summaryrefslogtreecommitdiff
path: root/arch/x86/crypto/poly1305_glue.c
diff options
context:
space:
mode:
Diffstat (limited to 'arch/x86/crypto/poly1305_glue.c')
-rw-r--r--arch/x86/crypto/poly1305_glue.c199
1 files changed, 119 insertions, 80 deletions
diff --git a/arch/x86/crypto/poly1305_glue.c b/arch/x86/crypto/poly1305_glue.c
index 4a1c05dce950..370cd88068ec 100644
--- a/arch/x86/crypto/poly1305_glue.c
+++ b/arch/x86/crypto/poly1305_glue.c
@@ -7,47 +7,23 @@
#include <crypto/algapi.h>
#include <crypto/internal/hash.h>
+#include <crypto/internal/poly1305.h>
#include <crypto/internal/simd.h>
-#include <crypto/poly1305.h>
#include <linux/crypto.h>
+#include <linux/jump_label.h>
#include <linux/kernel.h>
#include <linux/module.h>
#include <asm/simd.h>
-struct poly1305_simd_desc_ctx {
- struct poly1305_desc_ctx base;
- /* derived key u set? */
- bool uset;
-#ifdef CONFIG_AS_AVX2
- /* derived keys r^3, r^4 set? */
- bool wset;
-#endif
- /* derived Poly1305 key r^2 */
- u32 u[5];
- /* ... silently appended r^3 and r^4 when using AVX2 */
-};
-
asmlinkage void poly1305_block_sse2(u32 *h, const u8 *src,
const u32 *r, unsigned int blocks);
asmlinkage void poly1305_2block_sse2(u32 *h, const u8 *src, const u32 *r,
unsigned int blocks, const u32 *u);
-#ifdef CONFIG_AS_AVX2
asmlinkage void poly1305_4block_avx2(u32 *h, const u8 *src, const u32 *r,
unsigned int blocks, const u32 *u);
-static bool poly1305_use_avx2;
-#endif
-
-static int poly1305_simd_init(struct shash_desc *desc)
-{
- struct poly1305_simd_desc_ctx *sctx = shash_desc_ctx(desc);
- sctx->uset = false;
-#ifdef CONFIG_AS_AVX2
- sctx->wset = false;
-#endif
-
- return crypto_poly1305_init(desc);
-}
+static __ro_after_init DEFINE_STATIC_KEY_FALSE(poly1305_use_simd);
+static __ro_after_init DEFINE_STATIC_KEY_FALSE(poly1305_use_avx2);
static void poly1305_simd_mult(u32 *a, const u32 *b)
{
@@ -60,72 +36,85 @@ static void poly1305_simd_mult(u32 *a, const u32 *b)
poly1305_block_sse2(a, m, b, 1);
}
+static unsigned int poly1305_scalar_blocks(struct poly1305_desc_ctx *dctx,
+ const u8 *src, unsigned int srclen)
+{
+ unsigned int datalen;
+
+ if (unlikely(!dctx->sset)) {
+ datalen = crypto_poly1305_setdesckey(dctx, src, srclen);
+ src += srclen - datalen;
+ srclen = datalen;
+ }
+ if (srclen >= POLY1305_BLOCK_SIZE) {
+ poly1305_core_blocks(&dctx->h, dctx->r, src,
+ srclen / POLY1305_BLOCK_SIZE, 1);
+ srclen %= POLY1305_BLOCK_SIZE;
+ }
+ return srclen;
+}
+
static unsigned int poly1305_simd_blocks(struct poly1305_desc_ctx *dctx,
const u8 *src, unsigned int srclen)
{
- struct poly1305_simd_desc_ctx *sctx;
unsigned int blocks, datalen;
- BUILD_BUG_ON(offsetof(struct poly1305_simd_desc_ctx, base));
- sctx = container_of(dctx, struct poly1305_simd_desc_ctx, base);
-
if (unlikely(!dctx->sset)) {
datalen = crypto_poly1305_setdesckey(dctx, src, srclen);
src += srclen - datalen;
srclen = datalen;
}
-#ifdef CONFIG_AS_AVX2
- if (poly1305_use_avx2 && srclen >= POLY1305_BLOCK_SIZE * 4) {
- if (unlikely(!sctx->wset)) {
- if (!sctx->uset) {
- memcpy(sctx->u, dctx->r.r, sizeof(sctx->u));
- poly1305_simd_mult(sctx->u, dctx->r.r);
- sctx->uset = true;
+ if (IS_ENABLED(CONFIG_AS_AVX2) &&
+ static_branch_likely(&poly1305_use_avx2) &&
+ srclen >= POLY1305_BLOCK_SIZE * 4) {
+ if (unlikely(dctx->rset < 4)) {
+ if (dctx->rset < 2) {
+ dctx->r[1] = dctx->r[0];
+ poly1305_simd_mult(dctx->r[1].r, dctx->r[0].r);
}
- memcpy(sctx->u + 5, sctx->u, sizeof(sctx->u));
- poly1305_simd_mult(sctx->u + 5, dctx->r.r);
- memcpy(sctx->u + 10, sctx->u + 5, sizeof(sctx->u));
- poly1305_simd_mult(sctx->u + 10, dctx->r.r);
- sctx->wset = true;
+ dctx->r[2] = dctx->r[1];
+ poly1305_simd_mult(dctx->r[2].r, dctx->r[0].r);
+ dctx->r[3] = dctx->r[2];
+ poly1305_simd_mult(dctx->r[3].r, dctx->r[0].r);
+ dctx->rset = 4;
}
blocks = srclen / (POLY1305_BLOCK_SIZE * 4);
- poly1305_4block_avx2(dctx->h.h, src, dctx->r.r, blocks,
- sctx->u);
+ poly1305_4block_avx2(dctx->h.h, src, dctx->r[0].r, blocks,
+ dctx->r[1].r);
src += POLY1305_BLOCK_SIZE * 4 * blocks;
srclen -= POLY1305_BLOCK_SIZE * 4 * blocks;
}
-#endif
+
if (likely(srclen >= POLY1305_BLOCK_SIZE * 2)) {
- if (unlikely(!sctx->uset)) {
- memcpy(sctx->u, dctx->r.r, sizeof(sctx->u));
- poly1305_simd_mult(sctx->u, dctx->r.r);
- sctx->uset = true;
+ if (unlikely(dctx->rset < 2)) {
+ dctx->r[1] = dctx->r[0];
+ poly1305_simd_mult(dctx->r[1].r, dctx->r[0].r);
+ dctx->rset = 2;
}
blocks = srclen / (POLY1305_BLOCK_SIZE * 2);
- poly1305_2block_sse2(dctx->h.h, src, dctx->r.r, blocks,
- sctx->u);
+ poly1305_2block_sse2(dctx->h.h, src, dctx->r[0].r,
+ blocks, dctx->r[1].r);
src += POLY1305_BLOCK_SIZE * 2 * blocks;
srclen -= POLY1305_BLOCK_SIZE * 2 * blocks;
}
if (srclen >= POLY1305_BLOCK_SIZE) {
- poly1305_block_sse2(dctx->h.h, src, dctx->r.r, 1);
+ poly1305_block_sse2(dctx->h.h, src, dctx->r[0].r, 1);
srclen -= POLY1305_BLOCK_SIZE;
}
return srclen;
}
-static int poly1305_simd_update(struct shash_desc *desc,
- const u8 *src, unsigned int srclen)
+void poly1305_init_arch(struct poly1305_desc_ctx *desc, const u8 *key)
{
- struct poly1305_desc_ctx *dctx = shash_desc_ctx(desc);
- unsigned int bytes;
-
- /* kernel_fpu_begin/end is costly, use fallback for small updates */
- if (srclen <= 288 || !crypto_simd_usable())
- return crypto_poly1305_update(desc, src, srclen);
+ poly1305_init_generic(desc, key);
+}
+EXPORT_SYMBOL(poly1305_init_arch);
- kernel_fpu_begin();
+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);
@@ -135,34 +124,84 @@ static int poly1305_simd_update(struct shash_desc *desc,
dctx->buflen += bytes;
if (dctx->buflen == POLY1305_BLOCK_SIZE) {
- poly1305_simd_blocks(dctx, dctx->buf,
- POLY1305_BLOCK_SIZE);
+ if (static_branch_likely(&poly1305_use_simd) &&
+ likely(crypto_simd_usable())) {
+ kernel_fpu_begin();
+ poly1305_simd_blocks(dctx, dctx->buf,
+ POLY1305_BLOCK_SIZE);
+ kernel_fpu_end();
+ } else {
+ poly1305_scalar_blocks(dctx, dctx->buf,
+ POLY1305_BLOCK_SIZE);
+ }
dctx->buflen = 0;
}
}
if (likely(srclen >= POLY1305_BLOCK_SIZE)) {
- bytes = poly1305_simd_blocks(dctx, src, srclen);
+ if (static_branch_likely(&poly1305_use_simd) &&
+ likely(crypto_simd_usable())) {
+ kernel_fpu_begin();
+ bytes = poly1305_simd_blocks(dctx, src, srclen);
+ kernel_fpu_end();
+ } else {
+ bytes = poly1305_scalar_blocks(dctx, src, srclen);
+ }
src += srclen - bytes;
srclen = bytes;
}
- kernel_fpu_end();
-
if (unlikely(srclen)) {
dctx->buflen = srclen;
memcpy(dctx->buf, src, srclen);
}
+}
+EXPORT_SYMBOL(poly1305_update_arch);
+
+void poly1305_final_arch(struct poly1305_desc_ctx *desc, u8 *digest)
+{
+ poly1305_final_generic(desc, digest);
+}
+EXPORT_SYMBOL(poly1305_final_arch);
+
+static int crypto_poly1305_init(struct shash_desc *desc)
+{
+ struct poly1305_desc_ctx *dctx = shash_desc_ctx(desc);
+
+ poly1305_core_init(&dctx->h);
+ dctx->buflen = 0;
+ dctx->rset = 0;
+ dctx->sset = false;
+
+ return 0;
+}
+
+static int crypto_poly1305_final(struct shash_desc *desc, u8 *dst)
+{
+ struct poly1305_desc_ctx *dctx = shash_desc_ctx(desc);
+
+ if (unlikely(!dctx->sset))
+ return -ENOKEY;
+
+ poly1305_final_generic(dctx, dst);
+ return 0;
+}
+
+static int poly1305_simd_update(struct shash_desc *desc,
+ const u8 *src, unsigned int srclen)
+{
+ struct poly1305_desc_ctx *dctx = shash_desc_ctx(desc);
+ poly1305_update_arch(dctx, src, srclen);
return 0;
}
static struct shash_alg alg = {
.digestsize = POLY1305_DIGEST_SIZE,
- .init = poly1305_simd_init,
+ .init = crypto_poly1305_init,
.update = poly1305_simd_update,
.final = crypto_poly1305_final,
- .descsize = sizeof(struct poly1305_simd_desc_ctx),
+ .descsize = sizeof(struct poly1305_desc_ctx),
.base = {
.cra_name = "poly1305",
.cra_driver_name = "poly1305-simd",
@@ -175,16 +214,16 @@ static struct shash_alg alg = {
static int __init poly1305_simd_mod_init(void)
{
if (!boot_cpu_has(X86_FEATURE_XMM2))
- return -ENODEV;
-
-#ifdef CONFIG_AS_AVX2
- poly1305_use_avx2 = boot_cpu_has(X86_FEATURE_AVX) &&
- boot_cpu_has(X86_FEATURE_AVX2) &&
- cpu_has_xfeatures(XFEATURE_MASK_SSE | XFEATURE_MASK_YMM, NULL);
- alg.descsize = sizeof(struct poly1305_simd_desc_ctx);
- if (poly1305_use_avx2)
- alg.descsize += 10 * sizeof(u32);
-#endif
+ return 0;
+
+ static_branch_enable(&poly1305_use_simd);
+
+ if (IS_ENABLED(CONFIG_AS_AVX2) &&
+ 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);
+
return crypto_register_shash(&alg);
}