X-Git-Url: https://git.whamcloud.com/?a=blobdiff_plain;f=lustre%2Flmv%2Flmv_obd.c;h=792cfd5aa19eaef64941cb465cfaa349bc3b646a;hb=b9cbb29935faa388d36b16140c7ef76a47754bec;hp=888a86f4fee02d1b44b2fc4c1f9e0985df70fe74;hpb=bfa6343413c5c6423e96b6db387ae616d756a27f;p=fs%2Flustre-release.git diff --git a/lustre/lmv/lmv_obd.c b/lustre/lmv/lmv_obd.c index 888a86f..792cfd5 100644 --- a/lustre/lmv/lmv_obd.c +++ b/lustre/lmv/lmv_obd.c @@ -27,7 +27,7 @@ * Copyright (c) 2004, 2010, Oracle and/or its affiliates. All rights reserved. * Use is subject to license terms. * - * Copyright (c) 2011, 2013, Intel Corporation. + * Copyright (c) 2011, 2014, Intel Corporation. */ /* * This file is part of Lustre, http://www.lustre.org/ @@ -35,28 +35,31 @@ */ #define DEBUG_SUBSYSTEM S_LMV -#ifdef __KERNEL__ #include #include #include +#include +#ifdef HAVE_UIDGID_HEADER +# include +#endif #include #include #include -#include +#include #include #include -#else -#include -#endif #include #include #include #include #include +#include #include -#include +#include #include +#include +#include #include "lmv_internal.h" static void lmv_activate_target(struct lmv_obd *lmv, @@ -68,6 +71,8 @@ static void lmv_activate_target(struct lmv_obd *lmv, tgt->ltd_active = activate; lmv->desc.ld_active_tgt_count += (activate ? 1 : -1); + + tgt->ltd_exp->exp_obd->obd_inactive = !activate; } /** @@ -204,9 +209,6 @@ static int lmv_connect(const struct lu_env *env, struct obd_uuid *cluuid, struct obd_connect_data *data, void *localdata) { -#ifdef __KERNEL__ - struct proc_dir_entry *lmv_proc_dir; -#endif struct lmv_obd *lmv = &obd->u.lmv; struct lustre_handle conn = { 0 }; int rc = 0; @@ -238,21 +240,18 @@ static int lmv_connect(const struct lu_env *env, if (data) lmv->conn_data = *data; -#ifdef __KERNEL__ - if (obd->obd_proc_private != NULL) { - lmv_proc_dir = obd->obd_proc_private; - } else { - lmv_proc_dir = lprocfs_seq_register("target_obds", - obd->obd_proc_entry, - NULL, NULL); - if (IS_ERR(lmv_proc_dir)) { - CERROR("could not register /proc/fs/lustre/%s/%s/target_obds.", - obd->obd_type->typ_name, obd->obd_name); - lmv_proc_dir = NULL; + if (lmv->targets_proc_entry == NULL) { + lmv->targets_proc_entry = lprocfs_register("target_obds", + obd->obd_proc_entry, + NULL, NULL); + if (IS_ERR(lmv->targets_proc_entry)) { + CERROR("%s: cannot register " + "/proc/fs/lustre/%s/%s/target_obds\n", + obd->obd_name, obd->obd_type->typ_name, + obd->obd_name); + lmv->targets_proc_entry = NULL; } - obd->obd_proc_private = lmv_proc_dir; } -#endif /* * All real clients should perform actual connection right away, because @@ -263,40 +262,13 @@ static int lmv_connect(const struct lu_env *env, if (data != NULL && (data->ocd_connect_flags & OBD_CONNECT_REAL)) rc = lmv_check_connect(obd); -#ifdef __KERNEL__ - if (rc && lmv_proc_dir) { - lprocfs_remove(&lmv_proc_dir); - obd->obd_proc_private = NULL; - } -#endif - RETURN(rc); -} - -static void lmv_set_timeouts(struct obd_device *obd) -{ - struct lmv_obd *lmv; - __u32 i; - - lmv = &obd->u.lmv; - if (lmv->server_timeout == 0) - return; - - if (lmv->connected == 0) - return; - - for (i = 0; i < lmv->desc.ld_tgt_count; i++) { - struct lmv_tgt_desc *tgt = lmv->tgts[i]; - - if (tgt == NULL || tgt->ltd_exp == NULL || !tgt->ltd_active) - continue; - - obd_set_info_async(NULL, tgt->ltd_exp, sizeof(KEY_INTERMDS), - KEY_INTERMDS, 0, NULL, NULL); - } + if (rc && lmv->targets_proc_entry != NULL) + lprocfs_remove(&lmv->targets_proc_entry); + RETURN(rc); } -static int lmv_init_ea_size(struct obd_export *exp, int easize, - int def_easize, int cookiesize) +static int lmv_init_ea_size(struct obd_export *exp, __u32 easize, + __u32 def_easize) { struct obd_device *obd = exp->exp_obd; struct lmv_obd *lmv = &obd->u.lmv; @@ -313,15 +285,12 @@ static int lmv_init_ea_size(struct obd_export *exp, int easize, lmv->max_def_easize = def_easize; change = 1; } - if (lmv->max_cookiesize < cookiesize) { - lmv->max_cookiesize = cookiesize; - change = 1; - } - if (change == 0) - RETURN(0); - if (lmv->connected == 0) - RETURN(0); + if (change == 0) + RETURN(0); + + if (lmv->connected == 0) + RETURN(0); for (i = 0; i < lmv->desc.ld_tgt_count; i++) { struct lmv_tgt_desc *tgt = lmv->tgts[i]; @@ -331,11 +300,10 @@ static int lmv_init_ea_size(struct obd_export *exp, int easize, continue; } - rc = md_init_ea_size(tgt->ltd_exp, easize, def_easize, - cookiesize); + rc = md_init_ea_size(tgt->ltd_exp, easize, def_easize); if (rc) { CERROR("%s: obd_init_ea_size() failed on MDT target %d:" - " rc = %d.\n", obd->obd_name, i, rc); + " rc = %d\n", obd->obd_name, i, rc); break; } } @@ -346,9 +314,6 @@ static int lmv_init_ea_size(struct obd_export *exp, int easize, int lmv_connect_mdc(struct obd_device *obd, struct lmv_tgt_desc *tgt) { -#ifdef __KERNEL__ - struct proc_dir_entry *lmv_proc_dir; -#endif struct lmv_obd *lmv = &obd->u.lmv; struct obd_uuid *cluuid = &lmv->cluuid; struct obd_uuid lmv_mdc_uuid = { "LMV_MDC_UUID" }; @@ -416,39 +381,33 @@ int lmv_connect_mdc(struct obd_device *obd, struct lmv_tgt_desc *tgt) } } - tgt->ltd_active = 1; - tgt->ltd_exp = mdc_exp; - lmv->desc.ld_active_tgt_count++; + tgt->ltd_active = 1; + tgt->ltd_exp = mdc_exp; + lmv->desc.ld_active_tgt_count++; - md_init_ea_size(tgt->ltd_exp, lmv->max_easize, - lmv->max_def_easize, lmv->max_cookiesize); + md_init_ea_size(tgt->ltd_exp, lmv->max_easize, lmv->max_def_easize); - CDEBUG(D_CONFIG, "Connected to %s(%s) successfully (%d)\n", - mdc_obd->obd_name, mdc_obd->obd_uuid.uuid, - cfs_atomic_read(&obd->obd_refcount)); + CDEBUG(D_CONFIG, "Connected to %s(%s) successfully (%d)\n", + mdc_obd->obd_name, mdc_obd->obd_uuid.uuid, + atomic_read(&obd->obd_refcount)); -#ifdef __KERNEL__ - lmv_proc_dir = obd->obd_proc_private; - if (lmv_proc_dir) { + if (lmv->targets_proc_entry != NULL) { struct proc_dir_entry *mdc_symlink; LASSERT(mdc_obd->obd_type != NULL); LASSERT(mdc_obd->obd_type->typ_name != NULL); mdc_symlink = lprocfs_add_symlink(mdc_obd->obd_name, - lmv_proc_dir, + lmv->targets_proc_entry, "../../../%s/%s", mdc_obd->obd_type->typ_name, mdc_obd->obd_name); if (mdc_symlink == NULL) { - CERROR("Could not register LMV target " - "/proc/fs/lustre/%s/%s/target_obds/%s.", + CERROR("cannot register LMV target " + "/proc/fs/lustre/%s/%s/target_obds/%s\n", obd->obd_type->typ_name, obd->obd_name, mdc_obd->obd_name); - lprocfs_remove(&lmv_proc_dir); - obd->obd_proc_private = NULL; } } -#endif RETURN(0); } @@ -465,34 +424,29 @@ static void lmv_del_target(struct lmv_obd *lmv, int index) static int lmv_add_target(struct obd_device *obd, struct obd_uuid *uuidp, __u32 index, int gen) { + struct obd_device *mdc_obd; struct lmv_obd *lmv = &obd->u.lmv; struct lmv_tgt_desc *tgt; + int orig_tgt_count = 0; int rc = 0; ENTRY; CDEBUG(D_CONFIG, "Target uuid: %s. index %d\n", uuidp->uuid, index); - - lmv_init_lock(lmv); - - if (lmv->desc.ld_tgt_count == 0) { - struct obd_device *mdc_obd; - - mdc_obd = class_find_client_obd(uuidp, LUSTRE_MDC_NAME, - &obd->obd_uuid); - if (!mdc_obd) { - lmv_init_unlock(lmv); - CERROR("%s: Target %s not attached: rc = %d\n", - obd->obd_name, uuidp->uuid, -EINVAL); - RETURN(-EINVAL); - } + mdc_obd = class_find_client_obd(uuidp, LUSTRE_MDC_NAME, + &obd->obd_uuid); + if (!mdc_obd) { + CERROR("%s: Target %s not attached: rc = %d\n", + obd->obd_name, uuidp->uuid, -EINVAL); + RETURN(-EINVAL); } + mutex_lock(&lmv->lmv_init_mutex); if ((index < lmv->tgts_size) && (lmv->tgts[index] != NULL)) { tgt = lmv->tgts[index]; CERROR("%s: UUID %s already assigned at LOV target index %d:" " rc = %d\n", obd->obd_name, obd_uuid2str(&tgt->ltd_uuid), index, -EEXIST); - lmv_init_unlock(lmv); + mutex_unlock(&lmv->lmv_init_mutex); RETURN(-EEXIST); } @@ -506,7 +460,7 @@ static int lmv_add_target(struct obd_device *obd, struct obd_uuid *uuidp, newsize = newsize << 1; OBD_ALLOC(newtgts, sizeof(*newtgts) * newsize); if (newtgts == NULL) { - lmv_init_unlock(lmv); + mutex_unlock(&lmv->lmv_init_mutex); RETURN(-ENOMEM); } @@ -529,7 +483,7 @@ static int lmv_add_target(struct obd_device *obd, struct obd_uuid *uuidp, OBD_ALLOC_PTR(tgt); if (!tgt) { - lmv_init_unlock(lmv); + mutex_unlock(&lmv->lmv_init_mutex); RETURN(-ENOMEM); } @@ -538,26 +492,33 @@ static int lmv_add_target(struct obd_device *obd, struct obd_uuid *uuidp, tgt->ltd_uuid = *uuidp; tgt->ltd_active = 0; lmv->tgts[index] = tgt; - if (index >= lmv->desc.ld_tgt_count) + if (index >= lmv->desc.ld_tgt_count) { + orig_tgt_count = lmv->desc.ld_tgt_count; lmv->desc.ld_tgt_count = index + 1; + } - if (lmv->connected) { - rc = lmv_connect_mdc(obd, tgt); - if (rc) { - spin_lock(&lmv->lmv_lock); - lmv->desc.ld_tgt_count--; - memset(tgt, 0, sizeof(*tgt)); - spin_unlock(&lmv->lmv_lock); - } else { - int easize = sizeof(struct lmv_stripe_md) + - lmv->desc.ld_tgt_count * - sizeof(struct lu_fid); - lmv_init_ea_size(obd->obd_self_export, easize, 0, 0); - } - } + if (lmv->connected == 0) { + /* lmv_check_connect() will connect this target. */ + mutex_unlock(&lmv->lmv_init_mutex); + RETURN(0); + } - lmv_init_unlock(lmv); - RETURN(rc); + /* Otherwise let's connect it ourselves */ + mutex_unlock(&lmv->lmv_init_mutex); + rc = lmv_connect_mdc(obd, tgt); + if (rc != 0) { + spin_lock(&lmv->lmv_lock); + if (lmv->desc.ld_tgt_count == index + 1) + lmv->desc.ld_tgt_count = orig_tgt_count; + memset(tgt, 0, sizeof(*tgt)); + spin_unlock(&lmv->lmv_lock); + } else { + int easize = sizeof(struct lmv_stripe_md) + + lmv->desc.ld_tgt_count * sizeof(struct lu_fid); + lmv_init_ea_size(obd->obd_self_export, easize, 0); + } + + RETURN(rc); } int lmv_check_connect(struct obd_device *obd) @@ -572,14 +533,14 @@ int lmv_check_connect(struct obd_device *obd) if (lmv->connected) RETURN(0); - lmv_init_lock(lmv); + mutex_lock(&lmv->lmv_init_mutex); if (lmv->connected) { - lmv_init_unlock(lmv); + mutex_unlock(&lmv->lmv_init_mutex); RETURN(0); } if (lmv->desc.ld_tgt_count == 0) { - lmv_init_unlock(lmv); + mutex_unlock(&lmv->lmv_init_mutex); CERROR("%s: no targets configured.\n", obd->obd_name); RETURN(-EINVAL); } @@ -587,7 +548,7 @@ int lmv_check_connect(struct obd_device *obd) LASSERT(lmv->tgts != NULL); if (lmv->tgts[0] == NULL) { - lmv_init_unlock(lmv); + mutex_unlock(&lmv->lmv_init_mutex); CERROR("%s: no target configured for index 0.\n", obd->obd_name); RETURN(-EINVAL); @@ -605,13 +566,12 @@ int lmv_check_connect(struct obd_device *obd) GOTO(out_disc, rc); } - lmv_set_timeouts(obd); - class_export_put(lmv->exp); - lmv->connected = 1; - easize = lmv_get_easize(lmv); - lmv_init_ea_size(obd->obd_self_export, easize, 0, 0); - lmv_init_unlock(lmv); - RETURN(0); + class_export_put(lmv->exp); + lmv->connected = 1; + easize = lmv_mds_md_size(lmv->desc.ld_tgt_count, LMV_MAGIC); + lmv_init_ea_size(obd->obd_self_export, easize, 0); + mutex_unlock(&lmv->lmv_init_mutex); + RETURN(0); out_disc: while (i-- > 0) { @@ -631,15 +591,12 @@ int lmv_check_connect(struct obd_device *obd) } } class_disconnect(lmv->exp); - lmv_init_unlock(lmv); + mutex_unlock(&lmv->lmv_init_mutex); RETURN(rc); } static int lmv_disconnect_mdc(struct obd_device *obd, struct lmv_tgt_desc *tgt) { -#ifdef __KERNEL__ - struct proc_dir_entry *lmv_proc_dir; -#endif struct lmv_obd *lmv = &obd->u.lmv; struct obd_device *mdc_obd; int rc; @@ -654,13 +611,12 @@ static int lmv_disconnect_mdc(struct obd_device *obd, struct lmv_tgt_desc *tgt) mdc_obd->obd_force = obd->obd_force; mdc_obd->obd_fail = obd->obd_fail; mdc_obd->obd_no_recov = obd->obd_no_recov; - } -#ifdef __KERNEL__ - lmv_proc_dir = obd->obd_proc_private; - if (lmv_proc_dir) - lprocfs_remove_proc_entry(mdc_obd->obd_name, lmv_proc_dir); -#endif + if (lmv->targets_proc_entry != NULL) + lprocfs_remove_proc_entry(mdc_obd->obd_name, + lmv->targets_proc_entry); + } + rc = obd_fid_fini(tgt->ltd_exp->exp_obd); if (rc) CERROR("Can't finanize fids factory\n"); @@ -708,13 +664,11 @@ static int lmv_disconnect(struct obd_export *exp) lmv_disconnect_mdc(obd, lmv->tgts[i]); } -#ifdef __KERNEL__ - if (obd->obd_proc_private) - lprocfs_remove((struct proc_dir_entry **)&obd->obd_proc_private); + if (lmv->targets_proc_entry != NULL) + lprocfs_remove(&lmv->targets_proc_entry); else CERROR("/proc/fs/lustre/%s/%s/target_obds missing\n", obd->obd_type->typ_name, obd->obd_name); -#endif out_local: /* @@ -1078,6 +1032,19 @@ static int lmv_iocontrol(unsigned int cmd, struct obd_export *exp, rc = obd_iocontrol(cmd, tgt->ltd_exp, len, karg, uarg); break; } + case LL_IOC_FID2MDTIDX: { + struct lu_fid *fid = karg; + int mdt_index; + + rc = lmv_fld_lookup(lmv, fid, &mdt_index); + if (rc != 0) + RETURN(rc); + + /* Note: this is from llite(see ll_dir_ioctl()), @uarg does not + * point to user space memory for FID2MDTIDX. */ + *(__u32 *)uarg = mdt_index; + break; + } case OBD_IOC_FID2PATH: { rc = lmv_fid2path(exp, len, karg, uarg); break; @@ -1199,9 +1166,7 @@ static int lmv_iocontrol(unsigned int cmd, struct obd_export *exp, mdc_obd = class_exp2obd(tgt->ltd_exp); mdc_obd->obd_force = obddev->obd_force; err = obd_iocontrol(cmd, tgt->ltd_exp, len, karg, uarg); - if (err == -ENODATA && cmd == OBD_IOC_POLL_QUOTACHECK) { - RETURN(err); - } else if (err) { + if (err) { if (tgt->ltd_active) { CERROR("error: iocontrol MDC %s on MDT" " idx %d cmd %x: err = %d\n", @@ -1267,8 +1232,7 @@ static int lmv_choose_mds(struct lmv_obd *lmv, struct md_op_data *op_data, * This is _inode_ placement policy function (not name). */ static int lmv_placement_policy(struct obd_device *obd, - struct md_op_data *op_data, - mdsno_t *mds) + struct md_op_data *op_data, u32 *mds) { struct lmv_obd *lmv = &obd->u.lmv; ENTRY; @@ -1280,42 +1244,44 @@ static int lmv_placement_policy(struct obd_device *obd, RETURN(0); } + if (op_data->op_default_stripe_offset != -1) { + *mds = op_data->op_default_stripe_offset; + RETURN(0); + } + /** * If stripe_offset is provided during setdirstripe * (setdirstripe -i xx), xx MDS will be choosen. */ - if (op_data->op_cli_flags & CLI_SET_MEA) { + if (op_data->op_cli_flags & CLI_SET_MEA && op_data->op_data != NULL) { struct lmv_user_md *lum; - lum = (struct lmv_user_md *)op_data->op_data; - if (lum->lum_type == LMV_STRIPE_TYPE && - lum->lum_stripe_offset != -1) { - if (lum->lum_stripe_offset >= lmv->desc.ld_tgt_count) { - CERROR("%s: Stripe_offset %d > MDT count %d:" - " rc = %d\n", obd->obd_name, - lum->lum_stripe_offset, - lmv->desc.ld_tgt_count, -ERANGE); - RETURN(-ERANGE); - } - *mds = lum->lum_stripe_offset; - RETURN(0); + lum = op_data->op_data; + + if (le32_to_cpu(lum->lum_stripe_offset) != (__u32)-1) { + *mds = le32_to_cpu(lum->lum_stripe_offset); + } else { + /* -1 means default, which will be in the same MDT with + * the stripe */ + *mds = op_data->op_mds; + lum->lum_stripe_offset = cpu_to_le32(op_data->op_mds); } + } else { + /* Allocate new fid on target according to operation type and + * parent home mds. */ + *mds = op_data->op_mds; } - /* Allocate new fid on target according to operation type and parent - * home mds. */ - *mds = op_data->op_mds; RETURN(0); } -int __lmv_fid_alloc(struct lmv_obd *lmv, struct lu_fid *fid, - mdsno_t mds) +int __lmv_fid_alloc(struct lmv_obd *lmv, struct lu_fid *fid, u32 mds) { struct lmv_tgt_desc *tgt; int rc; ENTRY; - tgt = lmv_get_target(lmv, mds); + tgt = lmv_get_target(lmv, mds, NULL); if (IS_ERR(tgt)) RETURN(PTR_ERR(tgt)); @@ -1328,14 +1294,14 @@ int __lmv_fid_alloc(struct lmv_obd *lmv, struct lu_fid *fid, if (tgt->ltd_active == 0 || tgt->ltd_exp == NULL) GOTO(out, rc = -ENODEV); - /* - * Asking underlaying tgt layer to allocate new fid. - */ - rc = obd_fid_alloc(tgt->ltd_exp, fid, NULL); - if (rc > 0) { - LASSERT(fid_is_sane(fid)); - rc = 0; - } + /* + * Asking underlying tgt layer to allocate new fid. + */ + rc = obd_fid_alloc(NULL, tgt->ltd_exp, fid, NULL); + if (rc > 0) { + LASSERT(fid_is_sane(fid)); + rc = 0; + } EXIT; out: @@ -1343,12 +1309,12 @@ out: return rc; } -int lmv_fid_alloc(struct obd_export *exp, struct lu_fid *fid, - struct md_op_data *op_data) +int lmv_fid_alloc(const struct lu_env *env, struct obd_export *exp, + struct lu_fid *fid, struct md_op_data *op_data) { struct obd_device *obd = class_exp2obd(exp); struct lmv_obd *lmv = &obd->u.lmv; - mdsno_t mds = 0; + u32 mds = 0; int rc; ENTRY; @@ -1390,25 +1356,24 @@ static int lmv_setup(struct obd_device *obd, struct lustre_cfg *lcfg) RETURN(-EINVAL); } - OBD_ALLOC(lmv->tgts, sizeof(*lmv->tgts) * 32); + lmv->tgts_size = 32U; + OBD_ALLOC(lmv->tgts, sizeof(*lmv->tgts) * lmv->tgts_size); if (lmv->tgts == NULL) RETURN(-ENOMEM); - lmv->tgts_size = 32; obd_str2uuid(&lmv->desc.ld_uuid, desc->ld_uuid.uuid); lmv->desc.ld_tgt_count = 0; lmv->desc.ld_active_tgt_count = 0; - lmv->max_cookiesize = 0; lmv->max_def_easize = 0; lmv->max_easize = 0; lmv->lmv_placement = PLACEMENT_CHAR_POLICY; spin_lock_init(&lmv->lmv_lock); - mutex_init(&lmv->init_mutex); + mutex_init(&lmv->lmv_init_mutex); -#ifdef LPROCFS +#ifdef CONFIG_PROC_FS obd->obd_vars = lprocfs_lmv_obd_vars; - lprocfs_seq_obd_setup(obd); + lprocfs_obd_setup(obd); lprocfs_alloc_md_stats(obd, 0); rc = lprocfs_seq_create(obd->obd_proc_entry, "target_obd", 0444, &lmv_proc_target_fops, obd); @@ -1448,7 +1413,7 @@ static int lmv_cleanup(struct obd_device *obd) RETURN(0); } -static int lmv_process_config(struct obd_device *obd, obd_count len, void *buf) +static int lmv_process_config(struct obd_device *obd, size_t len, void *buf) { struct lustre_cfg *lcfg = buf; struct obd_uuid obd_uuid; @@ -1466,7 +1431,7 @@ static int lmv_process_config(struct obd_device *obd, obd_count len, void *buf) obd_str2uuid(&obd_uuid, lustre_cfg_buf(lcfg, 1)); - if (sscanf(lustre_cfg_buf(lcfg, 2), "%d", &index) != 1) + if (sscanf(lustre_cfg_buf(lcfg, 2), "%u", &index) != 1) GOTO(out, rc = -EINVAL); if (sscanf(lustre_cfg_buf(lcfg, 3), "%d", &gen) != 1) GOTO(out, rc = -EINVAL); @@ -1535,9 +1500,7 @@ out_free_temp: return rc; } -static int lmv_getstatus(struct obd_export *exp, - struct lu_fid *fid, - struct obd_capa **pc) +static int lmv_getstatus(struct obd_export *exp, struct lu_fid *fid) { struct obd_device *obd = exp->exp_obd; struct lmv_obd *lmv = &obd->u.lmv; @@ -1548,14 +1511,14 @@ static int lmv_getstatus(struct obd_export *exp, if (rc) RETURN(rc); - rc = md_getstatus(lmv->tgts[0]->ltd_exp, fid, pc); + rc = md_getstatus(lmv->tgts[0]->ltd_exp, fid); RETURN(rc); } static int lmv_getxattr(struct obd_export *exp, const struct lu_fid *fid, - struct obd_capa *oc, obd_valid valid, const char *name, - const char *input, int input_size, int output_size, - int flags, struct ptlrpc_request **request) + u64 valid, const char *name, + const char *input, int input_size, int output_size, + int flags, struct ptlrpc_request **request) { struct obd_device *obd = exp->exp_obd; struct lmv_obd *lmv = &obd->u.lmv; @@ -1571,17 +1534,17 @@ static int lmv_getxattr(struct obd_export *exp, const struct lu_fid *fid, if (IS_ERR(tgt)) RETURN(PTR_ERR(tgt)); - rc = md_getxattr(tgt->ltd_exp, fid, oc, valid, name, input, - input_size, output_size, flags, request); + rc = md_getxattr(tgt->ltd_exp, fid, valid, name, input, + input_size, output_size, flags, request); - RETURN(rc); + RETURN(rc); } static int lmv_setxattr(struct obd_export *exp, const struct lu_fid *fid, - struct obd_capa *oc, obd_valid valid, const char *name, - const char *input, int input_size, int output_size, - int flags, __u32 suppgid, - struct ptlrpc_request **request) + u64 valid, const char *name, + const char *input, int input_size, int output_size, + int flags, __u32 suppgid, + struct ptlrpc_request **request) { struct obd_device *obd = exp->exp_obd; struct lmv_obd *lmv = &obd->u.lmv; @@ -1597,11 +1560,11 @@ static int lmv_setxattr(struct obd_export *exp, const struct lu_fid *fid, if (IS_ERR(tgt)) RETURN(PTR_ERR(tgt)); - rc = md_setxattr(tgt->ltd_exp, fid, oc, valid, name, input, - input_size, output_size, flags, suppgid, - request); + rc = md_setxattr(tgt->ltd_exp, fid, valid, name, input, + input_size, output_size, flags, suppgid, + request); - RETURN(rc); + RETURN(rc); } static int lmv_getattr(struct obd_export *exp, struct md_op_data *op_data, @@ -1662,27 +1625,39 @@ static int lmv_null_inode(struct obd_export *exp, const struct lu_fid *fid) static int lmv_find_cbdata(struct obd_export *exp, const struct lu_fid *fid, ldlm_iterator_t it, void *data) { - struct obd_device *obd = exp->exp_obd; - struct lmv_obd *lmv = &obd->u.lmv; - __u32 i; - int rc; - ENTRY; + struct obd_device *obd = exp->exp_obd; + struct lmv_obd *lmv = &obd->u.lmv; + int i; + int tgt; + int rc; + ENTRY; - rc = lmv_check_connect(obd); - if (rc) - RETURN(rc); + rc = lmv_check_connect(obd); + if (rc) + RETURN(rc); - CDEBUG(D_INODE, "CBDATA for "DFID"\n", PFID(fid)); + CDEBUG(D_INODE, "CBDATA for "DFID"\n", PFID(fid)); /* * With DNE every object can have two locks in different namespaces: * lookup lock in space of MDT storing direntry and update/open lock in - * space of MDT storing inode. + * space of MDT storing inode. Try the MDT that the FID maps to first, + * since this can be easily found, and only try others if that fails. */ - for (i = 0; i < lmv->desc.ld_tgt_count; i++) { - if (lmv->tgts[i] == NULL || lmv->tgts[i]->ltd_exp == NULL) + for (i = 0, tgt = lmv_find_target_index(lmv, fid); + i < lmv->desc.ld_tgt_count; + i++, tgt = (tgt + 1) % lmv->desc.ld_tgt_count) { + if (tgt < 0) { + CDEBUG(D_HA, "%s: "DFID" is inaccessible: rc = %d\n", + obd->obd_name, PFID(fid), tgt); + tgt = 0; + } + + if (lmv->tgts[tgt] == NULL || + lmv->tgts[tgt]->ltd_exp == NULL) continue; - rc = md_find_cbdata(lmv->tgts[i]->ltd_exp, fid, it, data); + + rc = md_find_cbdata(lmv->tgts[tgt]->ltd_exp, fid, it, data); if (rc) RETURN(rc); } @@ -1713,25 +1688,112 @@ static int lmv_close(struct obd_export *exp, struct md_op_data *op_data, RETURN(rc); } -struct lmv_tgt_desc -*lmv_locate_mds(struct lmv_obd *lmv, struct md_op_data *op_data, - struct lu_fid *fid) +/** + * Choosing the MDT by name or FID in @op_data. + * For non-striped directory, it will locate MDT by fid. + * For striped-directory, it will locate MDT by name. And also + * it will reset op_fid1 with the FID of the choosen stripe. + **/ +struct lmv_tgt_desc * +lmv_locate_target_for_name(struct lmv_obd *lmv, struct lmv_stripe_md *lsm, + const char *name, int namelen, struct lu_fid *fid, + u32 *mds) { - struct lmv_tgt_desc *tgt; + struct lmv_tgt_desc *tgt; + const struct lmv_oinfo *oinfo; - tgt = lmv_find_target(lmv, fid); - if (IS_ERR(tgt)) - return tgt; + if (OBD_FAIL_CHECK(OBD_FAIL_LFSCK_BAD_NAME_HASH)) { + if (cfs_fail_val >= lsm->lsm_md_stripe_count) + RETURN(ERR_PTR(-EBADF)); + oinfo = &lsm->lsm_md_oinfo[cfs_fail_val]; + } else { + oinfo = lsm_name_to_stripe_info(lsm, name, namelen); + if (IS_ERR(oinfo)) + RETURN(ERR_CAST(oinfo)); + } + + if (fid != NULL) + *fid = oinfo->lmo_fid; + if (mds != NULL) + *mds = oinfo->lmo_mds; - op_data->op_mds = tgt->ltd_idx; + tgt = lmv_get_target(lmv, oinfo->lmo_mds, NULL); + CDEBUG(D_INFO, "locate on mds %u "DFID"\n", oinfo->lmo_mds, + PFID(&oinfo->lmo_fid)); return tgt; } +/** + * Locate mds by fid or name + * + * For striped directory (lsm != NULL), it will locate the stripe + * by name hash (see lsm_name_to_stripe_info()). Note: if the hash_type + * is unknown, it will return -EBADFD, and lmv_intent_lookup might need + * walk through all of stripes to locate the entry. + * + * For normal direcotry, it will locate MDS by FID directly. + * \param[in] lmv LMV device + * \param[in] op_data client MD stack parameters, name, namelen + * mds_num etc. + * \param[in] fid object FID used to locate MDS. + * + * retval pointer to the lmv_tgt_desc if succeed. + * ERR_PTR(errno) if failed. + */ +struct lmv_tgt_desc* +lmv_locate_mds(struct lmv_obd *lmv, struct md_op_data *op_data, + struct lu_fid *fid) +{ + struct lmv_stripe_md *lsm = op_data->op_mea1; + struct lmv_tgt_desc *tgt; + + /* During creating VOLATILE file, it should honor the mdt + * index if the file under striped dir is being restored, see + * ct_restore(). */ + if (op_data->op_bias & MDS_CREATE_VOLATILE && + (int)op_data->op_mds != -1 && lsm != NULL) { + int i; + tgt = lmv_get_target(lmv, op_data->op_mds, NULL); + if (IS_ERR(tgt)) + return tgt; + + /* refill the right parent fid */ + for (i = 0; i < lsm->lsm_md_stripe_count; i++) { + struct lmv_oinfo *oinfo; + + oinfo = &lsm->lsm_md_oinfo[i]; + if (oinfo->lmo_mds == op_data->op_mds) { + *fid = oinfo->lmo_fid; + break; + } + } + + /* Hmm, can not find the stripe by mdt_index(op_mds) */ + if (i == lsm->lsm_md_stripe_count) + tgt = ERR_PTR(-EINVAL); + + return tgt; + } + + if (lsm == NULL || op_data->op_namelen == 0) { + tgt = lmv_find_target(lmv, fid); + if (IS_ERR(tgt)) + return tgt; + + op_data->op_mds = tgt->ltd_idx; + return tgt; + } + + return lmv_locate_target_for_name(lmv, lsm, op_data->op_name, + op_data->op_namelen, fid, + &op_data->op_mds); +} + int lmv_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) + const void *data, size_t datalen, umode_t mode, uid_t uid, + gid_t gid, cfs_cap_t cap_effective, __u64 rdev, + struct ptlrpc_request **request) { struct obd_device *obd = exp->exp_obd; struct lmv_obd *lmv = &obd->u.lmv; @@ -1750,18 +1812,31 @@ int lmv_create(struct obd_export *exp, struct md_op_data *op_data, if (IS_ERR(tgt)) RETURN(PTR_ERR(tgt)); - rc = lmv_fid_alloc(exp, &op_data->op_fid2, op_data); + CDEBUG(D_INODE, "CREATE name '%.*s' on "DFID" -> mds #%x\n", + (int)op_data->op_namelen, op_data->op_name, + PFID(&op_data->op_fid1), op_data->op_mds); + + rc = lmv_fid_alloc(NULL, exp, &op_data->op_fid2, op_data); if (rc) RETURN(rc); + if (exp_connect_flags(exp) & OBD_CONNECT_DIR_STRIPE) { + /* Send the create request to the MDT where the object + * will be located */ + tgt = lmv_find_target(lmv, &op_data->op_fid2); + if (IS_ERR(tgt)) + RETURN(PTR_ERR(tgt)); + + op_data->op_mds = tgt->ltd_idx; + } else { + CDEBUG(D_CONFIG, "Server doesn't support striped dirs\n"); + } - CDEBUG(D_INODE, "CREATE '%*s' on "DFID" -> mds #%x\n", - op_data->op_namelen, op_data->op_name, PFID(&op_data->op_fid1), - op_data->op_mds); + CDEBUG(D_INODE, "CREATE obj "DFID" -> mds #%x\n", + PFID(&op_data->op_fid2), op_data->op_mds); op_data->op_flags |= MF_MDC_CANCEL_FID1; rc = md_create(tgt->ltd_exp, op_data, data, datalen, mode, uid, gid, cap_effective, rdev, request); - if (rc == 0) { if (*request == NULL) RETURN(rc); @@ -1770,92 +1845,11 @@ int lmv_create(struct obd_export *exp, struct md_op_data *op_data, RETURN(rc); } -static int lmv_done_writing(struct obd_export *exp, - struct md_op_data *op_data, - struct md_open_data *mod) -{ - struct obd_device *obd = exp->exp_obd; - struct lmv_obd *lmv = &obd->u.lmv; - struct lmv_tgt_desc *tgt; - int rc; - ENTRY; - - rc = lmv_check_connect(obd); - if (rc) - RETURN(rc); - - tgt = lmv_find_target(lmv, &op_data->op_fid1); - if (IS_ERR(tgt)) - RETURN(PTR_ERR(tgt)); - - rc = md_done_writing(tgt->ltd_exp, op_data, mod); - RETURN(rc); -} - -static int -lmv_enqueue_remote(struct obd_export *exp, struct ldlm_enqueue_info *einfo, - struct lookup_intent *it, struct md_op_data *op_data, - struct lustre_handle *lockh, void *lmm, int lmmsize, - __u64 extra_lock_flags) -{ - struct ptlrpc_request *req = it->d.lustre.it_data; - struct obd_device *obd = exp->exp_obd; - struct lmv_obd *lmv = &obd->u.lmv; - struct lustre_handle plock; - struct lmv_tgt_desc *tgt; - struct md_op_data *rdata; - struct lu_fid fid1; - struct mdt_body *body; - int rc = 0; - int pmode; - ENTRY; - - body = req_capsule_server_get(&req->rq_pill, &RMF_MDT_BODY); - LASSERT(body != NULL); - - if (!(body->valid & OBD_MD_MDS)) - RETURN(0); - - CDEBUG(D_INODE, "REMOTE_ENQUEUE '%s' on "DFID" -> "DFID"\n", - LL_IT2STR(it), PFID(&op_data->op_fid1), PFID(&body->fid1)); - - /* - * We got LOOKUP lock, but we really need attrs. - */ - pmode = it->d.lustre.it_lock_mode; - LASSERT(pmode != 0); - memcpy(&plock, lockh, sizeof(plock)); - it->d.lustre.it_lock_mode = 0; - it->d.lustre.it_data = NULL; - fid1 = body->fid1; - - ptlrpc_req_finished(req); - - tgt = lmv_find_target(lmv, &fid1); - if (IS_ERR(tgt)) - GOTO(out, rc = PTR_ERR(tgt)); - - OBD_ALLOC_PTR(rdata); - if (rdata == NULL) - GOTO(out, rc = -ENOMEM); - - rdata->op_fid1 = fid1; - rdata->op_bias = MDS_CROSS_REF; - - rc = md_enqueue(tgt->ltd_exp, einfo, it, rdata, lockh, - lmm, lmmsize, NULL, extra_lock_flags); - OBD_FREE_PTR(rdata); - EXIT; -out: - ldlm_lock_decref(&plock, pmode); - return rc; -} - static int lmv_enqueue(struct obd_export *exp, struct ldlm_enqueue_info *einfo, - struct lookup_intent *it, struct md_op_data *op_data, - struct lustre_handle *lockh, void *lmm, int lmmsize, - struct ptlrpc_request **req, __u64 extra_lock_flags) + const union ldlm_policy_data *policy, + struct lookup_intent *it, struct md_op_data *op_data, + struct lustre_handle *lockh, __u64 extra_lock_flags) { struct obd_device *obd = exp->exp_obd; struct lmv_obd *lmv = &obd->u.lmv; @@ -1874,22 +1868,18 @@ lmv_enqueue(struct obd_export *exp, struct ldlm_enqueue_info *einfo, if (IS_ERR(tgt)) RETURN(PTR_ERR(tgt)); - CDEBUG(D_INODE, "ENQUEUE '%s' on "DFID" -> mds #%d\n", + CDEBUG(D_INODE, "ENQUEUE '%s' on "DFID" -> mds #%u\n", LL_IT2STR(it), PFID(&op_data->op_fid1), tgt->ltd_idx); - rc = md_enqueue(tgt->ltd_exp, einfo, it, op_data, lockh, - lmm, lmmsize, req, extra_lock_flags); + rc = md_enqueue(tgt->ltd_exp, einfo, policy, it, op_data, lockh, + extra_lock_flags); - if (rc == 0 && it && it->it_op == IT_OPEN) { - rc = lmv_enqueue_remote(exp, einfo, it, op_data, lockh, - lmm, lmmsize, extra_lock_flags); - } RETURN(rc); } static int lmv_getattr_name(struct obd_export *exp,struct md_op_data *op_data, - struct ptlrpc_request **request) + struct ptlrpc_request **preq) { struct ptlrpc_request *req = NULL; struct obd_device *obd = exp->exp_obd; @@ -1908,25 +1898,25 @@ lmv_getattr_name(struct obd_export *exp,struct md_op_data *op_data, RETURN(PTR_ERR(tgt)); CDEBUG(D_INODE, "GETATTR_NAME for %*s on "DFID" -> mds #%d\n", - op_data->op_namelen, op_data->op_name, PFID(&op_data->op_fid1), - tgt->ltd_idx); + (int)op_data->op_namelen, op_data->op_name, + PFID(&op_data->op_fid1), tgt->ltd_idx); - rc = md_getattr_name(tgt->ltd_exp, op_data, request); + rc = md_getattr_name(tgt->ltd_exp, op_data, preq); if (rc != 0) RETURN(rc); - body = req_capsule_server_get(&(*request)->rq_pill, - &RMF_MDT_BODY); + body = req_capsule_server_get(&(*preq)->rq_pill, &RMF_MDT_BODY); LASSERT(body != NULL); - if (body->valid & OBD_MD_MDS) { - struct lu_fid rid = body->fid1; + if (body->mbo_valid & OBD_MD_MDS) { + struct lu_fid rid = body->mbo_fid1; CDEBUG(D_INODE, "Request attrs for "DFID"\n", PFID(&rid)); tgt = lmv_find_target(lmv, &rid); if (IS_ERR(tgt)) { - ptlrpc_req_finished(*request); + ptlrpc_req_finished(*preq); + preq = NULL; RETURN(PTR_ERR(tgt)); } @@ -1935,8 +1925,8 @@ lmv_getattr_name(struct obd_export *exp,struct md_op_data *op_data, op_data->op_namelen = 0; op_data->op_name = NULL; rc = md_getattr_name(tgt->ltd_exp, op_data, &req); - ptlrpc_req_finished(*request); - *request = req; + ptlrpc_req_finished(*preq); + *preq = req; } RETURN(rc); @@ -1949,23 +1939,25 @@ lmv_getattr_name(struct obd_export *exp,struct md_op_data *op_data, fl == MF_MDC_CANCEL_FID4 ? &op_data->op_fid4 : \ NULL) -static int lmv_early_cancel(struct obd_export *exp, struct md_op_data *op_data, - int op_tgt, ldlm_mode_t mode, int bits, int flag) +static int lmv_early_cancel(struct obd_export *exp, struct lmv_tgt_desc *tgt, + struct md_op_data *op_data, + __u32 op_tgt, ldlm_mode_t mode, int bits, int flag) { - struct lu_fid *fid = md_op_data_fid(op_data, flag); - struct obd_device *obd = exp->exp_obd; - struct lmv_obd *lmv = &obd->u.lmv; - struct lmv_tgt_desc *tgt; - ldlm_policy_data_t policy = {{0}}; - int rc = 0; - ENTRY; + struct lu_fid *fid = md_op_data_fid(op_data, flag); + struct obd_device *obd = exp->exp_obd; + struct lmv_obd *lmv = &obd->u.lmv; + ldlm_policy_data_t policy = {{ 0 }}; + int rc = 0; + ENTRY; - if (!fid_is_sane(fid)) - RETURN(0); + if (!fid_is_sane(fid)) + RETURN(0); - tgt = lmv_find_target(lmv, fid); - if (IS_ERR(tgt)) - RETURN(PTR_ERR(tgt)); + if (tgt == NULL) { + tgt = lmv_find_target(lmv, fid); + if (IS_ERR(tgt)) + RETURN(PTR_ERR(tgt)); + } if (tgt->ltd_idx != op_tgt) { CDEBUG(D_INODE, "EARLY_CANCEL on "DFID"\n", PFID(fid)); @@ -2003,12 +1995,24 @@ static int lmv_link(struct obd_export *exp, struct md_op_data *op_data, LASSERT(op_data->op_namelen != 0); CDEBUG(D_INODE, "LINK "DFID":%*s to "DFID"\n", - PFID(&op_data->op_fid2), op_data->op_namelen, + PFID(&op_data->op_fid2), (int)op_data->op_namelen, op_data->op_name, PFID(&op_data->op_fid1)); - op_data->op_fsuid = current_fsuid(); - op_data->op_fsgid = current_fsgid(); + op_data->op_fsuid = from_kuid(&init_user_ns, current_fsuid()); + op_data->op_fsgid = from_kgid(&init_user_ns, current_fsgid()); op_data->op_cap = cfs_curproc_cap_pack(); + if (op_data->op_mea2 != NULL) { + struct lmv_stripe_md *lsm = op_data->op_mea2; + const struct lmv_oinfo *oinfo; + + oinfo = lsm_name_to_stripe_info(lsm, op_data->op_name, + op_data->op_namelen); + if (IS_ERR(oinfo)) + RETURN(PTR_ERR(oinfo)); + + op_data->op_fid2 = oinfo->lmo_fid; + } + tgt = lmv_locate_mds(lmv, op_data, &op_data->op_fid2); if (IS_ERR(tgt)) RETURN(PTR_ERR(tgt)); @@ -2017,7 +2021,7 @@ static int lmv_link(struct obd_export *exp, struct md_op_data *op_data, * Cancel UPDATE lock on child (fid1). */ op_data->op_flags |= MF_MDC_CANCEL_FID2; - rc = lmv_early_cancel(exp, op_data, tgt->ltd_idx, LCK_EX, + rc = lmv_early_cancel(exp, NULL, op_data, tgt->ltd_idx, LCK_EX, MDS_INODELOCK_UPDATE, MF_MDC_CANCEL_FID1); if (rc != 0) RETURN(rc); @@ -2028,36 +2032,92 @@ static int lmv_link(struct obd_export *exp, struct md_op_data *op_data, } static int lmv_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) + const char *old, size_t oldlen, + const char *new, size_t newlen, + struct ptlrpc_request **request) { - struct obd_device *obd = exp->exp_obd; - struct lmv_obd *lmv = &obd->u.lmv; - struct lmv_tgt_desc *src_tgt; + struct obd_device *obd = exp->exp_obd; + struct lmv_obd *lmv = &obd->u.lmv; + struct lmv_tgt_desc *src_tgt; struct lmv_tgt_desc *tgt_tgt; + struct obd_export *target_exp; + struct mdt_body *body; int rc; ENTRY; - LASSERT(oldlen != 0); + LASSERT(oldlen != 0); - CDEBUG(D_INODE, "RENAME %*s in "DFID" to %*s in "DFID"\n", - oldlen, old, PFID(&op_data->op_fid1), - newlen, new, PFID(&op_data->op_fid2)); + CDEBUG(D_INODE, "RENAME %.*s in "DFID":%d to %.*s in "DFID":%d\n", + (int)oldlen, old, PFID(&op_data->op_fid1), + op_data->op_mea1 ? op_data->op_mea1->lsm_md_stripe_count : 0, + (int)newlen, new, PFID(&op_data->op_fid2), + op_data->op_mea2 ? op_data->op_mea2->lsm_md_stripe_count : 0); - rc = lmv_check_connect(obd); - if (rc) - RETURN(rc); + rc = lmv_check_connect(obd); + if (rc) + RETURN(rc); - op_data->op_fsuid = current_fsuid(); - op_data->op_fsgid = current_fsgid(); + op_data->op_fsuid = from_kuid(&init_user_ns, current_fsuid()); + op_data->op_fsgid = from_kgid(&init_user_ns, current_fsgid()); op_data->op_cap = cfs_curproc_cap_pack(); - src_tgt = lmv_locate_mds(lmv, op_data, &op_data->op_fid1); - if (IS_ERR(src_tgt)) - RETURN(PTR_ERR(src_tgt)); + if (op_data->op_cli_flags & CLI_MIGRATE) { + LASSERTF(fid_is_sane(&op_data->op_fid3), "invalid FID "DFID"\n", + PFID(&op_data->op_fid3)); + + if (op_data->op_mea1 != NULL) { + struct lmv_stripe_md *lsm = op_data->op_mea1; + struct lmv_tgt_desc *tmp; + + /* Fix the parent fid for striped dir */ + tmp = lmv_locate_target_for_name(lmv, lsm, old, + oldlen, + &op_data->op_fid1, + NULL); + if (IS_ERR(tmp)) + RETURN(PTR_ERR(tmp)); + } + + rc = lmv_fid_alloc(NULL, exp, &op_data->op_fid2, op_data); + if (rc != 0) + RETURN(rc); + + src_tgt = lmv_find_target(lmv, &op_data->op_fid3); + if (IS_ERR(src_tgt)) + RETURN(PTR_ERR(src_tgt)); + + target_exp = src_tgt->ltd_exp; + } else { + if (op_data->op_mea1 != NULL) { + struct lmv_stripe_md *lsm = op_data->op_mea1; + + src_tgt = lmv_locate_target_for_name(lmv, lsm, old, + oldlen, + &op_data->op_fid1, + &op_data->op_mds); + } else { + src_tgt = lmv_find_target(lmv, &op_data->op_fid1); + } + if (IS_ERR(src_tgt)) + RETURN(PTR_ERR(src_tgt)); + + + if (op_data->op_mea2 != NULL) { + struct lmv_stripe_md *lsm = op_data->op_mea2; + + tgt_tgt = lmv_locate_target_for_name(lmv, lsm, new, + newlen, + &op_data->op_fid2, + &op_data->op_mds); + } else { + tgt_tgt = lmv_find_target(lmv, &op_data->op_fid2); + + } + if (IS_ERR(tgt_tgt)) + RETURN(PTR_ERR(tgt_tgt)); + + target_exp = tgt_tgt->ltd_exp; + } - tgt_tgt = lmv_locate_mds(lmv, op_data, &op_data->op_fid2); - if (IS_ERR(tgt_tgt)) - RETURN(PTR_ERR(tgt_tgt)); /* * LOOKUP lock on src child (fid3) should also be cancelled for * src_tgt in mdc_rename. @@ -2068,37 +2128,86 @@ static int lmv_rename(struct obd_export *exp, struct md_op_data *op_data, * Cancel UPDATE locks on tgt parent (fid2), tgt_tgt is its * own target. */ - rc = lmv_early_cancel(exp, op_data, src_tgt->ltd_idx, + rc = lmv_early_cancel(exp, NULL, op_data, src_tgt->ltd_idx, LCK_EX, MDS_INODELOCK_UPDATE, MF_MDC_CANCEL_FID2); + if (rc != 0) + RETURN(rc); /* - * Cancel LOOKUP locks on tgt child (fid4) for parent tgt_tgt. + * Cancel LOOKUP locks on source child (fid3) for parent tgt_tgt. */ - if (rc == 0) { - rc = lmv_early_cancel(exp, op_data, src_tgt->ltd_idx, + if (fid_is_sane(&op_data->op_fid3)) { + struct lmv_tgt_desc *tgt; + + tgt = lmv_find_target(lmv, &op_data->op_fid1); + if (IS_ERR(tgt)) + RETURN(PTR_ERR(tgt)); + + /* Cancel LOOKUP lock on its parent */ + rc = lmv_early_cancel(exp, tgt, op_data, src_tgt->ltd_idx, LCK_EX, MDS_INODELOCK_LOOKUP, - MF_MDC_CANCEL_FID4); + MF_MDC_CANCEL_FID3); + if (rc != 0) + RETURN(rc); + + rc = lmv_early_cancel(exp, NULL, op_data, src_tgt->ltd_idx, + LCK_EX, MDS_INODELOCK_FULL, + MF_MDC_CANCEL_FID3); + if (rc != 0) + RETURN(rc); } +retry_rename: /* * Cancel all the locks on tgt child (fid4). */ - if (rc == 0) - rc = lmv_early_cancel(exp, op_data, src_tgt->ltd_idx, + if (fid_is_sane(&op_data->op_fid4)) { + struct lmv_tgt_desc *tgt; + + rc = lmv_early_cancel(exp, NULL, op_data, src_tgt->ltd_idx, LCK_EX, MDS_INODELOCK_FULL, MF_MDC_CANCEL_FID4); + if (rc != 0) + RETURN(rc); - if (rc == 0) - rc = md_rename(src_tgt->ltd_exp, op_data, old, oldlen, - new, newlen, request); - RETURN(rc); + tgt = lmv_find_target(lmv, &op_data->op_fid4); + if (IS_ERR(tgt)) + RETURN(PTR_ERR(tgt)); + + /* Since the target child might be destroyed, and it might + * become orphan, and we can only check orphan on the local + * MDT right now, so we send rename request to the MDT where + * target child is located. If target child does not exist, + * then it will send the request to the target parent */ + target_exp = tgt->ltd_exp; + } + + rc = md_rename(target_exp, op_data, old, oldlen, new, newlen, + request); + + if (rc != 0 && rc != -EXDEV) + RETURN(rc); + + body = req_capsule_server_get(&(*request)->rq_pill, &RMF_MDT_BODY); + if (body == NULL) + RETURN(-EPROTO); + + /* Not cross-ref case, just get out of here. */ + if (likely(!(body->mbo_valid & OBD_MD_MDS))) + RETURN(rc); + + CDEBUG(D_INODE, "%s: try rename to another MDT for "DFID"\n", + exp->exp_obd->obd_name, PFID(&body->mbo_fid1)); + + op_data->op_fid4 = body->mbo_fid1; + ptlrpc_req_finished(*request); + *request = NULL; + goto retry_rename; } static int lmv_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) + void *ea, size_t ealen, struct ptlrpc_request **request) { struct obd_device *obd = exp->exp_obd; struct lmv_obd *lmv = &obd->u.lmv; @@ -2118,14 +2227,13 @@ static int lmv_setattr(struct obd_export *exp, struct md_op_data *op_data, if (IS_ERR(tgt)) RETURN(PTR_ERR(tgt)); - rc = md_setattr(tgt->ltd_exp, op_data, ea, ealen, ea2, - ea2len, request, mod); + rc = md_setattr(tgt->ltd_exp, op_data, ea, ealen, request); RETURN(rc); } static int lmv_fsync(struct obd_export *exp, const struct lu_fid *fid, - struct obd_capa *oc, struct ptlrpc_request **request) + struct ptlrpc_request **request) { struct obd_device *obd = exp->exp_obd; struct lmv_obd *lmv = &obd->u.lmv; @@ -2141,178 +2249,414 @@ static int lmv_fsync(struct obd_export *exp, const struct lu_fid *fid, if (IS_ERR(tgt)) RETURN(PTR_ERR(tgt)); - rc = md_fsync(tgt->ltd_exp, fid, oc, request); + rc = md_fsync(tgt->ltd_exp, fid, request); RETURN(rc); } -/* - * Adjust a set of pages, each page containing an array of lu_dirpages, - * so that each page can be used as a single logical lu_dirpage. - * - * A lu_dirpage is laid out as follows, where s = ldp_hash_start, - * e = ldp_hash_end, f = ldp_flags, p = padding, and each "ent" is a - * struct lu_dirent. It has size up to LU_PAGE_SIZE. The ldp_hash_end - * value is used as a cookie to request the next lu_dirpage in a - * directory listing that spans multiple pages (two in this example): - * ________ - * | | - * .|--------v------- -----. - * |s|e|f|p|ent|ent| ... |ent| - * '--|-------------- -----' Each CFS_PAGE contains a single - * '------. lu_dirpage. - * .---------v------- -----. - * |s|e|f|p|ent| 0 | ... | 0 | - * '----------------- -----' - * - * However, on hosts where the native VM page size (PAGE_CACHE_SIZE) is - * larger than LU_PAGE_SIZE, a single host page may contain multiple - * lu_dirpages. After reading the lu_dirpages from the MDS, the - * ldp_hash_end of the first lu_dirpage refers to the one immediately - * after it in the same CFS_PAGE (arrows simplified for brevity, but - * in general e0==s1, e1==s2, etc.): - * - * .-------------------- -----. - * |s0|e0|f0|p|ent|ent| ... |ent| - * |---v---------------- -----| - * |s1|e1|f1|p|ent|ent| ... |ent| - * |---v---------------- -----| Here, each CFS_PAGE contains - * ... multiple lu_dirpages. - * |---v---------------- -----| - * |s'|e'|f'|p|ent|ent| ... |ent| - * '---|---------------- -----' - * v - * .----------------------------. - * | next CFS_PAGE | +/** + * Get current minimum entry from striped directory * - * This structure is transformed into a single logical lu_dirpage as follows: + * This function will search the dir entry, whose hash value is the + * closest(>=) to @hash_offset, from all of sub-stripes, and it is + * only being called for striped directory. * - * - Replace e0 with e' so the request for the next lu_dirpage gets the page - * labeled 'next CFS_PAGE'. + * \param[in] exp export of LMV + * \param[in] op_data parameters transferred beween client MD stack + * stripe_information will be included in this + * parameter + * \param[in] cb_op ldlm callback being used in enqueue in + * mdc_read_page + * \param[in] hash_offset the hash value, which is used to locate + * minum(closet) dir entry + * \param[in|out] stripe_offset the caller use this to indicate the stripe + * index of last entry, so to avoid hash conflict + * between stripes. It will also be used to + * return the stripe index of current dir entry. + * \param[in|out] entp the minum entry and it also is being used + * to input the last dir entry to resolve the + * hash conflict * - * - Copy the LDF_COLLIDE flag from f' to f0 to correctly reflect whether - * a hash collision with the next page exists. + * \param[out] ppage the page which holds the minum entry * - * - Adjust the lde_reclen of the ending entry of each lu_dirpage to span - * to the first entry of the next lu_dirpage. + * \retval = 0 get the entry successfully + * negative errno (< 0) does not get the entry */ -#if PAGE_CACHE_SIZE > LU_PAGE_SIZE -static void lmv_adjust_dirpages(struct page **pages, int ncfspgs, int nlupgs) +static int lmv_get_min_striped_entry(struct obd_export *exp, + struct md_op_data *op_data, + struct md_callback *cb_op, + __u64 hash_offset, int *stripe_offset, + struct lu_dirent **entp, + struct page **ppage) { - int i; + struct obd_device *obd = exp->exp_obd; + struct lmv_obd *lmv = &obd->u.lmv; + struct lmv_stripe_md *lsm = op_data->op_mea1; + struct lmv_tgt_desc *tgt; + int stripe_count; + struct lu_dirent *min_ent = NULL; + struct page *min_page = NULL; + int min_idx = 0; + int i; + int rc = 0; + ENTRY; - for (i = 0; i < ncfspgs; i++) { - struct lu_dirpage *dp = kmap(pages[i]); - struct lu_dirpage *first = dp; - struct lu_dirent *end_dirent = NULL; - struct lu_dirent *ent; - __u64 hash_end = dp->ldp_hash_end; - __u32 flags = dp->ldp_flags; + stripe_count = lsm->lsm_md_stripe_count; + for (i = 0; i < stripe_count; i++) { + struct lu_dirent *ent = NULL; + struct page *page = NULL; + struct lu_dirpage *dp; + __u64 stripe_hash = hash_offset; - while (--nlupgs > 0) { - ent = lu_dirent_start(dp); - for (end_dirent = ent; ent != NULL; - end_dirent = ent, ent = lu_dirent_next(ent)); + tgt = lmv_get_target(lmv, lsm->lsm_md_oinfo[i].lmo_mds, NULL); + if (IS_ERR(tgt)) + GOTO(out, rc = PTR_ERR(tgt)); + + /* op_data will be shared by each stripe, so we need + * reset these value for each stripe */ + op_data->op_fid1 = lsm->lsm_md_oinfo[i].lmo_fid; + op_data->op_fid2 = lsm->lsm_md_oinfo[i].lmo_fid; + op_data->op_data = lsm->lsm_md_oinfo[i].lmo_root; +next: + rc = md_read_page(tgt->ltd_exp, op_data, cb_op, stripe_hash, + &page); + if (rc != 0) + GOTO(out, rc); + + dp = page_address(page); + for (ent = lu_dirent_start(dp); ent != NULL; + ent = lu_dirent_next(ent)) { + /* Skip dummy entry */ + if (le16_to_cpu(ent->lde_namelen) == 0) + continue; - /* Advance dp to next lu_dirpage. */ - dp = (struct lu_dirpage *)((char *)dp + LU_PAGE_SIZE); + if (le64_to_cpu(ent->lde_hash) < hash_offset) + continue; - /* Check if we've reached the end of the CFS_PAGE. */ - if (!((unsigned long)dp & ~CFS_PAGE_MASK)) - break; + if (le64_to_cpu(ent->lde_hash) == hash_offset && + (*entp == ent || i < *stripe_offset)) + continue; - /* Save the hash and flags of this lu_dirpage. */ - hash_end = dp->ldp_hash_end; - flags = dp->ldp_flags; + /* skip . and .. for other stripes */ + if (i != 0 && + (strncmp(ent->lde_name, ".", + le16_to_cpu(ent->lde_namelen)) == 0 || + strncmp(ent->lde_name, "..", + le16_to_cpu(ent->lde_namelen)) == 0)) + continue; + break; + } - /* Check if lu_dirpage contains no entries. */ - if (!end_dirent) - break; + if (ent == NULL) { + stripe_hash = le64_to_cpu(dp->ldp_hash_end); + + kunmap(page); + page_cache_release(page); + page = NULL; - /* Enlarge the end entry lde_reclen from 0 to - * first entry of next lu_dirpage. */ - LASSERT(le16_to_cpu(end_dirent->lde_reclen) == 0); - end_dirent->lde_reclen = - cpu_to_le16((char *)(dp->ldp_entries) - - (char *)end_dirent); + /* reach the end of current stripe, go to next stripe */ + if (stripe_hash == MDS_DIR_END_OFF) + continue; + else + goto next; } - first->ldp_hash_end = hash_end; - first->ldp_flags &= ~cpu_to_le32(LDF_COLLIDE); - first->ldp_flags |= flags & cpu_to_le32(LDF_COLLIDE); + if (min_ent != NULL) { + if (le64_to_cpu(min_ent->lde_hash) > + le64_to_cpu(ent->lde_hash)) { + min_ent = ent; + kunmap(min_page); + page_cache_release(min_page); + min_idx = i; + min_page = page; + } else { + kunmap(page); + page_cache_release(page); + page = NULL; + } + } else { + min_ent = ent; + min_page = page; + min_idx = i; + } + } - kunmap(pages[i]); +out: + if (*ppage != NULL) { + kunmap(*ppage); + page_cache_release(*ppage); } - LASSERTF(nlupgs == 0, "left = %d", nlupgs); + *stripe_offset = min_idx; + *entp = min_ent; + *ppage = min_page; + RETURN(rc); } -#else -#define lmv_adjust_dirpages(pages, ncfspgs, nlupgs) do {} while (0) -#endif /* PAGE_CACHE_SIZE > LU_PAGE_SIZE */ -static int lmv_readpage(struct obd_export *exp, struct md_op_data *op_data, - struct page **pages, struct ptlrpc_request **request) +/** + * Build dir entry page from a striped directory + * + * This function gets one entry by @offset from a striped directory. It will + * read entries from all of stripes, and choose one closest to the required + * offset(&offset). A few notes + * 1. skip . and .. for non-zero stripes, because there can only have one . + * and .. in a directory. + * 2. op_data will be shared by all of stripes, instead of allocating new + * one, so need to restore before reusing. + * 3. release the entry page if that is not being chosen. + * + * \param[in] exp obd export refer to LMV + * \param[in] op_data hold those MD parameters of read_entry + * \param[in] cb_op ldlm callback being used in enqueue in mdc_read_entry + * \param[out] ldp the entry being read + * \param[out] ppage the page holding the entry. Note: because the entry + * will be accessed in upper layer, so we need hold the + * page until the usages of entry is finished, see + * ll_dir_entry_next. + * + * retval =0 if get entry successfully + * <0 cannot get entry + */ +static int lmv_read_striped_page(struct obd_export *exp, + struct md_op_data *op_data, + struct md_callback *cb_op, + __u64 offset, struct page **ppage) { struct obd_device *obd = exp->exp_obd; - struct lmv_obd *lmv = &obd->u.lmv; - __u64 offset = op_data->op_offset; + struct lu_fid master_fid = op_data->op_fid1; + struct inode *master_inode = op_data->op_data; + __u64 hash_offset = offset; + struct lu_dirpage *dp; + struct page *min_ent_page = NULL; + struct page *ent_page = NULL; + struct lu_dirent *ent; + void *area; + int ent_idx = 0; + struct lu_dirent *min_ent = NULL; + struct lu_dirent *last_ent; + size_t left_bytes; int rc; - int ncfspgs; /* pages read in PAGE_CACHE_SIZE */ - int nlupgs; /* pages read in LU_PAGE_SIZE */ - struct lmv_tgt_desc *tgt; ENTRY; rc = lmv_check_connect(obd); if (rc) RETURN(rc); - CDEBUG(D_INODE, "READPAGE at "LPX64" from "DFID"\n", - offset, PFID(&op_data->op_fid1)); + /* Allocate a page and read entries from all of stripes and fill + * the page by hash order */ + ent_page = alloc_page(GFP_KERNEL); + if (ent_page == NULL) + RETURN(-ENOMEM); - tgt = lmv_find_target(lmv, &op_data->op_fid1); - if (IS_ERR(tgt)) - RETURN(PTR_ERR(tgt)); + /* Initialize the entry page */ + dp = kmap(ent_page); + memset(dp, 0, sizeof(*dp)); + dp->ldp_hash_start = cpu_to_le64(offset); + dp->ldp_flags |= LDF_COLLIDE; + + area = dp + 1; + left_bytes = PAGE_CACHE_SIZE - sizeof(*dp); + ent = area; + last_ent = ent; + do { + __u16 ent_size; + + /* Find the minum entry from all sub-stripes */ + rc = lmv_get_min_striped_entry(exp, op_data, cb_op, hash_offset, + &ent_idx, &min_ent, + &min_ent_page); + if (rc != 0) + GOTO(out, rc); + + /* If it can not get minum entry, it means it already reaches + * the end of this directory */ + if (min_ent == NULL) { + last_ent->lde_reclen = 0; + hash_offset = MDS_DIR_END_OFF; + GOTO(out, rc); + } + + ent_size = le16_to_cpu(min_ent->lde_reclen); + + /* the last entry lde_reclen is 0, but it might not + * the end of this entry of this temporay entry */ + if (ent_size == 0) + ent_size = lu_dirent_calc_size( + le16_to_cpu(min_ent->lde_namelen), + le32_to_cpu(min_ent->lde_attrs)); + if (ent_size > left_bytes) { + last_ent->lde_reclen = cpu_to_le16(0); + hash_offset = le64_to_cpu(min_ent->lde_hash); + GOTO(out, rc); + } + + memcpy(ent, min_ent, ent_size); + + /* Replace . with master FID and Replace .. with the parent FID + * of master object */ + if (strncmp(ent->lde_name, ".", + le16_to_cpu(ent->lde_namelen)) == 0 && + le16_to_cpu(ent->lde_namelen) == 1) + fid_cpu_to_le(&ent->lde_fid, &master_fid); + else if (strncmp(ent->lde_name, "..", + le16_to_cpu(ent->lde_namelen)) == 0 && + le16_to_cpu(ent->lde_namelen) == 2) + fid_cpu_to_le(&ent->lde_fid, &op_data->op_fid3); + + left_bytes -= ent_size; + ent->lde_reclen = cpu_to_le16(ent_size); + last_ent = ent; + ent = (void *)ent + ent_size; + hash_offset = le64_to_cpu(min_ent->lde_hash); + if (hash_offset == MDS_DIR_END_OFF) { + last_ent->lde_reclen = 0; + break; + } + } while (1); +out: + if (min_ent_page != NULL) { + kunmap(min_ent_page); + page_cache_release(min_ent_page); + } + + if (unlikely(rc != 0)) { + __free_page(ent_page); + ent_page = NULL; + } else { + if (ent == area) + dp->ldp_flags |= LDF_EMPTY; + dp->ldp_flags = cpu_to_le32(dp->ldp_flags); + dp->ldp_hash_end = cpu_to_le64(hash_offset); + } + + /* We do not want to allocate md_op_data during each + * dir entry reading, so op_data will be shared by every stripe, + * then we need to restore it back to original value before + * return to the upper layer */ + op_data->op_fid1 = master_fid; + op_data->op_fid2 = master_fid; + op_data->op_data = master_inode; + + *ppage = ent_page; - rc = md_readpage(tgt->ltd_exp, op_data, pages, request); + RETURN(rc); +} + +int lmv_read_page(struct obd_export *exp, struct md_op_data *op_data, + struct md_callback *cb_op, __u64 offset, + struct page **ppage) +{ + struct obd_device *obd = exp->exp_obd; + struct lmv_obd *lmv = &obd->u.lmv; + struct lmv_stripe_md *lsm = op_data->op_mea1; + struct lmv_tgt_desc *tgt; + int rc; + ENTRY; + + rc = lmv_check_connect(obd); if (rc != 0) RETURN(rc); - ncfspgs = ((*request)->rq_bulk->bd_nob_transferred + - PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; - nlupgs = (*request)->rq_bulk->bd_nob_transferred >> LU_PAGE_SHIFT; - LASSERT(!((*request)->rq_bulk->bd_nob_transferred & ~LU_PAGE_MASK)); - LASSERT(ncfspgs > 0 && ncfspgs <= op_data->op_npages); + if (unlikely(lsm != NULL)) { + rc = lmv_read_striped_page(exp, op_data, cb_op, offset, ppage); + RETURN(rc); + } - CDEBUG(D_INODE, "read %d(%d)/%d pages\n", ncfspgs, nlupgs, - op_data->op_npages); + tgt = lmv_find_target(lmv, &op_data->op_fid1); + if (IS_ERR(tgt)) + RETURN(PTR_ERR(tgt)); - lmv_adjust_dirpages(pages, ncfspgs, nlupgs); + rc = md_read_page(tgt->ltd_exp, op_data, cb_op, offset, ppage); RETURN(rc); } +/** + * Unlink a file/directory + * + * Unlink a file or directory under the parent dir. The unlink request + * usually will be sent to the MDT where the child is located, but if + * the client does not have the child FID then request will be sent to the + * MDT where the parent is located. + * + * If the parent is a striped directory then it also needs to locate which + * stripe the name of the child is located, and replace the parent FID + * (@op->op_fid1) with the stripe FID. Note: if the stripe is unknown, + * it will walk through all of sub-stripes until the child is being + * unlinked finally. + * + * \param[in] exp export refer to LMV + * \param[in] op_data different parameters transferred beween client + * MD stacks, name, namelen, FIDs etc. + * op_fid1 is the parent FID, op_fid2 is the child + * FID. + * \param[out] request point to the request of unlink. + * + * retval 0 if succeed + * negative errno if failed. + */ static int lmv_unlink(struct obd_export *exp, struct md_op_data *op_data, struct ptlrpc_request **request) { struct obd_device *obd = exp->exp_obd; struct lmv_obd *lmv = &obd->u.lmv; struct lmv_tgt_desc *tgt = NULL; + struct lmv_tgt_desc *parent_tgt = NULL; struct mdt_body *body; int rc; + int stripe_index = 0; + struct lmv_stripe_md *lsm = op_data->op_mea1; ENTRY; rc = lmv_check_connect(obd); if (rc) RETURN(rc); -retry: +retry_unlink: + /* For striped dir, we need to locate the parent as well */ + if (lsm != NULL) { + struct lmv_tgt_desc *tmp; + + LASSERT(op_data->op_name != NULL && + op_data->op_namelen != 0); + + tmp = lmv_locate_target_for_name(lmv, lsm, + op_data->op_name, + op_data->op_namelen, + &op_data->op_fid1, + &op_data->op_mds); + + /* return -EBADFD means unknown hash type, might + * need try all sub-stripe here */ + if (IS_ERR(tmp) && PTR_ERR(tmp) != -EBADFD) + RETURN(PTR_ERR(tmp)); + + /* Note: both migrating dir and unknown hash dir need to + * try all of sub-stripes, so we need start search the + * name from stripe 0, but migrating dir is already handled + * inside lmv_locate_target_for_name(), so we only check + * unknown hash type directory here */ + if (!lmv_is_known_hash_type(lsm->lsm_md_hash_type)) { + struct lmv_oinfo *oinfo; + + oinfo = &lsm->lsm_md_oinfo[stripe_index]; + + op_data->op_fid1 = oinfo->lmo_fid; + op_data->op_mds = oinfo->lmo_mds; + } + } + +try_next_stripe: /* Send unlink requests to the MDT where the child is located */ if (likely(!fid_is_zero(&op_data->op_fid2))) - tgt = lmv_locate_mds(lmv, op_data, &op_data->op_fid2); + tgt = lmv_find_target(lmv, &op_data->op_fid2); + else if (lsm != NULL) + tgt = lmv_get_target(lmv, op_data->op_mds, NULL); else tgt = lmv_locate_mds(lmv, op_data, &op_data->op_fid1); + if (IS_ERR(tgt)) RETURN(PTR_ERR(tgt)); - op_data->op_fsuid = current_fsuid(); - op_data->op_fsgid = current_fsgid(); + op_data->op_fsuid = from_kuid(&init_user_ns, current_fsuid()); + op_data->op_fsgid = from_kgid(&init_user_ns, current_fsgid()); op_data->op_cap = cfs_curproc_cap_pack(); /* @@ -2327,29 +2671,57 @@ retry: /* * Cancel FULL locks on child (fid3). */ - rc = lmv_early_cancel(exp, op_data, tgt->ltd_idx, LCK_EX, - MDS_INODELOCK_FULL, MF_MDC_CANCEL_FID3); + parent_tgt = lmv_find_target(lmv, &op_data->op_fid1); + if (IS_ERR(parent_tgt)) + RETURN(PTR_ERR(parent_tgt)); + if (parent_tgt != tgt) { + rc = lmv_early_cancel(exp, parent_tgt, op_data, tgt->ltd_idx, + LCK_EX, MDS_INODELOCK_LOOKUP, + MF_MDC_CANCEL_FID3); + } + + rc = lmv_early_cancel(exp, NULL, op_data, tgt->ltd_idx, LCK_EX, + MDS_INODELOCK_FULL, MF_MDC_CANCEL_FID3); if (rc != 0) RETURN(rc); - CDEBUG(D_INODE, "unlink with fid="DFID"/"DFID" -> mds #%d\n", + CDEBUG(D_INODE, "unlink with fid="DFID"/"DFID" -> mds #%u\n", PFID(&op_data->op_fid1), PFID(&op_data->op_fid2), tgt->ltd_idx); rc = md_unlink(tgt->ltd_exp, op_data, request); - if (rc != 0 && rc != -EREMOTE) + if (rc != 0 && rc != -EREMOTE && rc != -ENOENT) RETURN(rc); + /* Try next stripe if it is needed. */ + if (rc == -ENOENT && lsm != NULL && lmv_need_try_all_stripes(lsm)) { + struct lmv_oinfo *oinfo; + + stripe_index++; + if (stripe_index >= lsm->lsm_md_stripe_count) + RETURN(rc); + + oinfo = &lsm->lsm_md_oinfo[stripe_index]; + + op_data->op_fid1 = oinfo->lmo_fid; + op_data->op_mds = oinfo->lmo_mds; + + ptlrpc_req_finished(*request); + *request = NULL; + + goto try_next_stripe; + } + body = req_capsule_server_get(&(*request)->rq_pill, &RMF_MDT_BODY); if (body == NULL) RETURN(-EPROTO); /* Not cross-ref case, just get out of here. */ - if (likely(!(body->valid & OBD_MD_MDS))) - RETURN(0); + if (likely(!(body->mbo_valid & OBD_MD_MDS))) + RETURN(rc); CDEBUG(D_INODE, "%s: try unlink to another MDT for "DFID"\n", - exp->exp_obd->obd_name, PFID(&body->fid1)); + exp->exp_obd->obd_name, PFID(&body->mbo_fid1)); /* This is a remote object, try remote MDT, Note: it may * try more than 1 time here, Considering following case @@ -2370,11 +2742,11 @@ retry: * * In theory, it might try unlimited time here, but it should * be very rare case. */ - op_data->op_fid2 = body->fid1; + op_data->op_fid2 = body->mbo_fid1; ptlrpc_req_finished(*request); *request = NULL; - goto retry; + goto retry_unlink; } static int lmv_precleanup(struct obd_device *obd, enum obd_cleanup_stage stage) @@ -2398,9 +2770,24 @@ static int lmv_precleanup(struct obd_device *obd, enum obd_cleanup_stage stage) RETURN(rc); } +/** + * Get by key a value associated with a LMV device. + * + * Dispatch request to lower-layer devices as needed. + * + * \param[in] env execution environment for this thread + * \param[in] exp export for the LMV device + * \param[in] keylen length of key identifier + * \param[in] key identifier of key to get value for + * \param[in] vallen size of \a val + * \param[out] val pointer to storage location for value + * \param[in] lsm optional striping metadata of object + * + * \retval 0 on success + * \retval negative negated errno on failure + */ static int lmv_get_info(const struct lu_env *env, struct obd_export *exp, - __u32 keylen, void *key, __u32 *vallen, void *val, - struct lov_stripe_md *lsm) + __u32 keylen, void *key, __u32 *vallen, void *val) { struct obd_device *obd; struct lmv_obd *lmv; @@ -2432,21 +2819,23 @@ static int lmv_get_info(const struct lu_env *env, struct obd_export *exp, continue; if (!obd_get_info(env, tgt->ltd_exp, keylen, key, - vallen, val, NULL)) + vallen, val)) RETURN(0); - } - RETURN(-EINVAL); - } else if (KEY_IS(KEY_MAX_EASIZE) || KEY_IS(KEY_CONN_DATA)) { - rc = lmv_check_connect(obd); - if (rc) - RETURN(rc); + } + RETURN(-EINVAL); + } else if (KEY_IS(KEY_MAX_EASIZE) || + KEY_IS(KEY_DEFAULT_EASIZE) || + KEY_IS(KEY_CONN_DATA)) { + rc = lmv_check_connect(obd); + if (rc) + RETURN(rc); /* * Forwarding this request to first MDS, it should know LOV * desc. */ rc = obd_get_info(env, lmv->tgts[0]->ltd_exp, keylen, key, - vallen, val, NULL); + vallen, val); if (!rc && KEY_IS(KEY_CONN_DATA)) exp->exp_connect_data = *(struct obd_connect_data *)val; RETURN(rc); @@ -2459,26 +2848,43 @@ static int lmv_get_info(const struct lu_env *env, struct obd_export *exp, RETURN(-EINVAL); } +/** + * Asynchronously set by key a value associated with a LMV device. + * + * Dispatch request to lower-layer devices as needed. + * + * \param[in] env execution environment for this thread + * \param[in] exp export for the LMV device + * \param[in] keylen length of key identifier + * \param[in] key identifier of key to store value for + * \param[in] vallen size of value to store + * \param[in] val pointer to data to be stored + * \param[in] set optional list of related ptlrpc requests + * + * \retval 0 on success + * \retval negative negated errno on failure + */ int lmv_set_info_async(const struct lu_env *env, struct obd_export *exp, - obd_count keylen, void *key, obd_count vallen, - void *val, struct ptlrpc_request_set *set) + __u32 keylen, void *key, __u32 vallen, void *val, + struct ptlrpc_request_set *set) { - struct lmv_tgt_desc *tgt = NULL; - struct obd_device *obd; - struct lmv_obd *lmv; - int rc = 0; - ENTRY; + struct lmv_tgt_desc *tgt = NULL; + struct obd_device *obd; + struct lmv_obd *lmv; + int rc = 0; + ENTRY; - obd = class_exp2obd(exp); - if (obd == NULL) { - CDEBUG(D_IOCTL, "Invalid client cookie "LPX64"\n", - exp->exp_handle.h_cookie); - RETURN(-EINVAL); - } - lmv = &obd->u.lmv; + obd = class_exp2obd(exp); + if (obd == NULL) { + CDEBUG(D_IOCTL, "Invalid client cookie "LPX64"\n", + exp->exp_handle.h_cookie); + RETURN(-EINVAL); + } + lmv = &obd->u.lmv; - if (KEY_IS(KEY_READ_ONLY) || KEY_IS(KEY_FLUSH_CTX)) { - int i, err = 0; + if (KEY_IS(KEY_READ_ONLY) || KEY_IS(KEY_FLUSH_CTX) || + KEY_IS(KEY_DEFAULT_EASIZE)) { + int i, err = 0; for (i = 0; i < lmv->desc.ld_tgt_count; i++) { tgt = lmv->tgts[i]; @@ -2486,121 +2892,145 @@ int lmv_set_info_async(const struct lu_env *env, struct obd_export *exp, if (tgt == NULL || tgt->ltd_exp == NULL) continue; - err = obd_set_info_async(env, tgt->ltd_exp, - keylen, key, vallen, val, set); - if (err && rc == 0) - rc = err; - } + err = obd_set_info_async(env, tgt->ltd_exp, + keylen, key, vallen, val, set); + if (err && rc == 0) + rc = err; + } - RETURN(rc); - } + RETURN(rc); + } - RETURN(-EINVAL); + RETURN(-EINVAL); } -int lmv_packmd(struct obd_export *exp, struct lov_mds_md **lmmp, - struct lov_stripe_md *lsm) +static int lmv_unpack_md_v1(struct obd_export *exp, struct lmv_stripe_md *lsm, + const struct lmv_mds_md_v1 *lmm1) { - struct obd_device *obd = class_exp2obd(exp); - struct lmv_obd *lmv = &obd->u.lmv; - struct lmv_stripe_md *meap; - struct lmv_stripe_md *lsmp; - int mea_size; - __u32 i; - ENTRY; - - mea_size = lmv_get_easize(lmv); - if (!lmmp) - RETURN(mea_size); - - if (*lmmp && !lsm) { - OBD_FREE_LARGE(*lmmp, mea_size); - *lmmp = NULL; - RETURN(0); - } - - if (*lmmp == NULL) { - OBD_ALLOC_LARGE(*lmmp, mea_size); - if (*lmmp == NULL) - RETURN(-ENOMEM); - } - - if (!lsm) - RETURN(mea_size); - - lsmp = (struct lmv_stripe_md *)lsm; - meap = (struct lmv_stripe_md *)*lmmp; - - if (lsmp->mea_magic != MEA_MAGIC_LAST_CHAR && - lsmp->mea_magic != MEA_MAGIC_ALL_CHARS) - RETURN(-EINVAL); - - meap->mea_magic = cpu_to_le32(lsmp->mea_magic); - meap->mea_count = cpu_to_le32(lsmp->mea_count); - meap->mea_master = cpu_to_le32(lsmp->mea_master); + struct lmv_obd *lmv = &exp->exp_obd->u.lmv; + int stripe_count; + int cplen; + int i; + int rc = 0; + ENTRY; - for (i = 0; i < lmv->desc.ld_tgt_count; i++) { - meap->mea_ids[i] = lsmp->mea_ids[i]; - fid_cpu_to_le(&meap->mea_ids[i], &lsmp->mea_ids[i]); + lsm->lsm_md_magic = le32_to_cpu(lmm1->lmv_magic); + lsm->lsm_md_stripe_count = le32_to_cpu(lmm1->lmv_stripe_count); + lsm->lsm_md_master_mdt_index = le32_to_cpu(lmm1->lmv_master_mdt_index); + if (OBD_FAIL_CHECK(OBD_FAIL_UNKNOWN_LMV_STRIPE)) + lsm->lsm_md_hash_type = LMV_HASH_TYPE_UNKNOWN; + else + lsm->lsm_md_hash_type = le32_to_cpu(lmm1->lmv_hash_type); + lsm->lsm_md_layout_version = le32_to_cpu(lmm1->lmv_layout_version); + cplen = strlcpy(lsm->lsm_md_pool_name, lmm1->lmv_pool_name, + sizeof(lsm->lsm_md_pool_name)); + + if (cplen >= sizeof(lsm->lsm_md_pool_name)) + RETURN(-E2BIG); + + CDEBUG(D_INFO, "unpack lsm count %d, master %d hash_type %d" + "layout_version %d\n", lsm->lsm_md_stripe_count, + lsm->lsm_md_master_mdt_index, lsm->lsm_md_hash_type, + lsm->lsm_md_layout_version); + + stripe_count = le32_to_cpu(lmm1->lmv_stripe_count); + for (i = 0; i < stripe_count; i++) { + fid_le_to_cpu(&lsm->lsm_md_oinfo[i].lmo_fid, + &lmm1->lmv_stripe_fids[i]); + rc = lmv_fld_lookup(lmv, &lsm->lsm_md_oinfo[i].lmo_fid, + &lsm->lsm_md_oinfo[i].lmo_mds); + if (rc != 0) + RETURN(rc); + CDEBUG(D_INFO, "unpack fid #%d "DFID"\n", i, + PFID(&lsm->lsm_md_oinfo[i].lmo_fid)); } - RETURN(mea_size); + RETURN(rc); } -int lmv_unpackmd(struct obd_export *exp, struct lov_stripe_md **lsmp, - struct lov_mds_md *lmm, int lmm_size) +static int lmv_unpackmd(struct obd_export *exp, struct lmv_stripe_md **lsmp, + const union lmv_mds_md *lmm, size_t lmm_size) { - struct obd_device *obd = class_exp2obd(exp); - struct lmv_stripe_md **tmea = (struct lmv_stripe_md **)lsmp; - struct lmv_stripe_md *mea = (struct lmv_stripe_md *)lmm; - struct lmv_obd *lmv = &obd->u.lmv; - int mea_size; - __u32 i; - __u32 magic; - ENTRY; + struct lmv_stripe_md *lsm; + int lsm_size; + int rc; + bool allocated = false; + ENTRY; - mea_size = lmv_get_easize(lmv); - if (lsmp == NULL) - return mea_size; + LASSERT(lsmp != NULL); - if (*lsmp != NULL && lmm == NULL) { - OBD_FREE_LARGE(*tmea, mea_size); - *lsmp = NULL; - RETURN(0); - } + lsm = *lsmp; + /* Free memmd */ + if (lsm != NULL && lmm == NULL) { + int i; + for (i = 0; i < lsm->lsm_md_stripe_count; i++) { + /* For migrating inode, the master stripe and master + * object will be the same, so do not need iput, see + * ll_update_lsm_md */ + if (!(lsm->lsm_md_hash_type & LMV_HASH_FLAG_MIGRATION && + i == 0) && lsm->lsm_md_oinfo[i].lmo_root != NULL) + iput(lsm->lsm_md_oinfo[i].lmo_root); + } + lsm_size = lmv_stripe_md_size(lsm->lsm_md_stripe_count); + OBD_FREE(lsm, lsm_size); + *lsmp = NULL; + RETURN(0); + } - LASSERT(mea_size == lmm_size); + if (le32_to_cpu(lmm->lmv_magic) == LMV_MAGIC_STRIPE) + RETURN(-EPERM); - OBD_ALLOC_LARGE(*tmea, mea_size); - if (*tmea == NULL) - RETURN(-ENOMEM); + /* Unpack memmd */ + if (le32_to_cpu(lmm->lmv_magic) != LMV_MAGIC_V1 && + le32_to_cpu(lmm->lmv_magic) != LMV_USER_MAGIC) { + CERROR("%s: invalid lmv magic %x: rc = %d\n", + exp->exp_obd->obd_name, le32_to_cpu(lmm->lmv_magic), + -EIO); + RETURN(-EIO); + } + + if (le32_to_cpu(lmm->lmv_magic) == LMV_MAGIC_V1) + lsm_size = lmv_stripe_md_size(lmv_mds_md_stripe_count_get(lmm)); + else + /** + * Unpack default dirstripe(lmv_user_md) to lmv_stripe_md, + * stripecount should be 0 then. + */ + lsm_size = lmv_stripe_md_size(0); - if (!lmm) - RETURN(mea_size); + lsm_size = lmv_stripe_md_size(lmv_mds_md_stripe_count_get(lmm)); + if (lsm == NULL) { + OBD_ALLOC(lsm, lsm_size); + if (lsm == NULL) + RETURN(-ENOMEM); + allocated = true; + *lsmp = lsm; + } - if (mea->mea_magic == MEA_MAGIC_LAST_CHAR || - mea->mea_magic == MEA_MAGIC_ALL_CHARS || - mea->mea_magic == MEA_MAGIC_HASH_SEGMENT) - { - magic = le32_to_cpu(mea->mea_magic); - } else { - /* - * Old mea is not handled here. - */ - CERROR("Old not supportable EA is found\n"); - LBUG(); - } + switch (le32_to_cpu(lmm->lmv_magic)) { + case LMV_MAGIC_V1: + rc = lmv_unpack_md_v1(exp, lsm, &lmm->lmv_md_v1); + break; + default: + CERROR("%s: unrecognized magic %x\n", exp->exp_obd->obd_name, + le32_to_cpu(lmm->lmv_magic)); + rc = -EINVAL; + break; + } - (*tmea)->mea_magic = magic; - (*tmea)->mea_count = le32_to_cpu(mea->mea_count); - (*tmea)->mea_master = le32_to_cpu(mea->mea_master); + if (rc != 0 && allocated) { + OBD_FREE(lsm, lsm_size); + *lsmp = NULL; + lsm_size = rc; + } + RETURN(lsm_size); +} - for (i = 0; i < (*tmea)->mea_count; i++) { - (*tmea)->mea_ids[i] = mea->mea_ids[i]; - fid_le_to_cpu(&(*tmea)->mea_ids[i], &(*tmea)->mea_ids[i]); - } - RETURN(mea_size); +void lmv_free_memmd(struct lmv_stripe_md *lsm) +{ + lmv_unpackmd(NULL, &lsm, NULL, 0); } +EXPORT_SYMBOL(lmv_free_memmd); static int lmv_cancel_unused(struct obd_export *exp, const struct lu_fid *fid, ldlm_policy_data_t *policy, ldlm_mode_t mode, @@ -2648,45 +3078,55 @@ ldlm_mode_t lmv_lock_match(struct obd_export *exp, __u64 flags, ldlm_policy_data_t *policy, ldlm_mode_t mode, struct lustre_handle *lockh) { - struct obd_device *obd = exp->exp_obd; - struct lmv_obd *lmv = &obd->u.lmv; - ldlm_mode_t rc; - __u32 i; - ENTRY; + struct obd_device *obd = exp->exp_obd; + struct lmv_obd *lmv = &obd->u.lmv; + ldlm_mode_t rc; + int tgt; + int i; + ENTRY; - CDEBUG(D_INODE, "Lock match for "DFID"\n", PFID(fid)); + CDEBUG(D_INODE, "Lock match for "DFID"\n", PFID(fid)); /* - * With CMD every object can have two locks in different namespaces: - * lookup lock in space of mds storing direntry and update/open lock in - * space of mds storing inode. Thus we check all targets, not only that - * one fid was created in. - */ - for (i = 0; i < lmv->desc.ld_tgt_count; i++) { - struct lmv_tgt_desc *tgt = lmv->tgts[i]; + * With DNE every object can have two locks in different namespaces: + * lookup lock in space of MDT storing direntry and update/open lock in + * space of MDT storing inode. Try the MDT that the FID maps to first, + * since this can be easily found, and only try others if that fails. + */ + for (i = 0, tgt = lmv_find_target_index(lmv, fid); + i < lmv->desc.ld_tgt_count; + i++, tgt = (tgt + 1) % lmv->desc.ld_tgt_count) { + if (tgt < 0) { + CDEBUG(D_HA, "%s: "DFID" is inaccessible: rc = %d\n", + obd->obd_name, PFID(fid), tgt); + tgt = 0; + } - if (tgt == NULL || tgt->ltd_exp == NULL || !tgt->ltd_active) + if (lmv->tgts[tgt] == NULL || + lmv->tgts[tgt]->ltd_exp == NULL || + lmv->tgts[tgt]->ltd_active == 0) continue; - rc = md_lock_match(tgt->ltd_exp, flags, fid, type, policy, mode, - lockh); - if (rc) - RETURN(rc); - } + rc = md_lock_match(lmv->tgts[tgt]->ltd_exp, flags, fid, + type, policy, mode, lockh); + if (rc) + RETURN(rc); + } - RETURN(0); + RETURN(0); } int lmv_get_lustre_md(struct obd_export *exp, struct ptlrpc_request *req, struct obd_export *dt_exp, struct obd_export *md_exp, struct lustre_md *md) { - struct lmv_obd *lmv = &exp->exp_obd->u.lmv; + struct lmv_obd *lmv = &exp->exp_obd->u.lmv; struct lmv_tgt_desc *tgt = lmv->tgts[0]; if (tgt == NULL || tgt->ltd_exp == NULL) RETURN(-EINVAL); - return md_get_lustre_md(tgt->ltd_exp, req, dt_exp, md_exp, md); + + return md_get_lustre_md(lmv->tgts[0]->ltd_exp, req, dt_exp, md_exp, md); } int lmv_free_lustre_md(struct obd_export *exp, struct lustre_md *md) @@ -2696,11 +3136,13 @@ int lmv_free_lustre_md(struct obd_export *exp, struct lustre_md *md) struct lmv_tgt_desc *tgt = lmv->tgts[0]; ENTRY; - if (md->mea) - obd_free_memmd(exp, (void *)&md->mea); + if (md->lmv != NULL) { + lmv_free_memmd(md->lmv); + md->lmv = NULL; + } if (tgt == NULL || tgt->ltd_exp == NULL) RETURN(-EINVAL); - RETURN(md_free_lustre_md(tgt->ltd_exp, md)); + RETURN(md_free_lustre_md(lmv->tgts[0]->ltd_exp, md)); } int lmv_set_open_replay_data(struct obd_export *exp, @@ -2734,10 +3176,8 @@ int lmv_clear_open_replay_data(struct obd_export *exp, RETURN(md_clear_open_replay_data(tgt->ltd_exp, och)); } -static int lmv_get_remote_perm(struct obd_export *exp, - const struct lu_fid *fid, - struct obd_capa *oc, __u32 suppgid, - struct ptlrpc_request **request) +static int lmv_get_remote_perm(struct obd_export *exp, const struct lu_fid *fid, + u32 suppgid, struct ptlrpc_request **request) { struct obd_device *obd = exp->exp_obd; struct lmv_obd *lmv = &obd->u.lmv; @@ -2753,40 +3193,8 @@ static int lmv_get_remote_perm(struct obd_export *exp, if (IS_ERR(tgt)) RETURN(PTR_ERR(tgt)); - rc = md_get_remote_perm(tgt->ltd_exp, fid, oc, suppgid, request); - RETURN(rc); -} - -static int lmv_renew_capa(struct obd_export *exp, struct obd_capa *oc, - renew_capa_cb_t cb) -{ - struct obd_device *obd = exp->exp_obd; - struct lmv_obd *lmv = &obd->u.lmv; - struct lmv_tgt_desc *tgt; - int rc; - ENTRY; - - rc = lmv_check_connect(obd); - if (rc) - RETURN(rc); - - tgt = lmv_find_target(lmv, &oc->c_capa.lc_fid); - if (IS_ERR(tgt)) - RETURN(PTR_ERR(tgt)); - - rc = md_renew_capa(tgt->ltd_exp, oc, cb); - RETURN(rc); -} - -int lmv_unpack_capa(struct obd_export *exp, struct ptlrpc_request *req, - const struct req_msg_field *field, struct obd_capa **oc) -{ - struct lmv_obd *lmv = &exp->exp_obd->u.lmv; - struct lmv_tgt_desc *tgt = lmv->tgts[0]; - - if (tgt == NULL || tgt->ltd_exp == NULL) - RETURN(-EINVAL); - return md_unpack_capa(tgt->ltd_exp, req, field, oc); + rc = md_get_remote_perm(tgt->ltd_exp, fid, suppgid, request); + RETURN(rc); } int lmv_intent_getattr_async(struct obd_export *exp, @@ -2804,7 +3212,7 @@ int lmv_intent_getattr_async(struct obd_export *exp, if (rc) RETURN(rc); - tgt = lmv_find_target(lmv, &op_data->op_fid1); + tgt = lmv_locate_mds(lmv, op_data, &op_data->op_fid1); if (IS_ERR(tgt)) RETURN(PTR_ERR(tgt)); @@ -2833,6 +3241,22 @@ int lmv_revalidate_lock(struct obd_export *exp, struct lookup_intent *it, RETURN(rc); } +int lmv_get_fid_from_lsm(struct obd_export *exp, + const struct lmv_stripe_md *lsm, + const char *name, int namelen, struct lu_fid *fid) +{ + const struct lmv_oinfo *oinfo; + + LASSERT(lsm != NULL); + oinfo = lsm_name_to_stripe_info(lsm, name, namelen); + if (IS_ERR(oinfo)) + return PTR_ERR(oinfo); + + *fid = oinfo->lmo_fid; + + RETURN(0); +} + /** * For lmv, only need to send request to master MDT, and the master MDT will * process with other slave MDTs. The only exception is Q_GETOQUOTA for which @@ -2886,30 +3310,47 @@ int lmv_quotactl(struct obd_device *unused, struct obd_export *exp, RETURN(rc); } -int lmv_quotacheck(struct obd_device *unused, struct obd_export *exp, - struct obd_quotactl *oqctl) +static int lmv_merge_attr(struct obd_export *exp, + const struct lmv_stripe_md *lsm, + struct cl_attr *attr, + ldlm_blocking_callback cb_blocking) { - struct obd_device *obd = class_exp2obd(exp); - struct lmv_obd *lmv = &obd->u.lmv; - struct lmv_tgt_desc *tgt; - __u32 i; - int rc = 0; - ENTRY; + int rc; + int i; - for (i = 0; i < lmv->desc.ld_tgt_count; i++) { - int err; - tgt = lmv->tgts[i]; - if (tgt == NULL || tgt->ltd_exp == NULL || !tgt->ltd_active) { - CERROR("lmv idx %d inactive\n", i); - RETURN(-EIO); - } + rc = lmv_revalidate_slaves(exp, lsm, cb_blocking, 0); + if (rc < 0) + return rc; - err = obd_quotacheck(tgt->ltd_exp, oqctl); - if (err && !rc) - rc = err; - } + for (i = 0; i < lsm->lsm_md_stripe_count; i++) { + struct inode *inode = lsm->lsm_md_oinfo[i].lmo_root; - RETURN(rc); + CDEBUG(D_INFO, ""DFID" size %llu, blocks %llu nlink %u," + " atime %lu ctime %lu, mtime %lu.\n", + PFID(&lsm->lsm_md_oinfo[i].lmo_fid), + i_size_read(inode), (unsigned long long)inode->i_blocks, + inode->i_nlink, LTIME_S(inode->i_atime), + LTIME_S(inode->i_ctime), LTIME_S(inode->i_mtime)); + + /* for slave stripe, it needs to subtract nlink for . and .. */ + if (i != 0) + attr->cat_nlink += inode->i_nlink - 2; + else + attr->cat_nlink = inode->i_nlink; + + attr->cat_size += i_size_read(inode); + attr->cat_blocks += inode->i_blocks; + + if (attr->cat_atime < LTIME_S(inode->i_atime)) + attr->cat_atime = LTIME_S(inode->i_atime); + + if (attr->cat_ctime < LTIME_S(inode->i_ctime)) + attr->cat_ctime = LTIME_S(inode->i_ctime); + + if (attr->cat_mtime < LTIME_S(inode->i_mtime)) + attr->cat_mtime = LTIME_S(inode->i_mtime); + } + return 0; } struct obd_ops lmv_obd_ops = { @@ -2923,12 +3364,9 @@ struct obd_ops lmv_obd_ops = { .o_statfs = lmv_statfs, .o_get_info = lmv_get_info, .o_set_info_async = lmv_set_info_async, - .o_packmd = lmv_packmd, - .o_unpackmd = lmv_unpackmd, .o_notify = lmv_notify, .o_get_uuid = lmv_get_uuid, .o_iocontrol = lmv_iocontrol, - .o_quotacheck = lmv_quotacheck, .o_quotactl = lmv_quotactl }; @@ -2938,7 +3376,6 @@ struct md_ops lmv_md_ops = { .m_find_cbdata = lmv_find_cbdata, .m_close = lmv_close, .m_create = lmv_create, - .m_done_writing = lmv_done_writing, .m_enqueue = lmv_enqueue, .m_getattr = lmv_getattr, .m_getxattr = lmv_getxattr, @@ -2949,33 +3386,30 @@ struct md_ops lmv_md_ops = { .m_setattr = lmv_setattr, .m_setxattr = lmv_setxattr, .m_fsync = lmv_fsync, - .m_readpage = lmv_readpage, + .m_read_page = lmv_read_page, .m_unlink = lmv_unlink, .m_init_ea_size = lmv_init_ea_size, .m_cancel_unused = lmv_cancel_unused, .m_set_lock_data = lmv_set_lock_data, .m_lock_match = lmv_lock_match, - .m_get_lustre_md = lmv_get_lustre_md, - .m_free_lustre_md = lmv_free_lustre_md, + .m_get_lustre_md = lmv_get_lustre_md, + .m_free_lustre_md = lmv_free_lustre_md, + .m_merge_attr = lmv_merge_attr, .m_set_open_replay_data = lmv_set_open_replay_data, .m_clear_open_replay_data = lmv_clear_open_replay_data, - .m_renew_capa = lmv_renew_capa, - .m_unpack_capa = lmv_unpack_capa, .m_get_remote_perm = lmv_get_remote_perm, .m_intent_getattr_async = lmv_intent_getattr_async, - .m_revalidate_lock = lmv_revalidate_lock + .m_revalidate_lock = lmv_revalidate_lock, + .m_get_fid_from_lsm = lmv_get_fid_from_lsm, + .m_unpackmd = lmv_unpackmd, }; int __init lmv_init(void) { - return class_register_type(&lmv_obd_ops, &lmv_md_ops, NULL, -#ifndef HAVE_ONLY_PROCFS_SEQ - NULL, -#endif - LUSTRE_LMV_NAME, NULL); + return class_register_type(&lmv_obd_ops, &lmv_md_ops, true, NULL, + LUSTRE_LMV_NAME, NULL); } -#ifdef __KERNEL__ static void lmv_exit(void) { class_unregister_type(LUSTRE_LMV_NAME); @@ -2987,4 +3421,3 @@ MODULE_LICENSE("GPL"); module_init(lmv_init); module_exit(lmv_exit); -#endif