crypto: arm64/chacha - expose arm64 ChaCha routine as library function

Expose the accelerated NEON ChaCha routine directly as a symbol
export so that users of the ChaCha library API can use it directly.

Given that calls into the library API will always go through the
routines in this module if it is enabled, switch to static keys
to select the optimal implementation available (which may be none
at all, in which case we defer to the generic implementation for
all invocations).
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

arch/arm64/crypto/Kconfig

@@ -104,6 +104,7 @@ config CRYPTO_CHACHA20_NEON
 	depends on KERNEL_MODE_NEON
 	select CRYPTO_SKCIPHER
 	select CRYPTO_LIB_CHACHA_GENERIC
+	select CRYPTO_ARCH_HAVE_LIB_CHACHA
 
 config CRYPTO_NHPOLY1305_NEON
 	tristate "NHPoly1305 hash function using NEON instructions (for Adiantum)"

arch/arm64/crypto/chacha-neon-glue.c

@@ -23,6 +23,7 @@
 #include <crypto/internal/chacha.h>
 #include <crypto/internal/simd.h>
 #include <crypto/internal/skcipher.h>
+#include <linux/jump_label.h>
 #include <linux/kernel.h>
 #include <linux/module.h>
 
@@ -36,6 +37,8 @@ asmlinkage void chacha_4block_xor_neon(u32 *state, u8 *dst, const u8 *src,
 				       int nrounds, int bytes);
 asmlinkage void hchacha_block_neon(const u32 *state, u32 *out, int nrounds);
 
+static __ro_after_init DEFINE_STATIC_KEY_FALSE(have_neon);
+
 static void chacha_doneon(u32 *state, u8 *dst, const u8 *src,
 			  int bytes, int nrounds)
 {
@@ -59,6 +62,37 @@ static void chacha_doneon(u32 *state, u8 *dst, const u8 *src,
 	}
 }
 
+void hchacha_block_arch(const u32 *state, u32 *stream, int nrounds)
+{
+	if (!static_branch_likely(&have_neon) || !crypto_simd_usable()) {
+		hchacha_block_generic(state, stream, nrounds);
+	} else {
+		kernel_neon_begin();
+		hchacha_block_neon(state, stream, nrounds);
+		kernel_neon_end();
+	}
+}
+EXPORT_SYMBOL(hchacha_block_arch);
+
+void chacha_init_arch(u32 *state, const u32 *key, const u8 *iv)
+{
+	chacha_init_generic(state, key, iv);
+}
+EXPORT_SYMBOL(chacha_init_arch);
+
+void chacha_crypt_arch(u32 *state, u8 *dst, const u8 *src, unsigned int bytes,
+		       int nrounds)
+{
+	if (!static_branch_likely(&have_neon) || bytes <= CHACHA_BLOCK_SIZE ||
+	    !crypto_simd_usable())
+		return chacha_crypt_generic(state, dst, src, bytes, nrounds);
+
+	kernel_neon_begin();
+	chacha_doneon(state, dst, src, bytes, nrounds);
+	kernel_neon_end();
+}
+EXPORT_SYMBOL(chacha_crypt_arch);
+
 static int chacha_neon_stream_xor(struct skcipher_request *req,
 				  const struct chacha_ctx *ctx, const u8 *iv)
 {
@@ -76,7 +110,8 @@ static int chacha_neon_stream_xor(struct skcipher_request *req,
 		if (nbytes < walk.total)
 			nbytes = rounddown(nbytes, walk.stride);
 
-		if (!crypto_simd_usable()) {
+		if (!static_branch_likely(&have_neon) ||
+		    !crypto_simd_usable()) {
 			chacha_crypt_generic(state, walk.dst.virt.addr,
 					     walk.src.virt.addr, nbytes,
 					     ctx->nrounds);
@@ -109,14 +144,7 @@ static int xchacha_neon(struct skcipher_request *req)
 	u8 real_iv[16];
 
 	chacha_init_generic(state, ctx->key, req->iv);
-
-	if (crypto_simd_usable()) {
-		kernel_neon_begin();
-		hchacha_block_neon(state, subctx.key, ctx->nrounds);
-		kernel_neon_end();
-	} else {
-		hchacha_block_generic(state, subctx.key, ctx->nrounds);
-	}
+	hchacha_block_arch(state, subctx.key, ctx->nrounds);
 	subctx.nrounds = ctx->nrounds;
 
 	memcpy(&real_iv[0], req->iv + 24, 8);
@@ -179,13 +207,16 @@ static struct skcipher_alg algs[] = {
 static int __init chacha_simd_mod_init(void)
 {
 	if (!cpu_have_named_feature(ASIMD))
-		return -ENODEV;
+		return 0;
+
+	static_branch_enable(&have_neon);
 
 	return crypto_register_skciphers(algs, ARRAY_SIZE(algs));
 }
 
 static void __exit chacha_simd_mod_fini(void)
 {
+	if (cpu_have_named_feature(ASIMD))
 		crypto_unregister_skciphers(algs, ARRAY_SIZE(algs));
 }