Merge tag 'ceph-for-5.8-rc1' of git://github.com/ceph/ceph-client

Pull ceph updates from Ilya Dryomov:
 "The highlights are:

   - OSD/MDS latency and caps cache metrics infrastructure for the
     filesytem (Xiubo Li). Currently available through debugfs and will
     be periodically sent to the MDS in the future.

   - support for replica reads (balanced and localized reads) for rbd
     and the filesystem (myself). The default remains to always read
     from primary, users can opt-in with the new crush_location and
     read_from_replica options. Note that reading from replica is safe
     for general use only since Octopus.

   - support for RADOS allocation hint flags (myself). Currently used by
     rbd to propagate the compressible/incompressible hint given with
     the new compression_hint map option and ready for passing on more
     advanced hints, e.g. based on fadvise() from the filesystem.

   - support for efficient cross-quota-realm renames (Luis Henriques)

   - assorted cap handling improvements and cleanups, particularly
     untangling some of the locking (Jeff Layton)"

* tag 'ceph-for-5.8-rc1' of git://github.com/ceph/ceph-client: (29 commits)
  rbd: compression_hint option
  libceph: support for alloc hint flags
  libceph: read_from_replica option
  libceph: support for balanced and localized reads
  libceph: crush_location infrastructure
  libceph: decode CRUSH device/bucket types and names
  libceph: add non-asserting rbtree insertion helper
  ceph: skip checking caps when session reconnecting and releasing reqs
  ceph: make sure mdsc->mutex is nested in s->s_mutex to fix dead lock
  ceph: don't return -ESTALE if there's still an open file
  libceph, rbd: replace zero-length array with flexible-array
  ceph: allow rename operation under different quota realms
  ceph: normalize 'delta' parameter usage in check_quota_exceeded
  ceph: ceph_kick_flushing_caps needs the s_mutex
  ceph: request expedited service on session's last cap flush
  ceph: convert mdsc->cap_dirty to a per-session list
  ceph: reset i_requested_max_size if file write is not wanted
  ceph: throw a warning if we destroy session with mutex still locked
  ceph: fix potential race in ceph_check_caps
  ceph: document what protects i_dirty_item and i_flushing_item
  ...

6 files changed, 64 insertions, 6 deletions

diff --git a/include/linux/ceph/libceph.h b/include/linux/ceph/libceph.h
index 525b7c3f1c81..2247e71beb83 100644
--- a/include/linux/ceph/libceph.h
+++ b/include/linux/ceph/libceph.h
@@ -53,6 +53,8 @@ struct ceph_options {
 	unsigned long osd_keepalive_timeout;	/* jiffies */
 	unsigned long osd_request_timeout;	/* jiffies */
 
+	u32 osd_req_flags;  /* CEPH_OSD_FLAG_*, applied to each OSD request */
+
 	/*
 	 * any type that can't be simply compared or doesn't need
 	 * to be compared should go beyond this point,
@@ -64,6 +66,7 @@ struct ceph_options {
 	int num_mon;
 	char *name;
 	struct ceph_crypto_key *key;
+	struct rb_root crush_locs;
 };
 
 /*
@@ -188,7 +191,7 @@ static inline int calc_pages_for(u64 off, u64 len)
 #define RB_CMP3WAY(a, b) ((a) < (b) ? -1 : (a) > (b))
 
 #define DEFINE_RB_INSDEL_FUNCS2(name, type, keyfld, cmpexp, keyexp, nodefld) \
-static void insert_##name(struct rb_root *root, type *t)		\
+static bool __insert_##name(struct rb_root *root, type *t)		\
 {									\
 	struct rb_node **n = &root->rb_node;				\
 	struct rb_node *parent = NULL;					\
@@ -206,11 +209,17 @@ static void insert_##name(struct rb_root *root, type *t)		\
 		else if (cmp > 0)					\
 			n = &(*n)->rb_right;				\
 		else							\
-			BUG();						\
+			return false;					\
 	}								\
 									\
 	rb_link_node(&t->nodefld, parent, n);				\
 	rb_insert_color(&t->nodefld, root);				\
+	return true;							\
+}									\
+static void __maybe_unused insert_##name(struct rb_root *root, type *t)	\
+{									\
+	if (!__insert_##name(root, t))					\
+		BUG();							\
 }									\
 static void erase_##name(struct rb_root *root, type *t)			\
 {									\
diff --git a/include/linux/ceph/mon_client.h b/include/linux/ceph/mon_client.h
index dbb8a6959a73..ce4ffeb384d7 100644
--- a/include/linux/ceph/mon_client.h
+++ b/include/linux/ceph/mon_client.h
@@ -19,7 +19,7 @@ struct ceph_monmap {
 	struct ceph_fsid fsid;
 	u32 epoch;
 	u32 num_mon;
-	struct ceph_entity_inst mon_inst[0];
+	struct ceph_entity_inst mon_inst[];
 };
 
 struct ceph_mon_client;
diff --git a/include/linux/ceph/osd_client.h b/include/linux/ceph/osd_client.h
index 9d9f745b98a1..c60b59e9291b 100644
--- a/include/linux/ceph/osd_client.h
+++ b/include/linux/ceph/osd_client.h
@@ -8,6 +8,7 @@
 #include <linux/mempool.h>
 #include <linux/rbtree.h>
 #include <linux/refcount.h>
+#include <linux/ktime.h>
 
 #include <linux/ceph/types.h>
 #include <linux/ceph/osdmap.h>
@@ -135,6 +136,7 @@ struct ceph_osd_req_op {
 		struct {
 			u64 expected_object_size;
 			u64 expected_write_size;
+			u32 flags;  /* CEPH_OSD_OP_ALLOC_HINT_FLAG_* */
 		} alloc_hint;
 		struct {
 			u64 snapid;
@@ -164,6 +166,7 @@ struct ceph_osd_request_target {
 	bool recovery_deletes;
 
 	unsigned int flags;                /* CEPH_OSD_FLAG_* */
+	bool used_replica;
 	bool paused;
 
 	u32 epoch;
@@ -213,6 +216,8 @@ struct ceph_osd_request {
 	/* internal */
 	unsigned long r_stamp;                /* jiffies, send or check time */
 	unsigned long r_start_stamp;          /* jiffies */
+	ktime_t r_start_latency;              /* ktime_t */
+	ktime_t r_end_latency;                /* ktime_t */
 	int r_attempts;
 	u32 r_map_dne_bound;
 
@@ -468,7 +473,8 @@ extern int osd_req_op_xattr_init(struct ceph_osd_request *osd_req, unsigned int
 extern void osd_req_op_alloc_hint_init(struct ceph_osd_request *osd_req,
 				       unsigned int which,
 				       u64 expected_object_size,
-				       u64 expected_write_size);
+				       u64 expected_write_size,
+				       u32 flags);
 
 extern struct ceph_osd_request *ceph_osdc_alloc_request(struct ceph_osd_client *osdc,
 					       struct ceph_snap_context *snapc,
diff --git a/include/linux/ceph/osdmap.h b/include/linux/ceph/osdmap.h
index 5e601975745f..3f4498fef6ad 100644
--- a/include/linux/ceph/osdmap.h
+++ b/include/linux/ceph/osdmap.h
@@ -302,9 +302,26 @@ bool ceph_pg_to_primary_shard(struct ceph_osdmap *osdmap,
 int ceph_pg_to_acting_primary(struct ceph_osdmap *osdmap,
 			      const struct ceph_pg *raw_pgid);
 
+struct crush_loc {
+	char *cl_type_name;
+	char *cl_name;
+};
+
+struct crush_loc_node {
+	struct rb_node cl_node;
+	struct crush_loc cl_loc;  /* pointers into cl_data */
+	char cl_data[];
+};
+
+int ceph_parse_crush_location(char *crush_location, struct rb_root *locs);
+int ceph_compare_crush_locs(struct rb_root *locs1, struct rb_root *locs2);
+void ceph_clear_crush_locs(struct rb_root *locs);
+
+int ceph_get_crush_locality(struct ceph_osdmap *osdmap, int id,
+			    struct rb_root *locs);
+
 extern struct ceph_pg_pool_info *ceph_pg_pool_by_id(struct ceph_osdmap *map,
 						    u64 id);
-
 extern const char *ceph_pg_pool_name_by_id(struct ceph_osdmap *map, u64 id);
 extern int ceph_pg_poolid_by_name(struct ceph_osdmap *map, const char *name);
 u64 ceph_pg_pool_flags(struct ceph_osdmap *map, u64 id);
diff --git a/include/linux/ceph/rados.h b/include/linux/ceph/rados.h
index 88ed3c5c04c5..3a518fd0eaad 100644
--- a/include/linux/ceph/rados.h
+++ b/include/linux/ceph/rados.h
@@ -465,6 +465,19 @@ enum {
 const char *ceph_osd_watch_op_name(int o);
 
 enum {
+	CEPH_OSD_ALLOC_HINT_FLAG_SEQUENTIAL_WRITE = 1,
+	CEPH_OSD_ALLOC_HINT_FLAG_RANDOM_WRITE = 2,
+	CEPH_OSD_ALLOC_HINT_FLAG_SEQUENTIAL_READ = 4,
+	CEPH_OSD_ALLOC_HINT_FLAG_RANDOM_READ = 8,
+	CEPH_OSD_ALLOC_HINT_FLAG_APPEND_ONLY = 16,
+	CEPH_OSD_ALLOC_HINT_FLAG_IMMUTABLE = 32,
+	CEPH_OSD_ALLOC_HINT_FLAG_SHORTLIVED = 64,
+	CEPH_OSD_ALLOC_HINT_FLAG_LONGLIVED = 128,
+	CEPH_OSD_ALLOC_HINT_FLAG_COMPRESSIBLE = 256,
+	CEPH_OSD_ALLOC_HINT_FLAG_INCOMPRESSIBLE = 512,
+};
+
+enum {
 	CEPH_OSD_BACKOFF_OP_BLOCK = 1,
 	CEPH_OSD_BACKOFF_OP_ACK_BLOCK = 2,
 	CEPH_OSD_BACKOFF_OP_UNBLOCK = 3,
@@ -517,6 +530,7 @@ struct ceph_osd_op {
 		struct {
 			__le64 expected_object_size;
 			__le64 expected_write_size;
+			__le32 flags;  /* CEPH_OSD_OP_ALLOC_HINT_FLAG_* */
 		} __attribute__ ((packed)) alloc_hint;
 		struct {
 			__le64 snapid;
diff --git a/include/linux/crush/crush.h b/include/linux/crush/crush.h
index 54741295c70b..33c16f2de7f6 100644
--- a/include/linux/crush/crush.h
+++ b/include/linux/crush/crush.h
@@ -87,7 +87,7 @@ struct crush_rule_mask {
 struct crush_rule {
 	__u32 len;
 	struct crush_rule_mask mask;
-	struct crush_rule_step steps[0];
+	struct crush_rule_step steps[];
 };
 
 #define crush_rule_size(len) (sizeof(struct crush_rule) + \
@@ -301,6 +301,12 @@ struct crush_map {
 
 	__u32 *choose_tries;
 #else
+	/* device/bucket type id -> type name (CrushWrapper::type_map) */
+	struct rb_root type_names;
+
+	/* device/bucket id -> name (CrushWrapper::name_map) */
+	struct rb_root names;
+
 	/* CrushWrapper::choose_args */
 	struct rb_root choose_args;
 #endif
@@ -342,4 +348,10 @@ struct crush_work {
 	struct crush_work_bucket **work; /* Per-bucket working store */
 };
 
+#ifdef __KERNEL__
+/* osdmap.c */
+void clear_crush_names(struct rb_root *root);
+void clear_choose_args(struct crush_map *c);
+#endif
+
 #endif