X-Git-Url: https://git.whamcloud.com/?p=fs%2Flustre-release.git;a=blobdiff_plain;f=lustre%2Finclude%2Flustre%2Flustre_idl.h;h=beeaf4b03d6b73ee706f5af1c4b142b4561f2453;hp=02ace2ff6d1f1ffa86cb8dfe5bb2677a2df23d10;hb=68234a15f0ca73b035f91e5bcc9399ca0bca1287;hpb=831972ed6f8db022dc03f2a26e3c783e355fe635 diff --git a/lustre/include/lustre/lustre_idl.h b/lustre/include/lustre/lustre_idl.h index 02ace2f..beeaf4b 100644 --- a/lustre/include/lustre/lustre_idl.h +++ b/lustre/include/lustre/lustre_idl.h @@ -1,6 +1,4 @@ -/* -*- mode: c; c-basic-offset: 8; indent-tabs-mode: nil; -*- - * vim:expandtab:shiftwidth=8:tabstop=8: - * +/* * GPL HEADER START * * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. @@ -28,6 +26,8 @@ /* * Copyright (c) 2007, 2010, Oracle and/or its affiliates. All rights reserved. * Use is subject to license terms. + * + * Copyright (c) 2011, 2012, Whamcloud, Inc. */ /* * This file is part of Lustre, http://www.lustre.org/ @@ -91,6 +91,10 @@ #ifndef _LUSTRE_IDL_H_ #define _LUSTRE_IDL_H_ +#if !defined(LASSERT) && !defined(LPU64) +#include /* for LASSERT, LPUX64, etc */ +#endif + /* Defn's shared with user-space. */ #include @@ -135,16 +139,10 @@ #define SEQ_METADATA_PORTAL 30 #define SEQ_DATA_PORTAL 31 #define SEQ_CONTROLLER_PORTAL 32 +#define MGS_BULK_PORTAL 33 /* Portal 63 is reserved for the Cray Inc DVS - nic@cray.com, roe@cray.com, n8851@cray.com */ -#define SVC_KILLED 1 -#define SVC_EVENT 2 -#define SVC_SIGNAL 4 -#define SVC_RUNNING 8 -#define SVC_STOPPING 16 -#define SVC_STOPPED 32 - /* packet types */ #define PTL_RPC_MSG_REQUEST 4711 #define PTL_RPC_MSG_ERR 4712 @@ -172,7 +170,7 @@ typedef __u32 mdsno_t; typedef __u64 seqno_t; typedef __u64 obd_id; typedef __u64 obd_seq; -typedef __u64 obd_time; +typedef __s64 obd_time; typedef __u64 obd_size; typedef __u64 obd_off; typedef __u64 obd_blocks; @@ -187,15 +185,18 @@ typedef __u32 obd_count; /** * Describes a range of sequence, lsr_start is included but lsr_end is * not in the range. - * Same structure is used in fld module where lsr_mdt field holds mdt id + * Same structure is used in fld module where lsr_index field holds mdt id * of the home mdt. */ +#define LU_SEQ_RANGE_MDT 0x0 +#define LU_SEQ_RANGE_OST 0x1 + struct lu_seq_range { __u64 lsr_start; __u64 lsr_end; - __u32 lsr_mdt; - __u32 lsr_padding; + __u32 lsr_index; + __u32 lsr_flags; }; /** @@ -213,7 +214,7 @@ static inline __u64 range_space(const struct lu_seq_range *range) static inline void range_init(struct lu_seq_range *range) { - range->lsr_start = range->lsr_end = range->lsr_mdt = 0; + range->lsr_start = range->lsr_end = range->lsr_index = 0; } /** @@ -242,12 +243,22 @@ static inline int range_is_exhausted(const struct lu_seq_range *range) return range_space(range) == 0; } -#define DRANGE "[%#16.16"LPF64"x-%#16.16"LPF64"x):%x" +/* return 0 if two range have the same location */ +static inline int range_compare_loc(const struct lu_seq_range *r1, + const struct lu_seq_range *r2) +{ + return r1->lsr_index != r2->lsr_index || + r1->lsr_flags != r2->lsr_flags; +} + +#define DRANGE "[%#16.16"LPF64"x-%#16.16"LPF64"x):%x:%s" #define PRANGE(range) \ - (range)->lsr_start, \ - (range)->lsr_end, \ - (range)->lsr_mdt + (range)->lsr_start, \ + (range)->lsr_end, \ + (range)->lsr_index, \ + (range)->lsr_flags == LU_SEQ_RANGE_MDT ? "mdt" : "ost" + /** \defgroup lu_fid lu_fid * @{ */ @@ -387,6 +398,10 @@ static inline obd_id fid_ver_oid(const struct lu_fid *fid) } /** + * Note that reserved SEQ numbers below 12 will conflict with ldiskfs + * inodes in the IGIF namespace, so these reserved SEQ numbers can be + * used for other purposes and not risk collisions with existing inodes. + * * Different FID Format * http://arch.lustre.org/index.php?title=Interoperability_fids_zfs#NEW.0 */ @@ -405,7 +420,12 @@ enum fid_seq { FID_SEQ_START = 0x200000000ULL, FID_SEQ_LOCAL_FILE = 0x200000001ULL, FID_SEQ_DOT_LUSTRE = 0x200000002ULL, - FID_SEQ_NORMAL = 0x200000400ULL + /* XXX 0x200000003ULL is reserved for FID_SEQ_LLOG_OBJ */ + FID_SEQ_SPECIAL = 0x200000004ULL, + FID_SEQ_QUOTA = 0x200000005ULL, + FID_SEQ_QUOTA_GLB = 0x200000006ULL, + FID_SEQ_NORMAL = 0x200000400ULL, + FID_SEQ_LOV_DEFAULT= 0xffffffffffffffffULL }; #define OBIF_OID_MAX_BITS 32 @@ -415,6 +435,17 @@ enum fid_seq { #define IDIF_MAX_OID (1ULL << IDIF_OID_MAX_BITS) #define IDIF_OID_MASK ((1ULL << IDIF_OID_MAX_BITS) - 1) +/** OID for FID_SEQ_SPECIAL */ +enum special_oid { + /* Big Filesystem Lock to serialize rename operations */ + FID_OID_SPECIAL_BFL = 1UL, +}; + +/** OID for FID_SEQ_DOT_LUSTRE */ +enum dot_lustre_oid { + FID_OID_DOT_LUSTRE = 1UL, + FID_OID_DOT_LUSTRE_OBF = 2UL, +}; static inline int fid_seq_is_mdt0(obd_seq seq) { @@ -434,7 +465,7 @@ static inline int fid_seq_is_mdt(const __u64 seq) static inline int fid_seq_is_rsvd(const __u64 seq) { - return seq <= FID_SEQ_RSVD; + return (seq > FID_SEQ_OST_MDT0 && seq <= FID_SEQ_RSVD); }; static inline int fid_is_mdt0(const struct lu_fid *fid) @@ -578,21 +609,24 @@ static inline int fid_ostid_unpack(struct lu_fid *fid, struct ost_id *ostid, } /* pack an IDIF FID into an ostid (id/seq) for the wire/disk */ -static inline void ostid_idif_pack(struct lu_fid *fid, struct ost_id *ostid) +static inline void ostid_idif_pack(const struct lu_fid *fid, + struct ost_id *ostid) { ostid->oi_seq = FID_SEQ_OST_MDT0; ostid->oi_id = fid_idif_id(fid->f_seq, fid->f_oid, fid->f_ver); } /* pack a non-IDIF FID into an ostid (id/seq) for the wire/disk */ -static inline void ostid_fid_pack(struct lu_fid *fid, struct ost_id *ostid) +static inline void ostid_fid_pack(const struct lu_fid *fid, + struct ost_id *ostid) { ostid->oi_seq = fid_seq(fid); ostid->oi_id = fid_ver_oid(fid); } /* pack any OST FID into an ostid (id/seq) for the wire/disk */ -static inline int fid_ostid_pack(struct lu_fid *fid, struct ost_id *ostid) +static inline int fid_ostid_pack(const struct lu_fid *fid, + struct ost_id *ostid) { if (unlikely(fid_seq_is_igif(fid->f_seq))) { CERROR("bad IGIF, "DFID"\n", PFID(fid)); @@ -728,7 +762,7 @@ static inline int fid_is_sane(const struct lu_fid *fid) fid != NULL && ((fid_seq(fid) >= FID_SEQ_START && fid_oid(fid) != 0 && fid_ver(fid) == 0) || - fid_is_igif(fid)); + fid_is_igif(fid) || fid_seq_is_rsvd(fid_seq(fid))); } static inline int fid_is_zero(const struct lu_fid *fid) @@ -745,8 +779,10 @@ static inline int lu_fid_eq(const struct lu_fid *f0, /* Check that there is no alignment padding. */ CLASSERT(sizeof *f0 == sizeof f0->f_seq + sizeof f0->f_oid + sizeof f0->f_ver); - LASSERTF(fid_is_igif(f0) || fid_ver(f0) == 0, DFID, PFID(f0)); - LASSERTF(fid_is_igif(f1) || fid_ver(f1) == 0, DFID, PFID(f1)); + LASSERTF((fid_is_igif(f0) || fid_is_idif(f0)) || + fid_ver(f0) == 0, DFID, PFID(f0)); + LASSERTF((fid_is_igif(f1) || fid_is_idif(f1)) || + fid_ver(f1) == 0, DFID, PFID(f1)); return memcmp(f0, f1, sizeof *f0) == 0; } @@ -779,8 +815,9 @@ static inline int lu_fid_cmp(const struct lu_fid *f0, * enumeration. */ enum lu_dirent_attrs { - LUDA_FID = 0x0001, - LUDA_TYPE = 0x0002, + LUDA_FID = 0x0001, + LUDA_TYPE = 0x0002, + LUDA_64BITHASH = 0x0004, }; /** @@ -843,7 +880,14 @@ struct lu_dirpage { }; enum lu_dirpage_flags { - LDF_EMPTY = 1 << 0 + /** + * dirpage contains no entry. + */ + LDF_EMPTY = 1 << 0, + /** + * last entry's lde_hash equals ldp_hash_end. + */ + LDF_COLLIDE = 1 << 1 }; static inline struct lu_dirent *lu_dirent_start(struct lu_dirpage *dp) @@ -889,7 +933,22 @@ static inline int lu_dirent_size(struct lu_dirent *ent) return le16_to_cpu(ent->lde_reclen); } -#define DIR_END_OFF 0xfffffffffffffffeULL +#define MDS_DIR_END_OFF 0xfffffffffffffffeULL + +/** + * MDS_READPAGE page size + * + * This is the directory page size packed in MDS_READPAGE RPC. + * It's different than CFS_PAGE_SIZE because the client needs to + * access the struct lu_dirpage header packed at the beginning of + * the "page" and without this there isn't any way to know find the + * lu_dirpage header is if client and server CFS_PAGE_SIZE differ. + */ +#define LU_PAGE_SHIFT 12 +#define LU_PAGE_SIZE (1UL << LU_PAGE_SHIFT) +#define LU_PAGE_MASK (~(LU_PAGE_SIZE - 1)) + +#define LU_PAGE_COUNT 1 << (CFS_PAGE_SHIFT - LU_PAGE_SHIFT) /** @} lu_dir */ @@ -903,8 +962,8 @@ static inline int lustre_handle_is_used(struct lustre_handle *lh) return lh->cookie != 0ull; } -static inline int lustre_handle_equal(struct lustre_handle *lh1, - struct lustre_handle *lh2) +static inline int lustre_handle_equal(const struct lustre_handle *lh1, + const struct lustre_handle *lh2) { return lh1->cookie == lh2->cookie; } @@ -936,7 +995,33 @@ struct lustre_msg_v2 { /* without gss, ptlrpc_body is put at the first buffer. */ #define PTLRPC_NUM_VERSIONS 4 -struct ptlrpc_body { +#define JOBSTATS_JOBID_SIZE 32 /* 32 bytes string */ +struct ptlrpc_body_v3 { + struct lustre_handle pb_handle; + __u32 pb_type; + __u32 pb_version; + __u32 pb_opc; + __u32 pb_status; + __u64 pb_last_xid; + __u64 pb_last_seen; + __u64 pb_last_committed; + __u64 pb_transno; + __u32 pb_flags; + __u32 pb_op_flags; + __u32 pb_conn_cnt; + __u32 pb_timeout; /* for req, the deadline, for rep, the service est */ + __u32 pb_service_time; /* for rep, actual service time */ + __u32 pb_limit; + __u64 pb_slv; + /* VBR: pre-versions */ + __u64 pb_pre_versions[PTLRPC_NUM_VERSIONS]; + /* padding for future needs */ + __u64 pb_padding[4]; + char pb_jobid[JOBSTATS_JOBID_SIZE]; +}; +#define ptlrpc_body ptlrpc_body_v3 + +struct ptlrpc_body_v2 { struct lustre_handle pb_handle; __u32 pb_type; __u32 pb_version; @@ -950,7 +1035,8 @@ struct ptlrpc_body { __u32 pb_op_flags; __u32 pb_conn_cnt; __u32 pb_timeout; /* for req, the deadline, for rep, the service est */ - __u32 pb_service_time; /* for rep, actual service time */ + __u32 pb_service_time; /* for rep, actual service time, also used for + net_latency of req */ __u32 pb_limit; __u64 pb_slv; /* VBR: pre-versions */ @@ -1058,8 +1144,33 @@ extern void lustre_swab_ptlrpc_body(struct ptlrpc_body *pb); #define OBD_CONNECT_SKIP_ORPHAN 0x400000000ULL /* don't reuse orphan objids */ #define OBD_CONNECT_MAX_EASIZE 0x800000000ULL /* preserved for large EA */ #define OBD_CONNECT_FULL20 0x1000000000ULL /* it is 2.0 client */ -/* also update obd_connect_names[] for lprocfs_rd_connect_flags() - * and lustre/utils/wirecheck.c */ +#define OBD_CONNECT_LAYOUTLOCK 0x2000000000ULL /* client uses layout lock */ +#define OBD_CONNECT_64BITHASH 0x4000000000ULL /* client supports 64-bits + * directory hash */ +#define OBD_CONNECT_MAXBYTES 0x8000000000ULL /* max stripe size */ +#define OBD_CONNECT_IMP_RECOV 0x10000000000ULL /* imp recovery support */ +#define OBD_CONNECT_JOBSTATS 0x20000000000ULL /* jobid in ptlrpc_body */ +#define OBD_CONNECT_UMASK 0x40000000000ULL /* create uses client umask */ +#define OBD_CONNECT_EINPROGRESS 0x80000000000ULL /* client handles -EINPROGRESS + * write RPC error properly */ +#define OBD_CONNECT_GRANT_PARAM 0x100000000000ULL/* extra grant params used for + * finer space reservation */ +#define OBD_CONNECT_NANOSECOND_TIMES 0x200000000000ULL /* nanosec resolution + * timestamps supported + */ +#define OBD_CONNECT_LVB_TYPE 0x400000000000ULL /* variable type of LVB */ + +/* XXX README XXX: + * Please DO NOT add flag values here before first ensuring that this same + * flag value is not in use on some other branch. Please clear any such + * changes with senior engineers before starting to use a new flag. Then, + * submit a small patch against EVERY branch that ONLY adds the new flag + * and updates obd_connect_names[] for lprocfs_rd_connect_flags(), so it + * can be approved and landed easily to reserve the flag for future use. */ + +#define OCD_HAS_FLAG(ocd, flg) \ + (!!((ocd)->ocd_connect_flags & OBD_CONNECT_##flg)) + #ifdef HAVE_LRU_RESIZE_SUPPORT #define LRU_RESIZE_CONNECT_FLAG OBD_CONNECT_LRU_RESIZE @@ -1069,16 +1180,18 @@ extern void lustre_swab_ptlrpc_body(struct ptlrpc_body *pb); #define MDT_CONNECT_SUPPORTED (OBD_CONNECT_RDONLY | OBD_CONNECT_VERSION | \ OBD_CONNECT_ACL | OBD_CONNECT_XATTR | \ - OBD_CONNECT_IBITS | OBD_CONNECT_JOIN | \ - OBD_CONNECT_NODEVOH |/* OBD_CONNECT_ATTRFID |*/\ + OBD_CONNECT_IBITS | \ + OBD_CONNECT_NODEVOH | OBD_CONNECT_ATTRFID | \ OBD_CONNECT_CANCELSET | OBD_CONNECT_AT | \ OBD_CONNECT_RMT_CLIENT | \ OBD_CONNECT_RMT_CLIENT_FORCE | \ - OBD_CONNECT_MDS_CAPA | OBD_CONNECT_OSS_CAPA | \ - OBD_CONNECT_MDS_MDS | OBD_CONNECT_FID | \ - LRU_RESIZE_CONNECT_FLAG | OBD_CONNECT_VBR | \ - OBD_CONNECT_LOV_V3 | OBD_CONNECT_SOM | \ - OBD_CONNECT_FULL20) + OBD_CONNECT_BRW_SIZE | OBD_CONNECT_MDS_CAPA | \ + OBD_CONNECT_OSS_CAPA | OBD_CONNECT_MDS_MDS | \ + OBD_CONNECT_FID | LRU_RESIZE_CONNECT_FLAG | \ + OBD_CONNECT_VBR | OBD_CONNECT_LOV_V3 | \ + OBD_CONNECT_SOM | OBD_CONNECT_FULL20 | \ + OBD_CONNECT_64BITHASH | \ + OBD_CONNECT_EINPROGRESS | OBD_CONNECT_JOBSTATS) #define OST_CONNECT_SUPPORTED (OBD_CONNECT_SRVLOCK | OBD_CONNECT_GRANT | \ OBD_CONNECT_REQPORTAL | OBD_CONNECT_VERSION | \ OBD_CONNECT_TRUNCLOCK | OBD_CONNECT_INDEX | \ @@ -1086,13 +1199,21 @@ extern void lustre_swab_ptlrpc_body(struct ptlrpc_body *pb); OBD_CONNECT_CANCELSET | OBD_CONNECT_AT | \ LRU_RESIZE_CONNECT_FLAG | OBD_CONNECT_CKSUM | \ OBD_CONNECT_CHANGE_QS | \ - OBD_CONNECT_OSS_CAPA | OBD_CONNECT_RMT_CLIENT | \ + OBD_CONNECT_OSS_CAPA | \ + OBD_CONNECT_RMT_CLIENT | \ OBD_CONNECT_RMT_CLIENT_FORCE | OBD_CONNECT_VBR | \ OBD_CONNECT_MDS | OBD_CONNECT_SKIP_ORPHAN | \ - OBD_CONNECT_GRANT_SHRINK | OBD_CONNECT_FULL20) + OBD_CONNECT_GRANT_SHRINK | OBD_CONNECT_FULL20 | \ + OBD_CONNECT_64BITHASH | OBD_CONNECT_MAXBYTES | \ + OBD_CONNECT_MAX_EASIZE | \ + OBD_CONNECT_EINPROGRESS | OBD_CONNECT_JOBSTATS) #define ECHO_CONNECT_SUPPORTED (0) #define MGS_CONNECT_SUPPORTED (OBD_CONNECT_VERSION | OBD_CONNECT_AT | \ - OBD_CONNECT_FULL20) + OBD_CONNECT_FULL20 | OBD_CONNECT_IMP_RECOV) + +/* Features required for this version of the client to work with server */ +#define CLIENT_CONNECT_MDT_REQD (OBD_CONNECT_IBITS | OBD_CONNECT_FID | \ + OBD_CONNECT_FULL20) #define OBD_OCD_VERSION(major,minor,patch,fix) (((major)<<24) + ((minor)<<16) +\ ((patch)<<8) + (fix)) @@ -1105,22 +1226,71 @@ extern void lustre_swab_ptlrpc_body(struct ptlrpc_body *pb); * * If we eventually have separate connect data for different types, which we * almost certainly will, then perhaps we stick a union in here. */ -struct obd_connect_data { +struct obd_connect_data_v1 { __u64 ocd_connect_flags; /* OBD_CONNECT_* per above */ __u32 ocd_version; /* lustre release version number */ __u32 ocd_grant; /* initial cache grant amount (bytes) */ __u32 ocd_index; /* LOV index to connect to */ __u32 ocd_brw_size; /* Maximum BRW size in bytes */ __u64 ocd_ibits_known; /* inode bits this client understands */ - __u32 ocd_nllu; /* non-local-lustre-user */ - __u32 ocd_nllg; /* non-local-lustre-group */ + __u8 ocd_blocksize; /* log2 of the backend filesystem blocksize */ + __u8 ocd_inodespace; /* log2 of the per-inode space consumption */ + __u16 ocd_grant_extent; /* per-extent grant overhead, in 1K blocks */ + __u32 ocd_unused; /* also fix lustre_swab_connect */ __u64 ocd_transno; /* first transno from client to be replayed */ __u32 ocd_group; /* MDS group on OST */ __u32 ocd_cksum_types; /* supported checksum algorithms */ - __u64 padding1; /* also fix lustre_swab_connect */ - __u64 padding2; /* also fix lustre_swab_connect */ + __u32 ocd_max_easize; /* How big LOV EA can be on MDS */ + __u32 ocd_instance; /* also fix lustre_swab_connect */ + __u64 ocd_maxbytes; /* Maximum stripe size in bytes */ }; +struct obd_connect_data { + __u64 ocd_connect_flags; /* OBD_CONNECT_* per above */ + __u32 ocd_version; /* lustre release version number */ + __u32 ocd_grant; /* initial cache grant amount (bytes) */ + __u32 ocd_index; /* LOV index to connect to */ + __u32 ocd_brw_size; /* Maximum BRW size in bytes */ + __u64 ocd_ibits_known; /* inode bits this client understands */ + __u8 ocd_blocksize; /* log2 of the backend filesystem blocksize */ + __u8 ocd_inodespace; /* log2 of the per-inode space consumption */ + __u16 ocd_grant_extent; /* per-extent grant overhead, in 1K blocks */ + __u32 ocd_unused; /* also fix lustre_swab_connect */ + __u64 ocd_transno; /* first transno from client to be replayed */ + __u32 ocd_group; /* MDS group on OST */ + __u32 ocd_cksum_types; /* supported checksum algorithms */ + __u32 ocd_max_easize; /* How big LOV EA can be on MDS */ + __u32 ocd_instance; /* instance # of this target */ + __u64 ocd_maxbytes; /* Maximum stripe size in bytes */ + /* Fields after ocd_maxbytes are only accessible by the receiver + * if the corresponding flag in ocd_connect_flags is set. Accessing + * any field after ocd_maxbytes on the receiver without a valid flag + * may result in out-of-bound memory access and kernel oops. */ + __u64 padding1; /* added 2.1.0. also fix lustre_swab_connect */ + __u64 padding2; /* added 2.1.0. also fix lustre_swab_connect */ + __u64 padding3; /* added 2.1.0. also fix lustre_swab_connect */ + __u64 padding4; /* added 2.1.0. also fix lustre_swab_connect */ + __u64 padding5; /* added 2.1.0. also fix lustre_swab_connect */ + __u64 padding6; /* added 2.1.0. also fix lustre_swab_connect */ + __u64 padding7; /* added 2.1.0. also fix lustre_swab_connect */ + __u64 padding8; /* added 2.1.0. also fix lustre_swab_connect */ + __u64 padding9; /* added 2.1.0. also fix lustre_swab_connect */ + __u64 paddingA; /* added 2.1.0. also fix lustre_swab_connect */ + __u64 paddingB; /* added 2.1.0. also fix lustre_swab_connect */ + __u64 paddingC; /* added 2.1.0. also fix lustre_swab_connect */ + __u64 paddingD; /* added 2.1.0. also fix lustre_swab_connect */ + __u64 paddingE; /* added 2.1.0. also fix lustre_swab_connect */ + __u64 paddingF; /* added 2.1.0. also fix lustre_swab_connect */ +}; +/* XXX README XXX: + * Please DO NOT use any fields here before first ensuring that this same + * field is not in use on some other branch. Please clear any such changes + * with senior engineers before starting to use a new field. Then, submit + * a small patch against EVERY branch that ONLY adds the new field along with + * the matching OBD_CONNECT flag, so that can be approved and landed easily to + * reserve the flag for future use. */ + + extern void lustre_swab_connect(struct obd_connect_data *ocd); /* @@ -1132,6 +1302,7 @@ extern void lustre_swab_connect(struct obd_connect_data *ocd); typedef enum { OBD_CKSUM_CRC32 = 0x00000001, OBD_CKSUM_ADLER = 0x00000002, + OBD_CKSUM_CRC32C= 0x00000004, } cksum_type_t; /* @@ -1177,17 +1348,23 @@ enum obdo_flags { OBD_FL_SRVLOCK = 0x00000800, /* delegate DLM locking to server */ OBD_FL_CKSUM_CRC32 = 0x00001000, /* CRC32 checksum type */ OBD_FL_CKSUM_ADLER = 0x00002000, /* ADLER checksum type */ - OBD_FL_CKSUM_RSVD1 = 0x00004000, /* for future cksum types */ + OBD_FL_CKSUM_CRC32C = 0x00004000, /* CRC32C checksum type */ OBD_FL_CKSUM_RSVD2 = 0x00008000, /* for future cksum types */ OBD_FL_CKSUM_RSVD3 = 0x00010000, /* for future cksum types */ OBD_FL_SHRINK_GRANT = 0x00020000, /* object shrink the grant */ + OBD_FL_MMAP = 0x00040000, /* object is mmapped on the client. + * XXX: obsoleted - reserved for old + * clients prior than 2.2 */ + OBD_FL_RECOV_RESEND = 0x00080000, /* recoverable resent */ + OBD_FL_NOSPC_BLK = 0x00100000, /* no more block space on OST */ - OBD_FL_CKSUM_ALL = OBD_FL_CKSUM_CRC32 | OBD_FL_CKSUM_ADLER, + /* Note that while these checksum values are currently separate bits, + * in 2.x we can actually allow all values from 1-31 if we wanted. */ + OBD_FL_CKSUM_ALL = OBD_FL_CKSUM_CRC32 | OBD_FL_CKSUM_ADLER | + OBD_FL_CKSUM_CRC32C, /* mask for local-only flag, which won't be sent over network */ OBD_FL_LOCAL_MASK = 0xF0000000, - /* temporary OBDO used by osc_brw_async (see bug 18364) */ - OBD_FL_TEMPORARY = 0x10000000, }; #define LOV_MAGIC_V1 0x0BD10BD0 @@ -1200,9 +1377,6 @@ enum obdo_flags { #define LOV_PATTERN_FIRST 0x100 /* first stripe is not in round-robin */ #define LOV_PATTERN_CMOBD 0x200 -#define LOV_OBJECT_GROUP_DEFAULT ~0ULL -#define LOV_OBJECT_GROUP_CLEAR 0ULL - #define lov_ost_data lov_ost_data_v1 struct lov_ost_data_v1 { /* per-stripe data structure (little-endian)*/ __u64 l_object_id; /* OST object ID */ @@ -1218,7 +1392,9 @@ struct lov_mds_md_v1 { /* LOV EA mds/wire data (little-endian) */ __u64 lmm_object_id; /* LOV object ID */ __u64 lmm_object_seq; /* LOV object seq number */ __u32 lmm_stripe_size; /* size of stripe in bytes */ - __u32 lmm_stripe_count; /* num stripes in use for this object */ + /* lmm_stripe_count used to be __u32 */ + __u16 lmm_stripe_count; /* num stripes in use for this object */ + __u16 lmm_layout_gen; /* layout generation number */ struct lov_ost_data_v1 lmm_objects[0]; /* per-stripe data */ }; @@ -1238,6 +1414,8 @@ struct lov_mds_md_v1 { /* LOV EA mds/wire data (little-endian) */ #define XATTR_NAME_LMA "trusted.lma" #define XATTR_NAME_LMV "trusted.lmv" #define XATTR_NAME_LINK "trusted.link" +#define XATTR_NAME_FID "trusted.fid" +#define XATTR_NAME_VERSION "trusted.version" struct lov_mds_md_v3 { /* LOV EA mds/wire data (little-endian) */ @@ -1246,7 +1424,9 @@ struct lov_mds_md_v3 { /* LOV EA mds/wire data (little-endian) */ __u64 lmm_object_id; /* LOV object ID */ __u64 lmm_object_seq; /* LOV object seq number */ __u32 lmm_stripe_size; /* size of stripe in bytes */ - __u32 lmm_stripe_count; /* num stripes in use for this object */ + /* lmm_stripe_count used to be __u32 */ + __u16 lmm_stripe_count; /* num stripes in use for this object */ + __u16 lmm_layout_gen; /* layout generation number */ char lmm_pool_name[LOV_MAXPOOLNAME]; /* must be 32bit aligned */ struct lov_ost_data_v1 lmm_objects[0]; /* per-stripe data */ }; @@ -1290,23 +1470,25 @@ struct lov_mds_md_v3 { /* LOV EA mds/wire data (little-endian) */ #define OBD_MD_MEA (0x0000000400000000ULL) /* CMD split EA */ #define OBD_MD_MDTIDX (0x0000000800000000ULL) /* Get MDT index */ -#define OBD_MD_FLXATTR (0x0000001000000000ULL) /* xattr */ -#define OBD_MD_FLXATTRLS (0x0000002000000000ULL) /* xattr list */ -#define OBD_MD_FLXATTRRM (0x0000004000000000ULL) /* xattr remove */ -#define OBD_MD_FLACL (0x0000008000000000ULL) /* ACL */ -#define OBD_MD_FLRMTPERM (0x0000010000000000ULL) /* remote permission */ -#define OBD_MD_FLMDSCAPA (0x0000020000000000ULL) /* MDS capability */ -#define OBD_MD_FLOSSCAPA (0x0000040000000000ULL) /* OSS capability */ -#define OBD_MD_FLCKSPLIT (0x0000080000000000ULL) /* Check split on server */ -#define OBD_MD_FLCROSSREF (0x0000100000000000ULL) /* Cross-ref case */ +#define OBD_MD_FLXATTR (0x0000001000000000ULL) /* xattr */ +#define OBD_MD_FLXATTRLS (0x0000002000000000ULL) /* xattr list */ +#define OBD_MD_FLXATTRRM (0x0000004000000000ULL) /* xattr remove */ +#define OBD_MD_FLACL (0x0000008000000000ULL) /* ACL */ +#define OBD_MD_FLRMTPERM (0x0000010000000000ULL) /* remote permission */ +#define OBD_MD_FLMDSCAPA (0x0000020000000000ULL) /* MDS capability */ +#define OBD_MD_FLOSSCAPA (0x0000040000000000ULL) /* OSS capability */ +#define OBD_MD_FLCKSPLIT (0x0000080000000000ULL) /* Check split on server */ +#define OBD_MD_FLCROSSREF (0x0000100000000000ULL) /* Cross-ref case */ #define OBD_MD_FLGETATTRLOCK (0x0000200000000000ULL) /* Get IOEpoch attributes * under lock */ -#define OBD_FL_TRUNC (0x0000200000000000ULL) /* for filter_truncate */ +#define OBD_MD_FLOBJCOUNT (0x0000400000000000ULL) /* for multiple destroy */ + +#define OBD_MD_FLRMTLSETFACL (0x0001000000000000ULL) /* lfs lsetfacl case */ +#define OBD_MD_FLRMTLGETFACL (0x0002000000000000ULL) /* lfs lgetfacl case */ +#define OBD_MD_FLRMTRSETFACL (0x0004000000000000ULL) /* lfs rsetfacl case */ +#define OBD_MD_FLRMTRGETFACL (0x0008000000000000ULL) /* lfs rgetfacl case */ -#define OBD_MD_FLRMTLSETFACL (0x0001000000000000ULL) /* lfs lsetfacl case */ -#define OBD_MD_FLRMTLGETFACL (0x0002000000000000ULL) /* lfs lgetfacl case */ -#define OBD_MD_FLRMTRSETFACL (0x0004000000000000ULL) /* lfs rsetfacl case */ -#define OBD_MD_FLRMTRGETFACL (0x0008000000000000ULL) /* lfs rgetfacl case */ +#define OBD_MD_FLDATAVERSION (0x0010000000000000ULL) /* iversion sum */ #define OBD_MD_FLGETATTR (OBD_MD_FLID | OBD_MD_FLATIME | OBD_MD_FLMTIME | \ OBD_MD_FLCTIME | OBD_MD_FLSIZE | OBD_MD_FLBLKSZ | \ @@ -1319,16 +1501,6 @@ struct lov_mds_md_v3 { /* LOV EA mds/wire data (little-endian) */ extern void lustre_swab_obd_statfs (struct obd_statfs *os); -#define OBD_STATFS_NODELAY 0x0001 /* requests should be send without delay - * and resends for avoid deadlocks */ -#define OBD_STATFS_FROM_CACHE 0x0002 /* the statfs callback should not update - * obd_osfs_age */ -#define OBD_STATFS_PTLRPCD 0x0004 /* requests will be sent via ptlrpcd - * instead of a specific set. This - * means that we cannot rely on the set - * interpret routine to be called. - * lov_statfs_fini() must thus be called - * by the request interpret routine */ /* ost_body.data values for OST_BRW */ @@ -1344,6 +1516,7 @@ extern void lustre_swab_obd_statfs (struct obd_statfs *os); #define OBD_BRW_NOCACHE 0x80 /* this page is a part of non-cached IO */ #define OBD_BRW_NOQUOTA 0x100 #define OBD_BRW_SRVLOCK 0x200 /* Client holds no lock over this page */ +#define OBD_BRW_ASYNC 0x400 /* Server may delay commit to disk */ #define OBD_BRW_MEMALLOC 0x800 /* Client runs in the "kswapd" context */ #define OBD_OBJECT_EOF 0xffffffffffffffffULL @@ -1382,15 +1555,13 @@ extern void lustre_swab_niobuf_remote (struct niobuf_remote *nbr); #define OST_LVB_GET_ERR(blocks) (int)(blocks - OST_LVB_ERR_INIT) struct ost_lvb { - __u64 lvb_size; - __u64 lvb_mtime; - __u64 lvb_atime; - __u64 lvb_ctime; - __u64 lvb_blocks; + __u64 lvb_size; + obd_time lvb_mtime; + obd_time lvb_atime; + obd_time lvb_ctime; + __u64 lvb_blocks; }; -extern void lustre_swab_ost_lvb(struct ost_lvb *); - /* * MDS REQ RECORDS */ @@ -1458,23 +1629,29 @@ extern void lustre_swab_generic_32s (__u32 *val); #define MDS_INODELOCK_LOOKUP 0x000001 /* dentry, mode, owner, group */ #define MDS_INODELOCK_UPDATE 0x000002 /* size, links, timestamps */ #define MDS_INODELOCK_OPEN 0x000004 /* For opened files */ +#define MDS_INODELOCK_LAYOUT 0x000008 /* for layout */ -/* Do not forget to increase MDS_INODELOCK_MAXSHIFT when adding new bits */ +/* Do not forget to increase MDS_INODELOCK_MAXSHIFT when adding new bits + * XXX: MDS_INODELOCK_MAXSHIFT should be increased to 3 once the layout lock is + * supported */ #define MDS_INODELOCK_MAXSHIFT 2 /* This FULL lock is useful to take on unlink sort of operations */ #define MDS_INODELOCK_FULL ((1<<(MDS_INODELOCK_MAXSHIFT+1))-1) extern void lustre_swab_ll_fid (struct ll_fid *fid); -#define MDS_STATUS_CONN 1 -#define MDS_STATUS_LOV 2 - -struct mds_status_req { - __u32 flags; - __u32 repbuf; +/* NOTE: until Lustre 1.8.7/2.1.1 the fid_ver() was packed into name[2], + * but was moved into name[1] along with the OID to avoid consuming the + * name[2,3] fields that need to be used for the quota id (also a FID). */ +enum { + LUSTRE_RES_ID_SEQ_OFF = 0, + LUSTRE_RES_ID_VER_OID_OFF = 1, + LUSTRE_RES_ID_WAS_VER_OFF = 2, /* see note above */ + LUSTRE_RES_ID_HSH_OFF = 3 }; -extern void lustre_swab_mds_status_req (struct mds_status_req *r); +#define MDS_STATUS_CONN 1 +#define MDS_STATUS_LOV 2 /* mdt_thread_info.mti_flags. */ enum md_op_flags { @@ -1495,110 +1672,68 @@ enum md_op_flags { #define MF_SOM_LOCAL_FLAGS (MF_SOM_CHANGE | MF_EPOCH_OPEN | MF_EPOCH_CLOSE) -#define MDS_BFLAG_UNCOMMITTED_WRITES 0x1 +#define LUSTRE_BFLAG_UNCOMMITTED_WRITES 0x1 -/* these should be identical to their EXT3_*_FL counterparts, and are - * redefined here only to avoid dragging in ext3_fs.h */ -#define MDS_SYNC_FL 0x00000008 /* Synchronous updates */ -#define MDS_IMMUTABLE_FL 0x00000010 /* Immutable file */ -#define MDS_APPEND_FL 0x00000020 /* writes to file may only append */ -#define MDS_NOATIME_FL 0x00000080 /* do not update atime */ -#define MDS_DIRSYNC_FL 0x00010000 /* dirsync behaviour (dir only) */ +/* these should be identical to their EXT4_*_FL counterparts, they are + * redefined here only to avoid dragging in fs/ext4/ext4.h */ +#define LUSTRE_SYNC_FL 0x00000008 /* Synchronous updates */ +#define LUSTRE_IMMUTABLE_FL 0x00000010 /* Immutable file */ +#define LUSTRE_APPEND_FL 0x00000020 /* writes to file may only append */ +#define LUSTRE_NOATIME_FL 0x00000080 /* do not update atime */ +#define LUSTRE_DIRSYNC_FL 0x00010000 /* dirsync behaviour (dir only) */ #ifdef __KERNEL__ -/* Convert wire MDS_*_FL to corresponding client local VFS S_* values - * for the client inode i_flags. The MDS_*_FL are the Lustre wire +/* Convert wire LUSTRE_*_FL to corresponding client local VFS S_* values + * for the client inode i_flags. The LUSTRE_*_FL are the Lustre wire * protocol equivalents of LDISKFS_*_FL values stored on disk, while * the S_* flags are kernel-internal values that change between kernel * versions. These flags are set/cleared via FSFILT_IOC_{GET,SET}_FLAGS. * See b=16526 for a full history. */ static inline int ll_ext_to_inode_flags(int flags) { - return (((flags & MDS_SYNC_FL) ? S_SYNC : 0) | - ((flags & MDS_NOATIME_FL) ? S_NOATIME : 0) | - ((flags & MDS_APPEND_FL) ? S_APPEND : 0) | + return (((flags & LUSTRE_SYNC_FL) ? S_SYNC : 0) | + ((flags & LUSTRE_NOATIME_FL) ? S_NOATIME : 0) | + ((flags & LUSTRE_APPEND_FL) ? S_APPEND : 0) | #if defined(S_DIRSYNC) - ((flags & MDS_DIRSYNC_FL) ? S_DIRSYNC : 0) | + ((flags & LUSTRE_DIRSYNC_FL) ? S_DIRSYNC : 0) | #endif - ((flags & MDS_IMMUTABLE_FL) ? S_IMMUTABLE : 0)); + ((flags & LUSTRE_IMMUTABLE_FL) ? S_IMMUTABLE : 0)); } static inline int ll_inode_to_ext_flags(int iflags) { - return (((iflags & S_SYNC) ? MDS_SYNC_FL : 0) | - ((iflags & S_NOATIME) ? MDS_NOATIME_FL : 0) | - ((iflags & S_APPEND) ? MDS_APPEND_FL : 0) | + return (((iflags & S_SYNC) ? LUSTRE_SYNC_FL : 0) | + ((iflags & S_NOATIME) ? LUSTRE_NOATIME_FL : 0) | + ((iflags & S_APPEND) ? LUSTRE_APPEND_FL : 0) | #if defined(S_DIRSYNC) - ((iflags & S_DIRSYNC) ? MDS_DIRSYNC_FL : 0) | + ((iflags & S_DIRSYNC) ? LUSTRE_DIRSYNC_FL : 0) | #endif - ((iflags & S_IMMUTABLE) ? MDS_IMMUTABLE_FL : 0)); + ((iflags & S_IMMUTABLE) ? LUSTRE_IMMUTABLE_FL : 0)); } #endif -/* - * while mds_body is to interact with 1.6, mdt_body is to interact with 2.0. - * both of them should have the same fields layout, because at client side - * one could be dynamically cast to the other. - * - * mdt_body has large size than mds_body, with unused padding (48 bytes) - * at the end. client always use size of mdt_body to prepare request/reply - * buffers, and actual data could be interepeted as mdt_body or mds_body - * accordingly. - */ -struct mds_body { - struct ll_fid fid1; - struct ll_fid fid2; - struct lustre_handle handle; - __u64 valid; - __u64 size; /* Offset, in the case of MDS_READPAGE */ - __u64 mtime; - __u64 atime; - __u64 ctime; - __u64 blocks; /* XID, in the case of MDS_READPAGE */ - __u64 io_epoch; - __u64 ino; - __u32 fsuid; - __u32 fsgid; - __u32 capability; - __u32 mode; - __u32 uid; - __u32 gid; - __u32 flags; /* from vfs for pin/unpin, MDS_BFLAG for close */ - __u32 rdev; - __u32 nlink; /* #bytes to read in the case of MDS_READPAGE */ - __u32 generation; - __u32 suppgid; - __u32 eadatasize; - __u32 aclsize; - __u32 max_mdsize; - __u32 max_cookiesize; - __u32 padding_4; /* also fix lustre_swab_mds_body */ -}; - -extern void lustre_swab_mds_body (struct mds_body *b); - struct mdt_body { struct lu_fid fid1; struct lu_fid fid2; struct lustre_handle handle; __u64 valid; __u64 size; /* Offset, in the case of MDS_READPAGE */ - __u64 mtime; - __u64 atime; - __u64 ctime; + obd_time mtime; + obd_time atime; + obd_time ctime; __u64 blocks; /* XID, in the case of MDS_READPAGE */ __u64 ioepoch; - __u64 ino; /* for 1.6 compatibility */ + __u64 ino; __u32 fsuid; __u32 fsgid; __u32 capability; __u32 mode; __u32 uid; __u32 gid; - __u32 flags; /* from vfs for pin/unpin, MDS_BFLAG for close */ + __u32 flags; /* from vfs for pin/unpin, LUSTRE_BFLAG close */ __u32 rdev; __u32 nlink; /* #bytes to read in the case of MDS_READPAGE */ - __u32 generation; /* for 1.6 compatibility */ + __u32 generation; __u32 suppgid; __u32 eadatasize; __u32 aclsize; @@ -1676,9 +1811,11 @@ extern void lustre_swab_quota_adjust_qunit(struct quota_adjust_qunit *q); #define LQUOTA_FLAGS_ADJINO 8UL /* adjust the inode qunit size */ #define LQUOTA_FLAGS_CHG_QS 16UL /* indicate whether it has capability of * OBD_CONNECT_CHANGE_QS */ +#define LQUOTA_FLAGS_RECOVERY 32UL /* recovery is going on a uid/gid */ +#define LQUOTA_FLAGS_SETQUOTA 64UL /* being setquota on a uid/gid */ /* flags is specific for quota_adjust_qunit */ -#define LQUOTA_QAQ_CREATE_LQS (1 << 31) /* when it is set, need create lqs */ +#define LQUOTA_QAQ_CREATE_LQS (1UL << 31) /* when it is set, need create lqs */ /* the status of lqs_flags in struct lustre_qunit_size */ #define LQUOTA_QUNIT_FLAGS (LQUOTA_FLAGS_GRP | LQUOTA_FLAGS_BLK) @@ -1693,16 +1830,6 @@ extern void lustre_swab_quota_adjust_qunit(struct quota_adjust_qunit *q); #define QAQ_SET_ADJINO(qaq) ((qaq)->qaq_flags |= LQUOTA_FLAGS_ADJINO) #define QAQ_SET_CREATE_LQS(qaq) ((qaq)->qaq_flags |= LQUOTA_QAQ_CREATE_LQS) -/* inode access permission for remote user, the inode info are omitted, - * for client knows them. */ -struct mds_remote_perm { - __u32 rp_uid; - __u32 rp_gid; - __u32 rp_fsuid; - __u32 rp_fsgid; - __u32 rp_access_perm; /* MAY_READ/WRITE/EXEC */ -}; - /* permissions for md_perm.mp_perm */ enum { CFS_SETUID_PERM = 0x01, @@ -1712,8 +1839,8 @@ enum { CFS_RMTOWN_PERM = 0x10 }; -extern void lustre_swab_mds_remote_perm(struct mds_remote_perm *p); - +/* inode access permission for remote user, the inode info are omitted, + * for client knows them. */ struct mdt_remote_perm { __u32 rp_uid; __u32 rp_gid; @@ -1722,6 +1849,7 @@ struct mdt_remote_perm { __u32 rp_fsgid; __u32 rp_fsgid_h; __u32 rp_access_perm; /* MAY_READ/WRITE/EXEC */ + __u32 rp_padding; }; extern void lustre_swab_mdt_remote_perm(struct mdt_remote_perm *p); @@ -1743,9 +1871,9 @@ struct mdt_rec_setattr { __u32 sa_gid; __u64 sa_size; __u64 sa_blocks; - __u64 sa_mtime; - __u64 sa_atime; - __u64 sa_ctime; + obd_time sa_mtime; + obd_time sa_atime; + obd_time sa_ctime; __u32 sa_attr_flags; __u32 sa_mode; __u32 sa_padding_2; @@ -1784,18 +1912,18 @@ extern void lustre_swab_mdt_rec_setattr (struct mdt_rec_setattr *sa); #define FMODE_WRITE 00000002 #endif +#define MDS_FMODE_CLOSED 00000000 +#define MDS_FMODE_EXEC 00000004 /* IO Epoch is opened on a closed file. */ -#define FMODE_EPOCH 01000000 +#define MDS_FMODE_EPOCH 01000000 /* IO Epoch is opened on a file truncate. */ -#define FMODE_TRUNC 02000000 +#define MDS_FMODE_TRUNC 02000000 /* Size-on-MDS Attribute Update is pending. */ -#define FMODE_SOM 04000000 -#define FMODE_CLOSED 0 +#define MDS_FMODE_SOM 04000000 #define MDS_OPEN_CREATED 00000010 #define MDS_OPEN_CROSS 00000020 -#define MDS_FMODE_EXEC 00000004 #define MDS_OPEN_CREAT 00000100 #define MDS_OPEN_EXCL 00000200 #define MDS_OPEN_TRUNC 00001000 @@ -1803,6 +1931,7 @@ extern void lustre_swab_mdt_rec_setattr (struct mdt_rec_setattr *sa); #define MDS_OPEN_SYNC 00010000 #define MDS_OPEN_DIRECTORY 00200000 +#define MDS_OPEN_BY_FID 040000000 /* open_by_fid for known object */ #define MDS_OPEN_DELAY_CREATE 0100000000 /* delay initial object create */ #define MDS_OPEN_OWNEROVERRIDE 0200000000 /* NFSD rw-reopen ro file for owner */ #define MDS_OPEN_JOIN_FILE 0400000000 /* open for join file. @@ -1818,6 +1947,9 @@ extern void lustre_swab_mdt_rec_setattr (struct mdt_rec_setattr *sa); #define MDS_OPEN_LOCK 04000000000 /* This open requires open lock */ #define MDS_OPEN_HAS_EA 010000000000 /* specify object create pattern */ #define MDS_OPEN_HAS_OBJS 020000000000 /* Just set the EA the obj exist */ +#define MDS_OPEN_NORESTORE 0100000000000ULL /* Do not restore file at open */ +#define MDS_OPEN_NEWSTRIPE 0200000000000ULL /* New stripe needed (restripe or + * hsm restore) */ /* permission for create non-directory file */ #define MAY_CREATE (1 << 7) @@ -1843,9 +1975,13 @@ enum { MDS_PERM_BYPASS = 1 << 3, MDS_SOM = 1 << 4, MDS_QUOTA_IGNORE = 1 << 5, - MDS_CLOSE_CLEANUP = 1 << 6 + MDS_CLOSE_CLEANUP = 1 << 6, + MDS_KEEP_ORPHAN = 1 << 7, + MDS_RECOV_OPEN = 1 << 8, + MDS_UNLINK_DESTROY = 1 << 9, /* Destory ost object in mdd_unlink */ }; +/* instance of mdt_reint_rec */ struct mdt_rec_create { __u32 cr_opcode; __u32 cr_cap; @@ -1860,18 +1996,33 @@ struct mdt_rec_create { struct lu_fid cr_fid1; struct lu_fid cr_fid2; struct lustre_handle cr_old_handle; /* handle in case of open replay */ - __u64 cr_time; + obd_time cr_time; __u64 cr_rdev; __u64 cr_ioepoch; - __u64 cr_padding_1; /* pad for 64 bits*/ + __u64 cr_padding_1; /* rr_blocks */ __u32 cr_mode; __u32 cr_bias; - __u32 cr_flags; /* for use with open */ - __u32 cr_padding_2; - __u32 cr_padding_3; - __u32 cr_padding_4; + /* use of helpers set/get_mrc_cr_flags() is needed to access + * 64 bits cr_flags [cr_flags_l, cr_flags_h], this is done to + * extend cr_flags size without breaking 1.8 compat */ + __u32 cr_flags_l; /* for use with open, low 32 bits */ + __u32 cr_flags_h; /* for use with open, high 32 bits */ + __u32 cr_padding_3; /* rr_padding_3 */ + __u32 cr_padding_4; /* rr_padding_4 */ }; +static inline void set_mrc_cr_flags(struct mdt_rec_create *mrc, __u64 flags) +{ + mrc->cr_flags_l = (__u32)(flags & 0xFFFFFFFFUll); + mrc->cr_flags_h = (__u32)(flags >> 32); +} + +static inline __u64 get_mrc_cr_flags(struct mdt_rec_create *mrc) +{ + return ((__u64)(mrc->cr_flags_l) | ((__u64)mrc->cr_flags_h << 32)); +} + +/* instance of mdt_reint_rec */ struct mdt_rec_link { __u32 lk_opcode; __u32 lk_cap; @@ -1885,19 +2036,20 @@ struct mdt_rec_link { __u32 lk_suppgid2_h; struct lu_fid lk_fid1; struct lu_fid lk_fid2; - __u64 lk_time; - __u64 lk_padding_1; - __u64 lk_padding_2; - __u64 lk_padding_3; - __u64 lk_padding_4; + obd_time lk_time; + __u64 lk_padding_1; /* rr_atime */ + __u64 lk_padding_2; /* rr_ctime */ + __u64 lk_padding_3; /* rr_size */ + __u64 lk_padding_4; /* rr_blocks */ __u32 lk_bias; - __u32 lk_padding_5; - __u32 lk_padding_6; - __u32 lk_padding_7; - __u32 lk_padding_8; - __u32 lk_padding_9; + __u32 lk_padding_5; /* rr_mode */ + __u32 lk_padding_6; /* rr_flags */ + __u32 lk_padding_7; /* rr_padding_2 */ + __u32 lk_padding_8; /* rr_padding_3 */ + __u32 lk_padding_9; /* rr_padding_4 */ }; +/* instance of mdt_reint_rec */ struct mdt_rec_unlink { __u32 ul_opcode; __u32 ul_cap; @@ -1911,19 +2063,20 @@ struct mdt_rec_unlink { __u32 ul_suppgid2_h; struct lu_fid ul_fid1; struct lu_fid ul_fid2; - __u64 ul_time; - __u64 ul_padding_2; - __u64 ul_padding_3; - __u64 ul_padding_4; - __u64 ul_padding_5; + obd_time ul_time; + __u64 ul_padding_2; /* rr_atime */ + __u64 ul_padding_3; /* rr_ctime */ + __u64 ul_padding_4; /* rr_size */ + __u64 ul_padding_5; /* rr_blocks */ __u32 ul_bias; __u32 ul_mode; - __u32 ul_padding_6; - __u32 ul_padding_7; - __u32 ul_padding_8; - __u32 ul_padding_9; + __u32 ul_padding_6; /* rr_flags */ + __u32 ul_padding_7; /* rr_padding_2 */ + __u32 ul_padding_8; /* rr_padding_3 */ + __u32 ul_padding_9; /* rr_padding_4 */ }; +/* instance of mdt_reint_rec */ struct mdt_rec_rename { __u32 rn_opcode; __u32 rn_cap; @@ -1937,19 +2090,20 @@ struct mdt_rec_rename { __u32 rn_suppgid2_h; struct lu_fid rn_fid1; struct lu_fid rn_fid2; - __u64 rn_time; - __u64 rn_padding_1; - __u64 rn_padding_2; - __u64 rn_padding_3; - __u64 rn_padding_4; - __u32 rn_bias; /* some operation flags */ - __u32 rn_mode; /* cross-ref rename has mode */ - __u32 rn_padding_5; - __u32 rn_padding_6; - __u32 rn_padding_7; - __u32 rn_padding_8; -}; - + obd_time rn_time; + __u64 rn_padding_1; /* rr_atime */ + __u64 rn_padding_2; /* rr_ctime */ + __u64 rn_padding_3; /* rr_size */ + __u64 rn_padding_4; /* rr_blocks */ + __u32 rn_bias; /* some operation flags */ + __u32 rn_mode; /* cross-ref rename has mode */ + __u32 rn_padding_5; /* rr_flags */ + __u32 rn_padding_6; /* rr_padding_2 */ + __u32 rn_padding_7; /* rr_padding_3 */ + __u32 rn_padding_8; /* rr_padding_4 */ +}; + +/* instance of mdt_reint_rec */ struct mdt_rec_setxattr { __u32 sx_opcode; __u32 sx_cap; @@ -1962,22 +2116,30 @@ struct mdt_rec_setxattr { __u32 sx_suppgid2; __u32 sx_suppgid2_h; struct lu_fid sx_fid; - __u64 sx_padding_1; /* These three members are lu_fid size */ + __u64 sx_padding_1; /* These three are rr_fid2 */ __u32 sx_padding_2; __u32 sx_padding_3; __u64 sx_valid; - __u64 sx_time; - __u64 sx_padding_5; - __u64 sx_padding_6; - __u64 sx_padding_7; + obd_time sx_time; + __u64 sx_padding_5; /* rr_ctime */ + __u64 sx_padding_6; /* rr_size */ + __u64 sx_padding_7; /* rr_blocks */ __u32 sx_size; __u32 sx_flags; - __u32 sx_padding_8; - __u32 sx_padding_9; - __u32 sx_padding_10; - __u32 sx_padding_11; + __u32 sx_padding_8; /* rr_flags */ + __u32 sx_padding_9; /* rr_padding_2 */ + __u32 sx_padding_10; /* rr_padding_3 */ + __u32 sx_padding_11; /* rr_padding_4 */ }; +/* + * mdt_rec_reint is the template for all mdt_reint_xxx structures. + * Do NOT change the size of various members, otherwise the value + * will be broken in lustre_swab_mdt_rec_reint(). + * + * If you add new members in other mdt_reint_xxx structres and need to use the + * rr_padding_x fields, then update lustre_swab_mdt_rec_reint() also. + */ struct mdt_rec_reint { __u32 rr_opcode; __u32 rr_cap; @@ -1991,14 +2153,14 @@ struct mdt_rec_reint { __u32 rr_suppgid2_h; struct lu_fid rr_fid1; struct lu_fid rr_fid2; - __u64 rr_mtime; - __u64 rr_atime; - __u64 rr_ctime; + obd_time rr_mtime; + obd_time rr_atime; + obd_time rr_ctime; __u64 rr_size; __u64 rr_blocks; __u32 rr_bias; __u32 rr_mode; - __u32 rr_padding_1; /* also fix lustre_swab_mdt_rec_reint */ + __u32 rr_flags; __u32 rr_padding_2; /* also fix lustre_swab_mdt_rec_reint */ __u32 rr_padding_3; /* also fix lustre_swab_mdt_rec_reint */ __u32 rr_padding_4; /* also fix lustre_swab_mdt_rec_reint */ @@ -2043,13 +2205,6 @@ extern void lustre_swab_lmv_stripe_md(struct lmv_stripe_md *mea); #define MAX_HASH_SIZE 0x7fffffffffffffffULL #define MAX_HASH_HIGHEST_BIT 0x1000000000000000ULL -struct md_fld { - seqno_t mf_seq; - mdsno_t mf_mds; -}; - -extern void lustre_swab_md_fld (struct md_fld *mf); - enum fld_rpc_opc { FLD_QUERY = 900, FLD_LAST_OPC, @@ -2071,11 +2226,6 @@ enum seq_op { * LOV data structures */ -#define LOV_MIN_STRIPE_BITS 16 /* maximum PAGE_SIZE (ia64), power of 2 */ -#define LOV_MIN_STRIPE_SIZE (1<start <= ex2->end) && (ex2->start <= ex1->end); } +/* check if @ex1 contains @ex2 */ +static inline int ldlm_extent_contain(struct ldlm_extent *ex1, + struct ldlm_extent *ex2) +{ + return (ex1->start <= ex2->start) && (ex1->end >= ex2->end); +} + struct ldlm_inodebits { __u64 bits; }; -struct ldlm_flock { - __u64 start; - __u64 end; - __u64 blocking_export; /* not actually used over the wire */ - __u32 blocking_pid; /* not actually used over the wire */ - __u32 pid; +struct ldlm_flock_wire { + __u64 lfw_start; + __u64 lfw_end; + __u64 lfw_owner; + __u32 lfw_padding; + __u32 lfw_pid; }; /* it's important that the fields of the ldlm_extent structure match @@ -2183,11 +2346,20 @@ struct ldlm_flock { typedef union { struct ldlm_extent l_extent; - struct ldlm_flock l_flock; + struct ldlm_flock_wire l_flock; struct ldlm_inodebits l_inodebits; -} ldlm_policy_data_t; +} ldlm_wire_policy_data_t; -extern void lustre_swab_ldlm_policy_data (ldlm_policy_data_t *d); +extern void lustre_swab_ldlm_policy_data (ldlm_wire_policy_data_t *d); + +/* Similarly to ldlm_wire_policy_data_t, there is one common swabber for all + * LVB types. As a result, any new LVB structure must match the fields of the + * ost_lvb structure. */ +union ldlm_wire_lvb { + struct ost_lvb l_ost; +}; + +extern void lustre_swab_lvb(union ldlm_wire_lvb *); struct ldlm_intent { __u64 opc; @@ -2207,7 +2379,7 @@ struct ldlm_lock_desc { struct ldlm_resource_desc l_resource; ldlm_mode_t l_req_mode; ldlm_mode_t l_granted_mode; - ldlm_policy_data_t l_policy_data; + ldlm_wire_policy_data_t l_policy_data; }; extern void lustre_swab_ldlm_lock_desc (struct ldlm_lock_desc *l); @@ -2256,6 +2428,7 @@ typedef enum { MGS_TARGET_REG, /* whenever target starts up */ MGS_TARGET_DEL, MGS_SET_INFO, + MGS_CONFIG_READ, MGS_LAST_OPC } mgs_cmd_t; #define MGS_FIRST_OPC MGS_CONNECT @@ -2268,25 +2441,55 @@ struct mgs_send_param { }; /* We pass this info to the MGS so it can write config logs */ -#define MTI_NAME_MAXLEN 64 +#define MTI_NAME_MAXLEN 64 #define MTI_PARAM_MAXLEN 4096 -#define MTI_NIDS_MAX 32 +#define MTI_NIDS_MAX 32 struct mgs_target_info { __u32 mti_lustre_ver; __u32 mti_stripe_index; __u32 mti_config_ver; __u32 mti_flags; __u32 mti_nid_count; - __u32 padding; /* 64 bit align */ + __u32 mti_instance; /* Running instance of target */ char mti_fsname[MTI_NAME_MAXLEN]; char mti_svname[MTI_NAME_MAXLEN]; char mti_uuid[sizeof(struct obd_uuid)]; __u64 mti_nids[MTI_NIDS_MAX]; /* host nids (lnet_nid_t)*/ char mti_params[MTI_PARAM_MAXLEN]; }; - extern void lustre_swab_mgs_target_info(struct mgs_target_info *oinfo); +struct mgs_nidtbl_entry { + __u64 mne_version; /* table version of this entry */ + __u32 mne_instance; /* target instance # */ + __u32 mne_index; /* target index */ + __u32 mne_length; /* length of this entry - by bytes */ + __u8 mne_type; /* target type LDD_F_SV_TYPE_OST/MDT */ + __u8 mne_nid_type; /* type of nid(mbz). for ipv6. */ + __u8 mne_nid_size; /* size of each NID, by bytes */ + __u8 mne_nid_count; /* # of NIDs in buffer */ + union { + lnet_nid_t nids[0]; /* variable size buffer for NIDs. */ + } u; +}; +extern void lustre_swab_mgs_nidtbl_entry(struct mgs_nidtbl_entry *oinfo); + +struct mgs_config_body { + char mcb_name[MTI_NAME_MAXLEN]; /* logname */ + __u64 mcb_offset; /* next index of config log to request */ + __u16 mcb_type; /* type of log: CONFIG_T_[CONFIG|RECOVER] */ + __u8 mcb_reserved; + __u8 mcb_bits; /* bits unit size of config log */ + __u32 mcb_units; /* # of units for bulk transfer */ +}; +extern void lustre_swab_mgs_config_body(struct mgs_config_body *body); + +struct mgs_config_res { + __u64 mcr_offset; /* index of last config log */ + __u64 mcr_size; /* size of the log */ +}; +extern void lustre_swab_mgs_config_res(struct mgs_config_res *body); + /* Config marker flags (in config log) */ #define CM_START 0x01 #define CM_END 0x02 @@ -2299,9 +2502,9 @@ struct cfg_marker { __u32 cm_step; /* aka config version */ __u32 cm_flags; __u32 cm_vers; /* lustre release version number */ - __u32 padding; /* 64 bit align */ - __u64 cm_createtime; /*when this record was first created */ - __u64 cm_canceltime; /*when this record is no longer valid*/ + __u32 cm_padding; /* 64 bit align */ + obd_time cm_createtime; /*when this record was first created */ + obd_time cm_canceltime; /*when this record is no longer valid*/ char cm_tgtname[MTI_NAME_MAXLEN]; char cm_comment[MTI_NAME_MAXLEN]; }; @@ -2379,7 +2582,7 @@ struct llog_rec_hdr { __u32 lrh_len; __u32 lrh_index; __u32 lrh_type; - __u32 padding; + __u32 lrh_padding; }; struct llog_rec_tail { @@ -2390,11 +2593,11 @@ struct llog_rec_tail { struct llog_logid_rec { struct llog_rec_hdr lid_hdr; struct llog_logid lid_id; - __u32 padding1; - __u32 padding2; - __u32 padding3; - __u32 padding4; - __u32 padding5; + __u32 lid_padding1; + __u32 lid_padding2; + __u32 lid_padding3; + __u32 lid_padding4; + __u32 lid_padding5; struct llog_rec_tail lid_tail; } __attribute__((packed)); @@ -2403,7 +2606,7 @@ struct llog_create_rec { struct ll_fid lcr_fid; obd_id lcr_oid; obd_count lcr_oseq; - __u32 padding; + __u32 lcr_padding; struct llog_rec_tail lcr_tail; } __attribute__((packed)); @@ -2411,7 +2614,7 @@ struct llog_orphan_rec { struct llog_rec_hdr lor_hdr; obd_id lor_oid; obd_count lor_ogen; - __u32 padding; + __u32 lor_padding; struct llog_rec_tail lor_tail; } __attribute__((packed)); @@ -2429,7 +2632,7 @@ struct llog_setattr_rec { obd_count lsr_oseq; __u32 lsr_uid; __u32 lsr_gid; - __u32 padding; + __u32 lsr_padding; struct llog_rec_tail lsr_tail; } __attribute__((packed)); @@ -2437,7 +2640,7 @@ struct llog_setattr64_rec { struct llog_rec_hdr lsr_hdr; obd_id lsr_oid; obd_count lsr_oseq; - __u32 padding; + __u32 lsr_padding; __u32 lsr_uid; __u32 lsr_uid_h; __u32 lsr_gid; @@ -2449,7 +2652,7 @@ struct llog_size_change_rec { struct llog_rec_hdr lsc_hdr; struct ll_fid lsc_fid; __u32 lsc_ioepoch; - __u32 padding; + __u32 lsc_padding; struct llog_rec_tail lsc_tail; } __attribute__((packed)); @@ -2460,7 +2663,7 @@ struct llog_size_change_rec { /** bits covering all \a changelog_rec_type's */ #define CHANGELOG_ALLMASK 0XFFFFFFFF /** default \a changelog_rec_type mask */ -#define CHANGELOG_DEFMASK CHANGELOG_ALLMASK +#define CHANGELOG_DEFMASK CHANGELOG_ALLMASK & ~(1 << CL_ATIME | 1 << CL_CLOSE) /* changelog llog name, needed by client replicators */ #define CHANGELOG_CATALOG "changelog_catalog" @@ -2477,6 +2680,12 @@ struct llog_changelog_rec { struct llog_rec_tail cr_tail; /**< for_sizezof_only */ } __attribute__((packed)); +struct llog_changelog_ext_rec { + struct llog_rec_hdr cr_hdr; + struct changelog_ext_rec cr; + struct llog_rec_tail cr_tail; /**< for_sizezof_only */ +} __attribute__((packed)); + #define CHANGELOG_USER_PREFIX "cl" struct llog_changelog_user_rec { @@ -2511,7 +2720,7 @@ struct llog_gen_rec { struct llog_log_hdr { struct llog_rec_hdr llh_hdr; - __u64 llh_timestamp; + obd_time llh_timestamp; __u32 llh_count; __u32 llh_bitmap_offset; __u32 llh_size; @@ -2544,7 +2753,7 @@ enum llogd_rpc_ops { LLOG_ORIGIN_HANDLE_WRITE_REC = 504, LLOG_ORIGIN_HANDLE_CLOSE = 505, LLOG_ORIGIN_CONNECT = 506, - LLOG_CATINFO = 507, /* for lfs catinfo */ + LLOG_CATINFO = 507, /* deprecated */ LLOG_ORIGIN_HANDLE_PREV_BLOCK = 508, LLOG_ORIGIN_HANDLE_DESTROY = 509, /* for destroy llog object*/ LLOG_LAST_OPC, @@ -2591,12 +2800,17 @@ struct obdo { __u64 o_ioepoch; /* epoch in ost writes */ __u32 o_stripe_idx; /* holds stripe idx */ __u32 o_parent_ver; - struct lustre_handle o_handle; /* brw: lock handle to prolong locks */ - struct llog_cookie o_lcookie; /* destroy: unlink cookie from MDS */ - + struct lustre_handle o_handle; /* brw: lock handle to prolong + * locks */ + struct llog_cookie o_lcookie; /* destroy: unlink cookie from + * MDS */ __u32 o_uid_h; __u32 o_gid_h; - __u64 o_padding_3; + + __u64 o_data_version; /* getattr: sum of iversion for + * each stripe. + * brw: grant space consumed on + * the client for the write */ __u64 o_padding_4; __u64 o_padding_5; __u64 o_padding_6; @@ -2608,6 +2822,7 @@ struct obdo { #define o_undirty o_mode #define o_dropped o_misc #define o_cksum o_nlink +#define o_grant_used o_data_version static inline void lustre_set_wire_obdo(struct obdo *wobdo, struct obdo *lobdo) { @@ -2617,13 +2832,19 @@ static inline void lustre_set_wire_obdo(struct obdo *wobdo, struct obdo *lobdo) static inline void lustre_get_wire_obdo(struct obdo *lobdo, struct obdo *wobdo) { - obd_flag local_flags = lobdo->o_flags & OBD_FL_LOCAL_MASK; + obd_flag local_flags = 0; + + if (lobdo->o_valid & OBD_MD_FLFLAGS) + local_flags = lobdo->o_flags & OBD_FL_LOCAL_MASK; LASSERT(!(wobdo->o_flags & OBD_FL_LOCAL_MASK)); memcpy(lobdo, wobdo, sizeof(*lobdo)); - lobdo->o_flags &= ~OBD_FL_LOCAL_MASK; - lobdo->o_flags |= local_flags; + if (local_flags != 0) { + lobdo->o_valid |= OBD_MD_FLFLAGS; + lobdo->o_flags &= ~OBD_FL_LOCAL_MASK; + lobdo->o_flags |= local_flags; + } } extern void lustre_swab_obdo (struct obdo *o);