crc32: add-slicing-by-8.diff

Index: infiniband/lib/crc32.c
===================================================================
@@ -45,6 +45,7 @@ MODULE_LICENSE("GPL");
 
 #if CRC_LE_BITS > 8 || CRC_BE_BITS > 8
 
+/* implements slicing-by-4 or slicing-by-8 algorithm */
 static inline u32
 crc32_body(u32 crc, unsigned char const *buf, size_t len, const u32 (*tab)[256])
 {

@@ -54,42 +55,81 @@ crc32_body(u32 crc, unsigned char const
 			t2[(crc >> 8) & 255] ^ \
 			t1[(crc >> 16) & 255] ^ \
 			t0[(crc >> 24) & 255]
+#  define DO_CRC8a (t7[(q) & 255] ^ \
+			t6[(q >> 8) & 255] ^ \
+			t5[(q >> 16) & 255] ^ \
+			t4[(q >> 24) & 255])
+#  define DO_CRC8b (t3[(q) & 255] ^ \
+			t2[(q >> 8) & 255] ^ \
+			t1[(q >> 16) & 255] ^ \
+			t0[(q >> 24) & 255])
 # else
 #  define DO_CRC(x) (crc = t0[((crc >> 24) ^ (x)) & 255] ^ (crc << 8))
 #  define DO_CRC4 crc = t0[(crc) & 255] ^ \
 			t1[(crc >> 8) & 255] ^ \
 			t2[(crc >> 16) & 255] ^ \
 			t3[(crc >> 24) & 255]
+#  define DO_CRC8a (t4[(q) & 255] ^ \
+			t5[(q >> 8) & 255] ^ \
+			t6[(q >> 16) & 255] ^ \
+			t7[(q >> 24) & 255])
+#  define DO_CRC8b (t0[(q) & 255] ^ \
+			t1[(q >> 8) & 255] ^ \
+			t2[(q >> 16) & 255] ^ \
+			t3[(q >> 24) & 255])
 # endif
 	const u32 *b;
-	size_t    rem_len;
 	const u32 *t0 = tab[0], *t1 = tab[1], *t2 = tab[2], *t3 = tab[3];
+	const u32 *t4 = tab[4], *t5 = tab[5], *t6 = tab[6], *t7 = tab[7];
+	size_t init_len;
+	size_t middle_len;
+	size_t rem_len;
+
+	/* break buf into init_len bytes before and
+	 * rem_len bytes after a middle section with
+	 * middle_len properly aligned words
+	 * note 8 byte loop need only be aligned on
+	 * a 4 byte boundary */
+	init_len = min((-((uintptr_t)buf)) & 3, len);
+# if CRC_LE_BITS == 32
+	middle_len = (len - init_len) >> 2;
+	rem_len = (len - init_len) & 3;
+# else
+	middle_len = (len - init_len) >> 3;
+	rem_len = (len - init_len) & 7;
+# endif
 
 	/* Align it */
-	if (unlikely((long)buf & 3 && len)) {
+	if (unlikely(init_len)) {
 		do {
 			DO_CRC(*buf++);
-		} while ((--len) && ((long)buf)&3);
+		} while (--init_len);
 	}
-	rem_len = len & 3;
-	/* load data 32 bits wide, xor data 32 bits wide. */
-	len = len >> 2;
 	b = (const u32 *)buf;
-	for (--b; len; --len) {
+	for (--b; middle_len; --middle_len) {
+# if CRC_LE_BITS == 32
 		crc ^= *++b; /* use pre increment for speed */
 		DO_CRC4;
+# else
+		u32 q;
+		q = crc ^ *++b;
+		crc = DO_CRC8a;
+		q = *++b;
+		crc ^= DO_CRC8b;
+# endif
 	}
-	len = rem_len;
 	/* And the last few bytes */
-	if (len) {
+	if (rem_len) {
 		u8 *p = (u8 *)(b + 1) - 1;
 		do {
 			DO_CRC(*++p); /* use pre increment for speed */
-		} while (--len);
+		} while (--rem_len);
 	}
 	return crc;
 #undef DO_CRC
 #undef DO_CRC4
+#undef DO_CRC8a
+#undef DO_CRC8b
 }
 #endif
 

Index: infiniband/lib/crc32defs.h
===================================================================
@@ -6,29 +6,29 @@
 #define CRCPOLY_LE 0xedb88320
 #define CRCPOLY_BE 0x04c11db7
 
-/* How many bits at a time to use.  Valid values are 1, 2, 4, 8, and 32. */
+/* How many bits at a time to use.  Valid values are 1, 2, 4, 8, 32 and 64. */
 /* For less performance-sensitive, use 4 or 8 */
 #ifndef CRC_LE_BITS
-# define CRC_LE_BITS 32
+# define CRC_LE_BITS 64
 #endif
 #ifndef CRC_BE_BITS
-# define CRC_BE_BITS 32
+# define CRC_BE_BITS 64
 #endif
 
 /*
  * Little-endian CRC computation.  Used with serial bit streams sent
  * lsbit-first.  Be sure to use cpu_to_le32() to append the computed CRC.
  */
-#if CRC_LE_BITS > 32 || CRC_LE_BITS < 1 || CRC_LE_BITS == 16 || \
+#if CRC_LE_BITS > 64 || CRC_LE_BITS < 1 || CRC_LE_BITS == 16 || \
 	CRC_LE_BITS & CRC_LE_BITS-1
-# error "CRC_LE_BITS must be one of {1, 2, 4, 8, 32}"
+# error "CRC_LE_BITS must be one of {1, 2, 4, 8, 64}"
 #endif
 
 /*
  * Big-endian CRC computation.  Used with serial bit streams sent
  * msbit-first.  Be sure to use cpu_to_be32() to append the computed CRC.
  */
-#if CRC_BE_BITS > 32 || CRC_BE_BITS < 1 || CRC_BE_BITS == 16 || \
+#if CRC_BE_BITS > 64 || CRC_BE_BITS < 1 || CRC_BE_BITS == 16 || \
 	CRC_BE_BITS & CRC_BE_BITS-1
-# error "CRC_BE_BITS must be one of {1, 2, 4, 8, 32}"
+# error "CRC_BE_BITS must be one of {1, 2, 4, 8, 64}"
 #endif

86 lines not on this page · raw message