Whamcloud - gitweb
LU-1303 lod: transfer default striping from parent/fs
[fs/lustre-release.git] / lustre / fid / fid_request.c
index 4266358..bc0f46b 100644 (file)
@@ -1,6 +1,4 @@
-/* -*- mode: c; c-basic-offset: 8; indent-tabs-mode: nil; -*-
- * vim:expandtab:shiftwidth=8:tabstop=8:
- *
+/*
  * GPL HEADER START
  *
  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
  * GPL HEADER END
  */
 /*
- * Copyright  2008 Sun Microsystems, Inc. All rights reserved
+ * Copyright (c) 2007, 2010, Oracle and/or its affiliates. All rights reserved.
  * Use is subject to license terms.
+ *
+ * Copyright (c) 2011, 2012, Whamcloud, Inc.
  */
 /*
  * This file is part of Lustre, http://www.lustre.org/
@@ -40,9 +40,6 @@
  * Author: Yury Umanets <umka@clusterfs.com>
  */
 
-#ifndef EXPORT_SYMTAB
-# define EXPORT_SYMTAB
-#endif
 #define DEBUG_SUBSYSTEM S_FID
 
 #ifdef __KERNEL__
 #include <lustre_mdc.h>
 #include "fid_internal.h"
 
-static int seq_client_rpc(struct lu_client_seq *seq, struct lu_range *input,
-                          struct lu_range *output, __u32 opc,
+static int seq_client_rpc(struct lu_client_seq *seq,
+                          struct lu_seq_range *output, __u32 opc,
                           const char *opcname)
 {
-        struct obd_export     *exp = seq->lcs_exp;
-        struct ptlrpc_request *req;
-        struct lu_range       *out, *in;
-        __u32                 *op;
-        int                    rc;
-        ENTRY;
-
-        req = ptlrpc_request_alloc_pack(class_exp2cliimp(exp), &RQF_SEQ_QUERY,
-                                        LUSTRE_MDS_VERSION, SEQ_QUERY);
-        if (req == NULL)
-                RETURN(-ENOMEM);
-
-        /* Init operation code */
-        op = req_capsule_client_get(&req->rq_pill, &RMF_SEQ_OPC);
-        *op = opc;
-
-        /* Zero out input range, this is not recovery yet. */
-        in = req_capsule_client_get(&req->rq_pill, &RMF_SEQ_RANGE);
-        if (input != NULL)
-                *in = *input;
-        else
-                range_zero(in);
-
-        ptlrpc_request_set_replen(req);
-
-        if (seq->lcs_type == LUSTRE_SEQ_METADATA) {
-                req->rq_request_portal = (opc == SEQ_ALLOC_SUPER) ?
-                        SEQ_CONTROLLER_PORTAL : SEQ_METADATA_PORTAL;
-        } else {
-                req->rq_request_portal = (opc == SEQ_ALLOC_SUPER) ?
-                        SEQ_CONTROLLER_PORTAL : SEQ_DATA_PORTAL;
-        }
-        ptlrpc_at_set_req_timeout(req);
-
-        mdc_get_rpc_lock(exp->exp_obd->u.cli.cl_rpc_lock, NULL);
-        rc = ptlrpc_queue_wait(req);
-        mdc_put_rpc_lock(exp->exp_obd->u.cli.cl_rpc_lock, NULL);
-
-        if (rc)
-                GOTO(out_req, rc);
-
-        out = req_capsule_server_get(&req->rq_pill, &RMF_SEQ_RANGE);
-        *output = *out;
-
-        if (!range_is_sane(output)) {
-                CERROR("%s: Invalid range received from server: "
-                       DRANGE"\n", seq->lcs_name, PRANGE(output));
-                GOTO(out_req, rc = -EINVAL);
-        }
-
-        if (range_is_exhausted(output)) {
-                CERROR("%s: Range received from server is exhausted: "
-                       DRANGE"]\n", seq->lcs_name, PRANGE(output));
-                GOTO(out_req, rc = -EINVAL);
-        }
-        *in = *out;
-
-        CDEBUG(D_INFO, "%s: Allocated %s-sequence "DRANGE"]\n",
-               seq->lcs_name, opcname, PRANGE(output));
-
-        EXIT;
+       struct obd_export     *exp = seq->lcs_exp;
+       struct ptlrpc_request *req;
+       struct lu_seq_range   *out, *in;
+       __u32                 *op;
+       unsigned int           debug_mask;
+       int                    rc;
+       ENTRY;
+
+       req = ptlrpc_request_alloc_pack(class_exp2cliimp(exp), &RQF_SEQ_QUERY,
+                                       LUSTRE_MDS_VERSION, SEQ_QUERY);
+       if (req == NULL)
+               RETURN(-ENOMEM);
+
+       /* Init operation code */
+       op = req_capsule_client_get(&req->rq_pill, &RMF_SEQ_OPC);
+       *op = opc;
+
+       /* Zero out input range, this is not recovery yet. */
+       in = req_capsule_client_get(&req->rq_pill, &RMF_SEQ_RANGE);
+       range_init(in);
+
+       ptlrpc_request_set_replen(req);
+
+       if (seq->lcs_type == LUSTRE_SEQ_METADATA) {
+               req->rq_request_portal = SEQ_METADATA_PORTAL;
+               in->lsr_flags = LU_SEQ_RANGE_MDT;
+       } else {
+               LASSERTF(seq->lcs_type == LUSTRE_SEQ_DATA,
+                        "unknown lcs_type %u\n", seq->lcs_type);
+               req->rq_request_portal = SEQ_DATA_PORTAL;
+               in->lsr_flags = LU_SEQ_RANGE_OST;
+       }
+
+       if (opc == SEQ_ALLOC_SUPER) {
+               /* Update index field of *in, it is required for
+                * FLD update on super sequence allocator node. */
+               in->lsr_index = seq->lcs_space.lsr_index;
+               req->rq_request_portal = SEQ_CONTROLLER_PORTAL;
+               debug_mask = D_CONSOLE;
+       } else {
+               debug_mask = D_INFO;
+               LASSERTF(opc == SEQ_ALLOC_META,
+                        "unknown opcode %u\n, opc", opc);
+       }
+
+       ptlrpc_at_set_req_timeout(req);
+
+       mdc_get_rpc_lock(exp->exp_obd->u.cli.cl_rpc_lock, NULL);
+       rc = ptlrpc_queue_wait(req);
+       mdc_put_rpc_lock(exp->exp_obd->u.cli.cl_rpc_lock, NULL);
+
+       if (rc)
+               GOTO(out_req, rc);
+
+       out = req_capsule_server_get(&req->rq_pill, &RMF_SEQ_RANGE);
+       *output = *out;
+
+       if (!range_is_sane(output)) {
+               CERROR("%s: Invalid range received from server: "
+                      DRANGE"\n", seq->lcs_name, PRANGE(output));
+               GOTO(out_req, rc = -EINVAL);
+       }
+
+       if (range_is_exhausted(output)) {
+               CERROR("%s: Range received from server is exhausted: "
+                      DRANGE"]\n", seq->lcs_name, PRANGE(output));
+               GOTO(out_req, rc = -EINVAL);
+       }
+
+       CDEBUG_LIMIT(debug_mask, "%s: Allocated %s-sequence "DRANGE"]\n",
+                    seq->lcs_name, opcname, PRANGE(output));
+
+       EXIT;
 out_req:
-        ptlrpc_req_finished(req);
-        return rc;
+       ptlrpc_req_finished(req);
+       return rc;
 }
 
 /* Request sequence-controller node to allocate new super-sequence. */
-int seq_client_replay_super(struct lu_client_seq *seq,
-                            struct lu_range *range,
-                            const struct lu_env *env)
+int seq_client_alloc_super(struct lu_client_seq *seq,
+                           const struct lu_env *env)
 {
         int rc;
         ENTRY;
 
-        down(&seq->lcs_sem);
+        cfs_mutex_lock(&seq->lcs_mutex);
 
 #ifdef __KERNEL__
         if (seq->lcs_srv) {
                 LASSERT(env != NULL);
-                rc = seq_server_alloc_super(seq->lcs_srv, range,
-                                            &seq->lcs_space, env);
+                rc = seq_server_alloc_super(seq->lcs_srv, &seq->lcs_space,
+                                            env);
         } else {
 #endif
-                rc = seq_client_rpc(seq, range, &seq->lcs_space,
+                rc = seq_client_rpc(seq, &seq->lcs_space,
                                     SEQ_ALLOC_SUPER, "super");
 #ifdef __KERNEL__
         }
 #endif
-        up(&seq->lcs_sem);
+        cfs_mutex_unlock(&seq->lcs_mutex);
         RETURN(rc);
 }
 
-/* Request sequence-controller node to allocate new super-sequence. */
-int seq_client_alloc_super(struct lu_client_seq *seq,
-                           const struct lu_env *env)
-{
-        ENTRY;
-        RETURN(seq_client_replay_super(seq, NULL, env));
-}
-
 /* Request sequence-controller node to allocate new meta-sequence. */
-static int seq_client_alloc_meta(struct lu_client_seq *seq,
-                                 const struct lu_env *env)
+static int seq_client_alloc_meta(const struct lu_env *env,
+                                 struct lu_client_seq *seq)
 {
         int rc;
         ENTRY;
@@ -177,11 +177,10 @@ static int seq_client_alloc_meta(struct lu_client_seq *seq,
 #ifdef __KERNEL__
         if (seq->lcs_srv) {
                 LASSERT(env != NULL);
-                rc = seq_server_alloc_meta(seq->lcs_srv, NULL,
-                                           &seq->lcs_space, env);
+                rc = seq_server_alloc_meta(seq->lcs_srv, &seq->lcs_space, env);
         } else {
 #endif
-                rc = seq_client_rpc(seq, NULL, &seq->lcs_space,
+                rc = seq_client_rpc(seq, &seq->lcs_space,
                                     SEQ_ALLOC_META, "meta");
 #ifdef __KERNEL__
         }
@@ -190,7 +189,8 @@ static int seq_client_alloc_meta(struct lu_client_seq *seq,
 }
 
 /* Allocate new sequence for client. */
-static int seq_client_alloc_seq(struct lu_client_seq *seq, seqno_t *seqnr)
+static int seq_client_alloc_seq(const struct lu_env *env,
+                                struct lu_client_seq *seq, seqno_t *seqnr)
 {
         int rc;
         ENTRY;
@@ -198,9 +198,9 @@ static int seq_client_alloc_seq(struct lu_client_seq *seq, seqno_t *seqnr)
         LASSERT(range_is_sane(&seq->lcs_space));
 
         if (range_is_exhausted(&seq->lcs_space)) {
-                rc = seq_client_alloc_meta(seq, NULL);
+                rc = seq_client_alloc_meta(env, seq);
                 if (rc) {
-                        CERROR("%s: Can't allocate new meta-sequence, "
+                        CERROR("%s: Can't allocate new meta-sequence,"
                                "rc %d\n", seq->lcs_name, rc);
                         RETURN(rc);
                 } else {
@@ -212,8 +212,8 @@ static int seq_client_alloc_seq(struct lu_client_seq *seq, seqno_t *seqnr)
         }
 
         LASSERT(!range_is_exhausted(&seq->lcs_space));
-        *seqnr = seq->lcs_space.lr_start;
-        seq->lcs_space.lr_start += 1;
+        *seqnr = seq->lcs_space.lsr_start;
+        seq->lcs_space.lsr_start += 1;
 
         CDEBUG(D_INFO, "%s: Allocated sequence ["LPX64"]\n", seq->lcs_name,
                *seqnr);
@@ -221,27 +221,118 @@ static int seq_client_alloc_seq(struct lu_client_seq *seq, seqno_t *seqnr)
         RETURN(rc);
 }
 
+static int seq_fid_alloc_prep(struct lu_client_seq *seq,
+                              cfs_waitlink_t *link)
+{
+        if (seq->lcs_update) {
+                cfs_waitq_add(&seq->lcs_waitq, link);
+                cfs_set_current_state(CFS_TASK_UNINT);
+                cfs_mutex_unlock(&seq->lcs_mutex);
+
+                cfs_waitq_wait(link, CFS_TASK_UNINT);
+
+                cfs_mutex_lock(&seq->lcs_mutex);
+                cfs_waitq_del(&seq->lcs_waitq, link);
+                cfs_set_current_state(CFS_TASK_RUNNING);
+                return -EAGAIN;
+        }
+        ++seq->lcs_update;
+        cfs_mutex_unlock(&seq->lcs_mutex);
+        return 0;
+}
+
+static void seq_fid_alloc_fini(struct lu_client_seq *seq)
+{
+        LASSERT(seq->lcs_update == 1);
+        cfs_mutex_lock(&seq->lcs_mutex);
+        --seq->lcs_update;
+        cfs_waitq_signal(&seq->lcs_waitq);
+}
+
+/* Allocate the whole seq to the caller*/
+int seq_client_get_seq(const struct lu_env *env,
+                       struct lu_client_seq *seq, seqno_t *seqnr)
+{
+        cfs_waitlink_t link;
+        int rc;
+
+        LASSERT(seqnr != NULL);
+        cfs_mutex_lock(&seq->lcs_mutex);
+        cfs_waitlink_init(&link);
+
+        while (1) {
+                rc = seq_fid_alloc_prep(seq, &link);
+                if (rc == 0)
+                        break;
+        }
+
+        rc = seq_client_alloc_seq(env, seq, seqnr);
+        if (rc) {
+                CERROR("%s: Can't allocate new sequence, "
+                       "rc %d\n", seq->lcs_name, rc);
+                seq_fid_alloc_fini(seq);
+                cfs_mutex_unlock(&seq->lcs_mutex);
+                return rc;
+        }
+
+        CDEBUG(D_INFO, "%s: allocate sequence "
+               "[0x%16.16"LPF64"x]\n", seq->lcs_name, *seqnr);
+
+        /*Since the caller require the whole seq,
+         *so marked this seq to be used*/
+        seq->lcs_fid.f_oid = LUSTRE_SEQ_MAX_WIDTH;
+        seq->lcs_fid.f_seq = *seqnr;
+        seq->lcs_fid.f_ver = 0;
+
+        /*
+         * Inform caller that sequence switch is performed to allow it
+         * to setup FLD for it.
+         */
+        seq_fid_alloc_fini(seq);
+        cfs_mutex_unlock(&seq->lcs_mutex);
+
+        return rc;
+}
+EXPORT_SYMBOL(seq_client_get_seq);
+
 /* Allocate new fid on passed client @seq and save it to @fid. */
-int seq_client_alloc_fid(struct lu_client_seq *seq, struct lu_fid *fid)
+int seq_client_alloc_fid(const struct lu_env *env,
+                         struct lu_client_seq *seq, struct lu_fid *fid)
 {
+        cfs_waitlink_t link;
         int rc;
         ENTRY;
 
         LASSERT(seq != NULL);
         LASSERT(fid != NULL);
 
-        down(&seq->lcs_sem);
+        cfs_waitlink_init(&link);
+        cfs_mutex_lock(&seq->lcs_mutex);
 
-        if (fid_is_zero(&seq->lcs_fid) ||
-            fid_oid(&seq->lcs_fid) >= seq->lcs_width)
-        {
+       if (OBD_FAIL_CHECK(OBD_FAIL_SEQ_EXHAUST))
+               seq->lcs_fid.f_oid = seq->lcs_width;
+
+        while (1) {
                 seqno_t seqnr;
 
-                rc = seq_client_alloc_seq(seq, &seqnr);
+                if (!fid_is_zero(&seq->lcs_fid) &&
+                    fid_oid(&seq->lcs_fid) < seq->lcs_width) {
+                        /* Just bump last allocated fid and return to caller. */
+                        seq->lcs_fid.f_oid += 1;
+                        rc = 0;
+                        break;
+                }
+
+                rc = seq_fid_alloc_prep(seq, &link);
+                if (rc)
+                        continue;
+
+                rc = seq_client_alloc_seq(env, seq, &seqnr);
                 if (rc) {
                         CERROR("%s: Can't allocate new sequence, "
                                "rc %d\n", seq->lcs_name, rc);
-                        up(&seq->lcs_sem);
+                        seq_fid_alloc_fini(seq);
+                        cfs_mutex_unlock(&seq->lcs_mutex);
                         RETURN(rc);
                 }
 
@@ -257,14 +348,13 @@ int seq_client_alloc_fid(struct lu_client_seq *seq, struct lu_fid *fid)
                  * to setup FLD for it.
                  */
                 rc = 1;
-        } else {
-                /* Just bump last allocated fid and return to caller. */
-                seq->lcs_fid.f_oid += 1;
-                rc = 0;
+
+                seq_fid_alloc_fini(seq);
+                break;
         }
 
         *fid = seq->lcs_fid;
-        up(&seq->lcs_sem);
+        cfs_mutex_unlock(&seq->lcs_mutex);
 
         CDEBUG(D_INFO, "%s: Allocated FID "DFID"\n", seq->lcs_name,  PFID(fid));
         RETURN(rc);
@@ -277,11 +367,34 @@ EXPORT_SYMBOL(seq_client_alloc_fid);
  */
 void seq_client_flush(struct lu_client_seq *seq)
 {
+        cfs_waitlink_t link;
+
         LASSERT(seq != NULL);
-        down(&seq->lcs_sem);
+        cfs_waitlink_init(&link);
+        cfs_mutex_lock(&seq->lcs_mutex);
+
+        while (seq->lcs_update) {
+                cfs_waitq_add(&seq->lcs_waitq, &link);
+                cfs_set_current_state(CFS_TASK_UNINT);
+                cfs_mutex_unlock(&seq->lcs_mutex);
+
+                cfs_waitq_wait(&link, CFS_TASK_UNINT);
+
+                cfs_mutex_lock(&seq->lcs_mutex);
+                cfs_waitq_del(&seq->lcs_waitq, &link);
+                cfs_set_current_state(CFS_TASK_RUNNING);
+        }
+
         fid_zero(&seq->lcs_fid);
-        range_zero(&seq->lcs_space);
-        up(&seq->lcs_sem);
+        /**
+         * this id shld not be used for seq range allocation.
+         * set to -1 for dgb check.
+         */
+
+        seq->lcs_space.lsr_index = -1;
+
+        range_init(&seq->lcs_space);
+        cfs_mutex_unlock(&seq->lcs_mutex);
 }
 EXPORT_SYMBOL(seq_client_flush);
 
@@ -356,8 +469,9 @@ int seq_client_init(struct lu_client_seq *seq,
         seq->lcs_exp = exp;
         seq->lcs_srv = srv;
         seq->lcs_type = type;
-        sema_init(&seq->lcs_sem, 1);
+        cfs_mutex_init(&seq->lcs_mutex);
         seq->lcs_width = LUSTRE_SEQ_MAX_WIDTH;
+        cfs_waitq_init(&seq->lcs_waitq);
 
         /* Make sure that things are clear before work is started. */
         seq_client_flush(seq);