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

Pull Ceph updates from Sage Weil:
 "There is the long-awaited discard support for RBD (Guangliang Zhao,
  Josh Durgin), a pile of RBD bug fixes that didn't belong in late -rc's
  (Ilya Dryomov, Li RongQing), a pile of fs/ceph bug fixes and
  performance and debugging improvements (Yan, Zheng, John Spray), and a
  smattering of cleanups (Chao Yu, Fabian Frederick, Joe Perches)"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client: (40 commits)
  ceph: fix divide-by-zero in __validate_layout()
  rbd: rbd workqueues need a resque worker
  libceph: ceph-msgr workqueue needs a resque worker
  ceph: fix bool assignments
  libceph: separate multiple ops with commas in debugfs output
  libceph: sync osd op definitions in rados.h
  libceph: remove redundant declaration
  ceph: additional debugfs output
  ceph: export ceph_session_state_name function
  ceph: include the initial ACL in create/mkdir/mknod MDS requests
  ceph: use pagelist to present MDS request data
  libceph: reference counting pagelist
  ceph: fix llistxattr on symlink
  ceph: send client metadata to MDS
  ceph: remove redundant code for max file size verification
  ceph: remove redundant io_iter_advance()
  ceph: move ceph_find_inode() outside the s_mutex
  ceph: request xattrs if xattr_version is zero
  rbd: set the remaining discard properties to enable support
  rbd: use helpers to handle discard for layered images correctly
  ...

1 files changed, 276 insertions, 120 deletions

diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c
index 4b97baf8afa3..0a54c588e433 100644
--- a/drivers/block/rbd.c
+++ b/drivers/block/rbd.c
@@ -210,6 +210,12 @@ enum obj_request_type {
         OBJ_REQUEST_NODATA, OBJ_REQUEST_BIO, OBJ_REQUEST_PAGES
 };
 
+enum obj_operation_type {
+        OBJ_OP_WRITE,
+        OBJ_OP_READ,
+        OBJ_OP_DISCARD,
+};
+
 enum obj_req_flags {
         OBJ_REQ_DONE,           /* completion flag: not done = 0, done = 1 */
         OBJ_REQ_IMG_DATA,       /* object usage: standalone = 0, image = 1 */
@@ -276,6 +282,7 @@ enum img_req_flags {
         IMG_REQ_WRITE,          /* I/O direction: read = 0, write = 1 */
         IMG_REQ_CHILD,          /* initiator: block = 0, child image = 1 */
         IMG_REQ_LAYERED,        /* ENOENT handling: normal = 0, layered = 1 */
+        IMG_REQ_DISCARD,        /* discard: normal = 0, discard request = 1 */
 };
 
 struct rbd_img_request {
@@ -785,6 +792,20 @@ static int parse_rbd_opts_token(char *c, void *private)
         return 0;
 }
 
+static char* obj_op_name(enum obj_operation_type op_type)
+{
+        switch (op_type) {
+        case OBJ_OP_READ:
+                return "read";
+        case OBJ_OP_WRITE:
+                return "write";
+        case OBJ_OP_DISCARD:
+                return "discard";
+        default:
+                return "???";
+        }
+}
+
 /*
  * Get a ceph client with specific addr and configuration, if one does
  * not exist create it.  Either way, ceph_opts is consumed by this
@@ -1600,6 +1621,21 @@ static bool img_request_write_test(struct rbd_img_request *img_request)
         return test_bit(IMG_REQ_WRITE, &img_request->flags) != 0;
 }
 
+/*
+ * Set the discard flag when the img_request is an discard request
+ */
+static void img_request_discard_set(struct rbd_img_request *img_request)
+{
+        set_bit(IMG_REQ_DISCARD, &img_request->flags);
+        smp_mb();
+}
+
+static bool img_request_discard_test(struct rbd_img_request *img_request)
+{
+        smp_mb();
+        return test_bit(IMG_REQ_DISCARD, &img_request->flags) != 0;
+}
+
 static void img_request_child_set(struct rbd_img_request *img_request)
 {
         set_bit(IMG_REQ_CHILD, &img_request->flags);
@@ -1636,6 +1672,17 @@ static bool img_request_layered_test(struct rbd_img_request *img_request)
         return test_bit(IMG_REQ_LAYERED, &img_request->flags) != 0;
 }
 
+static enum obj_operation_type
+rbd_img_request_op_type(struct rbd_img_request *img_request)
+{
+        if (img_request_write_test(img_request))
+                return OBJ_OP_WRITE;
+        else if (img_request_discard_test(img_request))
+                return OBJ_OP_DISCARD;
+        else
+                return OBJ_OP_READ;
+}
+
 static void
 rbd_img_obj_request_read_callback(struct rbd_obj_request *obj_request)
 {
@@ -1722,6 +1769,21 @@ static void rbd_osd_write_callback(struct rbd_obj_request *obj_request)
         obj_request_done_set(obj_request);
 }
 
+static void rbd_osd_discard_callback(struct rbd_obj_request *obj_request)
+{
+        dout("%s: obj %p result %d %llu\n", __func__, obj_request,
+                obj_request->result, obj_request->length);
+        /*
+         * There is no such thing as a successful short discard.  Set
+         * it to our originally-requested length.
+         */
+        obj_request->xferred = obj_request->length;
+        /* discarding a non-existent object is not a problem */
+        if (obj_request->result == -ENOENT)
+                obj_request->result = 0;
+        obj_request_done_set(obj_request);
+}
+
 /*
  * For a simple stat call there's nothing to do.  We'll do more if
  * this is part of a write sequence for a layered image.
@@ -1773,6 +1835,11 @@ static void rbd_osd_req_callback(struct ceph_osd_request *osd_req,
         case CEPH_OSD_OP_STAT:
                 rbd_osd_stat_callback(obj_request);
                 break;
+        case CEPH_OSD_OP_DELETE:
+        case CEPH_OSD_OP_TRUNCATE:
+        case CEPH_OSD_OP_ZERO:
+                rbd_osd_discard_callback(obj_request);
+                break;
         case CEPH_OSD_OP_CALL:
         case CEPH_OSD_OP_NOTIFY_ACK:
         case CEPH_OSD_OP_WATCH:
@@ -1823,7 +1890,7 @@ static void rbd_osd_req_format_write(struct rbd_obj_request *obj_request)
  */
 static struct ceph_osd_request *rbd_osd_req_create(
                                         struct rbd_device *rbd_dev,
-                                        bool write_request,
+                                        enum obj_operation_type op_type,
                                         unsigned int num_ops,
                                         struct rbd_obj_request *obj_request)
 {
@@ -1831,16 +1898,18 @@ static struct ceph_osd_request *rbd_osd_req_create(
         struct ceph_osd_client *osdc;
         struct ceph_osd_request *osd_req;
 
-        if (obj_request_img_data_test(obj_request)) {
+        if (obj_request_img_data_test(obj_request) &&
+                (op_type == OBJ_OP_DISCARD || op_type == OBJ_OP_WRITE)) {
                 struct rbd_img_request *img_request = obj_request->img_request;
-
-                rbd_assert(write_request ==
-                                img_request_write_test(img_request));
-                if (write_request)
-                        snapc = img_request->snapc;
+                if (op_type == OBJ_OP_WRITE) {
+                        rbd_assert(img_request_write_test(img_request));
+                } else {
+                        rbd_assert(img_request_discard_test(img_request));
+                }
+                snapc = img_request->snapc;
         }
 
-        rbd_assert(num_ops == 1 || (write_request && num_ops == 2));
+        rbd_assert(num_ops == 1 || ((op_type == OBJ_OP_WRITE) && num_ops == 2));
 
         /* Allocate and initialize the request, for the num_ops ops */
 
@@ -1850,7 +1919,7 @@ static struct ceph_osd_request *rbd_osd_req_create(
         if (!osd_req)
                 return NULL;    /* ENOMEM */
 
-        if (write_request)
+        if (op_type == OBJ_OP_WRITE || op_type == OBJ_OP_DISCARD)
                 osd_req->r_flags = CEPH_OSD_FLAG_WRITE | CEPH_OSD_FLAG_ONDISK;
         else
                 osd_req->r_flags = CEPH_OSD_FLAG_READ;
@@ -1865,9 +1934,10 @@ static struct ceph_osd_request *rbd_osd_req_create(
 }