libceph: support for alloc hint flags

Allow indicating future I/O pattern via flags.  This is supported since
Kraken (and bluestore persists flags together with expected_object_size
and expected_write_size).

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Reviewed-by: Jason Dillaman <dillaman@redhat.com>

diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c
index 67d65ac..97e102e 100644 (file)
--- a/drivers/block/rbd.c
+++ b/drivers/block/rbd.c
@@ -2253,7 +2253,8 @@ static void __rbd_osd_setup_write_ops(struct ceph_osd_request *osd_req,
            !(obj_req->flags & RBD_OBJ_FLAG_MAY_EXIST)) {
                osd_req_op_alloc_hint_init(osd_req, which++,
                                           rbd_dev->layout.object_size,
-                                          rbd_dev->layout.object_size);
+                                          rbd_dev->layout.object_size,
+                                          0);
        }
 
        if (rbd_obj_is_entire(obj_req))

diff --git a/include/linux/ceph/osd_client.h b/include/linux/ceph/osd_client.h
index 671fb93..c60b59e 100644 (file)
--- a/include/linux/ceph/osd_client.h
+++ b/include/linux/ceph/osd_client.h
@@ -136,6 +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;
@@ -472,7 +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/rados.h b/include/linux/ceph/rados.h
index 88ed3c5..3a518fd 100644 (file)
--- a/include/linux/ceph/rados.h
+++ b/include/linux/ceph/rados.h
@@ -464,6 +464,19 @@ 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,
@@ -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/net/ceph/osd_client.c b/net/ceph/osd_client.c
index 22733e8..4fea3c3 100644 (file)
--- a/net/ceph/osd_client.c
+++ b/net/ceph/osd_client.c
@@ -932,10 +932,14 @@ static void osd_req_op_watch_init(struct ceph_osd_request *req, int which,
        op->watch.gen = 0;
 }
 
+/*
+ * @flags: CEPH_OSD_OP_ALLOC_HINT_FLAG_*
+ */
 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)
 {
        struct ceph_osd_req_op *op = _osd_req_op_init(osd_req, which,
                                                      CEPH_OSD_OP_SETALLOCHINT,
@@ -943,6 +947,7 @@ void osd_req_op_alloc_hint_init(struct ceph_osd_request *osd_req,
 
        op->alloc_hint.expected_object_size = expected_object_size;
        op->alloc_hint.expected_write_size = expected_write_size;
+       op->alloc_hint.flags = flags;
 
        /*
         * CEPH_OSD_OP_SETALLOCHINT op is advisory and therefore deemed
@@ -1018,6 +1023,7 @@ static u32 osd_req_encode_op(struct ceph_osd_op *dst,
                    cpu_to_le64(src->alloc_hint.expected_object_size);
                dst->alloc_hint.expected_write_size =
                    cpu_to_le64(src->alloc_hint.expected_write_size);
+               dst->alloc_hint.flags = cpu_to_le32(src->alloc_hint.flags);
                break;
        case CEPH_OSD_OP_SETXATTR:
        case CEPH_OSD_OP_CMPXATTR: