--- /dev/null
+/* -*- mode: c; c-basic-offset: 8; indent-tabs-mode: nil; -*-
+ * vim:expandtab:shiftwidth=8:tabstop=8:
+ *
+ * Copyright (c) 2002 Cluster File Systems, Inc. <info@clusterfs.com>
+ *
+ * This file is part of Lustre, http://www.lustre.org.
+ *
+ * Lustre is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ *
+ * Lustre is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with Lustre; if not, write to the Free Software
+ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#define DEBUG_SUBSYSTEM S_CMOBD
+
+#include <linux/version.h>
+#include <linux/init.h>
+#include <linux/obd_support.h>
+#include <linux/lustre_lib.h>
+#include <linux/lustre_net.h>
+#include <linux/lustre_idl.h>
+#include <linux/obd_class.h>
+#include <linux/lustre_log.h>
+#include <linux/lustre_cmobd.h>
+#include <linux/lustre_fsfilt.h>
+#include <linux/lustre_smfs.h>
+
+#include "cm_internal.h"
+
+void lov_free_memmd(struct lov_stripe_md **lsmp);
+
+int lov_alloc_memmd(struct lov_stripe_md **lsmp, int stripe_count,
+ int pattern);
+
+int smfs_rec_unpack(struct smfs_proc_args *args, char *record,
+ char **pbuf, int *opcode);
+
+/* helper functions for cmobd to construct pseudo lsm */
+int cmobd_dummy_lsm(struct lov_stripe_md **lsmp, int stripe_cnt,
+ struct obdo *oa, __u32 stripe_size)
+{
+ int i, rc;
+ ENTRY;
+
+ rc = lov_alloc_memmd(lsmp, stripe_cnt, LOV_PATTERN_CMOBD);
+ if (rc < 0)
+ RETURN(rc);
+
+ for (i = 0; i < stripe_cnt; i++) {
+ (*lsmp)->lsm_oinfo[i].loi_id = oa->o_id;
+ (*lsmp)->lsm_object_id = oa->o_id;
+ if (oa->o_valid & OBD_MD_FLGROUP) {
+ (*lsmp)->lsm_oinfo[i].loi_gr = oa->o_gr;
+ (*lsmp)->lsm_object_gr = oa->o_gr;
+ }
+ (*lsmp)->lsm_oinfo[i].loi_ost_idx = i;
+ (*lsmp)->lsm_stripe_size = stripe_size;
+ }
+ RETURN(0);
+}
+
+void cmobd_free_lsm(struct lov_stripe_md **lsmp)
+{
+ ENTRY;
+ lov_free_memmd(lsmp);
+ EXIT;
+}
+
+/* reintegration functions */
+static int cmobd_setattr_reint(struct obd_device *obd, void *rec)
+{
+ struct cm_obd *cmobd = &obd->u.cm;
+ struct lov_obd *lov = &cmobd->master_obd->u.lov;
+ struct obd_export *exp = cmobd->master_exp;
+ struct lov_stripe_md *lsm;
+ struct obdo *oa = (struct obdo*)rec;
+ int rc;
+ ENTRY;
+
+ rc = cmobd_dummy_lsm(&lsm, lov->desc.ld_tgt_count, oa,
+ (__u32)lov->desc.ld_default_stripe_size);
+ if (rc)
+ GOTO(out, rc);
+
+ rc = obd_setattr(exp, oa, lsm, NULL);
+
+ cmobd_free_lsm(&lsm);
+out:
+ RETURN(rc);
+}
+
+static int cmobd_create_reint(struct obd_device *obd, void *rec)
+{
+ struct cm_obd *cmobd = &obd->u.cm;
+ struct lov_obd *lov = &cmobd->master_obd->u.lov;
+ struct obd_export *exp = cmobd->master_exp;
+ struct lov_stripe_md *lsm;
+ struct obd_trans_info oti = { 0 };
+ struct obdo *oa=(struct obdo*)rec;
+ int rc;
+ ENTRY;
+
+ rc = cmobd_dummy_lsm(&lsm, lov->desc.ld_tgt_count, oa,
+ (__u32)lov->desc.ld_default_stripe_size);
+ if (rc)
+ GOTO(out, rc);
+ if (cmobd->master_group != oa->o_gr) {
+ int group = oa->o_gr;
+ int valsize = sizeof(group);
+ rc = obd_set_info(exp, strlen("mds_conn"), "mds_conn",
+ valsize, &group);
+ if (rc)
+ GOTO(out, rc = -EINVAL);
+ cmobd->master_group = oa->o_gr;
+ }
+ rc = obd_create(exp, oa, &lsm, &oti);
+
+ cmobd_free_lsm(&lsm);
+out:
+ RETURN(rc);
+}
+
+/* direct cut-n-paste of filter_blocking_ast() */
+static int cache_blocking_ast(struct ldlm_lock *lock,
+ struct ldlm_lock_desc *desc,
+ void *data, int flag)
+{
+ int do_ast;
+ ENTRY;
+
+ if (flag == LDLM_CB_CANCELING) {
+ /* Don't need to do anything here. */
+ RETURN(0);
+ }
+
+ /* XXX layering violation! -phil */
+ l_lock(&lock->l_resource->lr_namespace->ns_lock);
+ /* Get this: if filter_blocking_ast is racing with ldlm_intent_policy,
+ * such that filter_blocking_ast is called just before l_i_p takes the
+ * ns_lock, then by the time we get the lock, we might not be the
+ * correct blocking function anymore. So check, and return early, if
+ * so. */
+ if (lock->l_blocking_ast != cache_blocking_ast) {
+ l_unlock(&lock->l_resource->lr_namespace->ns_lock);
+ RETURN(0);
+ }
+
+ lock->l_flags |= LDLM_FL_CBPENDING;
+ do_ast = (!lock->l_readers && !lock->l_writers);
+ l_unlock(&lock->l_resource->lr_namespace->ns_lock);
+
+ if (do_ast) {
+ struct lustre_handle lockh;
+ int rc;
+
+ LDLM_DEBUG(lock, "already unused, calling ldlm_cli_cancel");
+ ldlm_lock2handle(lock, &lockh);
+ rc = ldlm_cli_cancel(&lockh);
+ if (rc < 0)
+ CERROR("ldlm_cli_cancel: %d\n", rc);
+ } else {
+ LDLM_DEBUG(lock, "Lock still has references, will be "
+ "cancelled later");
+ }
+ RETURN(0);
+}
+
+static int master_blocking_ast(struct ldlm_lock *lock,
+ struct ldlm_lock_desc *desc,
+ void *data, int flag)
+{
+ int rc;
+ struct lustre_handle lockh;
+ ENTRY;
+
+ switch (flag) {
+ case LDLM_CB_BLOCKING:
+ ldlm_lock2handle(lock, &lockh);
+ rc = ldlm_cli_cancel(&lockh);
+ if (rc < 0) {
+ CDEBUG(D_INODE, "ldlm_cli_cancel: %d\n", rc);
+ RETURN(rc);
+ }
+ break;
+ case LDLM_CB_CANCELING:
+ /* do nothing here by now */
+ break;
+ default:
+ LBUG();
+ }
+ RETURN(0);
+}
+
+static int cmobd_write_extents(struct obd_device *obd, struct obdo *oa,
+ struct ldlm_extent *extent)
+{
+ struct cm_obd *cmobd = &obd->u.cm;
+ struct obd_device *cache = cmobd->cache_obd;
+ struct lov_obd *lov = &cmobd->master_obd->u.lov;
+ struct ldlm_res_id res_id;
+ ldlm_policy_data_t policy;
+ struct lustre_handle lockh_src = { 0 };
+ struct lustre_handle lockh_dst = { 0 };
+ struct lov_stripe_md *lsm;
+ int flags = 0, err, rc = 0;
+ ENTRY;
+
+ /* XXX for debug write replay without smfs and kml */
+ res_id.name[0]= oa->o_id;
+ res_id.name[1]= oa->o_gr;
+ policy.l_extent.start = extent->start;
+ policy.l_extent.end = extent->end;
+
+ /* get extent read lock on the source replay file */
+ rc = ldlm_cli_enqueue(NULL, NULL, cache->obd_namespace, res_id,
+ LDLM_EXTENT, &policy, LCK_PR,
+ &flags, cache_blocking_ast, ldlm_completion_ast,
+ NULL, NULL, NULL, 0, NULL, &lockh_src);
+ if (rc != ELDLM_OK)
+ RETURN(rc);
+
+ /* construct the pseudo lsm */
+ rc = cmobd_dummy_lsm(&lsm, lov->desc.ld_tgt_count, oa,
+ (__u32)lov->desc.ld_default_stripe_size);
+ if (rc)
+ GOTO(out_lock, rc);
+
+ rc = obd_enqueue(cmobd->master_exp, lsm, LDLM_EXTENT, &policy,
+ LCK_PW, &flags, master_blocking_ast,
+ ldlm_completion_ast, NULL,
+ NULL, 0, NULL, &lockh_dst);
+ if (rc != ELDLM_OK)
+ GOTO(out_lsm, rc);
+
+ err = cmobd_replay_write(obd, oa, &policy.l_extent);
+
+ rc = obd_cancel(cmobd->master_exp, lsm, LCK_PW, &lockh_dst);
+ if (rc)
+ GOTO(out_lsm, rc);
+ /* XXX in fact, I just want to cancel the only lockh_dst
+ * instantly. */
+ rc = obd_cancel_unused(cmobd->master_exp, lsm, 0, NULL);
+ if (err)
+ rc = err;
+out_lsm:
+ cmobd_free_lsm(&lsm);
+out_lock:
+ ldlm_lock_decref(&lockh_src, LCK_PR);
+ RETURN(rc);
+}
+
+static int cmobd_write_reint(struct obd_device *obd, void *rec)
+{
+ struct cm_obd *cmobd = &obd->u.cm;
+ struct obd_device *cache = cmobd->cache_obd;
+ struct obdo *oa = (struct obdo *)rec;
+ struct ldlm_extent *extent = NULL;
+ unsigned long csb, ino;
+ char *extents_buf = NULL;
+ int size = 0, rc = 0, ext_num = 0;
+ ENTRY;
+
+ size = sizeof(csb);
+
+ obd_get_info(cmobd->cache_exp, strlen("cache_sb") + 1,
+ "cache_sb", &size, &csb);
+
+ ino = *(int*)(&oa->o_inline[0]);
+ rc = fsfilt_get_ino_write_extents(cache, (struct super_block *)csb, ino,
+ &extents_buf, &ext_num);
+ if (rc)
+ GOTO(out, rc);
+ extent = (struct ldlm_extent *)extents_buf;
+ size = ext_num;
+ while (extent && size --) {
+ rc = cmobd_write_extents(obd, oa, extent);
+ if (rc)
+ GOTO(out, rc);
+ extent ++;
+ }
+out:
+ if (extents_buf)
+ fsfilt_free_write_extents(cache, (struct super_block *)csb,
+ ino, extents_buf, ext_num);
+ RETURN(rc);
+}
+
+int cmobd_reint_oss(struct obd_device *obd, void *record, int opcode)
+{
+ switch (opcode) {
+ case OST_CREATE:
+ return cmobd_create_reint(obd, record);
+ case OST_SETATTR:
+ return cmobd_setattr_reint(obd, record);
+ case OST_WRITE:
+ return cmobd_write_reint(obd, record);
+ default:
+ CERROR("unrecognized oss reint opcode %d\n",
+ opcode);
+ return -EINVAL;
+ }
+}
git://git.whamcloud.com / fs / lustre-release.git / blobdiff