/*
- * Copryright (C) 2001 Cluster File Systems, Inc.
+ * GPL HEADER START
*
+ * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 only,
+ * as published by the Free Software Foundation.
+ *
+ * This program 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 version 2 for more details (a copy is included
+ * in the LICENSE file that accompanied this code).
+ *
+ * You should have received a copy of the GNU General Public License
+ * version 2 along with this program; If not, see
+ * http://www.sun.com/software/products/lustre/docs/GPLv2.pdf
+ *
+ * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
+ * CA 95054 USA or visit www.sun.com if you need additional information or
+ * have any questions.
+ *
+ * GPL HEADER END
+ */
+/*
+ * Copyright (c) 2002, 2010, Oracle and/or its affiliates. All rights reserved.
+ * Use is subject to license terms.
+ *
+ * Copyright (c) 2012, Whamcloud, Inc.
+ */
+/*
+ * This file is part of Lustre, http://www.lustre.org/
+ * Lustre is a trademark of Sun Microsystems, Inc.
*/
-#define EXPORT_SYMTAB
-
-#include <linux/config.h>
-#include <linux/module.h>
-#include <linux/kernel.h>
-#include <linux/mm.h>
-#include <linux/string.h>
-#include <linux/stat.h>
-#include <linux/errno.h>
-#include <linux/locks.h>
-#include <linux/unistd.h>
-
-#include <asm/system.h>
-#include <asm/uaccess.h>
-#include <linux/module.h>
-
-#include <linux/fs.h>
-#include <linux/stat.h>
-#include <asm/uaccess.h>
-#include <linux/vmalloc.h>
-#include <asm/segment.h>
-#include <linux/miscdevice.h>
-
-#include <linux/obd_support.h>
-#include <linux/lustre_lib.h>
-#include <linux/lustre_idl.h>
-#include <linux/lustre_mds.h>
-
-extern int mdc_reint(struct mds_request *request);
-extern struct mds_request *mds_prep_req(int opcode, int namelen, char *name, int tgtlen, char *tgt);
-
-int mdc_setattr(struct inode *inode, struct iattr *iattr,
- struct mds_rep **rep, struct mds_rep_hdr **hdr)
+#define DEBUG_SUBSYSTEM S_MDC
+
+#ifdef __KERNEL__
+#ifndef AUTOCONF_INCLUDED
+# include <linux/config.h>
+#endif
+# include <linux/module.h>
+# include <linux/kernel.h>
+#else
+# include <liblustre.h>
+#endif
+
+#include <obd_class.h>
+#include "mdc_internal.h"
+#include <lustre_fid.h>
+
+/* mdc_setattr does its own semaphore handling */
+static int mdc_reint(struct ptlrpc_request *request,
+ struct mdc_rpc_lock *rpc_lock,
+ int level)
+{
+ int rc;
+
+ request->rq_send_state = level;
+
+ mdc_get_rpc_lock(rpc_lock, NULL);
+ rc = ptlrpc_queue_wait(request);
+ mdc_put_rpc_lock(rpc_lock, NULL);
+ if (rc)
+ CDEBUG(D_INFO, "error in handling %d\n", rc);
+ else if (!req_capsule_server_get(&request->rq_pill, &RMF_MDT_BODY)) {
+ rc = -EPROTO;
+ }
+ return rc;
+}
+
+/* Find and cancel locally locks matched by inode @bits & @mode in the resource
+ * found by @fid. Found locks are added into @cancel list. Returns the amount of
+ * locks added to @cancels list. */
+int mdc_resource_get_unused(struct obd_export *exp, struct lu_fid *fid,
+ cfs_list_t *cancels, ldlm_mode_t mode,
+ __u64 bits)
+{
+ ldlm_policy_data_t policy = {{0}};
+ struct ldlm_res_id res_id;
+ struct ldlm_resource *res;
+ int count;
+ ENTRY;
+
+ fid_build_reg_res_name(fid, &res_id);
+ res = ldlm_resource_get(exp->exp_obd->obd_namespace,
+ NULL, &res_id, 0, 0);
+ if (res == NULL)
+ RETURN(0);
+ LDLM_RESOURCE_ADDREF(res);
+ /* Initialize ibits lock policy. */
+ policy.l_inodebits.bits = bits;
+ count = ldlm_cancel_resource_local(res, cancels, &policy,
+ mode, 0, 0, NULL);
+ LDLM_RESOURCE_DELREF(res);
+ ldlm_resource_putref(res);
+ RETURN(count);
+}
+
+static int mdc_prep_elc_req(struct obd_export *exp, struct ptlrpc_request *req,
+ cfs_list_t *cancels, int count)
+{
+ return ldlm_prep_elc_req(exp, req, LUSTRE_MDS_VERSION, MDS_REINT,
+ 0, cancels, count);
+}
+
+int mdc_setattr(struct obd_export *exp, struct md_op_data *op_data,
+ void *ea, int ealen, void *ea2, int ea2len,
+ struct ptlrpc_request **request, struct md_open_data **mod)
+{
+ CFS_LIST_HEAD(cancels);
+ struct ptlrpc_request *req;
+ struct mdc_rpc_lock *rpc_lock;
+ struct obd_device *obd = exp->exp_obd;
+ int count = 0, rc;
+ __u64 bits;
+ ENTRY;
+
+ LASSERT(op_data != NULL);
+
+ bits = MDS_INODELOCK_UPDATE;
+ if (op_data->op_attr.ia_valid & (ATTR_MODE|ATTR_UID|ATTR_GID))
+ bits |= MDS_INODELOCK_LOOKUP;
+ if ((op_data->op_flags & MF_MDC_CANCEL_FID1) &&
+ (fid_is_sane(&op_data->op_fid1)) &&
+ !OBD_FAIL_CHECK(OBD_FAIL_LDLM_BL_CALLBACK))
+ count = mdc_resource_get_unused(exp, &op_data->op_fid1,
+ &cancels, LCK_EX, bits);
+ req = ptlrpc_request_alloc(class_exp2cliimp(exp),
+ &RQF_MDS_REINT_SETATTR);
+ if (req == NULL) {
+ ldlm_lock_list_put(&cancels, l_bl_ast, count);
+ RETURN(-ENOMEM);
+ }
+ mdc_set_capa_size(req, &RMF_CAPA1, op_data->op_capa1);
+ if ((op_data->op_flags & (MF_SOM_CHANGE | MF_EPOCH_OPEN)) == 0)
+ req_capsule_set_size(&req->rq_pill, &RMF_MDT_EPOCH, RCL_CLIENT,
+ 0);
+ req_capsule_set_size(&req->rq_pill, &RMF_EADATA, RCL_CLIENT, ealen);
+ req_capsule_set_size(&req->rq_pill, &RMF_LOGCOOKIES, RCL_CLIENT,
+ ea2len);
+
+ rc = mdc_prep_elc_req(exp, req, &cancels, count);
+ if (rc) {
+ ptlrpc_request_free(req);
+ RETURN(rc);
+ }
+
+ rpc_lock = obd->u.cli.cl_rpc_lock;
+
+ if (op_data->op_attr.ia_valid & (ATTR_MTIME | ATTR_CTIME))
+ CDEBUG(D_INODE, "setting mtime "CFS_TIME_T
+ ", ctime "CFS_TIME_T"\n",
+ LTIME_S(op_data->op_attr.ia_mtime),
+ LTIME_S(op_data->op_attr.ia_ctime));
+ mdc_setattr_pack(req, op_data, ea, ealen, ea2, ea2len);
+
+ ptlrpc_request_set_replen(req);
+ if (mod && (op_data->op_flags & MF_EPOCH_OPEN) &&
+ req->rq_import->imp_replayable)
+ {
+ LASSERT(*mod == NULL);
+
+ *mod = obd_mod_alloc();
+ if (*mod == NULL) {
+ DEBUG_REQ(D_ERROR, req, "Can't allocate "
+ "md_open_data");
+ } else {
+ req->rq_replay = 1;
+ req->rq_cb_data = *mod;
+ (*mod)->mod_open_req = req;
+ req->rq_commit_cb = mdc_commit_open;
+ /**
+ * Take an extra reference on \var mod, it protects \var
+ * mod from being freed on eviction (commit callback is
+ * called despite rq_replay flag).
+ * Will be put on mdc_done_writing().
+ */
+ obd_mod_get(*mod);
+ }
+ }
+
+ rc = mdc_reint(req, rpc_lock, LUSTRE_IMP_FULL);
+
+ /* Save the obtained info in the original RPC for the replay case. */
+ if (rc == 0 && (op_data->op_flags & MF_EPOCH_OPEN)) {
+ struct mdt_ioepoch *epoch;
+ struct mdt_body *body;
+
+ epoch = req_capsule_client_get(&req->rq_pill, &RMF_MDT_EPOCH);
+ body = req_capsule_server_get(&req->rq_pill, &RMF_MDT_BODY);
+ LASSERT(epoch != NULL);
+ LASSERT(body != NULL);
+ epoch->handle = body->handle;
+ epoch->ioepoch = body->ioepoch;
+ req->rq_replay_cb = mdc_replay_open;
+ /** bug 3633, open may be committed and estale answer is not error */
+ } else if (rc == -ESTALE && (op_data->op_flags & MF_SOM_CHANGE)) {
+ rc = 0;
+ } else if (rc == -ERESTARTSYS) {
+ rc = 0;
+ }
+ *request = req;
+ if (rc && req->rq_commit_cb) {
+ /* Put an extra reference on \var mod on error case. */
+ obd_mod_put(*mod);
+ req->rq_commit_cb(req);
+ }
+ RETURN(rc);
+}
+
+int mdc_create(struct obd_export *exp, struct md_op_data *op_data,
+ const void *data, int datalen, int mode, __u32 uid, __u32 gid,
+ cfs_cap_t cap_effective, __u64 rdev,
+ struct ptlrpc_request **request)
+{
+ struct ptlrpc_request *req;
+ int level, rc;
+ int count, resends = 0;
+ struct obd_import *import = exp->exp_obd->u.cli.cl_import;
+ int generation = import->imp_generation;
+ CFS_LIST_HEAD(cancels);
+ ENTRY;
+
+ /* For case if upper layer did not alloc fid, do it now. */
+ if (!fid_is_sane(&op_data->op_fid2)) {
+ /*
+ * mdc_fid_alloc() may return errno 1 in case of switch to new
+ * sequence, handle this.
+ */
+ rc = mdc_fid_alloc(exp, &op_data->op_fid2, op_data);
+ if (rc < 0) {
+ CERROR("Can't alloc new fid, rc %d\n", rc);
+ RETURN(rc);
+ }
+ }
+
+rebuild:
+ count = 0;
+ if ((op_data->op_flags & MF_MDC_CANCEL_FID1) &&
+ (fid_is_sane(&op_data->op_fid1)))
+ count = mdc_resource_get_unused(exp, &op_data->op_fid1,
+ &cancels, LCK_EX,
+ MDS_INODELOCK_UPDATE);
+
+ req = ptlrpc_request_alloc(class_exp2cliimp(exp),
+ &RQF_MDS_REINT_CREATE_RMT_ACL);
+ if (req == NULL) {
+ ldlm_lock_list_put(&cancels, l_bl_ast, count);
+ RETURN(-ENOMEM);
+ }
+ mdc_set_capa_size(req, &RMF_CAPA1, op_data->op_capa1);
+ req_capsule_set_size(&req->rq_pill, &RMF_NAME, RCL_CLIENT,
+ op_data->op_namelen + 1);
+ req_capsule_set_size(&req->rq_pill, &RMF_EADATA, RCL_CLIENT,
+ data && datalen ? datalen : 0);
+
+ rc = mdc_prep_elc_req(exp, req, &cancels, count);
+ if (rc) {
+ ptlrpc_request_free(req);
+ RETURN(rc);
+ }
+
+ /*
+ * mdc_create_pack() fills msg->bufs[1] with name and msg->bufs[2] with
+ * tgt, for symlinks or lov MD data.
+ */
+ mdc_create_pack(req, op_data, data, datalen, mode, uid,
+ gid, cap_effective, rdev);
+
+ ptlrpc_request_set_replen(req);
+
+ if (resends) {
+ req->rq_generation_set = 1;
+ req->rq_import_generation = generation;
+ req->rq_sent = cfs_time_current_sec() + resends;
+ }
+ level = LUSTRE_IMP_FULL;
+ resend:
+ rc = mdc_reint(req, exp->exp_obd->u.cli.cl_rpc_lock, level);
+
+ /* Resend if we were told to. */
+ if (rc == -ERESTARTSYS) {
+ level = LUSTRE_IMP_RECOVER;
+ goto resend;
+ } else if (rc == -EINPROGRESS) {
+ /* Retry create infinitely until succeed or get other
+ * error code. */
+ ptlrpc_req_finished(req);
+ resends++;
+
+ CDEBUG(D_HA, "%s: resend:%d create on "DFID"/"DFID"\n",
+ exp->exp_obd->obd_name, resends,
+ PFID(&op_data->op_fid1), PFID(&op_data->op_fid2));
+
+ if (generation == import->imp_generation) {
+ goto rebuild;
+ } else {
+ CDEBUG(D_HA, "resend cross eviction\n");
+ RETURN(-EIO);
+ }
+ } else if (rc == 0) {
+ struct mdt_body *body;
+ struct lustre_capa *capa;
+
+ body = req_capsule_server_get(&req->rq_pill, &RMF_MDT_BODY);
+ LASSERT(body);
+ if (body->valid & OBD_MD_FLMDSCAPA) {
+ capa = req_capsule_server_get(&req->rq_pill,
+ &RMF_CAPA1);
+ if (capa == NULL)
+ rc = -EPROTO;
+ }
+ }
+
+ *request = req;
+ RETURN(rc);
+}
+
+int mdc_unlink(struct obd_export *exp, struct md_op_data *op_data,
+ struct ptlrpc_request **request)
{
- int rc;
- struct mds_request *request;
- struct mds_rec_setattr *rec;
-
- request = mds_prep_req(MDS_REINT, 0, NULL, sizeof(*rec), NULL);
- if (!request) {
- printk("mdc request: cannot pack\n");
- return -ENOMEM;
- }
-
- rec = mds_req_tgt(request->rq_req);
- mds_setattr_pack(rec, inode, iattr);
- request->rq_req->opcode = HTON__u32(REINT_SETATTR);
-
- rc = mdc_reint(request);
-
- if (rep) {
- *rep = request->rq_rep;
- }
- if (hdr) {
- *hdr = request->rq_rephdr;
- }
-
- kfree(request);
- return rc;
+ CFS_LIST_HEAD(cancels);
+ struct obd_device *obd = class_exp2obd(exp);
+ struct ptlrpc_request *req = *request;
+ int count = 0, rc;
+ ENTRY;
+
+ LASSERT(req == NULL);
+
+ if ((op_data->op_flags & MF_MDC_CANCEL_FID1) &&
+ (fid_is_sane(&op_data->op_fid1)) &&
+ !OBD_FAIL_CHECK(OBD_FAIL_LDLM_BL_CALLBACK))
+ count = mdc_resource_get_unused(exp, &op_data->op_fid1,
+ &cancels, LCK_EX,
+ MDS_INODELOCK_UPDATE);
+ if ((op_data->op_flags & MF_MDC_CANCEL_FID3) &&
+ (fid_is_sane(&op_data->op_fid3)) &&
+ !OBD_FAIL_CHECK(OBD_FAIL_LDLM_BL_CALLBACK))
+ count += mdc_resource_get_unused(exp, &op_data->op_fid3,
+ &cancels, LCK_EX,
+ MDS_INODELOCK_FULL);
+ req = ptlrpc_request_alloc(class_exp2cliimp(exp),
+ &RQF_MDS_REINT_UNLINK);
+ if (req == NULL) {
+ ldlm_lock_list_put(&cancels, l_bl_ast, count);
+ RETURN(-ENOMEM);
+ }
+ mdc_set_capa_size(req, &RMF_CAPA1, op_data->op_capa1);
+ req_capsule_set_size(&req->rq_pill, &RMF_NAME, RCL_CLIENT,
+ op_data->op_namelen + 1);
+
+ rc = mdc_prep_elc_req(exp, req, &cancels, count);
+ if (rc) {
+ ptlrpc_request_free(req);
+ RETURN(rc);
+ }
+
+ mdc_unlink_pack(req, op_data);
+
+ req_capsule_set_size(&req->rq_pill, &RMF_MDT_MD, RCL_SERVER,
+ obd->u.cli.cl_max_mds_easize);
+ req_capsule_set_size(&req->rq_pill, &RMF_LOGCOOKIES, RCL_SERVER,
+ obd->u.cli.cl_max_mds_cookiesize);
+ ptlrpc_request_set_replen(req);
+
+ *request = req;
+
+ rc = mdc_reint(req, obd->u.cli.cl_rpc_lock, LUSTRE_IMP_FULL);
+ if (rc == -ERESTARTSYS)
+ rc = 0;
+ RETURN(rc);
+}
+
+int mdc_link(struct obd_export *exp, struct md_op_data *op_data,
+ struct ptlrpc_request **request)
+{
+ CFS_LIST_HEAD(cancels);
+ struct obd_device *obd = exp->exp_obd;
+ struct ptlrpc_request *req;
+ int count = 0, rc;
+ ENTRY;
+
+ if ((op_data->op_flags & MF_MDC_CANCEL_FID2) &&
+ (fid_is_sane(&op_data->op_fid2)))
+ count = mdc_resource_get_unused(exp, &op_data->op_fid2,
+ &cancels, LCK_EX,
+ MDS_INODELOCK_UPDATE);
+ if ((op_data->op_flags & MF_MDC_CANCEL_FID1) &&
+ (fid_is_sane(&op_data->op_fid1)))
+ count += mdc_resource_get_unused(exp, &op_data->op_fid1,
+ &cancels, LCK_EX,
+ MDS_INODELOCK_UPDATE);
+
+ req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_MDS_REINT_LINK);
+ if (req == NULL) {
+ ldlm_lock_list_put(&cancels, l_bl_ast, count);
+ RETURN(-ENOMEM);
+ }
+ mdc_set_capa_size(req, &RMF_CAPA1, op_data->op_capa1);
+ mdc_set_capa_size(req, &RMF_CAPA2, op_data->op_capa2);
+ req_capsule_set_size(&req->rq_pill, &RMF_NAME, RCL_CLIENT,
+ op_data->op_namelen + 1);
+
+ rc = mdc_prep_elc_req(exp, req, &cancels, count);
+ if (rc) {
+ ptlrpc_request_free(req);
+ RETURN(rc);
+ }
+
+ mdc_link_pack(req, op_data);
+ ptlrpc_request_set_replen(req);
+
+ rc = mdc_reint(req, obd->u.cli.cl_rpc_lock, LUSTRE_IMP_FULL);
+ *request = req;
+ if (rc == -ERESTARTSYS)
+ rc = 0;
+
+ RETURN(rc);
}
-int mdc_create(struct inode *dir, const char *name, int namelen,
- int mode, __u64 id, __u32 uid, __u32 gid, __u64 time,
- struct mds_rep **rep, struct mds_rep_hdr **hdr)
+int mdc_rename(struct obd_export *exp, struct md_op_data *op_data,
+ const char *old, int oldlen, const char *new, int newlen,
+ struct ptlrpc_request **request)
{
- int rc;
- struct mds_request *request;
- struct mds_rec_create *rec;
-
- request = mds_prep_req(MDS_REINT, 0, NULL,
- sizeof(*rec) + size_round(namelen + 1),
- NULL);
- if (!request) {
- printk("mdc_create: cannot pack\n");
- return -ENOMEM;
- }
-
- rec = mds_req_tgt(request->rq_req);
- mds_create_pack(rec, dir, name, namelen, mode, id, uid, gid, time);
-
- rc = mdc_reint(request);
-
- if (rep) {
- *rep = request->rq_rep;
- }
- if (hdr) {
- *hdr = request->rq_rephdr;
- }
-
- kfree(request);
- return rc;
+ CFS_LIST_HEAD(cancels);
+ struct obd_device *obd = exp->exp_obd;
+ struct ptlrpc_request *req;
+ int count = 0, rc;
+ ENTRY;
+
+ if ((op_data->op_flags & MF_MDC_CANCEL_FID1) &&
+ (fid_is_sane(&op_data->op_fid1)))
+ count = mdc_resource_get_unused(exp, &op_data->op_fid1,
+ &cancels, LCK_EX,
+ MDS_INODELOCK_UPDATE);
+ if ((op_data->op_flags & MF_MDC_CANCEL_FID2) &&
+ (fid_is_sane(&op_data->op_fid2)))
+ count += mdc_resource_get_unused(exp, &op_data->op_fid2,
+ &cancels, LCK_EX,
+ MDS_INODELOCK_UPDATE);
+ if ((op_data->op_flags & MF_MDC_CANCEL_FID3) &&
+ (fid_is_sane(&op_data->op_fid3)))
+ count += mdc_resource_get_unused(exp, &op_data->op_fid3,
+ &cancels, LCK_EX,
+ MDS_INODELOCK_LOOKUP);
+ if ((op_data->op_flags & MF_MDC_CANCEL_FID4) &&
+ (fid_is_sane(&op_data->op_fid4)))
+ count += mdc_resource_get_unused(exp, &op_data->op_fid4,
+ &cancels, LCK_EX,
+ MDS_INODELOCK_FULL);
+
+ req = ptlrpc_request_alloc(class_exp2cliimp(exp),
+ &RQF_MDS_REINT_RENAME);
+ if (req == NULL) {
+ ldlm_lock_list_put(&cancels, l_bl_ast, count);
+ RETURN(-ENOMEM);
+ }
+
+ mdc_set_capa_size(req, &RMF_CAPA1, op_data->op_capa1);
+ mdc_set_capa_size(req, &RMF_CAPA2, op_data->op_capa2);
+ req_capsule_set_size(&req->rq_pill, &RMF_NAME, RCL_CLIENT, oldlen + 1);
+ req_capsule_set_size(&req->rq_pill, &RMF_SYMTGT, RCL_CLIENT, newlen+1);
+
+ rc = mdc_prep_elc_req(exp, req, &cancels, count);
+ if (rc) {
+ ptlrpc_request_free(req);
+ RETURN(rc);
+ }
+
+ if (exp_connect_cancelset(exp) && req)
+ ldlm_cli_cancel_list(&cancels, count, req, 0);
+
+ mdc_rename_pack(req, op_data, old, oldlen, new, newlen);
+
+ req_capsule_set_size(&req->rq_pill, &RMF_MDT_MD, RCL_SERVER,
+ obd->u.cli.cl_max_mds_easize);
+ req_capsule_set_size(&req->rq_pill, &RMF_LOGCOOKIES, RCL_SERVER,
+ obd->u.cli.cl_max_mds_cookiesize);
+ ptlrpc_request_set_replen(req);
+
+ rc = mdc_reint(req, obd->u.cli.cl_rpc_lock, LUSTRE_IMP_FULL);
+ *request = req;
+ if (rc == -ERESTARTSYS)
+ rc = 0;
+
+ RETURN(rc);
}
git://git.whamcloud.com / fs / lustre-release.git / blobdiff