From c7c18c94a6becc42c71a17fdc6a551aa8efb2999 Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Tue, 22 Apr 2025 08:27:14 -0700 Subject: crypto: x86 - move library functions to arch/x86/lib/crypto/ Continue disentangling the crypto library functions from the generic crypto infrastructure by moving the x86 BLAKE2s, ChaCha, and Poly1305 library functions into a new directory arch/x86/lib/crypto/ that does not depend on CRYPTO. This mirrors the distinction between crypto/ and lib/crypto/. Signed-off-by: Eric Biggers Signed-off-by: Herbert Xu --- arch/x86/lib/crypto/poly1305_glue.c | 218 ++++++++++++++++++++++++++++++++++++ 1 file changed, 218 insertions(+) create mode 100644 arch/x86/lib/crypto/poly1305_glue.c (limited to 'arch/x86/lib/crypto/poly1305_glue.c') diff --git a/arch/x86/lib/crypto/poly1305_glue.c b/arch/x86/lib/crypto/poly1305_glue.c new file mode 100644 index 000000000000..cff35ca5822a --- /dev/null +++ b/arch/x86/lib/crypto/poly1305_glue.c @@ -0,0 +1,218 @@ +// SPDX-License-Identifier: GPL-2.0 OR MIT +/* + * Copyright (C) 2015-2019 Jason A. Donenfeld . All Rights Reserved. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include + +asmlinkage void poly1305_init_x86_64(void *ctx, + const u8 key[POLY1305_BLOCK_SIZE]); +asmlinkage void poly1305_blocks_x86_64(void *ctx, const u8 *inp, + const size_t len, const u32 padbit); +asmlinkage void poly1305_emit_x86_64(void *ctx, u8 mac[POLY1305_DIGEST_SIZE], + const u32 nonce[4]); +asmlinkage void poly1305_emit_avx(void *ctx, u8 mac[POLY1305_DIGEST_SIZE], + const u32 nonce[4]); +asmlinkage void poly1305_blocks_avx(void *ctx, const u8 *inp, const size_t len, + const u32 padbit); +asmlinkage void poly1305_blocks_avx2(void *ctx, const u8 *inp, const size_t len, + const u32 padbit); +asmlinkage void poly1305_blocks_avx512(void *ctx, const u8 *inp, + const size_t len, const u32 padbit); + +static __ro_after_init DEFINE_STATIC_KEY_FALSE(poly1305_use_avx); +static __ro_after_init DEFINE_STATIC_KEY_FALSE(poly1305_use_avx2); +static __ro_after_init DEFINE_STATIC_KEY_FALSE(poly1305_use_avx512); + +struct poly1305_arch_internal { + union { + struct { + u32 h[5]; + u32 is_base2_26; + }; + u64 hs[3]; + }; + u64 r[2]; + u64 pad; + struct { u32 r2, r1, r4, r3; } rn[9]; +}; + +/* The AVX code uses base 2^26, while the scalar code uses base 2^64. If we hit + * the unfortunate situation of using AVX and then having to go back to scalar + * -- because the user is silly and has called the update function from two + * separate contexts -- then we need to convert back to the original base before + * proceeding. It is possible to reason that the initial reduction below is + * sufficient given the implementation invariants. However, for an avoidance of + * doubt and because this is not performance critical, we do the full reduction + * anyway. Z3 proof of below function: https://xn--4db.cc/ltPtHCKN/py + */ +static void convert_to_base2_64(void *ctx) +{ + struct poly1305_arch_internal *state = ctx; + u32 cy; + + if (!state->is_base2_26) + return; + + cy = state->h[0] >> 26; state->h[0] &= 0x3ffffff; state->h[1] += cy; + cy = state->h[1] >> 26; state->h[1] &= 0x3ffffff; state->h[2] += cy; + cy = state->h[2] >> 26; state->h[2] &= 0x3ffffff; state->h[3] += cy; + cy = state->h[3] >> 26; state->h[3] &= 0x3ffffff; state->h[4] += cy; + state->hs[0] = ((u64)state->h[2] << 52) | ((u64)state->h[1] << 26) | state->h[0]; + state->hs[1] = ((u64)state->h[4] << 40) | ((u64)state->h[3] << 14) | (state->h[2] >> 12); + state->hs[2] = state->h[4] >> 24; +#define ULT(a, b) ((a ^ ((a ^ b) | ((a - b) ^ b))) >> (sizeof(a) * 8 - 1)) + cy = (state->hs[2] >> 2) + (state->hs[2] & ~3ULL); + state->hs[2] &= 3; + state->hs[0] += cy; + state->hs[1] += (cy = ULT(state->hs[0], cy)); + state->hs[2] += ULT(state->hs[1], cy); +#undef ULT + state->is_base2_26 = 0; +} + +static void poly1305_simd_init(void *ctx, const u8 key[POLY1305_BLOCK_SIZE]) +{ + poly1305_init_x86_64(ctx, key); +} + +static void poly1305_simd_blocks(void *ctx, const u8 *inp, size_t len, + const u32 padbit) +{ + struct poly1305_arch_internal *state = ctx; + + /* SIMD disables preemption, so relax after processing each page. */ + BUILD_BUG_ON(SZ_4K < POLY1305_BLOCK_SIZE || + SZ_4K % POLY1305_BLOCK_SIZE); + + if (!static_branch_likely(&poly1305_use_avx) || + (len < (POLY1305_BLOCK_SIZE * 18) && !state->is_base2_26) || + !crypto_simd_usable()) { + convert_to_base2_64(ctx); + poly1305_blocks_x86_64(ctx, inp, len, padbit); + return; + } + + do { + const size_t bytes = min_t(size_t, len, SZ_4K); + + kernel_fpu_begin(); + if (static_branch_likely(&poly1305_use_avx512)) + poly1305_blocks_avx512(ctx, inp, bytes, padbit); + else if (static_branch_likely(&poly1305_use_avx2)) + poly1305_blocks_avx2(ctx, inp, bytes, padbit); + else + poly1305_blocks_avx(ctx, inp, bytes, padbit); + kernel_fpu_end(); + + len -= bytes; + inp += bytes; + } while (len); +} + +static void poly1305_simd_emit(void *ctx, u8 mac[POLY1305_DIGEST_SIZE], + const u32 nonce[4]) +{ + if (!static_branch_likely(&poly1305_use_avx)) + poly1305_emit_x86_64(ctx, mac, nonce); + else + poly1305_emit_avx(ctx, mac, nonce); +} + +void poly1305_init_arch(struct poly1305_desc_ctx *dctx, const u8 key[POLY1305_KEY_SIZE]) +{ + poly1305_simd_init(&dctx->h, key); + dctx->s[0] = get_unaligned_le32(&key[16]); + dctx->s[1] = get_unaligned_le32(&key[20]); + dctx->s[2] = get_unaligned_le32(&key[24]); + dctx->s[3] = get_unaligned_le32(&key[28]); + dctx->buflen = 0; +} +EXPORT_SYMBOL(poly1305_init_arch); + +void poly1305_update_arch(struct poly1305_desc_ctx *dctx, const u8 *src, + unsigned int srclen) +{ + unsigned int bytes; + + if (unlikely(dctx->buflen)) { + bytes = min(srclen, POLY1305_BLOCK_SIZE - dctx->buflen); + memcpy(dctx->buf + dctx->buflen, src, bytes); + src += bytes; + srclen -= bytes; + dctx->buflen += bytes; + + if (dctx->buflen == POLY1305_BLOCK_SIZE) { + poly1305_simd_blocks(&dctx->h, dctx->buf, POLY1305_BLOCK_SIZE, 1); + dctx->buflen = 0; + } + } + + if (likely(srclen >= POLY1305_BLOCK_SIZE)) { + bytes = round_down(srclen, POLY1305_BLOCK_SIZE); + poly1305_simd_blocks(&dctx->h, src, bytes, 1); + src += bytes; + srclen -= bytes; + } + + if (unlikely(srclen)) { + dctx->buflen = srclen; + memcpy(dctx->buf, src, srclen); + } +} +EXPORT_SYMBOL(poly1305_update_arch); + +void poly1305_final_arch(struct poly1305_desc_ctx *dctx, u8 *dst) +{ + if (unlikely(dctx->buflen)) { + dctx->buf[dctx->buflen++] = 1; + memset(dctx->buf + dctx->buflen, 0, + POLY1305_BLOCK_SIZE - dctx->buflen); + poly1305_simd_blocks(&dctx->h, dctx->buf, POLY1305_BLOCK_SIZE, 0); + } + + poly1305_simd_emit(&dctx->h, dst, dctx->s); + memzero_explicit(dctx, sizeof(*dctx)); +} +EXPORT_SYMBOL(poly1305_final_arch); + +bool poly1305_is_arch_optimized(void) +{ + return static_key_enabled(&poly1305_use_avx); +} +EXPORT_SYMBOL(poly1305_is_arch_optimized); + +static int __init poly1305_simd_mod_init(void) +{ + if (boot_cpu_has(X86_FEATURE_AVX) && + cpu_has_xfeatures(XFEATURE_MASK_SSE | XFEATURE_MASK_YMM, NULL)) + static_branch_enable(&poly1305_use_avx); + if (boot_cpu_has(X86_FEATURE_AVX) && boot_cpu_has(X86_FEATURE_AVX2) && + cpu_has_xfeatures(XFEATURE_MASK_SSE | XFEATURE_MASK_YMM, NULL)) + static_branch_enable(&poly1305_use_avx2); + if (boot_cpu_has(X86_FEATURE_AVX) && boot_cpu_has(X86_FEATURE_AVX2) && + boot_cpu_has(X86_FEATURE_AVX512F) && + cpu_has_xfeatures(XFEATURE_MASK_SSE | XFEATURE_MASK_YMM | XFEATURE_MASK_AVX512, NULL) && + /* Skylake downclocks unacceptably much when using zmm, but later generations are fast. */ + boot_cpu_data.x86_vfm != INTEL_SKYLAKE_X) + static_branch_enable(&poly1305_use_avx512); + return 0; +} +arch_initcall(poly1305_simd_mod_init); + +static void __exit poly1305_simd_mod_exit(void) +{ +} +module_exit(poly1305_simd_mod_exit); + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Jason A. Donenfeld "); +MODULE_DESCRIPTION("Poly1305 authenticator"); -- cgit v1.2.3 From 318c53ae02f2abf1542062543741068278780d09 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Mon, 28 Apr 2025 12:56:18 +0800 Subject: crypto: x86/poly1305 - Add block-only interface Add block-only interface. Also remove the unnecessary SIMD fallback path. Signed-off-by: Herbert Xu --- arch/x86/lib/crypto/poly1305_glue.c | 121 ++++++++++++++---------------------- 1 file changed, 46 insertions(+), 75 deletions(-) (limited to 'arch/x86/lib/crypto/poly1305_glue.c') diff --git a/arch/x86/lib/crypto/poly1305_glue.c b/arch/x86/lib/crypto/poly1305_glue.c index cff35ca5822a..d98764ec3b47 100644 --- a/arch/x86/lib/crypto/poly1305_glue.c +++ b/arch/x86/lib/crypto/poly1305_glue.c @@ -3,34 +3,15 @@ * Copyright (C) 2015-2019 Jason A. Donenfeld . All Rights Reserved. */ -#include -#include +#include +#include +#include #include #include #include #include +#include #include -#include -#include - -asmlinkage void poly1305_init_x86_64(void *ctx, - const u8 key[POLY1305_BLOCK_SIZE]); -asmlinkage void poly1305_blocks_x86_64(void *ctx, const u8 *inp, - const size_t len, const u32 padbit); -asmlinkage void poly1305_emit_x86_64(void *ctx, u8 mac[POLY1305_DIGEST_SIZE], - const u32 nonce[4]); -asmlinkage void poly1305_emit_avx(void *ctx, u8 mac[POLY1305_DIGEST_SIZE], - const u32 nonce[4]); -asmlinkage void poly1305_blocks_avx(void *ctx, const u8 *inp, const size_t len, - const u32 padbit); -asmlinkage void poly1305_blocks_avx2(void *ctx, const u8 *inp, const size_t len, - const u32 padbit); -asmlinkage void poly1305_blocks_avx512(void *ctx, const u8 *inp, - const size_t len, const u32 padbit); - -static __ro_after_init DEFINE_STATIC_KEY_FALSE(poly1305_use_avx); -static __ro_after_init DEFINE_STATIC_KEY_FALSE(poly1305_use_avx2); -static __ro_after_init DEFINE_STATIC_KEY_FALSE(poly1305_use_avx512); struct poly1305_arch_internal { union { @@ -45,64 +26,50 @@ struct poly1305_arch_internal { struct { u32 r2, r1, r4, r3; } rn[9]; }; -/* The AVX code uses base 2^26, while the scalar code uses base 2^64. If we hit - * the unfortunate situation of using AVX and then having to go back to scalar - * -- because the user is silly and has called the update function from two - * separate contexts -- then we need to convert back to the original base before - * proceeding. It is possible to reason that the initial reduction below is - * sufficient given the implementation invariants. However, for an avoidance of - * doubt and because this is not performance critical, we do the full reduction - * anyway. Z3 proof of below function: https://xn--4db.cc/ltPtHCKN/py - */ -static void convert_to_base2_64(void *ctx) -{ - struct poly1305_arch_internal *state = ctx; - u32 cy; - - if (!state->is_base2_26) - return; - - cy = state->h[0] >> 26; state->h[0] &= 0x3ffffff; state->h[1] += cy; - cy = state->h[1] >> 26; state->h[1] &= 0x3ffffff; state->h[2] += cy; - cy = state->h[2] >> 26; state->h[2] &= 0x3ffffff; state->h[3] += cy; - cy = state->h[3] >> 26; state->h[3] &= 0x3ffffff; state->h[4] += cy; - state->hs[0] = ((u64)state->h[2] << 52) | ((u64)state->h[1] << 26) | state->h[0]; - state->hs[1] = ((u64)state->h[4] << 40) | ((u64)state->h[3] << 14) | (state->h[2] >> 12); - state->hs[2] = state->h[4] >> 24; -#define ULT(a, b) ((a ^ ((a ^ b) | ((a - b) ^ b))) >> (sizeof(a) * 8 - 1)) - cy = (state->hs[2] >> 2) + (state->hs[2] & ~3ULL); - state->hs[2] &= 3; - state->hs[0] += cy; - state->hs[1] += (cy = ULT(state->hs[0], cy)); - state->hs[2] += ULT(state->hs[1], cy); -#undef ULT - state->is_base2_26 = 0; -} +asmlinkage void poly1305_block_init_arch( + struct poly1305_block_state *state, + const u8 raw_key[POLY1305_BLOCK_SIZE]); +EXPORT_SYMBOL_GPL(poly1305_block_init_arch); +asmlinkage void poly1305_blocks_x86_64(struct poly1305_arch_internal *ctx, + const u8 *inp, + const size_t len, const u32 padbit); +asmlinkage void poly1305_emit_x86_64(const struct poly1305_state *ctx, + u8 mac[POLY1305_DIGEST_SIZE], + const u32 nonce[4]); +asmlinkage void poly1305_emit_avx(const struct poly1305_state *ctx, + u8 mac[POLY1305_DIGEST_SIZE], + const u32 nonce[4]); +asmlinkage void poly1305_blocks_avx(struct poly1305_arch_internal *ctx, + const u8 *inp, const size_t len, + const u32 padbit); +asmlinkage void poly1305_blocks_avx2(struct poly1305_arch_internal *ctx, + const u8 *inp, const size_t len, + const u32 padbit); +asmlinkage void poly1305_blocks_avx512(struct poly1305_arch_internal *ctx, + const u8 *inp, + const size_t len, const u32 padbit); -static void poly1305_simd_init(void *ctx, const u8 key[POLY1305_BLOCK_SIZE]) -{ - poly1305_init_x86_64(ctx, key); -} +static __ro_after_init DEFINE_STATIC_KEY_FALSE(poly1305_use_avx); +static __ro_after_init DEFINE_STATIC_KEY_FALSE(poly1305_use_avx2); +static __ro_after_init DEFINE_STATIC_KEY_FALSE(poly1305_use_avx512); -static void poly1305_simd_blocks(void *ctx, const u8 *inp, size_t len, - const u32 padbit) +void poly1305_blocks_arch(struct poly1305_block_state *state, const u8 *inp, + unsigned int len, u32 padbit) { - struct poly1305_arch_internal *state = ctx; + struct poly1305_arch_internal *ctx = + container_of(&state->h.h, struct poly1305_arch_internal, h); /* SIMD disables preemption, so relax after processing each page. */ BUILD_BUG_ON(SZ_4K < POLY1305_BLOCK_SIZE || SZ_4K % POLY1305_BLOCK_SIZE); - if (!static_branch_likely(&poly1305_use_avx) || - (len < (POLY1305_BLOCK_SIZE * 18) && !state->is_base2_26) || - !crypto_simd_usable()) { - convert_to_base2_64(ctx); + if (!static_branch_likely(&poly1305_use_avx)) { poly1305_blocks_x86_64(ctx, inp, len, padbit); return; } do { - const size_t bytes = min_t(size_t, len, SZ_4K); + const unsigned int bytes = min(len, SZ_4K); kernel_fpu_begin(); if (static_branch_likely(&poly1305_use_avx512)) @@ -117,24 +84,26 @@ static void poly1305_simd_blocks(void *ctx, const u8 *inp, size_t len, inp += bytes; } while (len); } +EXPORT_SYMBOL_GPL(poly1305_blocks_arch); -static void poly1305_simd_emit(void *ctx, u8 mac[POLY1305_DIGEST_SIZE], - const u32 nonce[4]) +void poly1305_emit_arch(const struct poly1305_state *ctx, + u8 mac[POLY1305_DIGEST_SIZE], const u32 nonce[4]) { if (!static_branch_likely(&poly1305_use_avx)) poly1305_emit_x86_64(ctx, mac, nonce); else poly1305_emit_avx(ctx, mac, nonce); } +EXPORT_SYMBOL_GPL(poly1305_emit_arch); void poly1305_init_arch(struct poly1305_desc_ctx *dctx, const u8 key[POLY1305_KEY_SIZE]) { - poly1305_simd_init(&dctx->h, key); dctx->s[0] = get_unaligned_le32(&key[16]); dctx->s[1] = get_unaligned_le32(&key[20]); dctx->s[2] = get_unaligned_le32(&key[24]); dctx->s[3] = get_unaligned_le32(&key[28]); dctx->buflen = 0; + poly1305_block_init_arch(&dctx->state, key); } EXPORT_SYMBOL(poly1305_init_arch); @@ -151,14 +120,15 @@ void poly1305_update_arch(struct poly1305_desc_ctx *dctx, const u8 *src, dctx->buflen += bytes; if (dctx->buflen == POLY1305_BLOCK_SIZE) { - poly1305_simd_blocks(&dctx->h, dctx->buf, POLY1305_BLOCK_SIZE, 1); + poly1305_blocks_arch(&dctx->state, dctx->buf, + POLY1305_BLOCK_SIZE, 1); dctx->buflen = 0; } } if (likely(srclen >= POLY1305_BLOCK_SIZE)) { bytes = round_down(srclen, POLY1305_BLOCK_SIZE); - poly1305_simd_blocks(&dctx->h, src, bytes, 1); + poly1305_blocks_arch(&dctx->state, src, bytes, 1); src += bytes; srclen -= bytes; } @@ -176,10 +146,11 @@ void poly1305_final_arch(struct poly1305_desc_ctx *dctx, u8 *dst) dctx->buf[dctx->buflen++] = 1; memset(dctx->buf + dctx->buflen, 0, POLY1305_BLOCK_SIZE - dctx->buflen); - poly1305_simd_blocks(&dctx->h, dctx->buf, POLY1305_BLOCK_SIZE, 0); + poly1305_blocks_arch(&dctx->state, dctx->buf, + POLY1305_BLOCK_SIZE, 0); } - poly1305_simd_emit(&dctx->h, dst, dctx->s); + poly1305_emit_arch(&dctx->h, dst, dctx->s); memzero_explicit(dctx, sizeof(*dctx)); } EXPORT_SYMBOL(poly1305_final_arch); -- cgit v1.2.3 From 10a6d72ea355b730aa9702da0fd36aef0898a80e Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Mon, 28 Apr 2025 12:56:28 +0800 Subject: crypto: lib/poly1305 - Use block-only interface Now that every architecture provides a block function, use that to implement the lib/poly1305 and remove the old per-arch code. Signed-off-by: Herbert Xu --- arch/x86/lib/crypto/poly1305_glue.c | 60 ------------------------------------- 1 file changed, 60 deletions(-) (limited to 'arch/x86/lib/crypto/poly1305_glue.c') diff --git a/arch/x86/lib/crypto/poly1305_glue.c b/arch/x86/lib/crypto/poly1305_glue.c index d98764ec3b47..f799828c5809 100644 --- a/arch/x86/lib/crypto/poly1305_glue.c +++ b/arch/x86/lib/crypto/poly1305_glue.c @@ -10,7 +10,6 @@ #include #include #include -#include #include struct poly1305_arch_internal { @@ -96,65 +95,6 @@ void poly1305_emit_arch(const struct poly1305_state *ctx, } EXPORT_SYMBOL_GPL(poly1305_emit_arch); -void poly1305_init_arch(struct poly1305_desc_ctx *dctx, const u8 key[POLY1305_KEY_SIZE]) -{ - dctx->s[0] = get_unaligned_le32(&key[16]); - dctx->s[1] = get_unaligned_le32(&key[20]); - dctx->s[2] = get_unaligned_le32(&key[24]); - dctx->s[3] = get_unaligned_le32(&key[28]); - dctx->buflen = 0; - poly1305_block_init_arch(&dctx->state, key); -} -EXPORT_SYMBOL(poly1305_init_arch); - -void poly1305_update_arch(struct poly1305_desc_ctx *dctx, const u8 *src, - unsigned int srclen) -{ - unsigned int bytes; - - if (unlikely(dctx->buflen)) { - bytes = min(srclen, POLY1305_BLOCK_SIZE - dctx->buflen); - memcpy(dctx->buf + dctx->buflen, src, bytes); - src += bytes; - srclen -= bytes; - dctx->buflen += bytes; - - if (dctx->buflen == POLY1305_BLOCK_SIZE) { - poly1305_blocks_arch(&dctx->state, dctx->buf, - POLY1305_BLOCK_SIZE, 1); - dctx->buflen = 0; - } - } - - if (likely(srclen >= POLY1305_BLOCK_SIZE)) { - bytes = round_down(srclen, POLY1305_BLOCK_SIZE); - poly1305_blocks_arch(&dctx->state, src, bytes, 1); - src += bytes; - srclen -= bytes; - } - - if (unlikely(srclen)) { - dctx->buflen = srclen; - memcpy(dctx->buf, src, srclen); - } -} -EXPORT_SYMBOL(poly1305_update_arch); - -void poly1305_final_arch(struct poly1305_desc_ctx *dctx, u8 *dst) -{ - if (unlikely(dctx->buflen)) { - dctx->buf[dctx->buflen++] = 1; - memset(dctx->buf + dctx->buflen, 0, - POLY1305_BLOCK_SIZE - dctx->buflen); - poly1305_blocks_arch(&dctx->state, dctx->buf, - POLY1305_BLOCK_SIZE, 0); - } - - poly1305_emit_arch(&dctx->h, dst, dctx->s); - memzero_explicit(dctx, sizeof(*dctx)); -} -EXPORT_SYMBOL(poly1305_final_arch); - bool poly1305_is_arch_optimized(void) { return static_key_enabled(&poly1305_use_avx); -- cgit v1.2.3 From ef93f1562803cd7bb8159e3abedaf7f47dce4e35 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Wed, 30 Apr 2025 16:17:02 +0800 Subject: Revert "crypto: run initcalls for generic implementations earlier" This reverts commit c4741b23059794bd99beef0f700103b0d983b3fd. Crypto API self-tests no longer run at registration time and now occur either at late_initcall or upon the first use. Therefore the premise of the above commit no longer exists. Revert it and subsequent additions of subsys_initcall and arch_initcall. Note that lib/crypto calls will stay at subsys_initcall (or rather downgraded from arch_initcall) because they may need to occur before Crypto API registration. Signed-off-by: Herbert Xu --- arch/x86/lib/crypto/poly1305_glue.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86/lib/crypto/poly1305_glue.c') diff --git a/arch/x86/lib/crypto/poly1305_glue.c b/arch/x86/lib/crypto/poly1305_glue.c index f799828c5809..b7e78a583e07 100644 --- a/arch/x86/lib/crypto/poly1305_glue.c +++ b/arch/x86/lib/crypto/poly1305_glue.c @@ -117,7 +117,7 @@ static int __init poly1305_simd_mod_init(void) static_branch_enable(&poly1305_use_avx512); return 0; } -arch_initcall(poly1305_simd_mod_init); +subsys_initcall(poly1305_simd_mod_init); static void __exit poly1305_simd_mod_exit(void) { -- cgit v1.2.3