Add basic RAID[56] support

David Woodhouse originally contributed this code, and Chris Mason
changed it around to reflect the current design goals for raid56.

The original code expected all metadata and data writes to be full
stripes.  This meant metadata block size == stripe size, and had a few
other restrictions.

This version allows metadata blocks smaller than the stripe size.  It
implements both raid5 and raid6, although it does not have code to
rebuild from parity if one of the drives is missing or incorrect.

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

1 files changed, 97 insertions, 0 deletions

diff --git a/raid6.c b/raid6.c
new file mode 100644
index 0000000..3a42bdf
--- /dev/null
+++ b/raid6.c
@@ -0,0 +1,97 @@
+/* -*- linux-c -*- ------------------------------------------------------- *
+ *
+ *   Copyright 2002-2004 H. Peter Anvin - All Rights Reserved
+ *
+ *   This program is free software; you can redistribute it and/or modify
+ *   it under the terms of the GNU General Public License as published by
+ *   the Free Software Foundation, Inc., 53 Temple Place Ste 330,
+ *   Boston MA 02111-1307, USA; either version 2 of the License, or
+ *   (at your option) any later version; incorporated herein by reference.
+ *
+ * ----------------------------------------------------------------------- */
+
+/*
+ * raid6int1.c
+ *
+ * 1-way unrolled portable integer math RAID-6 instruction set
+ *
+ * This file was postprocessed using unroll.pl and then ported to userspace
+ */
+#include <stdint.h>
+#include <unistd.h>
+/*
+ * This is the C data type to use
+ */
+
+/* Change this from BITS_PER_LONG if there is something better... */
+#if BITS_PER_LONG == 64
+# define NBYTES(x) ((x) * 0x0101010101010101UL)
+# define NSIZE  8
+# define NSHIFT 3
+typedef uint64_t unative_t;
+#else
+# define NBYTES(x) ((x) * 0x01010101U)
+# define NSIZE  4
+# define NSHIFT 2
+typedef uint32_t unative_t;
+#endif
+
+/*
+ * These sub-operations are separate inlines since they can sometimes be
+ * specially optimized using architecture-specific hacks.
+ */
+
+/*
+ * The SHLBYTE() operation shifts each byte left by 1, *not*
+ * rolling over into the next byte
+ */
+static inline __attribute_const__ unative_t SHLBYTE(unative_t v)
+{
+	unative_t vv;
+
+	vv = (v << 1) & NBYTES(0xfe);
+	return vv;
+}
+
+/*
+ * The MASK() operation returns 0xFF in any byte for which the high
+ * bit is 1, 0x00 for any byte for which the high bit is 0.
+ */
+static inline __attribute_const__ unative_t MASK(unative_t v)
+{
+	unative_t vv;
+
+	vv = v & NBYTES(0x80);
+	vv = (vv << 1) - (vv >> 7); /* Overflow on the top bit is OK */
+	return vv;
+}
+
+
+void raid6_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	uint8_t **dptr = (uint8_t **)ptrs;
+	uint8_t *p, *q;
+	int d, z, z0;
+
+	unative_t wd0, wq0, wp0, w10, w20;
+
+	z0 = disks - 3;		/* Highest data disk */
+	p = dptr[z0+1];		/* XOR parity */
+	q = dptr[z0+2];		/* RS syndrome */
+
+	for ( d = 0 ; d < bytes ; d += NSIZE*1 ) {
+		wq0 = wp0 = *(unative_t *)&dptr[z0][d+0*NSIZE];
+		for ( z = z0-1 ; z >= 0 ; z-- ) {
+			wd0 = *(unative_t *)&dptr[z][d+0*NSIZE];
+			wp0 ^= wd0;
+			w20 = MASK(wq0);
+			w10 = SHLBYTE(wq0);
+			w20 &= NBYTES(0x1d);
+			w10 ^= w20;
+			wq0 = w10 ^ wd0;
+		}
+		*(unative_t *)&p[d+NSIZE*0] = wp0;
+		*(unative_t *)&q[d+NSIZE*0] = wq0;
+	}
+}
+