Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client

Pull Ceph updates from Sage Weil:
 "We have a pile of bug fixes from Ilya, including a few patches that
  sync up the CRUSH code with the latest from userspace.

  There is also a long series from Zheng that fixes various issues with
  snapshots, inline data, and directory fsync, some simplification and
  improvement in the cap release code, and a rework of the caching of
  directory contents.

  To top it off there are a few small fixes and cleanups from Benoit and
  Hong"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client: (40 commits)
  rbd: use GFP_NOIO in rbd_obj_request_create()
  crush: fix a bug in tree bucket decode
  libceph: Fix ceph_tcp_sendpage()'s more boolean usage
  libceph: Remove spurious kunmap() of the zero page
  rbd: queue_depth map option
  rbd: store rbd_options in rbd_device
  rbd: terminate rbd_opts_tokens with Opt_err
  ceph: fix ceph_writepages_start()
  rbd: bump queue_max_segments
  ceph: rework dcache readdir
  crush: sync up with userspace
  crush: fix crash from invalid 'take' argument
  ceph: switch some GFP_NOFS memory allocation to GFP_KERNEL
  ceph: pre-allocate data structure that tracks caps flushing
  ceph: re-send flushing caps (which are revoked) in reconnect stage
  ceph: send TID of the oldest pending caps flush to MDS
  ceph: track pending caps flushing globally
  ceph: track pending caps flushing accurately
  libceph: fix wrong name "Ceph filesystem for Linux"
  ceph: fix directory fsync
  ...

net/ceph/ceph_common.c

@@ -352,8 +352,8 @@ ceph_parse_options(char *options, const char *dev_name,
 	/* start with defaults */
 	opt->flags = CEPH_OPT_DEFAULT;
 	opt->osd_keepalive_timeout = CEPH_OSD_KEEPALIVE_DEFAULT;
-	opt->mount_timeout = CEPH_MOUNT_TIMEOUT_DEFAULT; /* seconds */
-	opt->osd_idle_ttl = CEPH_OSD_IDLE_TTL_DEFAULT;   /* seconds */
+	opt->mount_timeout = CEPH_MOUNT_TIMEOUT_DEFAULT;
+	opt->osd_idle_ttl = CEPH_OSD_IDLE_TTL_DEFAULT;
 
 	/* get mon ip(s) */
 	/* ip1[:port1][,ip2[:port2]...] */
@@ -439,13 +439,32 @@ ceph_parse_options(char *options, const char *dev_name,
 			pr_warn("ignoring deprecated osdtimeout option\n");
 			break;
 		case Opt_osdkeepalivetimeout:
-			opt->osd_keepalive_timeout = intval;
+			/* 0 isn't well defined right now, reject it */
+			if (intval < 1 || intval > INT_MAX / 1000) {
+				pr_err("osdkeepalive out of range\n");
+				err = -EINVAL;
+				goto out;
+			}
+			opt->osd_keepalive_timeout =
+					msecs_to_jiffies(intval * 1000);
 			break;
 		case Opt_osd_idle_ttl:
-			opt->osd_idle_ttl = intval;
+			/* 0 isn't well defined right now, reject it */
+			if (intval < 1 || intval > INT_MAX / 1000) {
+				pr_err("osd_idle_ttl out of range\n");
+				err = -EINVAL;
+				goto out;
+			}
+			opt->osd_idle_ttl = msecs_to_jiffies(intval * 1000);
 			break;
 		case Opt_mount_timeout:
-			opt->mount_timeout = intval;
+			/* 0 is "wait forever" (i.e. infinite timeout) */
+			if (intval < 0 || intval > INT_MAX / 1000) {
+				pr_err("mount_timeout out of range\n");
+				err = -EINVAL;
+				goto out;
+			}
+			opt->mount_timeout = msecs_to_jiffies(intval * 1000);
 			break;
 
 		case Opt_share:
@@ -512,12 +531,14 @@ int ceph_print_client_options(struct seq_file *m, struct ceph_client *client)
 		seq_puts(m, "notcp_nodelay,");
 
 	if (opt->mount_timeout != CEPH_MOUNT_TIMEOUT_DEFAULT)
-		seq_printf(m, "mount_timeout=%d,", opt->mount_timeout);
+		seq_printf(m, "mount_timeout=%d,",
+			   jiffies_to_msecs(opt->mount_timeout) / 1000);
 	if (opt->osd_idle_ttl != CEPH_OSD_IDLE_TTL_DEFAULT)
-		seq_printf(m, "osd_idle_ttl=%d,", opt->osd_idle_ttl);
+		seq_printf(m, "osd_idle_ttl=%d,",
+			   jiffies_to_msecs(opt->osd_idle_ttl) / 1000);
 	if (opt->osd_keepalive_timeout != CEPH_OSD_KEEPALIVE_DEFAULT)
 		seq_printf(m, "osdkeepalivetimeout=%d,",
-			   opt->osd_keepalive_timeout);
+		    jiffies_to_msecs(opt->osd_keepalive_timeout) / 1000);
 
 	/* drop redundant comma */
 	if (m->count != pos)
@@ -626,8 +647,8 @@ static int have_mon_and_osd_map(struct ceph_client *client)
  */
 int __ceph_open_session(struct ceph_client *client, unsigned long started)
 {
-	int err;
-	unsigned long timeout = client->options->mount_timeout * HZ;
+	unsigned long timeout = client->options->mount_timeout;
+	long err;
 
 	/* open session, and wait for mon and osd maps */
 	err = ceph_monc_open_session(&client->monc);
@@ -635,16 +656,15 @@ int __ceph_open_session(struct ceph_client *client, unsigned long started)
 		return err;
 
 	while (!have_mon_and_osd_map(client)) {
-		err = -EIO;
 		if (timeout && time_after_eq(jiffies, started + timeout))
-			return err;
+			return -ETIMEDOUT;
 
 		/* wait */
 		dout("mount waiting for mon_map\n");
 		err = wait_event_interruptible_timeout(client->auth_wq,
 			have_mon_and_osd_map(client) || (client->auth_err < 0),
-			timeout);
-		if (err == -EINTR || err == -ERESTARTSYS)
+			ceph_timeout_jiffies(timeout));
+		if (err < 0)
 			return err;
 		if (client->auth_err < 0)
 			return client->auth_err;
@@ -721,5 +741,5 @@ module_exit(exit_ceph_lib);
 MODULE_AUTHOR("Sage Weil <sage@newdream.net>");
 MODULE_AUTHOR("Yehuda Sadeh <yehuda@hq.newdream.net>");
 MODULE_AUTHOR("Patience Warnick <patience@newdream.net>");
-MODULE_DESCRIPTION("Ceph filesystem for Linux");
+MODULE_DESCRIPTION("Ceph core library");
 MODULE_LICENSE("GPL");

net/ceph/crush/crush.c

@@ -1,15 +1,11 @@
-
 #ifdef __KERNEL__
 # include <linux/slab.h>
+# include <linux/crush/crush.h>
 #else
-# include <stdlib.h>
-# include <assert.h>
-# define kfree(x) do { if (x) free(x); } while (0)
-# define BUG_ON(x) assert(!(x))
+# include "crush_compat.h"
+# include "crush.h"
 #endif
 
-#include <linux/crush/crush.h>
-
 const char *crush_bucket_alg_name(int alg)
 {
 	switch (alg) {
@@ -134,6 +130,9 @@ void crush_destroy(struct crush_map *map)
 		kfree(map->rules);
 	}
 
+#ifndef __KERNEL__
+	kfree(map->choose_tries);
+#endif
 	kfree(map);
 }
 

net/ceph/crush/crush_ln_table.h

@@ -10,20 +10,20 @@
  *
  */
 
-#if defined(__linux__)
-#include <linux/types.h>
-#elif defined(__FreeBSD__)
-#include <sys/types.h>
-#endif
-
 #ifndef CEPH_CRUSH_LN_H
 #define CEPH_CRUSH_LN_H
 
+#ifdef __KERNEL__
+# include <linux/types.h>
+#else
+# include "crush_compat.h"
+#endif
 
-// RH_LH_tbl[2*k] = 2^48/(1.0+k/128.0)
-// RH_LH_tbl[2*k+1] = 2^48*log2(1.0+k/128.0)
-
-static int64_t __RH_LH_tbl[128*2+2] = {
+/*
+ * RH_LH_tbl[2*k] = 2^48/(1.0+k/128.0)
+ * RH_LH_tbl[2*k+1] = 2^48*log2(1.0+k/128.0)
+ */
+static __s64 __RH_LH_tbl[128*2+2] = {
   0x0001000000000000ll, 0x0000000000000000ll, 0x0000fe03f80fe040ll, 0x000002dfca16dde1ll,
   0x0000fc0fc0fc0fc1ll, 0x000005b9e5a170b4ll, 0x0000fa232cf25214ll, 0x0000088e68ea899all,
   0x0000f83e0f83e0f9ll, 0x00000b5d69bac77ell, 0x0000f6603d980f67ll, 0x00000e26fd5c8555ll,
@@ -89,11 +89,12 @@ static int64_t __RH_LH_tbl[128*2+2] = {
   0x0000820820820821ll, 0x0000fa2f045e7832ll, 0x000081848da8faf1ll, 0x0000fba577877d7dll,
   0x0000810204081021ll, 0x0000fd1a708bbe11ll, 0x0000808080808081ll, 0x0000fe8df263f957ll,
   0x0000800000000000ll, 0x0000ffff00000000ll,
-  };
+};
 
-
-    // LL_tbl[k] = 2^48*log2(1.0+k/2^15);
-static int64_t __LL_tbl[256] = {
+/*
+ * LL_tbl[k] = 2^48*log2(1.0+k/2^15)
+ */
+static __s64 __LL_tbl[256] = {
   0x0000000000000000ull, 0x00000002e2a60a00ull, 0x000000070cb64ec5ull, 0x00000009ef50ce67ull,
   0x0000000cd1e588fdull, 0x0000000fb4747e9cull, 0x0000001296fdaf5eull, 0x0000001579811b58ull,
   0x000000185bfec2a1ull, 0x0000001b3e76a552ull, 0x0000001e20e8c380ull, 0x0000002103551d43ull,
@@ -160,7 +161,4 @@ static int64_t __LL_tbl[256] = {
   0x000002d4562d2ec6ull, 0x000002d73330209dull, 0x000002da102d63b0ull, 0x000002dced24f814ull,
 };
 
-
-
-
 #endif

net/ceph/crush/hash.c

@@ -1,6 +1,8 @@
-
-#include <linux/types.h>
-#include <linux/crush/hash.h>
+#ifdef __KERNEL__
+# include <linux/crush/hash.h>
+#else
+# include "hash.h"
+#endif
 
 /*
  * Robert Jenkins' function for mixing 32-bit values

net/ceph/crush/mapper.c

@@ -1,27 +1,31 @@
+/*
+ * Ceph - scalable distributed file system
+ *
+ * Copyright (C) 2015 Intel Corporation All Rights Reserved
+ *
+ * This is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License version 2.1, as published by the Free Software
+ * Foundation.  See file COPYING.
+ *
+ */
 
 #ifdef __KERNEL__
 # include <linux/string.h>
 # include <linux/slab.h>
 # include <linux/bug.h>
 # include <linux/kernel.h>
-# ifndef dprintk
-#  define dprintk(args...)
-# endif
+# include <linux/crush/crush.h>
+# include <linux/crush/hash.h>
 #else
-# include <string.h>
-# include <stdio.h>
-# include <stdlib.h>
-# include <assert.h>
-# define BUG_ON(x) assert(!(x))
-# define dprintk(args...) /* printf(args) */
-# define kmalloc(x, f) malloc(x)
-# define kfree(x) free(x)
+# include "crush_compat.h"
+# include "crush.h"
+# include "hash.h"
 #endif
-
-#include <linux/crush/crush.h>
-#include <linux/crush/hash.h>
 #include "crush_ln_table.h"
 
+#define dprintk(args...) /* printf(args) */
+
 /*
  * Implement the core CRUSH mapping algorithm.
  */
@@ -139,7 +143,7 @@ static int bucket_list_choose(struct crush_bucket_list *bucket,
 	int i;
 
 	for (i = bucket->h.size-1; i >= 0; i--) {
-		__u64 w = crush_hash32_4(bucket->h.hash,x, bucket->h.items[i],
+		__u64 w = crush_hash32_4(bucket->h.hash, x, bucket->h.items[i],
 					 r, bucket->h.id);
 		w &= 0xffff;
 		dprintk("list_choose i=%d x=%d r=%d item %d weight %x "
@@ -238,43 +242,46 @@ static int bucket_straw_choose(struct crush_bucket_straw *bucket,
 	return bucket->h.items[high];
 }
 
-// compute 2^44*log2(input+1)
-uint64_t crush_ln(unsigned xin)
+/* compute 2^44*log2(input+1) */
+static __u64 crush_ln(unsigned int xin)
 {
-    unsigned x=xin, x1;
-    int iexpon, index1, index2;
-    uint64_t RH, LH, LL, xl64, result;
+	unsigned int x = xin, x1;
+	int iexpon, index1, index2;
+	__u64 RH, LH, LL, xl64, result;
 
-    x++;
+	x++;
 
-    // normalize input
-    iexpon = 15;
-    while(!(x&0x18000)) { x<<=1; iexpon--; }
+	/* normalize input */
+	iexpon = 15;
+	while (!(x & 0x18000)) {
+		x <<= 1;
+		iexpon--;
+	}
 
-    index1 = (x>>8)<<1;
-    // RH ~ 2^56/index1
-    RH = __RH_LH_tbl[index1 - 256];
-    // LH ~ 2^48 * log2(index1/256)
-    LH = __RH_LH_tbl[index1 + 1 - 256];
+	index1 = (x >> 8) << 1;
+	/* RH ~ 2^56/index1 */
+	RH = __RH_LH_tbl[index1 - 256];
+	/* LH ~ 2^48 * log2(index1/256) */
+	LH = __RH_LH_tbl[index1 + 1 - 256];
 
-    // RH*x ~ 2^48 * (2^15 + xf), xf<2^8
-    xl64 = (int64_t)x * RH;
-    xl64 >>= 48;
-    x1 = xl64;
+	/* RH*x ~ 2^48 * (2^15 + xf), xf<2^8 */
+	xl64 = (__s64)x * RH;
+	xl64 >>= 48;
+	x1 = xl64;
 
-    result = iexpon;
-    result <<= (12 + 32);
+	result = iexpon;
+	result <<= (12 + 32);
 
-    index2 = x1 & 0xff;
-    // LL ~ 2^48*log2(1.0+index2/2^15)
-    LL = __LL_tbl[index2];
+	index2 = x1 & 0xff;
+	/* LL ~ 2^48*log2(1.0+index2/2^15) */
+	LL = __LL_tbl[index2];
 
-    LH = LH + LL;
+	LH = LH + LL;
 
-    LH >>= (48-12 - 32);
-    result += LH;
+	LH >>= (48 - 12 - 32);
+	result += LH;
 
-    return result;
+	return result;
 }
 
 
@@ -290,9 +297,9 @@ uint64_t crush_ln(unsigned xin)
 static int bucket_straw2_choose(struct crush_bucket_straw2 *bucket,
 				int x, int r)
 {
-	unsigned i, high = 0;
-	unsigned u;
-	unsigned w;
+	unsigned int i, high = 0;
+	unsigned int u;
+	unsigned int w;
 	__s64 ln, draw, high_draw = 0;
 
 	for (i = 0; i < bucket->h.size; i++) {
@@ -567,6 +574,10 @@ reject:
 		out[outpos] = item;
 		outpos++;
 		count--;
+#ifndef __KERNEL__
+		if (map->choose_tries && ftotal <= map->choose_total_tries)
+			map->choose_tries[ftotal]++;
+#endif
 	}
 
 	dprintk("CHOOSE returns %d\n", outpos);
@@ -610,6 +621,20 @@ static void crush_choose_indep(const struct crush_map *map,
 	}
 
 	for (ftotal = 0; left > 0 && ftotal < tries; ftotal++) {
+#ifdef DEBUG_INDEP
+		if (out2 && ftotal) {
+			dprintk("%u %d a: ", ftotal, left);
+			for (rep = outpos; rep < endpos; rep++) {
+				dprintk(" %d", out[rep]);
+			}
+			dprintk("\n");
+			dprintk("%u %d b: ", ftotal, left);
+			for (rep = outpos; rep < endpos; rep++) {
+				dprintk(" %d", out2[rep]);
+			}
+			dprintk("\n");
+		}
+#endif
 		for (rep = outpos; rep < endpos; rep++) {
 			if (out[rep] != CRUSH_ITEM_UNDEF)
 				continue;
@@ -726,6 +751,24 @@ static void crush_choose_indep(const struct crush_map *map,
 			out2[rep] = CRUSH_ITEM_NONE;
 		}
 	}
+#ifndef __KERNEL__
+	if (map->choose_tries && ftotal <= map->choose_total_tries)
+		map->choose_tries[ftotal]++;
+#endif
+#ifdef DEBUG_INDEP
+	if (out2) {
+		dprintk("%u %d a: ", ftotal, left);
+		for (rep = outpos; rep < endpos; rep++) {
+			dprintk(" %d", out[rep]);
+		}
+		dprintk("\n");
+		dprintk("%u %d b: ", ftotal, left);
+		for (rep = outpos; rep < endpos; rep++) {
+			dprintk(" %d", out2[rep]);
+		}
+		dprintk("\n");
+	}
+#endif
 }
 
 /**
@@ -790,8 +833,15 @@ int crush_do_rule(const struct crush_map *map,
 
 		switch (curstep->op) {
 		case CRUSH_RULE_TAKE:
-			w[0] = curstep->arg1;
-			wsize = 1;
+			if ((curstep->arg1 >= 0 &&
+			     curstep->arg1 < map->max_devices) ||
+			    (-1-curstep->arg1 < map->max_buckets &&
+			     map->buckets[-1-curstep->arg1])) {
+				w[0] = curstep->arg1;
+				wsize = 1;
+			} else {
+				dprintk(" bad take value %d\n", curstep->arg1);
+			}
 			break;
 
 		case CRUSH_RULE_SET_CHOOSE_TRIES:
@@ -877,7 +927,7 @@ int crush_do_rule(const struct crush_map *map,
 						0);
 				} else {
 					out_size = ((numrep < (result_max-osize)) ?
-                                                    numrep : (result_max-osize));
+						    numrep : (result_max-osize));
 					crush_choose_indep(
 						map,
 						map->buckets[-1-w[i]],
@@ -923,5 +973,3 @@ int crush_do_rule(const struct crush_map *map,
 	}
 	return result_len;
 }
-
-

net/ceph/messenger.c

@@ -278,7 +278,6 @@ static void _ceph_msgr_exit(void)
 	ceph_msgr_slab_exit();
 
 	BUG_ON(zero_page == NULL);
-	kunmap(zero_page);
 	page_cache_release(zero_page);
 	zero_page = NULL;
 }
@@ -1545,7 +1544,7 @@ static int write_partial_message_data(struct ceph_connection *con)
 		page = ceph_msg_data_next(&msg->cursor, &page_offset, &length,
 							&last_piece);
 		ret = ceph_tcp_sendpage(con->sock, page, page_offset,
-				      length, last_piece);
+					length, !last_piece);
 		if (ret <= 0) {
 			if (do_datacrc)
 				msg->footer.data_crc = cpu_to_le32(crc);

net/ceph/mon_client.c

@@ -298,21 +298,28 @@ void ceph_monc_request_next_osdmap(struct ceph_mon_client *monc)
 }
 EXPORT_SYMBOL(ceph_monc_request_next_osdmap);
 
+/*
+ * Wait for an osdmap with a given epoch.
+ *
+ * @epoch: epoch to wait for
+ * @timeout: in jiffies, 0 means "wait forever"
+ */
 int ceph_monc_wait_osdmap(struct ceph_mon_client *monc, u32 epoch,
 			  unsigned long timeout)
 {
 	unsigned long started = jiffies;
-	int ret;
+	long ret;
 
 	mutex_lock(&monc->mutex);
 	while (monc->have_osdmap < epoch) {
 		mutex_unlock(&monc->mutex);
 
-		if (timeout != 0 && time_after_eq(jiffies, started + timeout))
+		if (timeout && time_after_eq(jiffies, started + timeout))
 			return -ETIMEDOUT;
 
 		ret = wait_event_interruptible_timeout(monc->client->auth_wq,
-					 monc->have_osdmap >= epoch, timeout);
+						monc->have_osdmap >= epoch,
+						ceph_timeout_jiffies(timeout));
 		if (ret < 0)
 			return ret;
 

net/ceph/osd_client.c

@@ -296,6 +296,9 @@ static void osd_req_op_data_release(struct ceph_osd_request *osd_req,
 	case CEPH_OSD_OP_CMPXATTR:
 		ceph_osd_data_release(&op->xattr.osd_data);
 		break;
+	case CEPH_OSD_OP_STAT:
+		ceph_osd_data_release(&op->raw_data_in);
+		break;
 	default:
 		break;
 	}
@@ -450,7 +453,7 @@ __CEPH_FORALL_OSD_OPS(GENERATE_CASE)
  */
 static struct ceph_osd_req_op *
 _osd_req_op_init(struct ceph_osd_request *osd_req, unsigned int which,
-				u16 opcode)
+		 u16 opcode, u32 flags)
 {
 	struct ceph_osd_req_op *op;
 
@@ -460,14 +463,15 @@ _osd_req_op_init(struct ceph_osd_request *osd_req, unsigned int which,
 	op = &osd_req->r_ops[which];
 	memset(op, 0, sizeof (*op));
 	op->op = opcode;
+	op->flags = flags;
 
 	return op;
 }
 
 void osd_req_op_init(struct ceph_osd_request *osd_req,
-				unsigned int which, u16 opcode)
+		     unsigned int which, u16 opcode, u32 flags)
 {
-	(void)_osd_req_op_init(osd_req, which, opcode);
+	(void)_osd_req_op_init(osd_req, which, opcode, flags);
 }
 EXPORT_SYMBOL(osd_req_op_init);
 
@@ -476,7 +480,8 @@ void osd_req_op_extent_init(struct ceph_osd_request *osd_req,
 				u64 offset, u64 length,
 				u64 truncate_size, u32 truncate_seq)
 {
-	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which, opcode);
+	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which,
+						      opcode, 0);
 	size_t payload_len = 0;
 
 	BUG_ON(opcode != CEPH_OSD_OP_READ && opcode != CEPH_OSD_OP_WRITE &&
@@ -515,7 +520,8 @@ EXPORT_SYMBOL(osd_req_op_extent_update);
 void osd_req_op_cls_init(struct ceph_osd_request *osd_req, unsigned int which,
 			u16 opcode, const char *class, const char *method)
 {
-	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which, opcode);
+	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which,
+						      opcode, 0);
 	struct ceph_pagelist *pagelist;
 	size_t payload_len = 0;
 	size_t size;
@@ -552,7 +558,8 @@ int osd_req_op_xattr_init(struct ceph_osd_request *osd_req, unsigned int which,
 			  u16 opcode, const char *name, const void *value,
 			  size_t size, u8 cmp_op, u8 cmp_mode)
 {
-	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which, opcode);
+	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which,
+						      opcode, 0);
 	struct ceph_pagelist *pagelist;
 	size_t payload_len;
 
@@ -585,7 +592,8 @@ void osd_req_op_watch_init(struct ceph_osd_request *osd_req,
 				unsigned int which, u16 opcode,
 				u64 cookie, u64 version, int flag)
 {
-	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which, opcode);
+	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which,
+						      opcode, 0);
 
 	BUG_ON(opcode != CEPH_OSD_OP_NOTIFY_ACK && opcode != CEPH_OSD_OP_WATCH);
 
@@ -602,7 +610,8 @@ void osd_req_op_alloc_hint_init(struct ceph_osd_request *osd_req,
 				u64 expected_write_size)
 {
 	struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which,
-						      CEPH_OSD_OP_SETALLOCHINT);
+						      CEPH_OSD_OP_SETALLOCHINT,
+						      0);
 
 	op->alloc_hint.expected_object_size = expected_object_size;
 	op->alloc_hint.expected_write_size = expected_write_size;
@@ -786,7 +795,7 @@ struct ceph_osd_request *ceph_osdc_new_request(struct ceph_osd_client *osdc,
 	}
 
 	if (opcode == CEPH_OSD_OP_CREATE || opcode == CEPH_OSD_OP_DELETE) {
-		osd_req_op_init(req, which, opcode);
+		osd_req_op_init(req, which, opcode, 0);
 	} else {
 		u32 object_size = le32_to_cpu(layout->fl_object_size);
 		u32 object_base = off - objoff;
@@ -1088,7 +1097,7 @@ static void __move_osd_to_lru(struct ceph_osd_client *osdc,
 	BUG_ON(!list_empty(&osd->o_osd_lru));
 
 	list_add_tail(&osd->o_osd_lru, &osdc->osd_lru);
-	osd->lru_ttl = jiffies + osdc->client->options->osd_idle_ttl * HZ;
+	osd->lru_ttl = jiffies + osdc->client->options->osd_idle_ttl;
 }
 
 static void maybe_move_osd_to_lru(struct ceph_osd_client *osdc,
@@ -1199,7 +1208,7 @@ static struct ceph_osd *__lookup_osd(struct ceph_osd_client *osdc, int o)
 static void __schedule_osd_timeout(struct ceph_osd_client *osdc)
 {
 	schedule_delayed_work(&osdc->timeout_work,
-			osdc->client->options->osd_keepalive_timeout * HZ);
+			      osdc->client->options->osd_keepalive_timeout);
 }
 
 static void __cancel_osd_timeout(struct ceph_osd_client *osdc)
@@ -1567,10 +1576,9 @@ static void handle_timeout(struct work_struct *work)
 {
 	struct ceph_osd_client *osdc =
 		container_of(work, struct ceph_osd_client, timeout_work.work);
+	struct ceph_options *opts = osdc->client->options;
 	struct ceph_osd_request *req;
 	struct ceph_osd *osd;
-	unsigned long keepalive =
-		osdc->client->options->osd_keepalive_timeout * HZ;
 	struct list_head slow_osds;
 	dout("timeout\n");
 	down_read(&osdc->map_sem);
@@ -1586,7 +1594,8 @@ static void handle_timeout(struct work_struct *work)
 	 */
 	INIT_LIST_HEAD(&slow_osds);
 	list_for_each_entry(req, &osdc->req_lru, r_req_lru_item) {
-		if (time_before(jiffies, req->r_stamp + keepalive))
+		if (time_before(jiffies,
+				req->r_stamp + opts->osd_keepalive_timeout))
 			break;
 
 		osd = req->r_osd;
@@ -1613,8 +1622,7 @@ static void handle_osds_timeout(struct work_struct *work)
 	struct ceph_osd_client *osdc =
 		container_of(work, struct ceph_osd_client,
 			     osds_timeout_work.work);
-	unsigned long delay =
-		osdc->client->options->osd_idle_ttl * HZ >> 2;
+	unsigned long delay = osdc->client->options->osd_idle_ttl / 4;
 
 	dout("osds timeout\n");
 	down_read(&osdc->map_sem);
@@ -2619,7 +2627,7 @@ int ceph_osdc_init(struct ceph_osd_client *osdc, struct ceph_client *client)
 	osdc->event_count = 0;
 
 	schedule_delayed_work(&osdc->osds_timeout_work,
-	   round_jiffies_relative(osdc->client->options->osd_idle_ttl * HZ));
+	    round_jiffies_relative(osdc->client->options->osd_idle_ttl));
 
 	err = -ENOMEM;
 	osdc->req_mempool = mempool_create_kmalloc_pool(10,

net/ceph/osdmap.c

@@ -89,7 +89,7 @@ static int crush_decode_tree_bucket(void **p, void *end,
 {
 	int j;
 	dout("crush_decode_tree_bucket %p to %p\n", *p, end);
-	ceph_decode_32_safe(p, end, b->num_nodes, bad);
+	ceph_decode_8_safe(p, end, b->num_nodes, bad);
 	b->node_weights = kcalloc(b->num_nodes, sizeof(u32), GFP_NOFS);
 	if (b->node_weights == NULL)
 		return -ENOMEM;

net/ceph/pagevec.c

@@ -51,10 +51,7 @@ void ceph_put_page_vector(struct page **pages, int num_pages, bool dirty)
 			set_page_dirty_lock(pages[i]);
 		put_page(pages[i]);
 	}
-	if (is_vmalloc_addr(pages))
-		vfree(pages);
-	else
-		kfree(pages);
+	kvfree(pages);
 }
 EXPORT_SYMBOL(ceph_put_page_vector);