X-Git-Url: https://git.whamcloud.com/?p=fs%2Flustre-release.git;a=blobdiff_plain;f=lustre%2Finclude%2Flustre_fid.h;h=44418c59b5c356ce3c5e7d2c4e56ec36b73a1424;hp=96ac595854edd15cfe67a8b7c756d83d1626b54a;hb=3ac7c325efaf3a2537fafc79168d87de67982acd;hpb=752c69506c4559cd5c97a8dcbe488b54822ca087 diff --git a/lustre/include/lustre_fid.h b/lustre/include/lustre_fid.h index 96ac595..44418c5 100644 --- a/lustre/include/lustre_fid.h +++ b/lustre/include/lustre_fid.h @@ -44,11 +44,114 @@ /** \defgroup fid fid * * @{ + * + * http://wiki.lustre.org/index.php/Architecture_-_Interoperability_fids_zfs + * describes the FID namespace and interoperability requirements for FIDs. + * The important parts of that document are included here for reference. + * + * FID + * File IDentifier generated by client from range allocated by the SEQuence + * service and stored in struct lu_fid. The FID is composed of three parts: + * SEQuence, ObjectID, and VERsion. The SEQ component is a filesystem + * unique 64-bit integer, and only one client is ever assigned any SEQ value. + * The first 0x400 FID_SEQ_NORMAL [2^33, 2^33 + 0x400] values are reserved + * for system use. The OID component is a 32-bit value generated by the + * client on a per-SEQ basis to allow creating many unique FIDs without + * communication with the server. The VER component is a 32-bit value that + * distinguishes between different FID instantiations, such as snapshots or + * separate subtrees within the filesystem. FIDs with the same VER field + * are considered part of the same namespace. + * + * OLD filesystems are those upgraded from Lustre 1.x that predate FIDs, and + * MDTs use 32-bit ldiskfs internal inode/generation numbers (IGIFs), while + * OSTs use 64-bit Lustre object IDs and generation numbers. + * + * NEW filesystems are those formatted since the introduction of FIDs. + * + * IGIF + * Inode and Generation In FID, a surrogate FID used to globally identify + * an existing object on OLD formatted MDT file system. This would only be + * used on MDT0 in a DNE filesystem, because there cannot be more than one + * MDT in an OLD formatted filesystem. Belongs to sequence in [12, 2^32 - 1] + * range, where inode number is stored in SEQ, and inode generation is in OID. + * NOTE: This assumes no more than 2^32-1 inodes exist in the MDT filesystem, + * which is the maximum possible for an ldiskfs backend. It also assumes + * that the reserved ext3/ext4/ldiskfs inode numbers [0-11] are never visible + * to clients, which has always been true. + * + * IDIF + * object ID In FID, a surrogate FID used to globally identify an existing + * OST object on OLD formatted OST file system. Belongs to a sequence in + * [2^32, 2^33 - 1]. Sequence number is calculated as: + * + * 1 << 32 | (ost_index << 16) | ((objid >> 32) & 0xffff) + * + * that is, SEQ consists of 16-bit OST index, and higher 16 bits of object + * ID. The generation of unique SEQ values per OST allows the IDIF FIDs to + * be identified in the FLD correctly. The OID field is calculated as: + * + * objid & 0xffffffff + * + * that is, it consists of lower 32 bits of object ID. For objects within + * the IDIF range, object ID extraction will be: + * + * o_id = (fid->f_seq & 0x7fff) << 16 | fid->f_oid; + * o_seq = 0; // formerly group number + * + * NOTE: This assumes that no more than 2^48-1 objects have ever been created + * on any OST, and that no more than 65535 OSTs are in use. Both are very + * reasonable assumptions, i.e. an IDIF can uniquely map all objects assuming + * a maximum creation rate of 1M objects per second for a maximum of 9 years, + * or combinations thereof. + * + * OST_MDT0 + * Surrogate FID used to identify an existing object on OLD formatted OST + * filesystem. Belongs to the reserved SEQuence 0, and is used prior to + * the introduction of FID-on-OST, at which point IDIF will be used to + * identify objects as residing on a specific OST. + * + * LLOG + * For Lustre Log objects the object sequence 1 is used. This is compatible + * with both OLD and NEW namespaces, as this SEQ number is in the + * ext3/ldiskfs reserved inode range and does not conflict with IGIF + * sequence numbers. + * + * ECHO + * For testing OST IO performance the object sequence 2 is used. This is + * compatible with both OLD and NEW namespaces, as this SEQ number is in + * the ext3/ldiskfs reserved inode range and does not conflict with IGIF + * sequence numbers. + * + * OST_MDT1 .. OST_MAX + * For testing with multiple MDTs the object sequence 3 through 9 is used, + * allowing direct mapping of MDTs 1 through 7 respectively, for a total + * of 8 MDTs including OST_MDT0. This matches the legacy CMD project "group" + * mappings. However, this SEQ range is only for testing prior to any + * production DNE release, as the objects in this range conflict across all + * OSTs, as the OST index is not part of the FID. For production DNE usage, + * OST objects created by MDT1+ will use FID_SEQ_NORMAL FIDs. + * + * DLM OST objid to IDIF mapping + * For compatibility with existing OLD OST network protocol structures, the + * FID must map onto the o_id and o_seq in a manner that ensures existing + * objects are identified consistently for IO, as well as onto the LDLM + * namespace to ensure IDIFs there is only a single resource name for any + * object in the DLM. The OLD OST object DLM resource mapping is: + * + * resource[] = {o_id, o_seq, 0, 0}; // o_seq == 0 for production releases + * + * The NEW OST object DLM resource mapping is the same for both MDT and OST: + * + * resource[] = {SEQ, OID, VER, HASH}; + * + * NOTE: for mapping IDIF values to DLM resource names the o_id may be + * larger than the 2^33 reserved sequence numbers for IDIF, so it is possible + * for the o_id numbers to overlap FID SEQ numbers in the resource. However, + * in all production releases the OLD o_seq field is always zero, and all + * valid FID OID values are non-zero, so the lock resources will not collide. + * Even so, the MDT and OST resources are also in different LDLM namespaces. */ -/* - * struct lu_fid - */ #include #include #include @@ -172,7 +275,7 @@ struct lu_server_seq; struct lu_client_seq { /* Sequence-controller export. */ struct obd_export *lcs_exp; - cfs_mutex_t lcs_mutex; + struct mutex lcs_mutex; /* * Range of allowed for allocation sequeces. When using lu_client_seq on @@ -238,7 +341,7 @@ struct lu_server_seq { struct lu_client_seq *lss_cli; /* Mutex for protecting allocation */ - cfs_mutex_t lss_mutex; + struct mutex lss_mutex; /* * Service uuid, passed from MDT + seq name to form unique seq name to @@ -333,6 +436,36 @@ fid_build_reg_res_name(const struct lu_fid *f, } /* + * Build (DLM) resource identifier from global quota FID and quota ID. + */ +static inline struct ldlm_res_id * +fid_build_quota_resid(const struct lu_fid *glb_fid, union lquota_id *qid, + struct ldlm_res_id *res) +{ + fid_build_reg_res_name(glb_fid, res); + res->name[LUSTRE_RES_ID_QUOTA_SEQ_OFF] = fid_seq(&qid->qid_fid); + res->name[LUSTRE_RES_ID_QUOTA_VER_OID_OFF] = fid_ver_oid(&qid->qid_fid); + return res; +} + +/* + * Extract global FID and quota ID from resource name + */ +static inline void fid_extract_quota_resid(struct ldlm_res_id *res, + struct lu_fid *glb_fid, + union lquota_id *qid) +{ + glb_fid->f_seq = res->name[LUSTRE_RES_ID_SEQ_OFF]; + glb_fid->f_oid = (__u32)res->name[LUSTRE_RES_ID_VER_OID_OFF]; + glb_fid->f_ver = (__u32)(res->name[LUSTRE_RES_ID_VER_OID_OFF] >> 32); + + qid->qid_fid.f_seq = res->name[LUSTRE_RES_ID_QUOTA_SEQ_OFF]; + qid->qid_fid.f_oid = (__u32)res->name[LUSTRE_RES_ID_QUOTA_VER_OID_OFF]; + qid->qid_fid.f_ver = + (__u32)(res->name[LUSTRE_RES_ID_QUOTA_VER_OID_OFF] >> 32); +} + +/* * Return true if resource is for object identified by fid. */ static inline int fid_res_name_eq(const struct lu_fid *f,