X-Git-Url: https://git.whamcloud.com/?p=fs%2Flustre-release.git;a=blobdiff_plain;f=lustre%2Finclude%2Flustre%2Flustre_idl.h;h=3be6942c789d20d3029d1dddcd08827f84b0c42d;hp=bc2cfd277b2703b0210ef446971100f25cac504d;hb=70e80ade90af09300396706b8910e196a7928520;hpb=f89a61cada93accbd02beee89f0610756ca595e9 diff --git a/lustre/include/lustre/lustre_idl.h b/lustre/include/lustre/lustre_idl.h index bc2cfd2..3be6942 100644 --- a/lustre/include/lustre/lustre_idl.h +++ b/lustre/include/lustre/lustre_idl.h @@ -1,7 +1,39 @@ /* -*- mode: c; c-basic-offset: 8; indent-tabs-mode: nil; -*- * vim:expandtab:shiftwidth=8:tabstop=8: * - * This file is part of Lustre, http://www.lustre.org + * GPL HEADER START + * + * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 only, + * as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License version 2 for more details (a copy is included + * in the LICENSE file that accompanied this code). + * + * You should have received a copy of the GNU General Public License + * version 2 along with this program; If not, see [sun.com URL with a + * copy of GPLv2]. + * + * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara, + * CA 95054 USA or visit www.sun.com if you need additional information or + * have any questions. + * + * GPL HEADER END + */ +/* + * Copyright 2008 Sun Microsystems, Inc. All rights reserved + * Use is subject to license terms. + */ +/* + * This file is part of Lustre, http://www.lustre.org/ + * Lustre is a trademark of Sun Microsystems, Inc. + * + * lustre/include/lustre/lustre_idl.h * * Lustre wire protocol definitions. * @@ -52,17 +84,9 @@ #ifndef _LUSTRE_IDL_H_ #define _LUSTRE_IDL_H_ -#include +#include -#if defined(__linux__) -#include -#elif defined(__APPLE__) -#include -#elif defined(__WINNT__) -#include -#else -#error Unsupported operating system. -#endif +#include /* Defn's shared with user-space. */ #include @@ -196,7 +220,7 @@ static inline int range_is_exhausted(struct lu_range *r) struct lu_fid { __u64 f_seq; /* holds fid sequence. Lustre should support 2 ^ 64 * objects, thus even if one sequence has one object we - * reach this value. */ + * will never reach this value. */ __u32 f_oid; /* fid number within its sequence. */ __u32 f_ver; /* holds fid version. */ }; @@ -205,8 +229,6 @@ struct lu_fid { * fid constants */ enum { - LUSTRE_ROOT_FID_SEQ = 1ULL, /* XXX: should go into mkfs. */ - /* initial fid id value */ LUSTRE_FID_INIT_OID = 1UL }; @@ -229,19 +251,55 @@ static inline __u32 fid_ver(const struct lu_fid *fid) return fid->f_ver; } -static inline int fid_seq_is_sane(__u64 seq) -{ - return seq != 0; -} - static inline void fid_zero(struct lu_fid *fid) { memset(fid, 0, sizeof(*fid)); } +/* Normal FID sequence starts from this value, i.e. 1<<33 */ +#define FID_SEQ_START 0x200000000ULL + +/* IDIF sequence starts from this value, i.e. 1<<32 */ +#define IDIF_SEQ_START 0x100000000ULL + +/** + * Check if a fid is igif or not. + * \param fid the fid to be tested. + * \return true if the fid is a igif; otherwise false. + */ static inline int fid_is_igif(const struct lu_fid *fid) { - return fid_seq(fid) == LUSTRE_ROOT_FID_SEQ; + return fid_seq(fid) > 0 && fid_seq(fid) < IDIF_SEQ_START; +} + +/** + * Check if a fid is idif or not. + * \param fid the fid to be tested. + * \return true if the fid is a idif; otherwise false. + */ +static inline int fid_is_idif(const struct lu_fid *fid) +{ + return fid_seq(fid) >= IDIF_SEQ_START && fid_seq(fid) < FID_SEQ_START; +} + +/** + * Get inode number from a igif. + * \param fid a igif to get inode number from. + * \return inode number for the igif. + */ +static inline ino_t lu_igif_ino(const struct lu_fid *fid) +{ + return fid_seq(fid); +} + +/** + * Get inode generation from a igif. + * \param fid a igif to get inode generation from. + * \return inode generation for the igif. + */ +static inline __u32 lu_igif_gen(const struct lu_fid *fid) +{ + return fid_oid(fid); } #define DFID "[0x%16.16"LPF64"x/0x%8.8x:0x%8.8x]" @@ -251,7 +309,30 @@ static inline int fid_is_igif(const struct lu_fid *fid) fid_oid(fid), \ fid_ver(fid) -#ifdef __KERNEL__ +static inline void fid_cpu_to_le(struct lu_fid *dst, const struct lu_fid *src) +{ + /* check that all fields are converted */ + CLASSERT(sizeof *src == + sizeof fid_seq(src) + + sizeof fid_oid(src) + sizeof fid_ver(src)); + LASSERTF(fid_is_igif(src) || fid_ver(src) == 0, DFID"\n", PFID(src)); + dst->f_seq = cpu_to_le64(fid_seq(src)); + dst->f_oid = cpu_to_le32(fid_oid(src)); + dst->f_ver = cpu_to_le32(fid_ver(src)); +} + +static inline void fid_le_to_cpu(struct lu_fid *dst, const struct lu_fid *src) +{ + /* check that all fields are converted */ + CLASSERT(sizeof *src == + sizeof fid_seq(src) + + sizeof fid_oid(src) + sizeof fid_ver(src)); + dst->f_seq = le64_to_cpu(fid_seq(src)); + dst->f_oid = le32_to_cpu(fid_oid(src)); + dst->f_ver = le32_to_cpu(fid_ver(src)); + LASSERTF(fid_is_igif(dst) || fid_ver(dst) == 0, DFID"\n", PFID(dst)); +} + static inline void fid_cpu_to_be(struct lu_fid *dst, const struct lu_fid *src) { /* check that all fields are converted */ @@ -276,6 +357,7 @@ static inline void fid_be_to_cpu(struct lu_fid *dst, const struct lu_fid *src) LASSERTF(fid_is_igif(dst) || fid_ver(dst) == 0, DFID"\n", PFID(dst)); } +#ifdef __KERNEL__ /* * Storage representation for fids. * @@ -298,7 +380,7 @@ static inline int fid_is_sane(const struct lu_fid *fid) { return fid != NULL && - ((fid_seq_is_sane(fid_seq(fid)) && fid_oid(fid) != 0 + ((fid_seq(fid) >= FID_SEQ_START && fid_oid(fid) != 0 && fid_ver(fid) == 0) || fid_is_igif(fid)); } @@ -327,16 +409,17 @@ static inline int lu_fid_eq(const struct lu_fid *f0, */ struct lu_dirent { struct lu_fid lde_fid; - __u32 lde_hash; + __u64 lde_hash; __u16 lde_reclen; __u16 lde_namelen; + __u32 lde_pad0; char lde_name[0]; }; struct lu_dirpage { - __u32 ldp_hash_start; - __u32 ldp_hash_end; - __u16 ldp_flags; + __u64 ldp_hash_start; + __u64 ldp_hash_end; + __u32 ldp_flags; __u32 ldp_pad0; struct lu_dirent ldp_entries[0]; }; @@ -347,7 +430,7 @@ enum lu_dirpage_flags { static inline struct lu_dirent *lu_dirent_start(struct lu_dirpage *dp) { - if (le16_to_cpu(dp->ldp_flags) & LDF_EMPTY) + if (le32_to_cpu(dp->ldp_flags) & LDF_EMPTY) return NULL; else return dp->ldp_entries; @@ -369,12 +452,12 @@ static inline int lu_dirent_size(struct lu_dirent *ent) { if (le16_to_cpu(ent->lde_reclen) == 0) { return (sizeof(*ent) + - le16_to_cpu(ent->lde_namelen) + 3) & ~3; + le16_to_cpu(ent->lde_namelen) + 7) & ~7; } return le16_to_cpu(ent->lde_reclen); } -#define DIR_END_OFF 0xfffffffeUL +#define DIR_END_OFF 0xfffffffffffffffeULL struct lustre_handle { __u64 cookie; @@ -398,23 +481,8 @@ static inline void lustre_handle_copy(struct lustre_handle *tgt, tgt->cookie = src->cookie; } -/* we depend on this structure to be 8-byte aligned */ -/* this type is only endian-adjusted in lustre_unpack_msg() */ -struct lustre_msg_v1 { - struct lustre_handle lm_handle; - __u32 lm_magic; - __u32 lm_type; - __u32 lm_version; - __u32 lm_opc; - __u64 lm_last_xid; - __u64 lm_last_committed; - __u64 lm_transno; - __u32 lm_status; - __u32 lm_flags; - __u32 lm_conn_cnt; - __u32 lm_bufcount; - __u32 lm_buflens[0]; -}; +/* flags for lm_flags */ +#define MSGHDR_AT_SUPPORT 0x1 #define lustre_msg lustre_msg_v2 /* we depend on this structure to be 8-byte aligned */ @@ -424,8 +492,8 @@ struct lustre_msg_v2 { __u32 lm_secflvr; __u32 lm_magic; __u32 lm_repsize; - __u32 lm_timeout; - __u32 lm_padding_1; + __u32 lm_cksum; + __u32 lm_flags; __u32 lm_padding_2; __u32 lm_padding_3; __u32 lm_buflens[0]; @@ -445,11 +513,10 @@ struct ptlrpc_body { __u32 pb_flags; __u32 pb_op_flags; __u32 pb_conn_cnt; - __u32 pb_padding_1; - __u32 pb_padding_2; - __u32 pb_padding_3; - __u32 pb_padding_4; - __u32 pb_padding_5; + __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; }; extern void lustre_swab_ptlrpc_body(struct ptlrpc_body *pb); @@ -479,12 +546,16 @@ extern void lustre_swab_ptlrpc_body(struct ptlrpc_body *pb); #define MSG_OP_FLAG_SHIFT 16 /* Flags that apply to all requests are in the bottom 16 bits */ -#define MSG_GEN_FLAG_MASK 0x0000ffff -#define MSG_LAST_REPLAY 1 -#define MSG_RESENT 2 -#define MSG_REPLAY 4 -#define MSG_REQ_REPLAY_DONE 8 -#define MSG_LOCK_REPLAY_DONE 16 +#define MSG_GEN_FLAG_MASK 0x0000ffff +#define MSG_LAST_REPLAY 0x0001 +#define MSG_RESENT 0x0002 +#define MSG_REPLAY 0x0004 +/* #define MSG_AT_SUPPORT 0x0008 + * This was used in early prototypes of adaptive timeouts, and while there + * shouldn't be any users of that code there also isn't a need for using this + * bits. Defer usage until at least 1.10 to avoid potential conflict. */ +#define MSG_REQ_REPLAY_DONE 0x0010 +#define MSG_LOCK_REPLAY_DONE 0x0020 /* * Flags for all connect opcodes (MDS_CONNECT, OST_CONNECT) @@ -526,9 +597,20 @@ extern void lustre_swab_ptlrpc_body(struct ptlrpc_body *pb); #define OBD_CONNECT_LRU_RESIZE 0x02000000ULL /* Lru resize feature. */ #define OBD_CONNECT_MDS_MDS 0x04000000ULL /* MDS-MDS connection*/ #define OBD_CONNECT_REAL 0x08000000ULL /* real connection */ +#define OBD_CONNECT_CHANGE_QS 0x10000000ULL /*shrink/enlarge qunit size + *b=10600 */ +#define OBD_CONNECT_CKSUM 0x20000000ULL /* support several cksum algos */ +#define OBD_CONNECT_FID 0x40000000ULL /* FID is supported by server */ + /* also update obd_connect_names[] for lprocfs_rd_connect_flags() * and lustre/utils/wirecheck.c */ +#ifdef HAVE_LRU_RESIZE_SUPPORT +#define LRU_RESIZE_CONNECT_FLAG OBD_CONNECT_LRU_RESIZE +#else +#define LRU_RESIZE_CONNECT_FLAG 0 +#endif + #define MDT_CONNECT_SUPPORTED (OBD_CONNECT_RDONLY | OBD_CONNECT_VERSION | \ OBD_CONNECT_ACL | OBD_CONNECT_XATTR | \ OBD_CONNECT_IBITS | OBD_CONNECT_JOIN | \ @@ -536,14 +618,18 @@ extern void lustre_swab_ptlrpc_body(struct ptlrpc_body *pb); OBD_CONNECT_LCL_CLIENT | \ OBD_CONNECT_RMT_CLIENT | \ OBD_CONNECT_MDS_CAPA | OBD_CONNECT_OSS_CAPA | \ - OBD_CONNECT_MDS_MDS | OBD_CONNECT_CANCELSET) + OBD_CONNECT_MDS_MDS | OBD_CONNECT_CANCELSET | \ + OBD_CONNECT_FID | \ + LRU_RESIZE_CONNECT_FLAG | OBD_CONNECT_AT) #define OST_CONNECT_SUPPORTED (OBD_CONNECT_SRVLOCK | OBD_CONNECT_GRANT | \ OBD_CONNECT_REQPORTAL | OBD_CONNECT_VERSION | \ OBD_CONNECT_TRUNCLOCK | OBD_CONNECT_INDEX | \ OBD_CONNECT_BRW_SIZE | OBD_CONNECT_QUOTA64 | \ - OBD_CONNECT_OSS_CAPA | OBD_CONNECT_CANCELSET) + OBD_CONNECT_OSS_CAPA | OBD_CONNECT_CANCELSET | \ + OBD_CONNECT_CKSUM | LRU_RESIZE_CONNECT_FLAG | \ + OBD_CONNECT_AT) #define ECHO_CONNECT_SUPPORTED (0) -#define MGS_CONNECT_SUPPORTED (OBD_CONNECT_VERSION) +#define MGS_CONNECT_SUPPORTED (OBD_CONNECT_VERSION | OBD_CONNECT_AT) #define MAX_QUOTA_COUNT32 (0xffffffffULL) @@ -554,9 +640,6 @@ extern void lustre_swab_ptlrpc_body(struct ptlrpc_body *pb); #define OBD_OCD_VERSION_PATCH(version) ((int)((version)>>8)&255) #define OBD_OCD_VERSION_FIX(version) ((int)(version)&255) -#define exp_connect_cancelset(exp) \ - ((exp) ? (exp)->exp_connect_flags & OBD_CONNECT_CANCELSET : 0) - /* This structure is used for both request and reply. * * If we eventually have separate connect data for different types, which we @@ -572,14 +655,25 @@ struct obd_connect_data { __u32 ocd_nllg; /* non-local-lustre-group */ __u64 ocd_transno; /* first transno from client to be replayed */ __u32 ocd_group; /* MDS group on OST */ - __u32 padding1; /* also fix lustre_swab_connect */ + __u32 ocd_cksum_types; /* supported checksum algorithms */ + __u64 padding1; /* also fix lustre_swab_connect */ __u64 padding2; /* also fix lustre_swab_connect */ - __u64 padding3; /* also fix lustre_swab_connect */ }; extern void lustre_swab_connect(struct obd_connect_data *ocd); /* + * Supported checksum algorithms. Up to 32 checksum types are supported. + * (32-bit mask stored in obd_connect_data::ocd_cksum_types) + * Please update DECLARE_CKSUM_NAME/OBD_CKSUM_ALL in obd.h when adding a new + * algorithm and also the OBD_FL_CKSUM* flags. + */ +typedef enum { + OBD_CKSUM_CRC32 = 0x00000001, + OBD_CKSUM_ADLER = 0x00000002, +} cksum_type_t; + +/* * OST requests: OBDO & OBD request records */ @@ -639,6 +733,13 @@ typedef uint32_t obd_count; #define OBD_FL_TRUNCLOCK (0x00000800) /* + * Checksum types + */ +#define OBD_FL_CKSUM_CRC32 (0x00001000) +#define OBD_FL_CKSUM_ADLER (0x00002000) +#define OBD_FL_CKSUM_ALL (OBD_FL_CKSUM_CRC32 | OBD_FL_CKSUM_ADLER) + +/* * This should not be smaller than sizeof(struct lustre_handle) + sizeof(struct * llog_cookie) + sizeof(struct ll_fid). Nevertheless struct ll_fid is not * longer stored in o_inline, we keep this just for case. @@ -692,6 +793,9 @@ extern void lustre_swab_obdo (struct obdo *o); #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 */ @@ -720,9 +824,6 @@ extern void lustre_swab_lov_mds_md(struct lov_mds_md *llm); #define XATTR_NAME_ACL_DEFAULT "system.posix_acl_default" #define XATTR_NAME_LOV "trusted.lov" -/* remote ACL */ -#define XATTR_NAME_LUSTRE_ACL "system.lustre_acl" - #define OBD_MD_FLID (0x00000001ULL) /* object ID */ #define OBD_MD_FLATIME (0x00000002ULL) /* access time */ #define OBD_MD_FLMTIME (0x00000004ULL) /* data modification time */ @@ -770,6 +871,11 @@ extern void lustre_swab_lov_mds_md(struct lov_mds_md *llm); #define OBD_MD_FLCKSPLIT (0x0000080000000000ULL) /* Check split on server */ #define OBD_MD_FLCROSSREF (0x0000100000000000ULL) /* Cross-ref 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_FLGETATTR (OBD_MD_FLID | OBD_MD_FLATIME | OBD_MD_FLMTIME | \ OBD_MD_FLCTIME | OBD_MD_FLSIZE | OBD_MD_FLBLKSZ | \ OBD_MD_FLMODE | OBD_MD_FLTYPE | OBD_MD_FLUID | \ @@ -813,6 +919,8 @@ struct obd_statfs { }; 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 */ /* ost_body.data values for OST_BRW */ @@ -861,6 +969,16 @@ extern void lustre_swab_ost_last_id(obd_id *id); /* lock value block communicated between the filter and llite */ +/* OST_LVB_ERR_INIT is needed because the return code in rc is + * negative, i.e. because ((MASK + rc) & MASK) != MASK. */ +#define OST_LVB_ERR_INIT 0xffbadbad80000000ULL +#define OST_LVB_ERR_MASK 0xffbadbad00000000ULL +#define OST_LVB_IS_ERR(blocks) \ + ((blocks & OST_LVB_ERR_MASK) == OST_LVB_ERR_MASK) +#define OST_LVB_SET_ERR(blocks, rc) \ + do { blocks = OST_LVB_ERR_INIT + rc; } while (0) +#define OST_LVB_GET_ERR(blocks) (int)(blocks - OST_LVB_ERR_INIT) + struct ost_lvb { __u64 lvb_size; __u64 lvb_mtime; @@ -894,7 +1012,7 @@ typedef enum { MDS_QUOTACHECK = 47, MDS_QUOTACTL = 48, MDS_GETXATTR = 49, - MDS_SETXATTR = 50, + MDS_SETXATTR = 50, /* obsolete, now it's MDS_REINT op */ MDS_WRITEPAGE = 51, MDS_IS_SUBDIR = 52, MDS_LAST_OPC @@ -913,6 +1031,9 @@ typedef enum { REINT_UNLINK = 4, REINT_RENAME = 5, REINT_OPEN = 6, + REINT_SETXATTR = 7, +// REINT_CLOSE = 8, +// REINT_WRITE = 9, REINT_MAX } mds_reint_t, mdt_reint_t; @@ -1031,6 +1152,7 @@ struct mdt_body { __u64 ctime; __u64 blocks; /* XID, in the case of MDS_READPAGE */ __u64 ioepoch; + __u64 ino; /* for 1.6 compatibility */ __u32 fsuid; __u32 fsgid; __u32 capability; @@ -1040,11 +1162,13 @@ struct mdt_body { __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; /* for 1.6 compatibility */ __u32 suppgid; __u32 eadatasize; __u32 aclsize; __u32 max_mdsize; - __u32 max_cookiesize; /* also fix lustre_swab_mdt_body */ + __u32 max_cookiesize; + __u32 padding_4; /* also fix lustre_swab_mdt_body */ }; struct mds_body { @@ -1073,7 +1197,7 @@ struct mds_body { __u32 eadatasize; __u32 aclsize; __u32 max_mdsize; - __u32 max_cookiesize; /* also fix lustre_swab_mds_body */ + __u32 max_cookiesize; __u32 padding_4; /* also fix lustre_swab_mds_body */ }; @@ -1121,10 +1245,13 @@ struct mds_remote_perm { __u32 rp_access_perm; /* MAY_READ/WRITE/EXEC */ }; -/* setxid permissions for mds_setxid_perm.mp_perm */ -#define LUSTRE_SETUID_PERM 0x01 -#define LUSTRE_SETGID_PERM 0x02 -#define LUSTRE_SETGRP_PERM 0x04 +/* permissions for md_perm.mp_perm */ +enum { + CFS_SETUID_PERM = 0x01, + CFS_SETGID_PERM = 0x02, + CFS_SETGRP_PERM = 0x04, + CFS_RMTACL_PERM = 0x08 +}; extern void lustre_swab_mds_remote_perm(struct mds_remote_perm *p); @@ -1146,7 +1273,7 @@ struct mds_rec_setattr { __u32 sa_suppgid; __u32 sa_mode; struct ll_fid sa_fid; - __u64 sa_valid; + __u64 sa_valid; /* MDS_ATTR_* attributes */ __u64 sa_size; __u64 sa_mtime; __u64 sa_atime; @@ -1165,26 +1292,48 @@ struct mdt_rec_setattr { __u32 sa_fsgid; __u32 sa_cap; __u32 sa_suppgid; - __u32 sa_mode; + __u32 sa_padding_1; struct lu_fid sa_fid; __u64 sa_valid; + __u32 sa_uid; + __u32 sa_gid; __u64 sa_size; __u64 sa_blocks; __u64 sa_mtime; __u64 sa_atime; __u64 sa_ctime; - __u32 sa_uid; - __u32 sa_gid; __u32 sa_attr_flags; - __u32 sa_padding; /* also fix lustre_swab_mds_rec_setattr */ + __u32 sa_mode; + __u32 sa_padding_2; + __u32 sa_padding_3; + __u32 sa_padding_4; + __u32 sa_padding_5; }; extern void lustre_swab_mdt_rec_setattr (struct mdt_rec_setattr *sa); -/* Remove this once we declare it in include/linux/fs.h (v21 kernel patch?) */ -#ifndef ATTR_CTIME_SET -#define ATTR_CTIME_SET 0x2000 -#endif +/* + * Attribute flags used in mds_rec_setattr::sa_valid. + * The kernel's #defines for ATTR_* should not be used over the network + * since the client and MDS may run different kernels (see bug 13828) + * Therefore, we should only use MDS_ATTR_* attributes for sa_valid. + */ +#define MDS_ATTR_MODE 0x1ULL /* = 1 */ +#define MDS_ATTR_UID 0x2ULL /* = 2 */ +#define MDS_ATTR_GID 0x4ULL /* = 4 */ +#define MDS_ATTR_SIZE 0x8ULL /* = 8 */ +#define MDS_ATTR_ATIME 0x10ULL /* = 16 */ +#define MDS_ATTR_MTIME 0x20ULL /* = 32 */ +#define MDS_ATTR_CTIME 0x40ULL /* = 64 */ +#define MDS_ATTR_ATIME_SET 0x80ULL /* = 128 */ +#define MDS_ATTR_MTIME_SET 0x100ULL /* = 256 */ +#define MDS_ATTR_FORCE 0x200ULL /* = 512, Not a change, but a change it */ +#define MDS_ATTR_ATTR_FLAG 0x400ULL /* = 1024 */ +#define MDS_ATTR_KILL_SUID 0x800ULL /* = 2048 */ +#define MDS_ATTR_KILL_SGID 0x1000ULL /* = 4096 */ +#define MDS_ATTR_CTIME_SET 0x2000ULL /* = 8192 */ +#define MDS_ATTR_FROM_OPEN 0x4000ULL /* = 16384, called from open path, ie O_TRUNC */ +#define MDS_ATTR_BLOCKS 0x8000ULL /* = 32768 */ #ifndef FMODE_READ #define FMODE_READ 00000001 @@ -1233,12 +1382,15 @@ extern void lustre_swab_mdt_rec_setattr (struct mdt_rec_setattr *sa); #define MAY_VTX_PART (1 << 12) /* full VTX permission check */ #define MAY_VTX_FULL (1 << 13) +/* lfs rgetfacl permission check */ +#define MAY_RGETFACL (1 << 14) enum { MDS_CHECK_SPLIT = 1 << 0, MDS_CROSS_REF = 1 << 1, MDS_VTX_BYPASS = 1 << 2, - MDS_PERM_BYPASS = 1 << 3 + MDS_PERM_BYPASS = 1 << 3, + MDS_SOM = 1 << 4 }; struct mds_rec_join { @@ -1281,18 +1433,21 @@ struct mdt_rec_create { __u32 cr_fsuid; __u32 cr_fsgid; __u32 cr_cap; - __u32 cr_flags; /* for use with open */ - __u32 cr_mode; - struct lustre_handle cr_old_handle; /* u64 handle in case of open replay */ + __u32 cr_suppgid1; + __u32 cr_suppgid2; struct lu_fid cr_fid1; struct lu_fid cr_fid2; + struct lustre_handle cr_old_handle; /* u64 handle in case of open replay */ __u64 cr_time; __u64 cr_rdev; __u64 cr_ioepoch; - __u32 cr_suppgid1; - __u32 cr_suppgid2; + __u64 cr_padding_1; /* pad for 64 bits*/ + __u32 cr_mode; __u32 cr_bias; - __u32 cr_padding_1; /* pad for 64 bits*/ + __u32 cr_flags; /* for use with open */ + __u32 cr_padding_2; + __u32 cr_padding_3; + __u32 cr_padding_4; }; extern void lustre_swab_mdt_rec_create (struct mdt_rec_create *cr); @@ -1325,14 +1480,18 @@ struct mdt_rec_link { 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; __u32 lk_bias; - __u32 lk_padding_2; /* also fix lustre_swab_mds_rec_link */ - __u32 lk_padding_3; /* also fix lustre_swab_mds_rec_link */ - __u32 lk_padding_4; /* also fix lustre_swab_mds_rec_link */ + __u32 lk_padding_5; + __u32 lk_padding_6; + __u32 lk_padding_7; + __u32 lk_padding_8; + __u32 lk_padding_9; }; -extern void lustre_swab_mdt_rec_link (struct mdt_rec_link *lk); - struct mds_rec_unlink { __u32 ul_opcode; __u32 ul_fsuid; @@ -1356,19 +1515,23 @@ struct mdt_rec_unlink { __u32 ul_fsuid; __u32 ul_fsgid; __u32 ul_cap; - __u32 ul_suppgid; - __u32 ul_mode; + __u32 ul_suppgid1; + __u32 ul_suppgid2; 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; __u32 ul_bias; - __u32 ul_padding_2; /* also fix lustre_swab_mds_rec_unlink */ - __u32 ul_padding_3; /* also fix lustre_swab_mds_rec_unlink */ - __u32 ul_padding_4; /* also fix lustre_swab_mds_rec_unlink */ + __u32 ul_mode; + __u32 ul_padding_6; + __u32 ul_padding_7; + __u32 ul_padding_8; + __u32 ul_padding_9; }; -extern void lustre_swab_mdt_rec_unlink (struct mdt_rec_unlink *ul); - struct mds_rec_rename { __u32 rn_opcode; __u32 rn_fsuid; @@ -1397,15 +1560,66 @@ struct mdt_rec_rename { struct lu_fid rn_fid1; struct lu_fid rn_fid2; __u64 rn_time; - __u32 rn_mode; /* cross-ref rename has mode */ + __u64 rn_padding_1; + __u64 rn_padding_2; + __u64 rn_padding_3; + __u64 rn_padding_4; __u32 rn_bias; /* some operation flags */ - __u32 rn_padding_3; /* also fix lustre_swab_mdt_rec_rename */ - __u32 rn_padding_4; /* also fix lustre_swab_mdt_rec_rename */ -}; - -extern void lustre_swab_mdt_rec_rename (struct mdt_rec_rename *rn); + __u32 rn_mode; /* cross-ref rename has mode */ + __u32 rn_padding_5; + __u32 rn_padding_6; + __u32 rn_padding_7; + __u32 rn_padding_8; +}; + +struct mdt_rec_setxattr { + __u32 sx_opcode; + __u32 sx_fsuid; + __u32 sx_fsgid; + __u32 sx_cap; + __u32 sx_suppgid1; + __u32 sx_suppgid2; + struct lu_fid sx_fid; + __u64 sx_padding_1; /* These three members are lu_fid size */ + __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; + __u32 sx_size; + __u32 sx_flags; + __u32 sx_padding_8; + __u32 sx_padding_9; + __u32 sx_padding_10; + __u32 sx_padding_11; +}; + +struct mdt_rec_reint { + __u32 rr_opcode; + __u32 rr_fsuid; + __u32 rr_fsgid; + __u32 rr_cap; + __u32 rr_suppgid1; + __u32 rr_suppgid2; + struct lu_fid rr_fid1; + struct lu_fid rr_fid2; + __u64 rr_mtime; + __u64 rr_atime; + __u64 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_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 */ +}; + +extern void lustre_swab_mdt_rec_reint(struct mdt_rec_reint *rr); -/* begin adding MDT by huanghua@clusterfs.com */ struct lmv_desc { __u32 ld_tgt_count; /* how many MDS's */ __u32 ld_active_tgt_count; /* how many active */ @@ -1413,7 +1627,6 @@ struct lmv_desc { }; extern void lustre_swab_lmv_desc (struct lmv_desc *ld); -/* end adding MDT by huanghua@clusterfs.com */ struct md_fld { seqno_t mf_seq; @@ -1445,6 +1658,7 @@ enum seq_op { #define LOV_MIN_STRIPE_SIZE 65536 /* maximum PAGE_SIZE (ia64), power of 2 */ #define LOV_MAX_STRIPE_COUNT 160 /* until bug 4424 is fixed */ +#define LOV_V1_INSANE_STRIPE_COUNT 65532 /* maximum stripe count bz13933 */ #define LOV_MAX_UUID_BUFFER_SIZE 8192 /* The size of the buffer the lov/mdc reserves for the @@ -1507,6 +1721,8 @@ typedef enum { LCK_MAXMODE } ldlm_mode_t; +#define LCK_MODE_NUM 7 + typedef enum { LDLM_PLAIN = 10, LDLM_EXTENT = 11, @@ -1523,6 +1739,12 @@ struct ldlm_extent { __u64 gid; }; +static inline int ldlm_extent_overlap(struct ldlm_extent *ex1, + struct ldlm_extent *ex2) +{ + return (ex1->start <= ex2->end) && (ex2->start <= ex1->end); +} + struct ldlm_inodebits { __u64 bits; }; @@ -1615,10 +1837,18 @@ typedef enum { MGS_EXCEPTION, /* node died, etc. */ MGS_TARGET_REG, /* whenever target starts up */ MGS_TARGET_DEL, + MGS_SET_INFO, MGS_LAST_OPC } mgs_cmd_t; #define MGS_FIRST_OPC MGS_CONNECT +#define MGS_PARAM_MAXLEN 1024 +#define KEY_SET_INFO "set_info" + +struct mgs_send_param { + char mgs_param[MGS_PARAM_MAXLEN]; +}; + /* We pass this info to the MGS so it can write config logs */ #define MTI_NAME_MAXLEN 64 #define MTI_PARAM_MAXLEN 4096