_ARCH)) + return crc32_le_arch(crc, p, len); + return crc32_le_base(crc, p, len); +} + +static inline u32 __pure crc32_be(u32 crc, const u8 *p, size_t len) +{ + if (IS_ENABLED(CONFIG_CRC32_ARCH)) + return crc32_be_arch(crc, p, len); + return crc32_be_base(crc, p, len); +} + +/* TODO: leading underscores should be dropped once callers have been updated */ +static inline u32 __pure __crc32c_le(u32 crc, const u8 *p, size_t len) +{ + if (IS_ENABLED(CONFIG_CRC32_ARCH)) + return crc32c_le_arch(crc, p, len); + return crc32c_le_base(crc, p, len); +} /** * crc32_le_combine - Combine two crc32 check values into one. For two @@ -38,9 +62,6 @@ static inline u32 crc32_le_combine(u32 crc1, u32 crc2, size_t len2) return crc32_le_shift(crc1, len2) ^ crc2; } -u32 __pure __crc32c_le(u32 crc, unsigned char const *p, size_t len); -u32 __pure crc32c_le_base(u32 crc, unsigned char const *p, size_t len); - /** * __crc32c_le_combine - Combine two crc32c check values into one. For two * sequences of bytes, seq1 and seq2 with lengths len1 diff --git a/lib/Kconfig b/lib/Kconfig index 5a318f753b2f..8858030bebae 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -190,6 +190,9 @@ config CRC32 the kernel tree does. Such modules that use library CRC32/CRC32c functions require M here. +config ARCH_HAS_CRC32 + bool + config CRC32_SELFTEST tristate "CRC32 perform self test on init" depends on CRC32 @@ -202,24 +205,39 @@ config CRC32_SELFTEST choice prompt "CRC32 implementation" depends on CRC32 - default CRC32_SLICEBY8 + default CRC32_IMPL_ARCH_PLUS_SLICEBY8 if ARCH_HAS_CRC32 + default CRC32_IMPL_SLICEBY8 if !ARCH_HAS_CRC32 help - This option allows a kernel builder to override the default choice - of CRC32 algorithm. Choose the default ("slice by 8") unless you - know that you need one of the others. + This option allows you to override the default choice of CRC32 + implementation. Choose the default unless you know that you need one + of the others. -config CRC32_SLICEBY8 +config CRC32_IMPL_ARCH_PLUS_SLICEBY8 + bool "Arch-optimized, with fallback to slice-by-8" if ARCH_HAS_CRC32 + help + Use architecture-optimized implementation of CRC32. Fall back to + slice-by-8 in cases where the arch-optimized implementation cannot be + used, e.g. if the CPU lacks support for the needed instructions. + + This is the default when an arch-optimized implementation exists. + +config CRC32_IMPL_ARCH_PLUS_SLICEBY1 + bool "Arch-optimized, with fallback to slice-by-1" if ARCH_HAS_CRC32 + help + Use architecture-optimized implementation of CRC32, but fall back to + slice-by-1 instead of slice-by-8 in order to reduce the binary size. + +config CRC32_IMPL_SLICEBY8 bool "Slice by 8 bytes" help Calculate checksum 8 bytes at a time with a clever slicing algorithm. - This is the fastest algorithm, but comes with a 8KiB lookup table. - Most modern processors have enough cache to hold this table without - thrashing the cache. - - This is the default implementation choice. Choose this one unless - you have a good reason not to. + This is much slower than the architecture-optimized implementation of + CRC32 (if the selected arch has one), but it is portable and is the + fastest implementation when no arch-optimized implementation is + available. It uses an 8KiB lookup table. Most modern processors have + enough cache to hold this table without thrashing the cache. -config CRC32_SLICEBY4 +config CRC32_IMPL_SLICEBY4 bool "Slice by 4 bytes" help Calculate checksum 4 bytes at a time with a clever slicing algorithm. @@ -228,15 +246,15 @@ config CRC32_SLICEBY4 Only choose this option if you know what you are doing. -config CRC32_SARWATE - bool "Sarwate's Algorithm (one byte at a time)" +config CRC32_IMPL_SLICEBY1 + bool "Slice by 1 byte (Sarwate's algorithm)" help Calculate checksum a byte at a time using Sarwate's algorithm. This - is not particularly fast, but has a small 256 byte lookup table. + is not particularly fast, but has a small 1KiB lookup table. Only choose this option if you know what you are doing. -config CRC32_BIT +config CRC32_IMPL_BIT bool "Classic Algorithm (one bit at a time)" help Calculate checksum one bit at a time. This is VERY slow, but has @@ -246,6 +264,26 @@ config CRC32_BIT endchoice +config CRC32_ARCH + tristate + default CRC32 if CRC32_IMPL_ARCH_PLUS_SLICEBY8 || CRC32_IMPL_ARCH_PLUS_SLICEBY1 + +config CRC32_SLICEBY8 + bool + default y if CRC32_IMPL_SLICEBY8 || CRC32_IMPL_ARCH_PLUS_SLICEBY8 + +config CRC32_SLICEBY4 + bool + default y if CRC32_IMPL_SLICEBY4 + +config CRC32_SARWATE + bool + default y if CRC32_IMPL_SLICEBY1 || CRC32_IMPL_ARCH_PLUS_SLICEBY1 + +config CRC32_BIT + bool + default y if CRC32_IMPL_BIT + config CRC64 tristate "CRC64 functions" help diff --git a/lib/crc32.c b/lib/crc32.c index c67059b0082b..47151624332e 100644 --- a/lib/crc32.c +++ b/lib/crc32.c @@ -183,35 +183,27 @@ static inline u32 __pure crc32_le_generic(u32 crc, unsigned char const *p, } #if CRC_LE_BITS == 1 -u32 __pure __weak crc32_le(u32 crc, unsigned char const *p, size_t len) +u32 __pure crc32_le_base(u32 crc, const u8 *p, size_t len) { return crc32_le_generic(crc, p, len, NULL, CRC32_POLY_LE); } -u32 __pure __weak __crc32c_le(u32 crc, unsigned char const *p, size_t len) +u32 __pure crc32c_le_base(u32 crc, const u8 *p, size_t len) { return crc32_le_generic(crc, p, len, NULL, CRC32C_POLY_LE); } #else -u32 __pure __weak crc32_le(u32 crc, unsigned char const *p, size_t len) +u32 __pure crc32_le_base(u32 crc, const u8 *p, size_t len) { return crc32_le_generic(crc, p, len, crc32table_le, CRC32_POLY_LE); } -u32 __pure __weak __crc32c_le(u32 crc, unsigned char const *p, size_t len) +u32 __pure crc32c_le_base(u32 crc, const u8 *p, size_t len) { return crc32_le_generic(crc, p, len, crc32ctable_le, CRC32C_POLY_LE); } #endif -EXPORT_SYMBOL(crc32_le); -EXPORT_SYMBOL(__crc32c_le); - -u32 __pure crc32_le_base(u32, unsigned char const *, size_t) __alias(crc32_le); EXPORT_SYMBOL(crc32_le_base); - -u32 __pure crc32c_le_base(u32, unsigned char const *, size_t) __alias(__crc32c_le); EXPORT_SYMBOL(crc32c_le_base); -u32 __pure crc32_be_base(u32, unsigned char const *, size_t) __alias(crc32_be); - /* * This multiplies the polynomials x and y modulo the given modulus. * This follows the "little-endian" CRC convention that the lsbit @@ -335,14 +327,14 @@ static inline u32 __pure crc32_be_generic(u32 crc, unsigned char const *p, } #if CRC_BE_BITS == 1 -u32 __pure __weak crc32_be(u32 crc, unsigned char const *p, size_t len) +u32 __pure crc32_be_base(u32 crc, const u8 *p, size_t len) { return crc32_be_generic(crc, p, len, NULL, CRC32_POLY_BE); } #else -u32 __pure __weak crc32_be(u32 crc, unsigned char const *p, size_t len) +u32 __pure crc32_be_base(u32 crc, const u8 *p, size_t len) { return crc32_be_generic(crc, p, len, crc32table_be, CRC32_POLY_BE); } #endif -EXPORT_SYMBOL(crc32_be); +EXPORT_SYMBOL(crc32_be_base); -- cgit v1.2.3 From b5ae12e0ee099e4c458f7814f0317f4e2cbf105e Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Sun, 1 Dec 2024 17:08:28 -0800 Subject: lib/crc32: expose whether the lib is really optimized at runtime Make the CRC32 library export a function crc32_optimizations() which returns flags that indicate which CRC32 functions are actually executing optimized code at runtime. This will be used to determine whether the crc32[c]-$arch shash algorithms should be registered in the crypto API. btrfs could also start using these flags instead of the hack that it currently uses where it parses the crypto_shash_driver_name. Reviewed-by: Ard Biesheuvel Link: https://lore.kernel.org/r/20241202010844.144356-4-ebiggers@kernel.org Signed-off-by: Eric Biggers --- arch/arm64/lib/crc32-glue.c | 10 ++++++++++ arch/riscv/lib/crc32-riscv.c | 10 ++++++++++ include/linux/crc32.h | 15 +++++++++++++++ 3 files changed, 35 insertions(+) (limited to 'include/linux') diff --git a/arch/arm64/lib/crc32-glue.c b/arch/arm64/lib/crc32-glue.c index d7f6e1cbf0d2..15c4c9db573e 100644 --- a/arch/arm64/lib/crc32-glue.c +++ b/arch/arm64/lib/crc32-glue.c @@ -85,5 +85,15 @@ u32 __pure crc32_be_arch(u32 crc, const u8 *p, size_t len) } EXPORT_SYMBOL(crc32_be_arch); +u32 crc32_optimizations(void) +{ + if (alternative_has_cap_likely(ARM64_HAS_CRC32)) + return CRC32_LE_OPTIMIZATION | + CRC32_BE_OPTIMIZATION | + CRC32C_OPTIMIZATION; + return 0; +} +EXPORT_SYMBOL(crc32_optimizations); + MODULE_LICENSE("GPL"); MODULE_DESCRIPTION("arm64-optimized CRC32 functions"); diff --git a/arch/riscv/lib/crc32-riscv.c b/arch/riscv/lib/crc32-riscv.c index a3ff7db2a1ce..53d56ab422c7 100644 --- a/arch/riscv/lib/crc32-riscv.c +++ b/arch/riscv/lib/crc32-riscv.c @@ -297,5 +297,15 @@ legacy: } EXPORT_SYMBOL(crc32_be_arch); +u32 crc32_optimizations(void) +{ + if (riscv_has_extension_likely(RISCV_ISA_EXT_ZBC)) + return CRC32_LE_OPTIMIZATION | + CRC32_BE_OPTIMIZATION | + CRC32C_OPTIMIZATION; + return 0; +} +EXPORT_SYMBOL(crc32_optimizations); + MODULE_LICENSE("GPL"); MODULE_DESCRIPTION("Accelerated CRC32 implementation with Zbc extension"); diff --git a/include/linux/crc32.h b/include/linux/crc32.h index 58c632533b08..e9bd40056687 100644 --- a/include/linux/crc32.h +++ b/include/linux/crc32.h @@ -37,6 +37,21 @@ static inline u32 __pure __crc32c_le(u32 crc, const u8 *p, size_t len) return crc32c_le_base(crc, p, len); } +/* + * crc32_optimizations() returns flags that indicate which CRC32 library + * functions are using architecture-specific optimizations. Unlike + * IS_ENABLED(CONFIG_CRC32_ARCH) it takes into account the different CRC32 + * variants and also whether any needed CPU features are available at runtime. + */ +#define CRC32_LE_OPTIMIZATION BIT(0) /* crc32_le() is optimized */ +#define CRC32_BE_OPTIMIZATION BIT(1) /* crc32_be() is optimized */ +#define CRC32C_OPTIMIZATION BIT(2) /* __crc32c_le() is optimized */ +#if IS_ENABLED(CONFIG_CRC32_ARCH) +u32 crc32_optimizations(void); +#else +static inline u32 crc32_optimizations(void) { return 0; } +#endif + /** * crc32_le_combine - Combine two crc32 check values into one. For two * sequences of bytes, seq1 and seq2 with lengths len1 -- cgit v1.2.3 From 38a9a5121c3bcf2ed857430a92e493568b247c35 Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Sun, 1 Dec 2024 17:08:40 -0800 Subject: lib/crc32: make crc32c() go directly to lib Now that the lower level __crc32c_le() library function is optimized for each architecture, make crc32c() just call that instead of taking an inefficient and error-prone detour through the shash API. Note: a future cleanup should make crc32c_le() be the actual library function instead of __crc32c_le(). That will require updating callers of __crc32c_le() to use crc32c_le() instead, and updating callers of crc32c_le() that expect a 'const void *' arg to expect 'const u8 *' instead. Similarly, a future cleanup should remove LIBCRC32C by making everyone who is selecting it just select CRC32 directly instead. Reviewed-by: Ard Biesheuvel Link: https://lore.kernel.org/r/20241202010844.144356-16-ebiggers@kernel.org Signed-off-by: Eric Biggers --- include/linux/crc32c.h | 7 +++-- lib/Kconfig | 10 ++----- lib/Makefile | 1 - lib/libcrc32c.c | 74 -------------------------------------------------- 4 files changed, 8 insertions(+), 84 deletions(-) delete mode 100644 lib/libcrc32c.c (limited to 'include/linux') diff --git a/include/linux/crc32c.h b/include/linux/crc32c.h index 357ae4611a45..47eb78003c26 100644 --- a/include/linux/crc32c.h +++ b/include/linux/crc32c.h @@ -2,9 +2,12 @@ #ifndef _LINUX_CRC32C_H #define _LINUX_CRC32C_H -#include +#include -extern u32 crc32c(u32 crc, const void *address, unsigned int length); +static inline u32 crc32c(u32 crc, const void *address, unsigned int length) +{ + return __crc32c_le(crc, address, length); +} /* This macro exists for backwards-compatibility. */ #define crc32c_le crc32c diff --git a/lib/Kconfig b/lib/Kconfig index 8858030bebae..d0c2dc0dac32 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -310,14 +310,10 @@ config CRC7 config LIBCRC32C tristate "CRC32c (Castagnoli, et al) Cyclic Redundancy-Check" - select CRYPTO - select CRYPTO_CRC32C + select CRC32 help - This option is provided for the case where no in-kernel-tree - modules require CRC32c functions, but a module built outside the - kernel tree does. Such modules that use library CRC32c functions - require M here. See Castagnoli93. - Module will be libcrc32c. + This option just selects CRC32 and is provided for compatibility + purposes until the users are updated to select CRC32 directly. config CRC8 tristate "CRC8 function" diff --git a/lib/Makefile b/lib/Makefile index a8155c972f02..6573163d24e4 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -167,7 +167,6 @@ obj-$(CONFIG_CRC64) += crc64.o obj-$(CONFIG_CRC32_SELFTEST) += crc32test.o obj-$(CONFIG_CRC4) += crc4.o obj-$(CONFIG_CRC7) += crc7.o -obj-$(CONFIG_LIBCRC32C) += libcrc32c.o obj-$(CONFIG_CRC8) += crc8.o obj-$(CONFIG_CRC64_ROCKSOFT) += crc64-rocksoft.o obj-$(CONFIG_XXHASH) += xxhash.o diff --git a/lib/libcrc32c.c b/lib/libcrc32c.c deleted file mode 100644 index 649e687413a0..000000000000 --- a/lib/libcrc32c.c +++ /dev/null @@ -1,74 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-or-later -/* - * CRC32C - *@Article{castagnoli-crc, - * author = { Guy Castagnoli and Stefan Braeuer and Martin Herrman}, - * title = {{Optimization of Cyclic Redundancy-Check Codes with 24 - * and 32 Parity Bits}}, - * journal = IEEE Transactions on Communication, - * year = {1993}, - * volume = {41}, - * number = {6}, - * pages = {}, - * month = {June}, - *} - * Used by the iSCSI driver, possibly others, and derived from - * the iscsi-crc.c module of the linux-iscsi driver at - * http://linux-iscsi.sourceforge.net. - * - * Following the example of lib/crc32, this function is intended to be - * flexible and useful for all users. Modules that currently have their - * own crc32c, but hopefully may be able to use this one are: - * net/sctp (please add all your doco to here if you change to - * use this one!) - * - * - * Copyright (c) 2004 Cisco Systems, Inc. - */ - -#include -#include -#include -#include -#include -#include - -static struct crypto_shash *tfm; - -u32 crc32c(u32 crc, const void *address, unsigned int length) -{ - SHASH_DESC_ON_STACK(shash, tfm); - u32 ret, *ctx = (u32 *)shash_desc_ctx(shash); - int err; - - shash->tfm = tfm; - *ctx = crc; - - err = crypto_shash_update(shash, address, length); - BUG_ON(err); - - ret = *ctx; - barrier_data(ctx); - return ret; -} - -EXPORT_SYMBOL(crc32c); - -static int __init libcrc32c_mod_init(void) -{ - tfm = crypto_alloc_shash("crc32c", 0, 0); - return PTR_ERR_OR_ZERO(tfm); -} - -static void __exit libcrc32c_mod_fini(void) -{ - crypto_free_shash(tfm); -} - -module_init(libcrc32c_mod_init); -module_exit(libcrc32c_mod_fini); - -MODULE_AUTHOR("Clay Haapala "); -MODULE_DESCRIPTION("CRC32c (Castagnoli) calculations"); -MODULE_LICENSE("GPL"); -MODULE_SOFTDEP("pre: crc32c"); -- cgit v1.2.3 From dd348f054b24a3f57cbcdc2c8e7ebc22c62eb72f Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Sun, 1 Dec 2024 17:08:42 -0800 Subject: jbd2: switch to using the crc32c library Now that the crc32c() library function directly takes advantage of architecture-specific optimizations, it is unnecessary to go through the crypto API. Just use crc32c(). This is much simpler, and it improves performance due to eliminating the crypto API overhead. Reviewed-by: Ard Biesheuvel Reviewed-by: Darrick J. Wong Acked-by: Theodore Ts'o Link: https://lore.kernel.org/r/20241202010844.144356-18-ebiggers@kernel.org Signed-off-by: Eric Biggers --- fs/jbd2/Kconfig | 2 -- fs/jbd2/journal.c | 30 +++--------------------------- include/linux/jbd2.h | 33 +++------------------------------ 3 files changed, 6 insertions(+), 59 deletions(-) (limited to 'include/linux') diff --git a/fs/jbd2/Kconfig b/fs/jbd2/Kconfig index 4ad2c67f93f1..9c19e1512101 100644 --- a/fs/jbd2/Kconfig +++ b/fs/jbd2/Kconfig @@ -2,8 +2,6 @@ config JBD2 tristate select CRC32 - select CRYPTO - select CRYPTO_CRC32C help This is a generic journaling layer for block devices that support both 32-bit and 64-bit block numbers. It is currently used by diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 7e49d912b091..d8084b31b361 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -1369,20 +1369,12 @@ static int journal_check_superblock(journal_t *journal) return err; } - /* Load the checksum driver */ if (jbd2_journal_has_csum_v2or3_feature(journal)) { if (sb->s_checksum_type != JBD2_CRC32C_CHKSUM) { printk(KERN_ERR "JBD2: Unknown checksum type\n"); return err; } - journal->j_chksum_driver = crypto_alloc_shash("crc32c", 0, 0); - if (IS_ERR(journal->j_chksum_driver)) { - printk(KERN_ERR "JBD2: Cannot load crc32c driver.\n"); - err = PTR_ERR(journal->j_chksum_driver); - journal->j_chksum_driver = NULL; - return err; - } /* Check superblock checksum */ if (sb->s_checksum != jbd2_superblock_csum(journal, sb)) { printk(KERN_ERR "JBD2: journal checksum error\n"); @@ -1608,8 +1600,6 @@ static journal_t *journal_init_common(struct block_device *bdev, err_cleanup: percpu_counter_destroy(&journal->j_checkpoint_jh_count); - if (journal->j_chksum_driver) - crypto_free_shash(journal->j_chksum_driver); kfree(journal->j_wbuf); jbd2_journal_destroy_revoke(journal); journal_fail_superblock(journal); @@ -2191,8 +2181,6 @@ int jbd2_journal_destroy(journal_t *journal) iput(journal->j_inode); if (journal->j_revoke) jbd2_journal_destroy_revoke(journal); - if (journal->j_chksum_driver) - crypto_free_shash(journal->j_chksum_driver); kfree(journal->j_fc_wbuf); kfree(journal->j_wbuf); kfree(journal); @@ -2337,27 +2325,15 @@ int jbd2_journal_set_features(journal_t *journal, unsigned long compat, } } - /* Load the checksum driver if necessary */ - if ((journal->j_chksum_driver == NULL) && - INCOMPAT_FEATURE_ON(JBD2_FEATURE_INCOMPAT_CSUM_V3)) { - journal->j_chksum_driver = crypto_alloc_shash("crc32c", 0, 0); - if (IS_ERR(journal->j_chksum_driver)) { - printk(KERN_ERR "JBD2: Cannot load crc32c driver.\n"); - journal->j_chksum_driver = NULL; - return 0; - } - /* Precompute checksum seed for all metadata */ - journal->j_csum_seed = jbd2_chksum(journal, ~0, sb->s_uuid, - sizeof(sb->s_uuid)); - } - lock_buffer(journal->j_sb_buffer); - /* If enabling v3 checksums, update superblock */ + /* If enabling v3 checksums, update superblock and precompute seed */ if (INCOMPAT_FEATURE_ON(JBD2_FEATURE_INCOMPAT_CSUM_V3)) { sb->s_checksum_type = JBD2_CRC32C_CHKSUM; sb->s_feature_compat &= ~cpu_to_be32(JBD2_FEATURE_COMPAT_CHECKSUM); + journal->j_csum_seed = jbd2_chksum(journal, ~0, sb->s_uuid, + sizeof(sb->s_uuid)); } /* If enabling v1 checksums, downgrade superblock */ diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h index 50f7ea8714bf..561025b4f3d9 100644 --- a/include/linux/jbd2.h +++ b/include/linux/jbd2.h @@ -28,7 +28,7 @@ #include #include #include -#include +#include #endif #define journal_oom_retry 1 @@ -1241,13 +1241,6 @@ struct journal_s */ void *j_private; - /** - * @j_chksum_driver: - * - * Reference to checksum algorithm driver via cryptoapi. - */ - struct crypto_shash *j_chksum_driver; - /** * @j_csum_seed: * @@ -1750,10 +1743,7 @@ static inline bool jbd2_journal_has_csum_v2or3_feature(journal_t *j) static inline int jbd2_journal_has_csum_v2or3(journal_t *journal) { - WARN_ON_ONCE(jbd2_journal_has_csum_v2or3_feature(journal) && - journal->j_chksum_driver == NULL); - - return journal->j_chksum_driver != NULL; + return jbd2_journal_has_csum_v2or3_feature(journal); } static inline int jbd2_journal_get_num_fc_blks(journal_superblock_t *jsb) @@ -1790,27 +1780,10 @@ static inline unsigned long jbd2_log_space_left(journal_t *journal) #define BJ_Reserved 4 /* Buffer is reserved for access by journal */ #define BJ_Types 5 -/* JBD uses a CRC32 checksum */ -#define JBD_MAX_CHECKSUM_SIZE 4 - static inline u32 jbd2_chksum(journal_t *journal, u32 crc, const void *address, unsigned int length) { - DEFINE_RAW_FLEX(struct shash_desc, desc, __ctx, - DIV_ROUND_UP(JBD_MAX_CHECKSUM_SIZE, - sizeof(*((struct shash_desc *)0)->__ctx))); - int err; - - BUG_ON(crypto_shash_descsize(journal->j_chksum_driver) > - JBD_MAX_CHECKSUM_SIZE); - - desc->tfm = journal->j_chksum_driver; - *(u32 *)desc->__ctx = crc; - - err = crypto_shash_update(desc, address, length); - BUG_ON(err); - - return *(u32 *)desc->__ctx; + return crc32c(crc, address, length); } /* Return most recent uncommitted transaction */ -- cgit v1.2.3 From be3c45b070cba3be4dd248b38d4798e3e2859451 Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Sun, 1 Dec 2024 17:20:45 -0800 Subject: lib/crc-t10dif: stop wrapping the crypto API In preparation for making the CRC-T10DIF library directly optimized for each architecture, like what has been done for CRC32, get rid of the weird layering where crc_t10dif_update() calls into the crypto API. Instead, move crc_t10dif_generic() into the crc-t10dif library module, and make crc_t10dif_update() just call crc_t10dif_generic(). Acceleration will be reintroduced via crc_t10dif_arch() in the following patches. Reviewed-by: Ard Biesheuvel Reviewed-by: Martin K. Petersen Link: https://lore.kernel.org/r/20241202012056.209768-2-ebiggers@kernel.org Signed-off-by: Eric Biggers --- crypto/Kconfig | 1 + crypto/Makefile | 2 +- crypto/crct10dif_common.c | 82 ------------------------ include/linux/crc-t10dif.h | 16 +++-- lib/Kconfig | 2 - lib/crc-t10dif.c | 156 +++++++++++++-------------------------------- 6 files changed, 58 insertions(+), 201 deletions(-) delete mode 100644 crypto/crct10dif_common.c (limited to 'include/linux') diff --git a/crypto/Kconfig b/crypto/Kconfig index 6b0bfbccac08..b459e8a23acc 100644 --- a/crypto/Kconfig +++ b/crypto/Kconfig @@ -1102,6 +1102,7 @@ config CRYPTO_CRC32 config CRYPTO_CRCT10DIF tristate "CRCT10DIF" select CRYPTO_HASH + select CRC_T10DIF help CRC16 CRC algorithm used for the T10 (SCSI) Data Integrity Field (DIF) diff --git a/crypto/Makefile b/crypto/Makefile index 77abca715445..0537df1719d3 100644 --- a/crypto/Makefile +++ b/crypto/Makefile @@ -157,7 +157,7 @@ obj-$(CONFIG_CRYPTO_CRC32C) += crc32c_generic.o obj-$(CONFIG_CRYPTO_CRC32) += crc32_generic.o CFLAGS_crc32c_generic.o += -DARCH=$(ARCH) CFLAGS_crc32_generic.o += -DARCH=$(ARCH) -obj-$(CONFIG_CRYPTO_CRCT10DIF) += crct10dif_common.o crct10dif_generic.o +obj-$(CONFIG_CRYPTO_CRCT10DIF) += crct10dif_generic.o obj-$(CONFIG_CRYPTO_CRC64_ROCKSOFT) += crc64_rocksoft_generic.o obj-$(CONFIG_CRYPTO_AUTHENC) += authenc.o authencesn.o obj-$(CONFIG_CRYPTO_LZO) += lzo.o lzo-rle.o diff --git a/crypto/crct10dif_common.c b/crypto/crct10dif_common.c deleted file mode 100644 index b2fab366f518..000000000000 --- a/crypto/crct10dif_common.c +++ /dev/null @@ -1,82 +0,0 @@ -/* - * Cryptographic API. - * - * T10 Data Integrity Field CRC16 Crypto Transform - * - * Copyright (c) 2007 Oracle Corporation. All rights reserved. - * Written by Martin K. Petersen - * Copyright (C) 2013 Intel Corporation - * Author: Tim Chen - * - * This program is free software; you can redistribute it and/or modify it - * under the terms of the GNU General Public License as published by the Free - * Software Foundation; either version 2 of the License, or (at your option) - * any later version. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, - * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF - * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND - * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS - * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN - * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN - * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE - * SOFTWARE. - * - */ - -#include -#include -#include - -/* Table generated using the following polynomium: - * x^16 + x^15 + x^11 + x^9 + x^8 + x^7 + x^5 + x^4 + x^2 + x + 1 - * gt: 0x8bb7 - */ -static const __u16 t10_dif_crc_table[256] = { - 0x0000, 0x8BB7, 0x9CD9, 0x176E, 0xB205, 0x39B2, 0x2EDC, 0xA56B, - 0xEFBD, 0x640A, 0x7364, 0xF8D3, 0x5DB8, 0xD60F, 0xC161, 0x4AD6, - 0x54CD, 0xDF7A, 0xC814, 0x43A3, 0xE6C8, 0x6D7F, 0x7A11, 0xF1A6, - 0xBB70, 0x30C7, 0x27A9, 0xAC1E, 0x0975, 0x82C2, 0x95AC, 0x1E1B, - 0xA99A, 0x222D, 0x3543, 0xBEF4, 0x1B9F, 0x9028, 0x8746, 0x0CF1, - 0x4627, 0xCD90, 0xDAFE, 0x5149, 0xF422, 0x7F95, 0x68FB, 0xE34C, - 0xFD57, 0x76E0, 0x618E, 0xEA39, 0x4F52, 0xC4E5, 0xD38B, 0x583C, - 0x12EA, 0x995D, 0x8E33, 0x0584, 0xA0EF, 0x2B58, 0x3C36, 0xB781, - 0xD883, 0x5334, 0x445A, 0xCFED, 0x6A86, 0xE131, 0xF65F, 0x7DE8, - 0x373E, 0xBC89, 0xABE7, 0x2050, 0x853B, 0x0E8C, 0x19E2, 0x9255, - 0x8C4E, 0x07F9, 0x1097, 0x9B20, 0x3E4B, 0xB5FC, 0xA292, 0x2925, - 0x63F3, 0xE844, 0xFF2A, 0x749D, 0xD1F6, 0x5A41, 0x4D2F, 0xC698, - 0x7119, 0xFAAE, 0xEDC0, 0x6677, 0xC31C, 0x48AB, 0x5FC5, 0xD472, - 0x9EA4, 0x1513, 0x027D, 0x89CA, 0x2CA1, 0xA716, 0xB078, 0x3BCF, - 0x25D4, 0xAE63, 0xB90D, 0x32BA, 0x97D1, 0x1C66, 0x0B08, 0x80BF, - 0xCA69, 0x41DE, 0x56B0, 0xDD07, 0x786C, 0xF3DB, 0xE4B5, 0x6F02, - 0x3AB1, 0xB106, 0xA668, 0x2DDF, 0x88B4, 0x0303, 0x146D, 0x9FDA, - 0xD50C, 0x5EBB, 0x49D5, 0xC262, 0x6709, 0xECBE, 0xFBD0, 0x7067, - 0x6E7C, 0xE5CB, 0xF2A5, 0x7912, 0xDC79, 0x57CE, 0x40A0, 0xCB17, - 0x81C1, 0x0A76, 0x1D18, 0x96AF, 0x33C4, 0xB873, 0xAF1D, 0x24AA, - 0x932B, 0x189C, 0x0FF2, 0x8445, 0x212E, 0xAA99, 0xBDF7, 0x3640, - 0x7C96, 0xF721, 0xE04F, 0x6BF8, 0xCE93, 0x4524, 0x524A, 0xD9FD, - 0xC7E6, 0x4C51, 0x5B3F, 0xD088, 0x75E3, 0xFE54, 0xE93A, 0x628D, - 0x285B, 0xA3EC, 0xB482, 0x3F35, 0x9A5E, 0x11E9, 0x0687, 0x8D30, - 0xE232, 0x6985, 0x7EEB, 0xF55C, 0x5037, 0xDB80, 0xCCEE, 0x4759, - 0x0D8F, 0x8638, 0x9156, 0x1AE1, 0xBF8A, 0x343D, 0x2353, 0xA8E4, - 0xB6FF, 0x3D48, 0x2A26, 0xA191, 0x04FA, 0x8F4D, 0x9823, 0x1394, - 0x5942, 0xD2F5, 0xC59B, 0x4E2C, 0xEB47, 0x60F0, 0x779E, 0xFC29, - 0x4BA8, 0xC01F, 0xD771, 0x5CC6, 0xF9AD, 0x721A, 0x6574, 0xEEC3, - 0xA415, 0x2FA2, 0x38CC, 0xB37B, 0x1610, 0x9DA7, 0x8AC9, 0x017E, - 0x1F65, 0x94D2, 0x83BC, 0x080B, 0xAD60, 0x26D7, 0x31B9, 0xBA0E, - 0xF0D8, 0x7B6F, 0x6C01, 0xE7B6, 0x42DD, 0xC96A, 0xDE04, 0x55B3 -}; - -__u16 crc_t10dif_generic(__u16 crc, const unsigned char *buffer, size_t len) -{ - unsigned int i; - - for (i = 0 ; i < len ; i++) - crc = (crc << 8) ^ t10_dif_crc_table[((crc >> 8) ^ buffer[i]) & 0xff]; - - return crc; -} -EXPORT_SYMBOL(crc_t10dif_generic); - -MODULE_DESCRIPTION("T10 DIF CRC calculation common code"); -MODULE_LICENSE("GPL"); diff --git a/include/linux/crc-t10dif.h b/include/linux/crc-t10dif.h index 6bb0c0bf357b..206ba2305483 100644 --- a/include/linux/crc-t10dif.h +++ b/include/linux/crc-t10dif.h @@ -6,11 +6,17 @@ #define CRC_T10DIF_DIGEST_SIZE 2 #define CRC_T10DIF_BLOCK_SIZE 1 -#define CRC_T10DIF_STRING "crct10dif" -extern __u16 crc_t10dif_generic(__u16 crc, const unsigned char *buffer, - size_t len); -extern __u16 crc_t10dif(unsigned char const *, size_t); -extern __u16 crc_t10dif_update(__u16 crc, unsigned char const *, size_t); +u16 crc_t10dif_generic(u16 crc, const u8 *p, size_t len); + +static inline u16 crc_t10dif_update(u16 crc, const u8 *p, size_t len) +{ + return crc_t10dif_generic(crc, p, len); +} + +static inline u16 crc_t10dif(const u8 *p, size_t len) +{ + return crc_t10dif_update(0, p, len); +} #endif diff --git a/lib/Kconfig b/lib/Kconfig index d0c2dc0dac32..be59f7cdf448 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -156,8 +156,6 @@ config CRC16 config CRC_T10DIF tristate "CRC calculation for the T10 Data Integrity Field" - select CRYPTO - select CRYPTO_CRCT10DIF help This option is only needed if a module that's not in the kernel tree needs to calculate CRC checks for use with the diff --git a/lib/crc-t10dif.c b/lib/crc-t10dif.c index 1ed2ed487097..311c2ab829f1 100644 --- a/lib/crc-t10dif.c +++ b/lib/crc-t10dif.c @@ -9,123 +9,57 @@ #include #include #include -#include -#include -#include -#include -#include -#include -static struct crypto_shash __rcu *crct10dif_tfm; -static DEFINE_STATIC_KEY_TRUE(crct10dif_fallback); -static DEFINE_MUTEX(crc_t10dif_mutex); -static struct work_struct crct10dif_rehash_work; - -static int crc_t10dif_notify(struct notifier_block *self, unsigned long val, void *data) -{ - struct crypto_alg *alg = data; - - if (val != CRYPTO_MSG_ALG_LOADED || - strcmp(alg->cra_name, CRC_T10DIF_STRING)) - return NOTIFY_DONE; - - schedule_work(&crct10dif_rehash_work); - return NOTIFY_OK; -} - -static void crc_t10dif_rehash(struct work_struct *work) -{ - struct crypto_shash *new, *old; - - mutex_lock(&crc_t10dif_mutex); - old = rcu_dereference_protected(crct10dif_tfm, - lockdep_is_held(&crc_t10dif_mutex)); - new = crypto_alloc_shash(CRC_T10DIF_STRING, 0, 0); - if (IS_ERR(new)) { - mutex_unlock(&crc_t10dif_mutex); - return; - } - rcu_assign_pointer(crct10dif_tfm, new); - mutex_unlock(&crc_t10dif_mutex); - - if (old) { - synchronize_rcu(); - crypto_free_shash(old); - } else { - static_branch_disable(&crct10dif_fallback); - } -} - -static struct notifier_block crc_t10dif_nb = { - .notifier_call = crc_t10dif_notify, +/* + * Table generated using the following polynomial: + * x^16 + x^15 + x^11 + x^9 + x^8 + x^7 + x^5 + x^4 + x^2 + x + 1 + * gt: 0x8bb7 + */ +static const u16 t10_dif_crc_table[256] = { + 0x0000, 0x8BB7, 0x9CD9, 0x176E, 0xB205, 0x39B2, 0x2EDC, 0xA56B, + 0xEFBD, 0x640A, 0x7364, 0xF8D3, 0x5DB8, 0xD60F, 0xC161, 0x4AD6, + 0x54CD, 0xDF7A, 0xC814, 0x43A3, 0xE6C8, 0x6D7F, 0x7A11, 0xF1A6, + 0xBB70, 0x30C7, 0x27A9, 0xAC1E, 0x0975, 0x82C2, 0x95AC, 0x1E1B, + 0xA99A, 0x222D, 0x3543, 0xBEF4, 0x1B9F, 0x9028, 0x8746, 0x0CF1, + 0x4627, 0xCD90, 0xDAFE, 0x5149, 0xF422, 0x7F95, 0x68FB, 0xE34C, + 0xFD57, 0x76E0, 0x618E, 0xEA39, 0x4F52, 0xC4E5, 0xD38B, 0x583C, + 0x12EA, 0x995D, 0x8E33, 0x0584, 0xA0EF, 0x2B58, 0x3C36, 0xB781, + 0xD883, 0x5334, 0x445A, 0xCFED, 0x6A86, 0xE131, 0xF65F, 0x7DE8, + 0x373E, 0xBC89, 0xABE7, 0x2050, 0x853B, 0x0E8C, 0x19E2, 0x9255, + 0x8C4E, 0x07F9, 0x1097, 0x9B20, 0x3E4B, 0xB5FC, 0xA292, 0x2925, + 0x63F3, 0xE844, 0xFF2A, 0x749D, 0xD1F6, 0x5A41, 0x4D2F, 0xC698, + 0x7119, 0xFAAE, 0xEDC0, 0x6677, 0xC31C, 0x48AB, 0x5FC5, 0xD472, + 0x9EA4, 0x1513, 0x027D, 0x89CA, 0x2CA1, 0xA716, 0xB078, 0x3BCF, + 0x25D4, 0xAE63, 0xB90D, 0x32BA, 0x97D1, 0x1C66, 0x0B08, 0x80BF, + 0xCA69, 0x41DE, 0x56B0, 0xDD07, 0x786C, 0xF3DB, 0xE4B5, 0x6F02, + 0x3AB1, 0xB106, 0xA668, 0x2DDF, 0x88B4, 0x0303, 0x146D, 0x9FDA, + 0xD50C, 0x5EBB, 0x49D5, 0xC262, 0x6709, 0xECBE, 0xFBD0, 0x7067, + 0x6E7C, 0xE5CB, 0xF2A5, 0x7912, 0xDC79, 0x57CE, 0x40A0, 0xCB17, + 0x81C1, 0x0A76, 0x1D18, 0x96AF, 0x33C4, 0xB873, 0xAF1D, 0x24AA, + 0x932B, 0x189C, 0x0FF2, 0x8445, 0x212E, 0xAA99, 0xBDF7, 0x3640, + 0x7C96, 0xF721, 0xE04F, 0x6BF8, 0xCE93, 0x4524, 0x524A, 0xD9FD, + 0xC7E6, 0x4C51, 0x5B3F, 0xD088, 0x75E3, 0xFE54, 0xE93A, 0x628D, + 0x285B, 0xA3EC, 0xB482, 0x3F35, 0x9A5E, 0x11E9, 0x0687, 0x8D30, + 0xE232, 0x6985, 0x7EEB, 0xF55C, 0x5037, 0xDB80, 0xCCEE, 0x4759, + 0x0D8F, 0x8638, 0x9156, 0x1AE1, 0xBF8A, 0x343D, 0x2353, 0xA8E4, + 0xB6FF, 0x3D48, 0x2A26, 0xA191, 0x04FA, 0x8F4D, 0x9823, 0x1394, + 0x5942, 0xD2F5, 0xC59B, 0x4E2C, 0xEB47, 0x60F0, 0x779E, 0xFC29, + 0x4BA8, 0xC01F, 0xD771, 0x5CC6, 0xF9AD, 0x721A, 0x6574, 0xEEC3, + 0xA415, 0x2FA2, 0x38CC, 0xB37B, 0x1610, 0x9DA7, 0x8AC9, 0x017E, + 0x1F65, 0x94D2, 0x83BC, 0x080B, 0xAD60, 0x26D7, 0x31B9, 0xBA0E, + 0xF0D8, 0x7B6F, 0x6C01, 0xE7B6, 0x42DD, 0xC96A, 0xDE04, 0x55B3 }; -__u16 crc_t10dif_update(__u16 crc, const unsigned char *buffer, size_t len) -{ - struct { - struct shash_desc shash; - __u16 crc; - } desc; - int err; - - if (static_branch_unlikely(&crct10dif_fallback)) - return crc_t10dif_generic(crc, buffer, len); - - rcu_read_lock(); - desc.shash.tfm = rcu_dereference(crct10dif_tfm); - desc.crc = crc; - err = crypto_shash_update(&desc.shash, buffer, len); - rcu_read_unlock(); - - BUG_ON(err); - - return desc.crc; -} -EXPORT_SYMBOL(crc_t10dif_update); - -__u16 crc_t10dif(const unsigned char *buffer, size_t len) -{ - return crc_t10dif_update(0, buffer, len); -} -EXPORT_SYMBOL(crc_t10dif); - -static int __init crc_t10dif_mod_init(void) -{ - INIT_WORK(&crct10dif_rehash_work, crc_t10dif_rehash); - crypto_register_notifier(&crc_t10dif_nb); - crc_t10dif_rehash(&crct10dif_rehash_work); - return 0; -} - -static void __exit crc_t10dif_mod_fini(void) -{ - crypto_unregister_notifier(&crc_t10dif_nb); - cancel_work_sync(&crct10dif_rehash_work); - crypto_free_shash(rcu_dereference_protected(crct10dif_tfm, 1)); -} - -module_init(crc_t10dif_mod_init); -module_exit(crc_t10dif_mod_fini); - -static int crc_t10dif_transform_show(char *buffer, const struct kernel_param *kp) +u16 crc_t10dif_generic(u16 crc, const u8 *p, size_t len) { - struct crypto_shash *tfm; - int len; + size_t i; - if (static_branch_unlikely(&crct10dif_fallback)) - return sprintf(buffer, "fallback\n"); + for (i = 0; i < len; i++) + crc = (crc << 8) ^ t10_dif_crc_table[(crc >> 8) ^ p[i]]; - rcu_read_lock(); - tfm = rcu_dereference(crct10dif_tfm); - len = snprintf(buffer, PAGE_SIZE, "%s\n", - crypto_shash_driver_name(tfm)); - rcu_read_unlock(); - - return len; + return crc; } +EXPORT_SYMBOL(crc_t10dif_generic); -module_param_call(transform, NULL, crc_t10dif_transform_show, NULL, 0444); - -MODULE_DESCRIPTION("T10 DIF CRC calculation (library API)"); +MODULE_DESCRIPTION("T10 DIF CRC calculation"); MODULE_LICENSE("GPL"); -MODULE_SOFTDEP("pre: crct10dif"); -- cgit v1.2.3 From 0961c3bcefa64d5f0999e2b703391862c733bb52 Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Sun, 1 Dec 2024 17:20:46 -0800 Subject: lib/crc-t10dif: add support for arch overrides Following what was done for CRC32, add support for architecture-specific override of the CRC-T10DIF library. This will allow the CRC-T10DIF library functions to access architecture-optimized code directly. Reviewed-by: Ard Biesheuvel Reviewed-by: Martin K. Petersen Link: https://lore.kernel.org/r/20241202012056.209768-3-ebiggers@kernel.org Signed-off-by: Eric Biggers --- include/linux/crc-t10dif.h | 12 ++++++++++++ lib/Kconfig | 32 ++++++++++++++++++++++++++++++++ 2 files changed, 44 insertions(+) (limited to 'include/linux') diff --git a/include/linux/crc-t10dif.h b/include/linux/crc-t10dif.h index 206ba2305483..16787c1cee21 100644 --- a/include/linux/crc-t10dif.h +++ b/include/linux/crc-t10dif.h @@ -7,10 +7,13 @@ #define CRC_T10DIF_DIGEST_SIZE 2 #define CRC_T10DIF_BLOCK_SIZE 1 +u16 crc_t10dif_arch(u16 crc, const u8 *p, size_t len); u16 crc_t10dif_generic(u16 crc, const u8 *p, size_t len); static inline u16 crc_t10dif_update(u16 crc, const u8 *p, size_t len) { + if (IS_ENABLED(CONFIG_CRC_T10DIF_ARCH)) + return crc_t10dif_arch(crc, p, len); return crc_t10dif_generic(crc, p, len); } @@ -19,4 +22,13 @@ static inline u16 crc_t10dif(const u8 *p, size_t len) return crc_t10dif_update(0, p, len); } +#if IS_ENABLED(CONFIG_CRC_T10DIF_ARCH) +bool crc_t10dif_is_optimized(void); +#else +static inline bool crc_t10dif_is_optimized(void) +{ + return false; +} +#endif + #endif diff --git a/lib/Kconfig b/lib/Kconfig index be59f7cdf448..e52a38d8d783 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -161,6 +161,38 @@ config CRC_T10DIF kernel tree needs to calculate CRC checks for use with the SCSI data integrity subsystem. +config ARCH_HAS_CRC_T10DIF + bool + +choice + prompt "CRC-T10DIF implementation" + depends on CRC_T10DIF + default CRC_T10DIF_IMPL_ARCH if ARCH_HAS_CRC_T10DIF + default CRC_T10DIF_IMPL_GENERIC if !ARCH_HAS_CRC_T10DIF + help + This option allows you to override the default choice of CRC-T10DIF + implementation. + +config CRC_T10DIF_IMPL_ARCH + bool "Architecture-optimized" if ARCH_HAS_CRC_T10DIF + help + Use the optimized implementation of CRC-T10DIF for the selected + architecture. It is recommended to keep this enabled, as it can + greatly improve CRC-T10DIF performance. + +config CRC_T10DIF_IMPL_GENERIC + bool "Generic implementation" + help + Use the generic table-based implementation of CRC-T10DIF. Selecting + this will reduce code size slightly but can greatly reduce CRC-T10DIF + performance. + +endchoice + +config CRC_T10DIF_ARCH + tristate + default CRC_T10DIF if CRC_T10DIF_IMPL_ARCH + config CRC64_ROCKSOFT tristate "CRC calculation for the Rocksoft model CRC64" select CRC64 -- cgit v1.2.3