crypto: arm/crct10dif - revert to C code for short inputs

The SIMD routine ported from x86 used to have a special code path
for inputs < 16 bytes, which got lost somewhere along the way.
Instead, the current glue code aligns the input pointer to permit
the NEON routine to use special versions of the vld1 instructions
that assume 16 byte alignment, but this could result in inputs of
less than 16 bytes to be passed in. This not only fails the new
extended tests that Eric has implemented, it also results in the
code reading past the end of the input, which could potentially
result in crashes when dealing with less than 16 bytes of input
at the end of a page which is followed by an unmapped page.

So update the glue code to only invoke the NEON routine if the
input is at least 16 bytes.
Reported-by: Eric Biggers <ebiggers@kernel.org>
Reviewed-by: Eric Biggers <ebiggers@kernel.org>
Fixes: 1d481f1c ("crypto: arm/crct10dif - port x86 SSE implementation to ARM")
Cc: <stable@vger.kernel.org> # v4.10+
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

arch/arm/crypto/crct10dif-ce-core.S

@@ -124,10 +124,10 @@ ENTRY(crc_t10dif_pmull)
 	vext.8		q10, qzr, q0, #4
 
 	// receive the initial 64B data, xor the initial crc value
-	vld1.64		{q0-q1}, [arg2, :128]!
-	vld1.64		{q2-q3}, [arg2, :128]!
-	vld1.64		{q4-q5}, [arg2, :128]!
-	vld1.64		{q6-q7}, [arg2, :128]!
+	vld1.64		{q0-q1}, [arg2]!
+	vld1.64		{q2-q3}, [arg2]!
+	vld1.64		{q4-q5}, [arg2]!
+	vld1.64		{q6-q7}, [arg2]!
 CPU_LE(	vrev64.8	q0, q0			)
 CPU_LE(	vrev64.8	q1, q1			)
 CPU_LE(	vrev64.8	q2, q2			)
@@ -167,7 +167,7 @@ CPU_LE(	vrev64.8	q7, q7			)
 _fold_64_B_loop:
 
 	.macro		fold64, reg1, reg2
-	vld1.64		{q11-q12}, [arg2, :128]!
+	vld1.64		{q11-q12}, [arg2]!
 
 	vmull.p64	q8, \reg1\()h, d21
 	vmull.p64	\reg1, \reg1\()l, d20
@@ -238,7 +238,7 @@ _16B_reduction_loop:
 	vmull.p64	q7, d15, d21
 	veor.8		q7, q7, q8
 
-	vld1.64		{q0}, [arg2, :128]!
+	vld1.64		{q0}, [arg2]!
 CPU_LE(	vrev64.8	q0, q0		)
 	vswp		d0, d1
 	veor.8		q7, q7, q0
@@ -335,7 +335,7 @@ _less_than_128:
 	vmov.i8		q0, #0
 	vmov		s3, arg1_low32		// get the initial crc value
 
-	vld1.64		{q7}, [arg2, :128]!
+	vld1.64		{q7}, [arg2]!
 CPU_LE(	vrev64.8	q7, q7		)
 	vswp		d14, d15
 	veor.8		q7, q7, q0

arch/arm/crypto/crct10dif-ce-glue.c

@@ -35,26 +35,15 @@ static int crct10dif_update(struct shash_desc *desc, const u8 *data,
 			    unsigned int length)
 {
 	u16 *crc = shash_desc_ctx(desc);
-	unsigned int l;
 
-	if (!may_use_simd()) {
-		*crc = crc_t10dif_generic(*crc, data, length);
+	if (length >= CRC_T10DIF_PMULL_CHUNK_SIZE && may_use_simd()) {
+		kernel_neon_begin();
+		*crc = crc_t10dif_pmull(*crc, data, length);
+		kernel_neon_end();
 	} else {
-		if (unlikely((u32)data % CRC_T10DIF_PMULL_CHUNK_SIZE)) {
-			l = min_t(u32, length, CRC_T10DIF_PMULL_CHUNK_SIZE -
-				  ((u32)data % CRC_T10DIF_PMULL_CHUNK_SIZE));
-
-			*crc = crc_t10dif_generic(*crc, data, l);
-
-			length -= l;
-			data += l;
-		}
-		if (length > 0) {
-			kernel_neon_begin();
-			*crc = crc_t10dif_pmull(*crc, data, length);
-			kernel_neon_end();
-		}
+		*crc = crc_t10dif_generic(*crc, data, length);
 	}
+
 	return 0;
 }