crypto: algapi - make crypto_xor() and crypto_inc() alignment agnostic

Instead of unconditionally forcing 4 byte alignment for all generic
chaining modes that rely on crypto_xor() or crypto_inc() (which may
result in unnecessary copying of data when the underlying hardware
can perform unaligned accesses efficiently), make those functions
deal with unaligned input explicitly, but only if the Kconfig symbol
HAVE_EFFICIENT_UNALIGNED_ACCESS is set. This will allow us to drop
the alignmasks from the CBC, CMAC, CTR, CTS, PCBC and SEQIV drivers.

For crypto_inc(), this simply involves making the 4-byte stride
conditional on HAVE_EFFICIENT_UNALIGNED_ACCESS being set, given that
it typically operates on 16 byte buffers.

For crypto_xor(), an algorithm is implemented that simply runs through
the input using the largest strides possible if unaligned accesses are
allowed. If they are not, an optimal sequence of memory accesses is
emitted that takes the relative alignment of the input buffers into
account, e.g., if the relative misalignment of dst and src is 4 bytes,
the entire xor operation will be completed using 4 byte loads and stores
(modulo unaligned bits at the start and end). Note that all expressions
involving misalign are simply eliminated by the compiler when
HAVE_EFFICIENT_UNALIGNED_ACCESS is defined.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

1 files changed, 50 insertions, 18 deletions

diff --git a/crypto/algapi.c b/crypto/algapi.c
index 1fad2a6b3bbb..6b52e8f0b95f 100644
--- a/crypto/algapi.c
+++ b/crypto/algapi.c
@@ -962,34 +962,66 @@ void crypto_inc(u8 *a, unsigned int size)
 	__be32 *b = (__be32 *)(a + size);
 	u32 c;
 
-	for (; size >= 4; size -= 4) {
-		c = be32_to_cpu(*--b) + 1;
-		*b = cpu_to_be32(c);
-		if (c)
-			return;
-	}
+	if (IS_ENABLED(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) ||
+	    !((unsigned long)b & (__alignof__(*b) - 1)))
+		for (; size >= 4; size -= 4) {
+			c = be32_to_cpu(*--b) + 1;
+			*b = cpu_to_be32(c);
+			if (c)
+				return;
+		}
 
 	crypto_inc_byte(a, size);
 }
 EXPORT_SYMBOL_GPL(crypto_inc);
 
-static inline void crypto_xor_byte(u8 *a, const u8 *b, unsigned int size)
+void __crypto_xor(u8 *dst, const u8 *src, unsigned int len)
 {
-	for (; size; size--)
-		*a++ ^= *b++;
-}
+	int relalign = 0;
+
+	if (!IS_ENABLED(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS)) {
+		int size = sizeof(unsigned long);
+		int d = ((unsigned long)dst ^ (unsigned long)src) & (size - 1);
+
+		relalign = d ? 1 << __ffs(d) : size;
+
+		/*
+		 * If we care about alignment, process as many bytes as
+		 * needed to advance dst and src to values whose alignments
+		 * equal their relative alignment. This will allow us to
+		 * process the remainder of the input using optimal strides.
+		 */
+		while (((unsigned long)dst & (relalign - 1)) && len > 0) {
+			*dst++ ^= *src++;
+			len--;
+		}
+	}
 
-void crypto_xor(u8 *dst, const u8 *src, unsigned int size)
-{
-	u32 *a = (u32 *)dst;
-	u32 *b = (u32 *)src;
+	while (IS_ENABLED(CONFIG_64BIT) && len >= 8 && !(relalign & 7)) {
+		*(u64 *)dst ^= *(u64 *)src;
+		dst += 8;
+		src += 8;
+		len -= 8;
+	}
 
-	for (; size >= 4; size -= 4)
-		*a++ ^= *b++;
+	while (len >= 4 && !(relalign & 3)) {
+		*(u32 *)dst ^= *(u32 *)src;
+		dst += 4;
+		src += 4;
+		len -= 4;
+	}
+
+	while (len >= 2 && !(relalign & 1)) {
+		*(u16 *)dst ^= *(u16 *)src;
+		dst += 2;
+		src += 2;
+		len -= 2;
+	}
 
-	crypto_xor_byte((u8 *)a, (u8 *)b, size);
+	while (len--)
+		*dst++ ^= *src++;
 }
-EXPORT_SYMBOL_GPL(crypto_xor);
+EXPORT_SYMBOL_GPL(__crypto_xor);
 
 unsigned int crypto_alg_extsize(struct crypto_alg *alg)
 {