diff options
Diffstat (limited to 'include/linux/ceph/osd_client.h')
| -rw-r--r-- | include/linux/ceph/osd_client.h | 136 |
1 files changed, 108 insertions, 28 deletions
diff --git a/include/linux/ceph/osd_client.h b/include/linux/ceph/osd_client.h index 83fa08a06507..50b14a5661c7 100644 --- a/include/linux/ceph/osd_client.h +++ b/include/linux/ceph/osd_client.h @@ -29,14 +29,63 @@ typedef void (*ceph_osdc_callback_t)(struct ceph_osd_request *); #define CEPH_HOMELESS_OSD -1 -/* a given osd we're communicating with */ +/* + * A single extent in a SPARSE_READ reply. + * + * Note that these come from the OSD as little-endian values. On BE arches, + * we convert them in-place after receipt. + */ +struct ceph_sparse_extent { + u64 off; + u64 len; +} __packed; + +/* Sparse read state machine state values */ +enum ceph_sparse_read_state { + CEPH_SPARSE_READ_HDR = 0, + CEPH_SPARSE_READ_EXTENTS, + CEPH_SPARSE_READ_DATA_LEN, + CEPH_SPARSE_READ_DATA_PRE, + CEPH_SPARSE_READ_DATA, +}; + +/* + * A SPARSE_READ reply is a 32-bit count of extents, followed by an array of + * 64-bit offset/length pairs, and then all of the actual file data + * concatenated after it (sans holes). + * + * Unfortunately, we don't know how long the extent array is until we've + * started reading the data section of the reply. The caller should send down + * a destination buffer for the array, but we'll alloc one if it's too small + * or if the caller doesn't. + */ +struct ceph_sparse_read { + enum ceph_sparse_read_state sr_state; /* state machine state */ + u64 sr_req_off; /* orig request offset */ + u64 sr_req_len; /* orig request length */ + u64 sr_pos; /* current pos in buffer */ + int sr_index; /* current extent index */ + u32 sr_datalen; /* length of actual data */ + u32 sr_count; /* extent count in reply */ + int sr_ext_len; /* length of extent array */ + struct ceph_sparse_extent *sr_extent; /* extent array */ +}; + +/* + * A given osd we're communicating with. + * + * Note that the o_requests tree can be searched while holding the "lock" mutex + * or the "o_requests_lock" spinlock. Insertion or removal requires both! + */ struct ceph_osd { refcount_t o_ref; + int o_sparse_op_idx; struct ceph_osd_client *o_osdc; int o_osd; int o_incarnation; struct rb_node o_node; struct ceph_connection o_con; + spinlock_t o_requests_lock; struct rb_root o_requests; struct rb_root o_linger_requests; struct rb_root o_backoff_mappings; @@ -46,6 +95,7 @@ struct ceph_osd { unsigned long lru_ttl; struct list_head o_keepalive_item; struct mutex lock; + struct ceph_sparse_read o_sparse_read; }; #define CEPH_OSD_SLAB_OPS 2 @@ -59,6 +109,7 @@ enum ceph_osd_data_type { CEPH_OSD_DATA_TYPE_BIO, #endif /* CONFIG_BLOCK */ CEPH_OSD_DATA_TYPE_BVECS, + CEPH_OSD_DATA_TYPE_ITER, }; struct ceph_osd_data { @@ -82,6 +133,7 @@ struct ceph_osd_data { struct ceph_bvec_iter bvec_pos; u32 num_bvecs; }; + struct iov_iter iter; }; }; @@ -98,6 +150,8 @@ struct ceph_osd_req_op { u64 offset, length; u64 truncate_size; u32 truncate_seq; + int sparse_ext_cnt; + struct ceph_sparse_extent *sparse_ext; struct ceph_osd_data osd_data; } extent; struct { @@ -145,6 +199,9 @@ struct ceph_osd_req_op { u32 src_fadvise_flags; struct ceph_osd_data osd_data; } copy_from; + struct { + u64 ver; + } assert_ver; }; }; @@ -199,6 +256,7 @@ struct ceph_osd_request { struct ceph_osd_client *r_osdc; struct kref r_kref; bool r_mempool; + bool r_linger; /* don't resend on failure */ struct completion r_completion; /* private to osd_client.c */ ceph_osdc_callback_t r_callback; @@ -211,9 +269,9 @@ struct ceph_osd_request { struct ceph_snap_context *r_snapc; /* for writes */ struct timespec64 r_mtime; /* ditto */ u64 r_data_offset; /* ditto */ - bool r_linger; /* don't resend on failure */ /* internal */ + u64 r_version; /* data version sent in reply */ unsigned long r_stamp; /* jiffies, send or check time */ unsigned long r_start_stamp; /* jiffies */ ktime_t r_start_latency; /* ktime_t */ @@ -221,7 +279,7 @@ struct ceph_osd_request { int r_attempts; u32 r_map_dne_bound; - struct ceph_osd_req_op r_ops[]; + struct ceph_osd_req_op r_ops[] __counted_by(r_num_ops); }; struct ceph_request_redirect { @@ -287,6 +345,9 @@ struct ceph_osd_linger_request { rados_watcherrcb_t errcb; void *data; + struct ceph_pagelist *request_pl; + struct page **notify_id_pages; + struct page ***preply_pages; size_t *preply_len; }; @@ -388,8 +449,6 @@ extern int ceph_osdc_init(struct ceph_osd_client *osdc, extern void ceph_osdc_stop(struct ceph_osd_client *osdc); extern void ceph_osdc_reopen_osds(struct ceph_osd_client *osdc); -extern void ceph_osdc_handle_reply(struct ceph_osd_client *osdc, - struct ceph_msg *msg); extern void ceph_osdc_handle_map(struct ceph_osd_client *osdc, struct ceph_msg *msg); void ceph_osdc_update_epoch_barrier(struct ceph_osd_client *osdc, u32 eb); @@ -431,9 +490,6 @@ extern void osd_req_op_extent_osd_data_pages(struct ceph_osd_request *, struct page **pages, u64 length, u32 alignment, bool pages_from_pool, bool own_pages); -extern void osd_req_op_extent_osd_data_pagelist(struct ceph_osd_request *, - unsigned int which, - struct ceph_pagelist *pagelist); #ifdef CONFIG_BLOCK void osd_req_op_extent_osd_data_bio(struct ceph_osd_request *osd_req, unsigned int which, @@ -447,10 +503,9 @@ void osd_req_op_extent_osd_data_bvecs(struct ceph_osd_request *osd_req, void osd_req_op_extent_osd_data_bvec_pos(struct ceph_osd_request *osd_req, unsigned int which, struct ceph_bvec_iter *bvec_pos); +void osd_req_op_extent_osd_iter(struct ceph_osd_request *osd_req, + unsigned int which, struct iov_iter *iter); -extern void osd_req_op_cls_request_data_pagelist(struct ceph_osd_request *, - unsigned int which, - struct ceph_pagelist *pagelist); extern void osd_req_op_cls_request_data_pages(struct ceph_osd_request *, unsigned int which, struct page **pages, u64 length, @@ -475,6 +530,14 @@ extern void osd_req_op_alloc_hint_init(struct ceph_osd_request *osd_req, u64 expected_object_size, u64 expected_write_size, u32 flags); +extern int osd_req_op_copy_from_init(struct ceph_osd_request *req, + u64 src_snapid, u64 src_version, + struct ceph_object_id *src_oid, + struct ceph_object_locator *src_oloc, + u32 src_fadvise_flags, + u32 dst_fadvise_flags, + u32 truncate_seq, u64 truncate_size, + u8 copy_from_flags); extern struct ceph_osd_request *ceph_osdc_alloc_request(struct ceph_osd_client *osdc, struct ceph_snap_context *snapc, @@ -493,12 +556,28 @@ extern struct ceph_osd_request *ceph_osdc_new_request(struct ceph_osd_client *, u32 truncate_seq, u64 truncate_size, bool use_mempool); +int __ceph_alloc_sparse_ext_map(struct ceph_osd_req_op *op, int cnt); + +/* + * How big an extent array should we preallocate for a sparse read? This is + * just a starting value. If we get more than this back from the OSD, the + * receiver will reallocate. + */ +#define CEPH_SPARSE_EXT_ARRAY_INITIAL 16 + +static inline int ceph_alloc_sparse_ext_map(struct ceph_osd_req_op *op, int cnt) +{ + if (!cnt) + cnt = CEPH_SPARSE_EXT_ARRAY_INITIAL; + + return __ceph_alloc_sparse_ext_map(op, cnt); +} + extern void ceph_osdc_get_request(struct ceph_osd_request *req); extern void ceph_osdc_put_request(struct ceph_osd_request *req); -extern int ceph_osdc_start_request(struct ceph_osd_client *osdc, - struct ceph_osd_request *req, - bool nofail); +void ceph_osdc_start_request(struct ceph_osd_client *osdc, + struct ceph_osd_request *req); extern void ceph_osdc_cancel_request(struct ceph_osd_request *req); extern int ceph_osdc_wait_request(struct ceph_osd_client *osdc, struct ceph_osd_request *req); @@ -515,17 +594,6 @@ int ceph_osdc_call(struct ceph_osd_client *osdc, struct page *req_page, size_t req_len, struct page **resp_pages, size_t *resp_len); -int ceph_osdc_copy_from(struct ceph_osd_client *osdc, - u64 src_snapid, u64 src_version, - struct ceph_object_id *src_oid, - struct ceph_object_locator *src_oloc, - u32 src_fadvise_flags, - struct ceph_object_id *dst_oid, - struct ceph_object_locator *dst_oloc, - u32 dst_fadvise_flags, - u32 truncate_seq, u64 truncate_size, - u8 copy_from_flags); - /* watch/notify */ struct ceph_osd_linger_request * ceph_osdc_watch(struct ceph_osd_client *osdc, @@ -552,12 +620,24 @@ int ceph_osdc_notify(struct ceph_osd_client *osdc, u32 timeout, struct page ***preply_pages, size_t *preply_len); -int ceph_osdc_watch_check(struct ceph_osd_client *osdc, - struct ceph_osd_linger_request *lreq); int ceph_osdc_list_watchers(struct ceph_osd_client *osdc, struct ceph_object_id *oid, struct ceph_object_locator *oloc, struct ceph_watch_item **watchers, u32 *num_watchers); -#endif +/* Find offset into the buffer of the end of the extent map */ +static inline u64 ceph_sparse_ext_map_end(struct ceph_osd_req_op *op) +{ + struct ceph_sparse_extent *ext; + + /* No extents? No data */ + if (op->extent.sparse_ext_cnt == 0) + return 0; + + ext = &op->extent.sparse_ext[op->extent.sparse_ext_cnt - 1]; + + return ext->off + ext->len - op->extent.offset; +} + +#endif |