Use intel hardware crc32c where available

Signed-off-by: Chris Mason <chris.mason@fusionio.com>

btrfs.c

@@ -258,6 +258,8 @@ int main(int argc, char **argv)
 {
 	const struct cmd_struct *cmd;
 
+	crc32c_optimization_init();
+
 	argc--;
 	argv++;
 	handle_options(&argc, &argv);

crc32c.c

@@ -9,6 +9,123 @@
  */
 #include "kerncompat.h"
 #include "crc32c.h"
+#include <inttypes.h>
+#include <string.h>
+#include <unistd.h>
+#include <stdlib.h>
+#include <signal.h>
+#include <sys/types.h>
+#include <sys/wait.h>
+
+u32 __crc32c_le(u32 crc, unsigned char const *data, size_t length);
+static u32 (*crc_function)(u32 crc, unsigned char const *data, size_t length) = __crc32c_le;
+
+#ifdef __x86_64__
+
+/*
+ * Based on a posting to lkml by Austin Zhang <austin.zhang@intel.com>
+ *
+ * Using hardware provided CRC32 instruction to accelerate the CRC32 disposal.
+ * CRC32C polynomial:0x1EDC6F41(BE)/0x82F63B78(LE)
+ * CRC32 is a new instruction in Intel SSE4.2, the reference can be found at:
+ * http://www.intel.com/products/processor/manuals/
+ * Intel(R) 64 and IA-32 Architectures Software Developer's Manual
+ * Volume 2A: Instruction Set Reference, A-M
+ */
+#if  __SIZEOF_LONG__ == 8
+#define REX_PRE "0x48, "
+#define SCALE_F 8
+#else
+#define REX_PRE
+#define SCALE_F 4
+#endif
+
+static int crc32c_probed = 0;
+static int crc32c_intel_available = 0;
+
+static uint32_t crc32c_intel_le_hw_byte(uint32_t crc, unsigned char const *data,
+					unsigned long length)
+{
+	while (length--) {
+		__asm__ __volatile__(
+			".byte 0xf2, 0xf, 0x38, 0xf0, 0xf1"
+			:"=S"(crc)
+			:"0"(crc), "c"(*data)
+		);
+		data++;
+	}
+
+	return crc;
+}
+
+/*
+ * Steps through buffer one byte at at time, calculates reflected 
+ * crc using table.
+ */
+uint32_t crc32c_intel(u32 crc, unsigned char const *data, unsigned long length)
+{
+	unsigned int iquotient = length / SCALE_F;
+	unsigned int iremainder = length % SCALE_F;
+	unsigned long *ptmp = (unsigned long *)data;
+
+	while (iquotient--) {
+		__asm__ __volatile__(
+			".byte 0xf2, " REX_PRE "0xf, 0x38, 0xf1, 0xf1;"
+			:"=S"(crc)
+			:"0"(crc), "c"(*ptmp)
+		);
+		ptmp++;
+	}
+
+	if (iremainder)
+		crc = crc32c_intel_le_hw_byte(crc, (unsigned char *)ptmp,
+				 iremainder);
+
+	return crc;
+}
+
+static void do_cpuid(unsigned int *eax, unsigned int *ebx, unsigned int *ecx,
+		     unsigned int *edx)
+{
+	int id = *eax;
+
+	asm("movl %4, %%eax;"
+	    "cpuid;"
+	    "movl %%eax, %0;"
+	    "movl %%ebx, %1;"
+	    "movl %%ecx, %2;"
+	    "movl %%edx, %3;"
+		: "=r" (*eax), "=r" (*ebx), "=r" (*ecx), "=r" (*edx)
+		: "r" (id)
+		: "eax", "ebx", "ecx", "edx");
+}
+
+void crc32c_intel_probe(void)
+{
+	if (!crc32c_probed) {
+		unsigned int eax, ebx, ecx, edx;
+
+		eax = 1;
+
+		do_cpuid(&eax, &ebx, &ecx, &edx);
+		crc32c_intel_available = (ecx & (1 << 20)) != 0;
+		crc32c_probed = 1;
+	}
+}
+
+int crc32c_optimization_init(void)
+{
+	crc32c_intel_probe();
+	if (crc32c_intel_available)
+		crc_function = crc32c_intel;
+}
+#else
+
+int crc32c_optimization_init(void)
+{
+}
+
+#endif /* __x86_64__ */
 
 /*
  * This is the CRC-32C table
@@ -91,11 +208,15 @@ static const u32 crc32c_table[256] = {
  * crc using table.
  */
 
-u32 crc32c_le(u32 crc, unsigned char const *data, size_t length)
+u32 __crc32c_le(u32 crc, unsigned char const *data, size_t length)
 {
 	while (length--)
 		crc =
 		    crc32c_table[(crc ^ *data++) & 0xFFL] ^ (crc >> 8);
-
 	return crc;
 }
+
+u32 crc32c_le(u32 crc, unsigned char const *data, size_t length)
+{
+	return crc_function(crc, data, length);
+}

crc32c.h

@@ -22,6 +22,7 @@
 #include "kerncompat.h"
 
 u32 crc32c_le(u32 seed, unsigned char const *data, size_t length);
+int crc32c_optimization_init(void);
 
 #define crc32c(seed, data, length) crc32c_le(seed, (unsigned char const *)data, length)
 #define btrfs_crc32c crc32c