From 7439cfed1c41e5a1f3d095d3a4ffecdb13b279c4 Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Sun, 1 Dec 2024 17:20:51 -0800 Subject: [PATCH] powerpc/crc-t10dif: expose CRC-T10DIF function through lib Move the powerpc CRC-T10DIF assembly code into the lib directory and wire it up to the library interface. This allows it to be used without going through the crypto API. It remains usable via the crypto API too via the shash algorithms that use the library interface. Thus all the arch-specific "shash" code becomes unnecessary and is removed. Note: to see the diff from arch/powerpc/crypto/crct10dif-vpmsum_glue.c to arch/powerpc/lib/crc-t10dif-glue.c, view this commit with 'git show -M10'. Reviewed-by: Ard Biesheuvel Reviewed-by: Martin K. Petersen Link: https://lore.kernel.org/r/20241202012056.209768-8-ebiggers@kernel.org Signed-off-by: Eric Biggers --- arch/powerpc/Kconfig | 1 + arch/powerpc/configs/powernv_defconfig | 1 - arch/powerpc/configs/ppc64_defconfig | 1 - arch/powerpc/crypto/Kconfig | 15 +--- arch/powerpc/crypto/Makefile | 2 - arch/powerpc/lib/Makefile | 3 + .../crc-t10dif-glue.c} | 69 +++++-------------- .../{crypto => lib}/crct10dif-vpmsum_asm.S | 2 +- 8 files changed, 23 insertions(+), 71 deletions(-) rename arch/powerpc/{crypto/crct10dif-vpmsum_glue.c => lib/crc-t10dif-glue.c} (50%) rename arch/powerpc/{crypto => lib}/crct10dif-vpmsum_asm.S (99%) diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig index b5050a09a27f8..da0ac6697ac57 100644 --- a/arch/powerpc/Kconfig +++ b/arch/powerpc/Kconfig @@ -128,6 +128,7 @@ config PPC select ARCH_ENABLE_MEMORY_HOTREMOVE select ARCH_HAS_COPY_MC if PPC64 select ARCH_HAS_CRC32 if PPC64 && ALTIVEC + select ARCH_HAS_CRC_T10DIF if PPC64 && ALTIVEC select ARCH_HAS_CURRENT_STACK_POINTER select ARCH_HAS_DEBUG_VIRTUAL select ARCH_HAS_DEBUG_VM_PGTABLE diff --git a/arch/powerpc/configs/powernv_defconfig b/arch/powerpc/configs/powernv_defconfig index 4a7ddea05b4db..6b6d7467fecfd 100644 --- a/arch/powerpc/configs/powernv_defconfig +++ b/arch/powerpc/configs/powernv_defconfig @@ -320,7 +320,6 @@ CONFIG_XMON=y CONFIG_CRYPTO_TEST=m CONFIG_CRYPTO_PCBC=m CONFIG_CRYPTO_HMAC=y -CONFIG_CRYPTO_CRCT10DIF_VPMSUM=m CONFIG_CRYPTO_MD5_PPC=m CONFIG_CRYPTO_MICHAEL_MIC=m CONFIG_CRYPTO_SHA1_PPC=m diff --git a/arch/powerpc/configs/ppc64_defconfig b/arch/powerpc/configs/ppc64_defconfig index 58e5f4488da46..15101f5c32381 100644 --- a/arch/powerpc/configs/ppc64_defconfig +++ b/arch/powerpc/configs/ppc64_defconfig @@ -389,7 +389,6 @@ CONFIG_CRYPTO_MICHAEL_MIC=m CONFIG_CRYPTO_SHA256=y CONFIG_CRYPTO_WP512=m CONFIG_CRYPTO_LZO=m -CONFIG_CRYPTO_CRCT10DIF_VPMSUM=m CONFIG_CRYPTO_VPMSUM_TESTER=m CONFIG_CRYPTO_MD5_PPC=m CONFIG_CRYPTO_SHA1_PPC=m diff --git a/arch/powerpc/crypto/Kconfig b/arch/powerpc/crypto/Kconfig index 2d89e35b3cddd..45376f1a43bdb 100644 --- a/arch/powerpc/crypto/Kconfig +++ b/arch/powerpc/crypto/Kconfig @@ -13,21 +13,10 @@ config CRYPTO_CURVE25519_PPC64 Architecture: PowerPC64 - Little-endian -config CRYPTO_CRCT10DIF_VPMSUM - tristate "CRC32T10DIF" - depends on PPC64 && ALTIVEC && CRC_T10DIF - select CRYPTO_HASH - help - CRC16 CRC algorithm used for the T10 (SCSI) Data Integrity Field (DIF) - - Architecture: powerpc64 using - - AltiVec extensions - - Enable on POWER8 and newer processors for improved performance. - config CRYPTO_VPMSUM_TESTER tristate "CRC32c and CRC32T10DIF hardware acceleration tester" - depends on CRYPTO_CRCT10DIF_VPMSUM && CRYPTO_CRC32C && CRC32_ARCH + depends on CRYPTO_CRC32C && CRC32_ARCH + depends on CRYPTO_CRCT10DIF && CRC_T10DIF_ARCH help Stress test for CRC32c and CRCT10DIF algorithms implemented with powerpc64 AltiVec extensions (POWER8 vpmsum instructions). diff --git a/arch/powerpc/crypto/Makefile b/arch/powerpc/crypto/Makefile index 54486192273c2..d2238ac7e52ce 100644 --- a/arch/powerpc/crypto/Makefile +++ b/arch/powerpc/crypto/Makefile @@ -10,7 +10,6 @@ obj-$(CONFIG_CRYPTO_MD5_PPC) += md5-ppc.o obj-$(CONFIG_CRYPTO_SHA1_PPC) += sha1-powerpc.o obj-$(CONFIG_CRYPTO_SHA1_PPC_SPE) += sha1-ppc-spe.o obj-$(CONFIG_CRYPTO_SHA256_PPC_SPE) += sha256-ppc-spe.o -obj-$(CONFIG_CRYPTO_CRCT10DIF_VPMSUM) += crct10dif-vpmsum.o obj-$(CONFIG_CRYPTO_VPMSUM_TESTER) += crc-vpmsum_test.o obj-$(CONFIG_CRYPTO_AES_GCM_P10) += aes-gcm-p10-crypto.o obj-$(CONFIG_CRYPTO_CHACHA20_P10) += chacha-p10-crypto.o @@ -23,7 +22,6 @@ md5-ppc-y := md5-asm.o md5-glue.o sha1-powerpc-y := sha1-powerpc-asm.o sha1.o sha1-ppc-spe-y := sha1-spe-asm.o sha1-spe-glue.o sha256-ppc-spe-y := sha256-spe-asm.o sha256-spe-glue.o -crct10dif-vpmsum-y := crct10dif-vpmsum_asm.o crct10dif-vpmsum_glue.o aes-gcm-p10-crypto-y := aes-gcm-p10-glue.o aes-gcm-p10.o ghashp10-ppc.o aesp10-ppc.o chacha-p10-crypto-y := chacha-p10-glue.o chacha-p10le-8x.o poly1305-p10-crypto-y := poly1305-p10-glue.o poly1305-p10le_64.o diff --git a/arch/powerpc/lib/Makefile b/arch/powerpc/lib/Makefile index da9381a1c95b3..dd8a4b52a0ccb 100644 --- a/arch/powerpc/lib/Makefile +++ b/arch/powerpc/lib/Makefile @@ -81,4 +81,7 @@ CFLAGS_xor_vmx.o += -isystem $(shell $(CC) -print-file-name=include) obj-$(CONFIG_CRC32_ARCH) += crc32-powerpc.o crc32-powerpc-y := crc32-glue.o crc32c-vpmsum_asm.o +obj-$(CONFIG_CRC_T10DIF_ARCH) += crc-t10dif-powerpc.o +crc-t10dif-powerpc-y := crc-t10dif-glue.o crct10dif-vpmsum_asm.o + obj-$(CONFIG_PPC64) += $(obj64-y) diff --git a/arch/powerpc/crypto/crct10dif-vpmsum_glue.c b/arch/powerpc/lib/crc-t10dif-glue.c similarity index 50% rename from arch/powerpc/crypto/crct10dif-vpmsum_glue.c rename to arch/powerpc/lib/crc-t10dif-glue.c index 1dc8b69151781..730850dbc51d7 100644 --- a/arch/powerpc/crypto/crct10dif-vpmsum_glue.c +++ b/arch/powerpc/lib/crc-t10dif-glue.c @@ -7,7 +7,6 @@ */ #include -#include #include #include #include @@ -22,15 +21,18 @@ #define VECTOR_BREAKPOINT 64 +static DEFINE_STATIC_KEY_FALSE(have_vec_crypto); + u32 __crct10dif_vpmsum(u32 crc, unsigned char const *p, size_t len); -static u16 crct10dif_vpmsum(u16 crci, unsigned char const *p, size_t len) +u16 crc_t10dif_arch(u16 crci, const u8 *p, size_t len) { unsigned int prealign; unsigned int tail; u32 crc = crci; - if (len < (VECTOR_BREAKPOINT + VMX_ALIGN) || !crypto_simd_usable()) + if (len < (VECTOR_BREAKPOINT + VMX_ALIGN) || + !static_branch_likely(&have_vec_crypto) || !crypto_simd_usable()) return crc_t10dif_generic(crc, p, len); if ((unsigned long)p & VMX_ALIGN_MASK) { @@ -60,67 +62,28 @@ static u16 crct10dif_vpmsum(u16 crci, unsigned char const *p, size_t len) return crc & 0xffff; } +EXPORT_SYMBOL(crc_t10dif_arch); -static int crct10dif_vpmsum_init(struct shash_desc *desc) -{ - u16 *crc = shash_desc_ctx(desc); - - *crc = 0; - return 0; -} - -static int crct10dif_vpmsum_update(struct shash_desc *desc, const u8 *data, - unsigned int length) -{ - u16 *crc = shash_desc_ctx(desc); - - *crc = crct10dif_vpmsum(*crc, data, length); - - return 0; -} - - -static int crct10dif_vpmsum_final(struct shash_desc *desc, u8 *out) +static int __init crc_t10dif_powerpc_init(void) { - u16 *crcp = shash_desc_ctx(desc); - - *(u16 *)out = *crcp; + if (cpu_has_feature(CPU_FTR_ARCH_207S) && + (cur_cpu_spec->cpu_user_features2 & PPC_FEATURE2_VEC_CRYPTO)) + static_branch_enable(&have_vec_crypto); return 0; } +arch_initcall(crc_t10dif_powerpc_init); -static struct shash_alg alg = { - .init = crct10dif_vpmsum_init, - .update = crct10dif_vpmsum_update, - .final = crct10dif_vpmsum_final, - .descsize = CRC_T10DIF_DIGEST_SIZE, - .digestsize = CRC_T10DIF_DIGEST_SIZE, - .base = { - .cra_name = "crct10dif", - .cra_driver_name = "crct10dif-vpmsum", - .cra_priority = 200, - .cra_blocksize = CRC_T10DIF_BLOCK_SIZE, - .cra_module = THIS_MODULE, - } -}; - -static int __init crct10dif_vpmsum_mod_init(void) +static void __exit crc_t10dif_powerpc_exit(void) { - if (!cpu_has_feature(CPU_FTR_ARCH_207S)) - return -ENODEV; - - return crypto_register_shash(&alg); } +module_exit(crc_t10dif_powerpc_exit); -static void __exit crct10dif_vpmsum_mod_fini(void) +bool crc_t10dif_is_optimized(void) { - crypto_unregister_shash(&alg); + return static_key_enabled(&have_vec_crypto); } - -module_cpu_feature_match(PPC_MODULE_FEATURE_VEC_CRYPTO, crct10dif_vpmsum_mod_init); -module_exit(crct10dif_vpmsum_mod_fini); +EXPORT_SYMBOL(crc_t10dif_is_optimized); MODULE_AUTHOR("Daniel Axtens "); MODULE_DESCRIPTION("CRCT10DIF using vector polynomial multiply-sum instructions"); MODULE_LICENSE("GPL"); -MODULE_ALIAS_CRYPTO("crct10dif"); -MODULE_ALIAS_CRYPTO("crct10dif-vpmsum"); diff --git a/arch/powerpc/crypto/crct10dif-vpmsum_asm.S b/arch/powerpc/lib/crct10dif-vpmsum_asm.S similarity index 99% rename from arch/powerpc/crypto/crct10dif-vpmsum_asm.S rename to arch/powerpc/lib/crct10dif-vpmsum_asm.S index 0a52261bf8599..f0b93a0fe168a 100644 --- a/arch/powerpc/crypto/crct10dif-vpmsum_asm.S +++ b/arch/powerpc/lib/crct10dif-vpmsum_asm.S @@ -842,4 +842,4 @@ .octa 0x0000000000000000000000018bb70000 #define CRC_FUNCTION_NAME __crct10dif_vpmsum -#include "../lib/crc32-vpmsum_core.S" +#include "crc32-vpmsum_core.S" -- 2.39.5