ceph: make object hash a pg_pool property

The object will be hashed to a placement seed (ps) based on the pg_pool's
hash function.  This allows new hashes to be introduced into an existing
object store, or selection of a hash appropriate to the objects that
will be stored in a particular pool.

Signed-off-by: Sage Weil <sage@newdream.net>

9 files changed, 137 insertions, 81 deletions

diff --git a/fs/ceph/Makefile b/fs/ceph/Makefile
index 8bad70aac36..bdd3e6fc160 100644
--- a/fs/ceph/Makefile
+++ b/fs/ceph/Makefile
@@ -13,7 +13,7 @@ ceph-objs := super.o inode.o dir.o file.o addr.o ioctl.o \
 	mon_client.o \
 	osd_client.o osdmap.o crush/crush.o crush/mapper.o crush/hash.o \
 	debugfs.o \
-	ceph_fs.o ceph_strings.o ceph_frag.o
+	ceph_fs.o ceph_strings.o ceph_hash.o ceph_frag.o
 
 else
 #Otherwise we were called directly from the command
diff --git a/fs/ceph/README b/fs/ceph/README
index 660e00074d5..18352fab37c 100644
--- a/fs/ceph/README
+++ b/fs/ceph/README
@@ -10,6 +10,8 @@ src/include/rados.h	    fs/ceph/rados.h
 src/include/ceph_strings.cc fs/ceph/ceph_strings.c
 src/include/ceph_frag.h	    fs/ceph/ceph_frag.h
 src/include/ceph_frag.cc    fs/ceph/ceph_frag.c
+src/include/ceph_hash.h	    fs/ceph/ceph_hash.h
+src/include/ceph_hash.cc    fs/ceph/ceph_hash.c
 src/crush/crush.c	    fs/ceph/crush/crush.c
 src/crush/crush.h	    fs/ceph/crush/crush.h
 src/crush/mapper.c	    fs/ceph/crush/mapper.c
diff --git a/fs/ceph/ceph_fs.c b/fs/ceph/ceph_fs.c
index b3ecf1b0752..79d76bc4303 100644
--- a/fs/ceph/ceph_fs.c
+++ b/fs/ceph/ceph_fs.c
@@ -72,80 +72,3 @@ int ceph_caps_for_mode(int mode)
 	}
 	return 0;
 }
-
-/*
- * Robert Jenkin's hash function.
- * http://burtleburtle.net/bob/hash/evahash.html
- * This is in the public domain.
- */
-#define mix(a, b, c)						\
-	do {							\
-		a = a - b;  a = a - c;  a = a ^ (c >> 13);	\
-		b = b - c;  b = b - a;  b = b ^ (a << 8);	\
-		c = c - a;  c = c - b;  c = c ^ (b >> 13);	\
-		a = a - b;  a = a - c;  a = a ^ (c >> 12);	\
-		b = b - c;  b = b - a;  b = b ^ (a << 16);	\
-		c = c - a;  c = c - b;  c = c ^ (b >> 5);	\
-		a = a - b;  a = a - c;  a = a ^ (c >> 3);	\
-		b = b - c;  b = b - a;  b = b ^ (a << 10);	\
-		c = c - a;  c = c - b;  c = c ^ (b >> 15);	\
-	} while (0)
-
-unsigned int ceph_full_name_hash(const char *str, unsigned int length)
-{
-	const unsigned char *k = (const unsigned char *)str;
-	__u32 a, b, c;  /* the internal state */
-	__u32 len;      /* how many key bytes still need mixing */
-
-	/* Set up the internal state */
-	len = length;
-	a = 0x9e3779b9;      /* the golden ratio; an arbitrary value */
-	b = a;
-	c = 0;               /* variable initialization of internal state */
-
-	/* handle most of the key */
-	while (len >= 12) {
-		a = a + (k[0] + ((__u32)k[1] << 8) + ((__u32)k[2] << 16) +
-			 ((__u32)k[3] << 24));
-		b = b + (k[4] + ((__u32)k[5] << 8) + ((__u32)k[6] << 16) +
-			 ((__u32)k[7] << 24));
-		c = c + (k[8] + ((__u32)k[9] << 8) + ((__u32)k[10] << 16) +
-			 ((__u32)k[11] << 24));
-		mix(a, b, c);
-		k = k + 12;
-		len = len - 12;
-	}
-
-	/* handle the last 11 bytes */
-	c = c + length;
-	switch (len) {            /* all the case statements fall through */
-	case 11:
-		c = c + ((__u32)k[10] << 24);
-	case 10:
-		c = c + ((__u32)k[9] << 16);
-	case 9:
-		c = c + ((__u32)k[8] << 8);
-		/* the first byte of c is reserved for the length */
-	case 8:
-		b = b + ((__u32)k[7] << 24);
-	case 7:
-		b = b + ((__u32)k[6] << 16);
-	case 6:
-		b = b + ((__u32)k[5] << 8);
-	case 5:
-		b = b + k[4];
-	case 4:
-		a = a + ((__u32)k[3] << 24);
-	case 3:
-		a = a + ((__u32)k[2] << 16);
-	case 2:
-		a = a + ((__u32)k[1] << 8);
-	case 1:
-		a = a + k[0];
-		/* case 0: nothing left to add */
-	}
-	mix(a, b, c);
-
-	return c;
-}
-
diff --git a/fs/ceph/ceph_fs.h b/fs/ceph/ceph_fs.h
index 25fc537f414..36becb02478 100644
--- a/fs/ceph/ceph_fs.h
+++ b/fs/ceph/ceph_fs.h
@@ -49,8 +49,6 @@
 #define CEPH_MAX_MON   31
 
 
-unsigned int ceph_full_name_hash(const char *name, unsigned int len);
-
 
 /*
  * ceph_file_layout - describe data layout for a file/inode
diff --git a/fs/ceph/ceph_hash.c b/fs/ceph/ceph_hash.c
new file mode 100644
index 00000000000..ac8be54631f
--- /dev/null
+++ b/fs/ceph/ceph_hash.c
@@ -0,0 +1,118 @@
+
+#include "types.h"
+
+/*
+ * Robert Jenkin's hash function.
+ * http://burtleburtle.net/bob/hash/evahash.html
+ * This is in the public domain.
+ */
+#define mix(a, b, c)						\
+	do {							\
+		a = a - b;  a = a - c;  a = a ^ (c >> 13);	\
+		b = b - c;  b = b - a;  b = b ^ (a << 8);	\
+		c = c - a;  c = c - b;  c = c ^ (b >> 13);	\
+		a = a - b;  a = a - c;  a = a ^ (c >> 12);	\
+		b = b - c;  b = b - a;  b = b ^ (a << 16);	\
+		c = c - a;  c = c - b;  c = c ^ (b >> 5);	\
+		a = a - b;  a = a - c;  a = a ^ (c >> 3);	\
+		b = b - c;  b = b - a;  b = b ^ (a << 10);	\
+		c = c - a;  c = c - b;  c = c ^ (b >> 15);	\
+	} while (0)
+
+unsigned ceph_str_hash_rjenkins(const char *str, unsigned length)
+{
+	const unsigned char *k = (const unsigned char *)str;
+	__u32 a, b, c;  /* the internal state */
+	__u32 len;      /* how many key bytes still need mixing */
+
+	/* Set up the internal state */
+	len = length;
+	a = 0x9e3779b9;      /* the golden ratio; an arbitrary value */
+	b = a;
+	c = 0;               /* variable initialization of internal state */
+
+	/* handle most of the key */
+	while (len >= 12) {
+		a = a + (k[0] + ((__u32)k[1] << 8) + ((__u32)k[2] << 16) +
+			 ((__u32)k[3] << 24));
+		b = b + (k[4] + ((__u32)k[5] << 8) + ((__u32)k[6] << 16) +
+			 ((__u32)k[7] << 24));
+		c = c + (k[8] + ((__u32)k[9] << 8) + ((__u32)k[10] << 16) +
+			 ((__u32)k[11] << 24));
+		mix(a, b, c);
+		k = k + 12;
+		len = len - 12;
+	}
+
+	/* handle the last 11 bytes */
+	c = c + length;
+	switch (len) {            /* all the case statements fall through */
+	case 11:
+		c = c + ((__u32)k[10] << 24);
+	case 10:
+		c = c + ((__u32)k[9] << 16);
+	case 9:
+		c = c + ((__u32)k[8] << 8);
+		/* the first byte of c is reserved for the length */
+	case 8:
+		b = b + ((__u32)k[7] << 24);
+	case 7:
+		b = b + ((__u32)k[6] << 16);
+	case 6:
+		b = b + ((__u32)k[5] << 8);
+	case 5:
+		b = b + k[4];
+	case 4:
+		a = a + ((__u32)k[3] << 24);
+	case 3:
+		a = a + ((__u32)k[2] << 16);
+	case 2:
+		a = a + ((__u32)k[1] << 8);
+	case 1:
+		a = a + k[0];
+		/* case 0: nothing left to add */
+	}
+	mix(a, b, c);
+
+	return c;
+}
+
+/*
+ * linux dcache hash
+ */
+unsigned ceph_str_hash_linux(const char *str, unsigned length)
+{
+        unsigned long hash = 0;
+	unsigned char c;
+
+        while (length-- > 0) {
+		c = *str++;
+		hash = (hash + (c << 4) + (c >> 4)) * 11;
+	}
+        return hash;
+}
+
+
+unsigned ceph_str_hash(int type, const char *s, unsigned len)
+{
+	switch (type) {
+	case CEPH_STR_HASH_LINUX:
+		return ceph_str_hash_linux(s, len);
+	case CEPH_STR_HASH_RJENKINS:
+		return ceph_str_hash_rjenkins(s, len);
+	default:
+		return -1;
+	}
+}
+
+const char *ceph_str_hash_name(int type) 
+{
+	switch (type) {
+	case CEPH_STR_HASH_LINUX:
+		return "linux";
+	case CEPH_STR_HASH_RJENKINS:
+		return "rjenkins";
+	default:
+		return "unknown";
+	}
+}
diff --git a/fs/ceph/ceph_hash.h b/fs/ceph/ceph_hash.h
new file mode 100644
index 00000000000..5ac470c433c
--- /dev/null
+++ b/fs/ceph/ceph_hash.h
@@ -0,0 +1,13 @@
+#ifndef _FS_CEPH_HASH_H
+#define _FS_CEPH_HASH_H
+
+#define CEPH_STR_HASH_LINUX      0x1  /* linux dcache hash */
+#define CEPH_STR_HASH_RJENKINS   0x2  /* robert jenkins' */
+
+extern unsigned ceph_str_hash_linux(const char *s, unsigned len);
+extern unsigned ceph_str_hash_rjenkins(const char *s, unsigned len);
+
+extern unsigned ceph_str_hash(int type, const char *s, unsigned len);
+extern const char *ceph_str_hash_name(int type);
+
+#endif
diff --git a/fs/ceph/osdmap.c b/fs/ceph/osdmap.c
index a025555b70a..68478270ad7 100644
--- a/fs/ceph/osdmap.c
+++ b/fs/ceph/osdmap.c
@@ -809,7 +809,7 @@ int ceph_calc_object_layout(struct ceph_object_layout *ol,
 		return -EIO;
 
 	pool = &osdmap->pg_pool[poolid];
-	ps = ceph_full_name_hash(oid, strlen(oid));
+	ps = ceph_str_hash(pool->v.object_hash, oid, strlen(oid));
 	if (preferred >= 0) {
 		ps += preferred;
 		num = le32_to_cpu(pool->v.lpg_num);
diff --git a/fs/ceph/rados.h b/fs/ceph/rados.h
index 85bdef78d14..fb23ff9297c 100644
--- a/fs/ceph/rados.h
+++ b/fs/ceph/rados.h
@@ -84,6 +84,7 @@ struct ceph_pg_pool {
 	__u8 type;                /* CEPH_PG_TYPE_* */
 	__u8 size;                /* number of osds in each pg */
 	__u8 crush_ruleset;       /* crush placement rule */
+	__u8 object_hash;         /* hash mapping object name to ps */
 	__le32 pg_num, pgp_num;   /* number of pg's */
 	__le32 lpg_num, lpgp_num; /* number of localized pg's */
 	__le32 last_change;       /* most recent epoch changed */
diff --git a/fs/ceph/types.h b/fs/ceph/types.h
index 8a514568cab..28b35a005ec 100644
--- a/fs/ceph/types.h
+++ b/fs/ceph/types.h
@@ -9,6 +9,7 @@
 
 #include "ceph_fs.h"
 #include "ceph_frag.h"
+#include "ceph_hash.h"
 
 /*
  * Identify inodes by both their ino AND snapshot id (a u64).