X-Git-Url: https://git.whamcloud.com/?a=blobdiff_plain;f=lustre%2Fllite%2Ffile.c;h=73c256ce94bd9b0d3bca6312592fcaeec8fb137b;hb=6bceb0030d15b70097d75f8da2d373310fe2c658;hp=aac8685f34bd2b159178333e3a263525daaa12aa;hpb=980f8741339f2bece106f6955b0b076d35211850;p=fs%2Flustre-release.git diff --git a/lustre/llite/file.c b/lustre/llite/file.c index aac8685f..73c256c 100644 --- a/lustre/llite/file.c +++ b/lustre/llite/file.c @@ -27,7 +27,6 @@ */ /* * This file is part of Lustre, http://www.lustre.org/ - * Lustre is a trademark of Sun Microsystems, Inc. * * lustre/llite/file.c * @@ -44,6 +43,7 @@ #include #include #include +#include #include #include @@ -104,19 +104,28 @@ static void ll_prepare_close(struct inode *inode, struct md_op_data *op_data, op_data->op_attr.ia_atime = inode->i_atime; op_data->op_attr.ia_mtime = inode->i_mtime; op_data->op_attr.ia_ctime = inode->i_ctime; - op_data->op_attr.ia_size = i_size_read(inode); + /* In case of encrypted file without the key, visible size was rounded + * up to next LUSTRE_ENCRYPTION_UNIT_SIZE, and clear text size was + * stored into lli_lazysize in ll_merge_attr(), so set proper file size + * now that we are closing. + */ + if (llcrypt_require_key(inode) == -ENOKEY && + ll_i2info(inode)->lli_attr_valid & OBD_MD_FLLAZYSIZE) + op_data->op_attr.ia_size = ll_i2info(inode)->lli_lazysize; + else + op_data->op_attr.ia_size = i_size_read(inode); op_data->op_attr.ia_valid |= (ATTR_MODE | ATTR_ATIME | ATTR_ATIME_SET | ATTR_MTIME | ATTR_MTIME_SET | ATTR_CTIME); op_data->op_xvalid |= OP_XVALID_CTIME_SET; op_data->op_attr_blocks = inode->i_blocks; op_data->op_attr_flags = ll_inode_to_ext_flags(inode->i_flags); - if (ll_file_test_flag(ll_i2info(inode), LLIF_PROJECT_INHERIT)) + if (test_bit(LLIF_PROJECT_INHERIT, &ll_i2info(inode)->lli_flags)) op_data->op_attr_flags |= LUSTRE_PROJINHERIT_FL; op_data->op_open_handle = och->och_open_handle; if (och->och_flags & FMODE_WRITE && - ll_file_test_and_clear_flag(ll_i2info(inode), LLIF_DATA_MODIFIED)) + test_and_clear_bit(LLIF_DATA_MODIFIED, &ll_i2info(inode)->lli_flags)) /* For HSM: if inode data has been modified, pack it so that * MDT can set data dirty flag in the archive. */ op_data->op_bias |= MDS_DATA_MODIFIED; @@ -162,7 +171,7 @@ static int ll_close_inode_openhandle(struct inode *inode, op_data->op_attr_blocks += ((struct inode *)data)->i_blocks; op_data->op_attr.ia_valid |= ATTR_SIZE; op_data->op_xvalid |= OP_XVALID_BLOCKS; - /* fallthrough */ + fallthrough; case MDS_CLOSE_LAYOUT_SPLIT: case MDS_CLOSE_LAYOUT_SWAP: { struct split_param *sp = data; @@ -317,27 +326,39 @@ static int ll_md_close(struct inode *inode, struct file *file) if (unlikely(fd->fd_flags & LL_FILE_GROUP_LOCKED)) ll_put_grouplock(inode, file, fd->fd_grouplock.lg_gid); + mutex_lock(&lli->lli_och_mutex); if (fd->fd_lease_och != NULL) { bool lease_broken; + struct obd_client_handle *lease_och; + + lease_och = fd->fd_lease_och; + fd->fd_lease_och = NULL; + mutex_unlock(&lli->lli_och_mutex); /* Usually the lease is not released when the * application crashed, we need to release here. */ - rc = ll_lease_close(fd->fd_lease_och, inode, &lease_broken); - CDEBUG(rc ? D_ERROR : D_INODE, "Clean up lease "DFID" %d/%d\n", - PFID(&lli->lli_fid), rc, lease_broken); + rc = ll_lease_close(lease_och, inode, &lease_broken); - fd->fd_lease_och = NULL; + mutex_lock(&lli->lli_och_mutex); + + CDEBUG_LIMIT(rc ? D_ERROR : D_INODE, + "Clean up lease "DFID" %d/%d\n", + PFID(&lli->lli_fid), rc, lease_broken); } if (fd->fd_och != NULL) { - rc = ll_close_inode_openhandle(inode, fd->fd_och, 0, NULL); + struct obd_client_handle *och; + + och = fd->fd_och; fd->fd_och = NULL; + mutex_unlock(&lli->lli_och_mutex); + + rc = ll_close_inode_openhandle(inode, och, 0, NULL); GOTO(out, rc); } /* Let's see if we have good enough OPEN lock on the file and if we can skip talking to MDS */ - mutex_lock(&lli->lli_och_mutex); if (fd->fd_omode & FMODE_WRITE) { lockmode = LCK_CW; LASSERT(lli->lli_open_fd_write_count); @@ -392,7 +413,7 @@ int ll_file_release(struct inode *inode, struct file *file) if (S_ISDIR(inode->i_mode) && lli->lli_opendir_key == fd) ll_deauthorize_statahead(inode, fd); - if (inode->i_sb->s_root == file_dentry(file)) { + if (is_root_inode(inode)) { file->private_data = NULL; ll_file_data_put(fd); GOTO(out, rc = 0); @@ -406,13 +427,15 @@ int ll_file_release(struct inode *inode, struct file *file) lli->lli_async_rc = 0; } + lli->lli_close_fd_time = ktime_get(); + rc = ll_md_close(inode, file); if (CFS_FAIL_TIMEOUT_MS(OBD_FAIL_PTLRPC_DUMP_LOG, cfs_fail_val)) libcfs_debug_dumplog(); out: - if (!rc && inode->i_sb->s_root != file_dentry(file)) + if (!rc && !is_root_inode(inode)) ll_stats_ops_tally(sbi, LPROC_LL_RELEASE, ktime_us_delta(ktime_get(), kstart)); RETURN(rc); @@ -420,8 +443,15 @@ out: static inline int ll_dom_readpage(void *data, struct page *page) { + /* since ll_dom_readpage is a page cache helper, it is safe to assume + * mapping and host pointers are set here + */ + struct inode *inode; struct niobuf_local *lnb = data; void *kaddr; + int rc = 0; + + inode = page2inode(page); kaddr = kmap_atomic(page); memcpy(kaddr, lnb->lnb_data, lnb->lnb_len); @@ -431,14 +461,40 @@ static inline int ll_dom_readpage(void *data, struct page *page) flush_dcache_page(page); SetPageUptodate(page); kunmap_atomic(kaddr); + + if (inode && IS_ENCRYPTED(inode) && S_ISREG(inode->i_mode)) { + if (!llcrypt_has_encryption_key(inode)) + CDEBUG(D_SEC, "no enc key for "DFID"\n", + PFID(ll_inode2fid(inode))); + else { + unsigned int offs = 0; + + while (offs < PAGE_SIZE) { + /* decrypt only if page is not empty */ + if (memcmp(page_address(page) + offs, + page_address(ZERO_PAGE(0)), + LUSTRE_ENCRYPTION_UNIT_SIZE) == 0) + break; + + rc = llcrypt_decrypt_pagecache_blocks(page, + LUSTRE_ENCRYPTION_UNIT_SIZE, + offs); + if (rc) + break; + + offs += LUSTRE_ENCRYPTION_UNIT_SIZE; + } + } + } unlock_page(page); - return 0; + return rc; } -void ll_dom_finish_open(struct inode *inode, struct ptlrpc_request *req, - struct lookup_intent *it) +void ll_dom_finish_open(struct inode *inode, struct ptlrpc_request *req) { + struct lu_env *env; + struct cl_io *io; struct ll_inode_info *lli = ll_i2info(inode); struct cl_object *obj = lli->lli_clob; struct address_space *mapping = inode->i_mapping; @@ -448,6 +504,8 @@ void ll_dom_finish_open(struct inode *inode, struct ptlrpc_request *req, char *data; unsigned long index, start; struct niobuf_local lnb; + __u16 refcheck; + int rc; ENTRY; @@ -474,13 +532,24 @@ void ll_dom_finish_open(struct inode *inode, struct ptlrpc_request *req, * buffer, in both cases total size should be equal to the file size. */ body = req_capsule_server_get(&req->rq_pill, &RMF_MDT_BODY); - if (rnb->rnb_offset + rnb->rnb_len != body->mbo_dom_size) { + if (rnb->rnb_offset + rnb->rnb_len != body->mbo_dom_size && + !(inode && IS_ENCRYPTED(inode))) { CERROR("%s: server returns off/len %llu/%u but size %llu\n", ll_i2sbi(inode)->ll_fsname, rnb->rnb_offset, rnb->rnb_len, body->mbo_dom_size); RETURN_EXIT; } + env = cl_env_get(&refcheck); + if (IS_ERR(env)) + RETURN_EXIT; + io = vvp_env_thread_io(env); + io->ci_obj = obj; + io->ci_ignore_layout = 1; + rc = cl_io_init(env, io, CIT_MISC, obj); + if (rc) + GOTO(out_io, rc); + CDEBUG(D_INFO, "Get data along with open at %llu len %i, size %llu\n", rnb->rnb_offset, rnb->rnb_len, body->mbo_dom_size); @@ -492,6 +561,8 @@ void ll_dom_finish_open(struct inode *inode, struct ptlrpc_request *req, LASSERT((lnb.lnb_file_offset & ~PAGE_MASK) == 0); lnb.lnb_page_offset = 0; do { + struct cl_page *page; + lnb.lnb_data = data + (index << PAGE_SHIFT); lnb.lnb_len = rnb->rnb_len - (index << PAGE_SHIFT); if (lnb.lnb_len > PAGE_SIZE) @@ -507,9 +578,33 @@ void ll_dom_finish_open(struct inode *inode, struct ptlrpc_request *req, PTR_ERR(vmpage)); break; } + lock_page(vmpage); + if (vmpage->mapping == NULL) { + unlock_page(vmpage); + put_page(vmpage); + /* page was truncated */ + break; + } + /* attach VM page to CL page cache */ + page = cl_page_find(env, obj, vmpage->index, vmpage, + CPT_CACHEABLE); + if (IS_ERR(page)) { + ClearPageUptodate(vmpage); + unlock_page(vmpage); + put_page(vmpage); + break; + } + SetPageUptodate(vmpage); + cl_page_put(env, page); + unlock_page(vmpage); put_page(vmpage); index++; } while (rnb->rnb_len > (index << PAGE_SHIFT)); + +out_io: + cl_io_fini(env, io); + cl_env_put(env, &refcheck); + EXIT; } @@ -556,7 +651,7 @@ retry: } op_data = ll_prep_md_op_data(NULL, parent->d_inode, de->d_inode, - name, len, 0, LUSTRE_OPC_ANY, NULL); + name, len, 0, LUSTRE_OPC_OPEN, NULL); if (IS_ERR(op_data)) { kfree(name); RETURN(PTR_ERR(op_data)); @@ -564,6 +659,8 @@ retry: op_data->op_data = lmm; op_data->op_data_size = lmmsize; + OBD_FAIL_TIMEOUT(OBD_FAIL_LLITE_OPEN_DELAY, cfs_fail_val); + rc = md_intent_lock(sbi->ll_md_exp, op_data, itp, &req, &ll_md_blocking_ast, 0); kfree(name); @@ -588,44 +685,47 @@ retry: GOTO(out, rc); } - rc = ll_prep_inode(&de->d_inode, req, NULL, itp); + rc = ll_prep_inode(&de->d_inode, &req->rq_pill, NULL, itp); if (!rc && itp->it_lock_mode) { - struct lustre_handle handle = {.cookie = itp->it_lock_handle}; - struct ldlm_lock *lock; - bool has_dom_bit = false; + __u64 bits = 0; /* If we got a lock back and it has a LOOKUP bit set, * make sure the dentry is marked as valid so we can find it. * We don't need to care about actual hashing since other bits * of kernel will deal with that later. */ - lock = ldlm_handle2lock(&handle); - if (lock) { - has_dom_bit = ldlm_has_dom(lock); - if (lock->l_policy_data.l_inodebits.bits & - MDS_INODELOCK_LOOKUP) - d_lustre_revalidate(de); - - LDLM_LOCK_PUT(lock); + ll_set_lock_data(sbi->ll_md_exp, de->d_inode, itp, &bits); + if (bits & MDS_INODELOCK_LOOKUP) { + d_lustre_revalidate(de); + ll_update_dir_depth(parent->d_inode, de->d_inode); } - ll_set_lock_data(sbi->ll_md_exp, de->d_inode, itp, NULL); - if (has_dom_bit) - ll_dom_finish_open(de->d_inode, req, itp); + + /* if DoM bit returned along with LAYOUT bit then there + * can be read-on-open data returned. + */ + if (bits & MDS_INODELOCK_DOM && bits & MDS_INODELOCK_LAYOUT) + ll_dom_finish_open(de->d_inode, req); } out: ptlrpc_req_finished(req); ll_intent_drop_lock(itp); - /* We did open by fid, but by the time we got to the server, - * the object disappeared. If this is a create, we cannot really - * tell the userspace that the file it was trying to create - * does not exist. Instead let's return -ESTALE, and the VFS will - * retry the create with LOOKUP_REVAL that we are going to catch - * in ll_revalidate_dentry() and use lookup then. + /* We did open by fid, but by the time we got to the server, the object + * disappeared. This is possible if the object was unlinked, but it's + * also possible if the object was unlinked by a rename. In the case + * of an object renamed over our existing one, we can't fail this open. + * O_CREAT also goes through this path if we had an existing dentry, + * and it's obviously wrong to return ENOENT for O_CREAT. + * + * Instead let's return -ESTALE, and the VFS will retry the open with + * LOOKUP_REVAL, which we catch in ll_revalidate_dentry and fail to + * revalidate, causing a lookup. This causes extra lookups in the case + * where we had a dentry in cache but the file is being unlinked and we + * lose the race with unlink, but this should be very rare. */ - if (rc == -ENOENT && itp->it_op & IT_CREAT) + if (rc == -ENOENT) rc = -ESTALE; RETURN(rc); @@ -668,13 +768,32 @@ static int ll_local_open(struct file *file, struct lookup_intent *it, ll_readahead_init(inode, &fd->fd_ras); fd->fd_omode = it->it_flags & (FMODE_READ | FMODE_WRITE | FMODE_EXEC); - /* ll_cl_context initialize */ - rwlock_init(&fd->fd_lock); - INIT_LIST_HEAD(&fd->fd_lccs); - RETURN(0); } +void ll_track_file_opens(struct inode *inode) +{ + struct ll_inode_info *lli = ll_i2info(inode); + struct ll_sb_info *sbi = ll_i2sbi(inode); + + /* do not skew results with delays from never-opened inodes */ + if (ktime_to_ns(lli->lli_close_fd_time)) + ll_stats_ops_tally(sbi, LPROC_LL_INODE_OPCLTM, + ktime_us_delta(ktime_get(), lli->lli_close_fd_time)); + + if (ktime_after(ktime_get(), + ktime_add_ms(lli->lli_close_fd_time, + sbi->ll_oc_max_ms))) { + lli->lli_open_fd_count = 1; + lli->lli_close_fd_time = ns_to_ktime(0); + } else { + lli->lli_open_fd_count++; + } + + ll_stats_ops_tally(ll_i2sbi(inode), LPROC_LL_INODE_OCOUNT, + lli->lli_open_fd_count); +} + /* Open a file, and (for the very first open) create objects on the OSTs at * this time. If opened with O_LOV_DELAY_CREATE, then we don't do the object * creation or open until ll_lov_setstripe() ioctl is called. @@ -706,6 +825,15 @@ int ll_file_open(struct inode *inode, struct file *file) it = file->private_data; /* XXX: compat macro */ file->private_data = NULL; /* prevent ll_local_open assertion */ + if (S_ISREG(inode->i_mode)) { + rc = ll_file_open_encrypt(inode, file); + if (rc) { + if (it && it->it_disposition) + ll_release_openhandle(file_dentry(file), it); + GOTO(out_nofiledata, rc); + } + } + fd = ll_file_data_get(); if (fd == NULL) GOTO(out_nofiledata, rc = -ENOMEM); @@ -714,7 +842,8 @@ int ll_file_open(struct inode *inode, struct file *file) if (S_ISDIR(inode->i_mode)) ll_authorize_statahead(inode, fd); - if (inode->i_sb->s_root == file_dentry(file)) { + ll_track_file_opens(inode); + if (is_root_inode(inode)) { file->private_data = fd; RETURN(0); } @@ -787,6 +916,7 @@ restart: LASSERT(*och_usecount == 0); if (!it->it_disposition) { struct dentry *dentry = file_dentry(file); + struct ll_sb_info *sbi = ll_i2sbi(inode); struct ll_dentry_data *ldd; /* We cannot just request lock handle now, new ELC code @@ -803,20 +933,43 @@ restart: * handle to be returned from LOOKUP|OPEN request, * for example if the target entry was a symlink. * - * Only fetch MDS_OPEN_LOCK if this is in NFS path, - * marked by a bit set in ll_iget_for_nfs. Clear the - * bit so that it's not confusing later callers. + * In NFS path we know there's pathologic behavior + * so we always enable open lock caching when coming + * from there. It's detected by setting a flag in + * ll_iget_for_nfs. * - * NB; when ldd is NULL, it must have come via normal - * lookup path only, since ll_iget_for_nfs always calls - * ll_d_init(). + * After reaching number of opens of this inode + * we always ask for an open lock on it to handle + * bad userspace actors that open and close files + * in a loop for absolutely no good reason */ + ldd = ll_d2d(dentry); - if (ldd && ldd->lld_nfs_dentry) { + if (filename_is_volatile(dentry->d_name.name, + dentry->d_name.len, + NULL)) { + /* There really is nothing here, but this + * make this more readable I think. + * We do not want openlock for volatile + * files under any circumstances + */ + } else if (ldd && ldd->lld_nfs_dentry) { + /* NFS path. This also happens to catch + * open by fh files I guess + */ + it->it_flags |= MDS_OPEN_LOCK; + /* clear the flag for future lookups */ ldd->lld_nfs_dentry = 0; - if (!filename_is_volatile(dentry->d_name.name, - dentry->d_name.len, - NULL)) + } else if (sbi->ll_oc_thrsh_count > 0) { + /* Take MDS_OPEN_LOCK with many opens */ + if (lli->lli_open_fd_count >= + sbi->ll_oc_thrsh_count) + it->it_flags |= MDS_OPEN_LOCK; + + /* If this is open after we just closed */ + else if (ktime_before(ktime_get(), + ktime_add_ms(lli->lli_close_fd_time, + sbi->ll_oc_thrsh_ms))) it->it_flags |= MDS_OPEN_LOCK; } @@ -861,9 +1014,6 @@ restart: mutex_unlock(&lli->lli_och_mutex); - /* lockless for direct IO so that it can do IO in parallel */ - if (file->f_flags & O_DIRECT) - fd->fd_flags |= LL_FILE_LOCKLESS_IO; fd = NULL; /* Must do this outside lli_och_mutex lock to prevent deadlock where @@ -1089,8 +1239,8 @@ ll_lease_open(struct inode *inode, struct file *file, fmode_t fmode, /* already get lease, handle lease lock */ ll_set_lock_data(sbi->ll_md_exp, inode, &it, NULL); - if (it.it_lock_mode == 0 || - it.it_lock_bits != MDS_INODELOCK_OPEN) { + if (!it.it_lock_mode || + !(it.it_lock_bits & MDS_INODELOCK_OPEN)) { /* open lock must return for lease */ CERROR(DFID "lease granted but no open lock, %d/%llu.\n", PFID(ll_inode2fid(inode)), it.it_lock_mode, @@ -1137,8 +1287,8 @@ static int ll_check_swap_layouts_validity(struct inode *inode1, if (!S_ISREG(inode1->i_mode) || !S_ISREG(inode2->i_mode)) return -EINVAL; - if (inode_permission(inode1, MAY_WRITE) || - inode_permission(inode2, MAY_WRITE)) + if (inode_permission(&init_user_ns, inode1, MAY_WRITE) || + inode_permission(&init_user_ns, inode2, MAY_WRITE)) return -EPERM; if (inode1->i_sb != inode2->i_sb) @@ -1297,7 +1447,7 @@ int ll_merge_attr(const struct lu_env *env, struct inode *inode) * POSIX. Solving this problem needs to send an RPC to MDT for each * read, this will hurt performance. */ - if (ll_file_test_and_clear_flag(lli, LLIF_UPDATE_ATIME) || + if (test_and_clear_bit(LLIF_UPDATE_ATIME, &lli->lli_flags) || inode->i_atime.tv_sec < lli->lli_atime) inode->i_atime.tv_sec = lli->lli_atime; @@ -1330,6 +1480,16 @@ int ll_merge_attr(const struct lu_env *env, struct inode *inode) CDEBUG(D_VFSTRACE, DFID" updating i_size %llu\n", PFID(&lli->lli_fid), attr->cat_size); + if (llcrypt_require_key(inode) == -ENOKEY) { + /* Without the key, round up encrypted file size to next + * LUSTRE_ENCRYPTION_UNIT_SIZE. Clear text size is put in + * lli_lazysize for proper file size setting at close time. + */ + lli->lli_attr_valid |= OBD_MD_FLLAZYSIZE; + lli->lli_lazysize = attr->cat_size; + attr->cat_size = round_up(attr->cat_size, + LUSTRE_ENCRYPTION_UNIT_SIZE); + } i_size_write(inode, attr->cat_size); inode->i_blocks = attr->cat_blocks; @@ -1412,11 +1572,17 @@ void ll_io_init(struct cl_io *io, struct file *file, enum cl_io_type iot, IS_SYNC(inode)); #ifdef HAVE_GENERIC_WRITE_SYNC_2ARGS io->u.ci_wr.wr_sync |= !!(args && - args->via_io_subtype == IO_NORMAL && - args->u.normal.via_iocb->ki_flags & IOCB_DSYNC); + (args->u.normal.via_iocb->ki_flags & + IOCB_DSYNC)); #endif } +#ifdef IOCB_NOWAIT + io->ci_iocb_nowait = !!(args && + (args->u.normal.via_iocb->ki_flags & + IOCB_NOWAIT)); +#endif + io->ci_obj = ll_i2info(inode)->lli_clob; io->ci_lockreq = CILR_MAYBE; if (ll_file_nolock(file)) { @@ -1471,16 +1637,24 @@ ll_file_io_generic(const struct lu_env *env, struct vvp_io_args *args, struct file *file, enum cl_io_type iot, loff_t *ppos, size_t count) { - struct vvp_io *vio = vvp_env_io(env); - struct inode *inode = file_inode(file); - struct ll_inode_info *lli = ll_i2info(inode); - struct ll_file_data *fd = file->private_data; - struct range_lock range; - struct cl_io *io; - ssize_t result = 0; - int rc = 0; - unsigned retried = 0; - unsigned ignore_lockless = 0; + struct vvp_io *vio = vvp_env_io(env); + struct inode *inode = file_inode(file); + struct ll_inode_info *lli = ll_i2info(inode); + struct ll_sb_info *sbi = ll_i2sbi(inode); + struct ll_file_data *fd = file->private_data; + struct range_lock range; + bool range_locked = false; + struct cl_io *io; + ssize_t result = 0; + int rc = 0; + int rc2 = 0; + unsigned int retried = 0, dio_lock = 0; + bool is_aio = false; + bool is_parallel_dio = false; + struct cl_dio_aio *ci_aio = NULL; + size_t per_bytes; + bool partial_io = false; + size_t max_io_pages, max_cached_pages; ENTRY; @@ -1488,73 +1662,146 @@ ll_file_io_generic(const struct lu_env *env, struct vvp_io_args *args, file_dentry(file)->d_name.name, iot == CIT_READ ? "read" : "write", *ppos, count); + max_io_pages = PTLRPC_MAX_BRW_PAGES * OBD_MAX_RIF_DEFAULT; + max_cached_pages = sbi->ll_cache->ccc_lru_max; + if (max_io_pages > (max_cached_pages >> 2)) + max_io_pages = max_cached_pages >> 2; + + io = vvp_env_thread_io(env); + if (file->f_flags & O_DIRECT) { + if (file->f_flags & O_APPEND) + dio_lock = 1; + if (!is_sync_kiocb(args->u.normal.via_iocb)) + is_aio = true; + + /* the kernel does not support AIO on pipes, and parallel DIO + * uses part of the AIO path, so we must not do parallel dio + * to pipes + */ + is_parallel_dio = !iov_iter_is_pipe(args->u.normal.via_iter) && + !is_aio; + + if (!ll_sbi_has_parallel_dio(sbi)) + is_parallel_dio = false; + + ci_aio = cl_aio_alloc(args->u.normal.via_iocb, + ll_i2info(inode)->lli_clob, NULL); + if (!ci_aio) + GOTO(out, rc = -ENOMEM); + } + restart: + /** + * IO block size need be aware of cached page limit, otherwise + * if we have small max_cached_mb but large block IO issued, io + * could not be finished and blocked whole client. + */ + if (file->f_flags & O_DIRECT) + per_bytes = count; + else + per_bytes = min(max_io_pages << PAGE_SHIFT, count); + partial_io = per_bytes < count; io = vvp_env_thread_io(env); ll_io_init(io, file, iot, args); - io->ci_ignore_lockless = ignore_lockless; + io->ci_aio = ci_aio; + io->ci_dio_lock = dio_lock; io->ci_ndelay_tried = retried; + io->ci_parallel_dio = is_parallel_dio; - if (cl_io_rw_init(env, io, iot, *ppos, count) == 0) { - bool range_locked = false; - + if (cl_io_rw_init(env, io, iot, *ppos, per_bytes) == 0) { if (file->f_flags & O_APPEND) range_lock_init(&range, 0, LUSTRE_EOF); else - range_lock_init(&range, *ppos, *ppos + count - 1); + range_lock_init(&range, *ppos, *ppos + per_bytes - 1); vio->vui_fd = file->private_data; - vio->vui_io_subtype = args->via_io_subtype; - - switch (vio->vui_io_subtype) { - case IO_NORMAL: - vio->vui_iter = args->u.normal.via_iter; - vio->vui_iocb = args->u.normal.via_iocb; - /* Direct IO reads must also take range lock, - * or multiple reads will try to work on the same pages - * See LU-6227 for details. */ - if (((iot == CIT_WRITE) || - (iot == CIT_READ && (file->f_flags & O_DIRECT))) && - !(vio->vui_fd->fd_flags & LL_FILE_GROUP_LOCKED)) { - CDEBUG(D_VFSTRACE, "Range lock "RL_FMT"\n", - RL_PARA(&range)); - rc = range_lock(&lli->lli_write_tree, &range); - if (rc < 0) - GOTO(out, rc); + vio->vui_iter = args->u.normal.via_iter; + vio->vui_iocb = args->u.normal.via_iocb; + /* Direct IO reads must also take range lock, + * or multiple reads will try to work on the same pages + * See LU-6227 for details. + */ + if (((iot == CIT_WRITE) || + (iot == CIT_READ && (file->f_flags & O_DIRECT))) && + !(vio->vui_fd->fd_flags & LL_FILE_GROUP_LOCKED)) { + CDEBUG(D_VFSTRACE, "Range lock "RL_FMT"\n", + RL_PARA(&range)); + rc = range_lock(&lli->lli_write_tree, &range); + if (rc < 0) + GOTO(out, rc); - range_locked = true; - } - break; - case IO_SPLICE: - vio->u.splice.vui_pipe = args->u.splice.via_pipe; - vio->u.splice.vui_flags = args->u.splice.via_flags; - break; - default: - CERROR("unknown IO subtype %u\n", vio->vui_io_subtype); - LBUG(); + range_locked = true; } - ll_cl_add(file, env, io, LCC_RW); + ll_cl_add(inode, env, io, LCC_RW); rc = cl_io_loop(env, io); - ll_cl_remove(file, env); + ll_cl_remove(inode, env); - if (range_locked) { + if (range_locked && !is_parallel_dio) { CDEBUG(D_VFSTRACE, "Range unlock "RL_FMT"\n", RL_PARA(&range)); range_unlock(&lli->lli_write_tree, &range); + range_locked = false; } } else { /* cl_io_rw_init() handled IO */ rc = io->ci_result; } + /* N/B: parallel DIO may be disabled during i/o submission; + * if that occurs, async RPCs are resolved before we get here, and this + * wait call completes immediately. + */ + if (is_parallel_dio) { + struct cl_sync_io *anchor = &io->ci_aio->cda_sync; + + /* for dio, EIOCBQUEUED is an implementation detail, + * and we don't return it to userspace + */ + if (rc == -EIOCBQUEUED) + rc = 0; + + rc2 = cl_sync_io_wait_recycle(env, anchor, 0, 0); + if (rc2 < 0) + rc = rc2; + + if (range_locked) { + range_unlock(&lli->lli_write_tree, &range); + range_locked = false; + } + } + + /* + * In order to move forward AIO, ci_nob was increased, + * but that doesn't mean io have been finished, it just + * means io have been submited, we will always return + * EIOCBQUEUED to the caller, So we could only return + * number of bytes in non-AIO case. + */ if (io->ci_nob > 0) { - result += io->ci_nob; - count -= io->ci_nob; - *ppos = io->u.ci_wr.wr.crw_pos; /* for splice */ + if (!is_aio) { + if (rc2 == 0) { + result += io->ci_nob; + *ppos = io->u.ci_wr.wr.crw_pos; /* for splice */ + } else if (rc2) { + result = 0; + } + } + count -= io->ci_nob; /* prepare IO restart */ - if (count > 0 && args->via_io_subtype == IO_NORMAL) + if (count > 0) args->u.normal.via_iter = vio->vui_iter; + + if (partial_io) { + /** + * Reexpand iov count because it was zero + * after IO finish. + */ + iov_iter_reexpand(vio->vui_iter, count); + if (per_bytes == io->ci_nob) + io->ci_need_restart = 1; + } } out: cl_io_fini(env, io); @@ -1573,10 +1820,31 @@ out: *ppos, count, result, rc); /* preserve the tried count for FLR */ retried = io->ci_ndelay_tried; - ignore_lockless = io->ci_ignore_lockless; + dio_lock = io->ci_dio_lock; goto restart; } + if (io->ci_aio) { + /* + * VFS will call aio_complete() if no -EIOCBQUEUED + * is returned for AIO, so we can not call aio_complete() + * in our end_io(). + */ + if (rc != -EIOCBQUEUED) + io->ci_aio->cda_no_aio_complete = 1; + /** + * Drop one extra reference so that end_io() could be + * called for this IO context, we could call it after + * we make sure all AIO requests have been proceed. + */ + cl_sync_io_note(env, &io->ci_aio->cda_sync, + rc == -EIOCBQUEUED ? 0 : rc); + if (!is_aio) { + cl_aio_free(env, io->ci_aio); + io->ci_aio = NULL; + } + } + if (iot == CIT_READ) { if (result > 0) ll_stats_ops_tally(ll_i2sbi(inode), @@ -1683,8 +1951,15 @@ static ssize_t ll_file_read_iter(struct kiocb *iocb, struct iov_iter *to) ktime_t kstart = ktime_get(); bool cached; + ENTRY; + + CDEBUG(D_VFSTRACE|D_IOTRACE, "file %s:"DFID", ppos: %lld, count: %zu\n", + file_dentry(file)->d_name.name, + PFID(ll_inode2fid(file_inode(file))), iocb->ki_pos, + iov_iter_count(to)); + if (!iov_iter_count(to)) - return 0; + RETURN(0); /** * Currently when PCC read failed, we do not fall back to the @@ -1709,9 +1984,9 @@ static ssize_t ll_file_read_iter(struct kiocb *iocb, struct iov_iter *to) env = cl_env_get(&refcheck); if (IS_ERR(env)) - return PTR_ERR(env); + RETURN(PTR_ERR(env)); - args = ll_env_args(env, IO_NORMAL); + args = ll_env_args(env); args->u.normal.via_iter = to; args->u.normal.via_iocb = iocb; @@ -1732,7 +2007,13 @@ out: ktime_us_delta(ktime_get(), kstart)); } - return result; + CDEBUG(D_IOTRACE, + "COMPLETED: file %s:"DFID", ppos: %lld, count: %zu\n", + file_dentry(file)->d_name.name, + PFID(ll_inode2fid(file_inode(file))), iocb->ki_pos, + iov_iter_count(to)); + + RETURN(result); } /** @@ -1784,9 +2065,7 @@ static ssize_t ll_do_tiny_write(struct kiocb *iocb, struct iov_iter *iter) if (result > 0) { ll_heat_add(inode, CIT_WRITE, result); - ll_stats_ops_tally(ll_i2sbi(inode), LPROC_LL_WRITE_BYTES, - result); - ll_file_set_flag(ll_i2info(inode), LLIF_DATA_MODIFIED); + set_bit(LLIF_DATA_MODIFIED, &ll_i2info(inode)->lli_flags); } CDEBUG(D_VFSTRACE, "result: %zu, original count %zu\n", result, count); @@ -1810,6 +2089,11 @@ static ssize_t ll_file_write_iter(struct kiocb *iocb, struct iov_iter *from) ENTRY; + CDEBUG(D_VFSTRACE|D_IOTRACE, "file %s:"DFID", ppos: %lld, count: %zu\n", + file_dentry(file)->d_name.name, + PFID(ll_inode2fid(file_inode(file))), iocb->ki_pos, + iov_iter_count(from)); + if (!iov_iter_count(from)) GOTO(out, rc_normal = 0); @@ -1844,9 +2128,9 @@ static ssize_t ll_file_write_iter(struct kiocb *iocb, struct iov_iter *from) env = cl_env_get(&refcheck); if (IS_ERR(env)) - return PTR_ERR(env); + RETURN(PTR_ERR(env)); - args = ll_env_args(env, IO_NORMAL); + args = ll_env_args(env); args->u.normal.via_iter = from; args->u.normal.via_iocb = iocb; @@ -1872,6 +2156,12 @@ out: ktime_us_delta(ktime_get(), kstart)); } + CDEBUG(D_IOTRACE, + "COMPLETED: file %s:"DFID", ppos: %lld, count: %zu\n", + file_dentry(file)->d_name.name, + PFID(ll_inode2fid(file_inode(file))), iocb->ki_pos, + iov_iter_count(from)); + RETURN(rc_normal); } @@ -2018,46 +2308,6 @@ static ssize_t ll_file_write(struct file *file, const char __user *buf, } #endif /* !HAVE_FILE_OPERATIONS_READ_WRITE_ITER */ -/* - * Send file content (through pagecache) somewhere with helper - */ -static ssize_t ll_file_splice_read(struct file *in_file, loff_t *ppos, - struct pipe_inode_info *pipe, size_t count, - unsigned int flags) -{ - struct lu_env *env; - struct vvp_io_args *args; - ssize_t result; - __u16 refcheck; - bool cached; - - ENTRY; - - result = pcc_file_splice_read(in_file, ppos, pipe, - count, flags, &cached); - if (cached) - RETURN(result); - - ll_ras_enter(in_file, *ppos, count); - - env = cl_env_get(&refcheck); - if (IS_ERR(env)) - RETURN(PTR_ERR(env)); - - args = ll_env_args(env, IO_SPLICE); - args->u.splice.via_pipe = pipe; - args->u.splice.via_flags = flags; - - result = ll_file_io_generic(env, args, in_file, CIT_READ, ppos, count); - cl_env_put(env, &refcheck); - - if (result > 0) - ll_rw_stats_tally(ll_i2sbi(file_inode(in_file)), current->pid, - in_file->private_data, *ppos, result, - READ); - RETURN(result); -} - int ll_lov_setstripe_ea_info(struct inode *inode, struct dentry *dentry, __u64 flags, struct lov_user_md *lum, int lum_size) { @@ -2092,44 +2342,45 @@ int ll_lov_getstripe_ea_info(struct inode *inode, const char *filename, struct lov_mds_md **lmmp, int *lmm_size, struct ptlrpc_request **request) { - struct ll_sb_info *sbi = ll_i2sbi(inode); - struct mdt_body *body; - struct lov_mds_md *lmm = NULL; - struct ptlrpc_request *req = NULL; - struct md_op_data *op_data; - int rc, lmmsize; + struct ll_sb_info *sbi = ll_i2sbi(inode); + struct mdt_body *body; + struct lov_mds_md *lmm = NULL; + struct ptlrpc_request *req = NULL; + struct md_op_data *op_data; + int rc, lmmsize; + + ENTRY; rc = ll_get_default_mdsize(sbi, &lmmsize); if (rc) RETURN(rc); - op_data = ll_prep_md_op_data(NULL, inode, NULL, filename, - strlen(filename), lmmsize, - LUSTRE_OPC_ANY, NULL); - if (IS_ERR(op_data)) - RETURN(PTR_ERR(op_data)); + op_data = ll_prep_md_op_data(NULL, inode, NULL, filename, + strlen(filename), lmmsize, + LUSTRE_OPC_ANY, NULL); + if (IS_ERR(op_data)) + RETURN(PTR_ERR(op_data)); - op_data->op_valid = OBD_MD_FLEASIZE | OBD_MD_FLDIREA; - rc = md_getattr_name(sbi->ll_md_exp, op_data, &req); - ll_finish_md_op_data(op_data); - if (rc < 0) { - CDEBUG(D_INFO, "md_getattr_name failed " - "on %s: rc %d\n", filename, rc); - GOTO(out, rc); - } + op_data->op_valid = OBD_MD_FLEASIZE | OBD_MD_FLDIREA; + rc = md_getattr_name(sbi->ll_md_exp, op_data, &req); + ll_finish_md_op_data(op_data); + if (rc < 0) { + CDEBUG(D_INFO, "md_getattr_name failed on %s: rc %d\n", + filename, rc); + GOTO(out, rc); + } - body = req_capsule_server_get(&req->rq_pill, &RMF_MDT_BODY); - LASSERT(body != NULL); /* checked by mdc_getattr_name */ + body = req_capsule_server_get(&req->rq_pill, &RMF_MDT_BODY); + LASSERT(body != NULL); /* checked by mdc_getattr_name */ lmmsize = body->mbo_eadatasize; if (!(body->mbo_valid & (OBD_MD_FLEASIZE | OBD_MD_FLDIREA)) || - lmmsize == 0) { - GOTO(out, rc = -ENODATA); - } + lmmsize == 0) + GOTO(out, rc = -ENODATA); - lmm = req_capsule_server_sized_get(&req->rq_pill, &RMF_MDT_MD, lmmsize); - LASSERT(lmm != NULL); + lmm = req_capsule_server_sized_get(&req->rq_pill, &RMF_MDT_MD, lmmsize); + LASSERT(lmm != NULL); if (lmm->lmm_magic != cpu_to_le32(LOV_MAGIC_V1) && lmm->lmm_magic != cpu_to_le32(LOV_MAGIC_V3) && @@ -2139,11 +2390,10 @@ int ll_lov_getstripe_ea_info(struct inode *inode, const char *filename, /* * This is coming from the MDS, so is probably in - * little endian. We convert it to host endian before + * little endian. We convert it to host endian before * passing it to userspace. */ - if ((lmm->lmm_magic & __swab32(LOV_MAGIC_MAGIC)) == - __swab32(LOV_MAGIC_MAGIC)) { + if (cpu_to_le32(LOV_MAGIC) != LOV_MAGIC) { int stripe_count = 0; if (lmm->lmm_magic == cpu_to_le32(LOV_MAGIC_V1) || @@ -2152,28 +2402,74 @@ int ll_lov_getstripe_ea_info(struct inode *inode, const char *filename, if (le32_to_cpu(lmm->lmm_pattern) & LOV_PATTERN_F_RELEASED) stripe_count = 0; - } + lustre_swab_lov_user_md((struct lov_user_md *)lmm, 0); - lustre_swab_lov_user_md((struct lov_user_md *)lmm, 0); - - /* if function called for directory - we should - * avoid swab not existent lsm objects */ - if (lmm->lmm_magic == LOV_MAGIC_V1 && S_ISREG(body->mbo_mode)) - lustre_swab_lov_user_md_objects( + /* if function called for directory - we should + * avoid swab not existent lsm objects + */ + if (lmm->lmm_magic == LOV_MAGIC_V1 && + S_ISREG(body->mbo_mode)) + lustre_swab_lov_user_md_objects( ((struct lov_user_md_v1 *)lmm)->lmm_objects, stripe_count); - else if (lmm->lmm_magic == LOV_MAGIC_V3 && - S_ISREG(body->mbo_mode)) - lustre_swab_lov_user_md_objects( + else if (lmm->lmm_magic == LOV_MAGIC_V3 && + S_ISREG(body->mbo_mode)) + lustre_swab_lov_user_md_objects( ((struct lov_user_md_v3 *)lmm)->lmm_objects, stripe_count); + } else if (lmm->lmm_magic == cpu_to_le32(LOV_MAGIC_COMP_V1)) { + lustre_swab_lov_comp_md_v1( + (struct lov_comp_md_v1 *)lmm); + } } + if (lmm->lmm_magic == LOV_MAGIC_COMP_V1) { + struct lov_comp_md_v1 *comp_v1 = NULL; + struct lov_comp_md_entry_v1 *ent; + struct lov_user_md_v1 *v1; + __u32 off; + int i = 0; + + comp_v1 = (struct lov_comp_md_v1 *)lmm; + /* Dump the striping information */ + for (; i < comp_v1->lcm_entry_count; i++) { + ent = &comp_v1->lcm_entries[i]; + off = ent->lcme_offset; + v1 = (struct lov_user_md_v1 *)((char *)lmm + off); + CDEBUG(D_INFO, + "comp[%d]: stripe_count=%u, stripe_size=%u\n", + i, v1->lmm_stripe_count, v1->lmm_stripe_size); + } + + /** + * Return valid stripe_count and stripe_size instead of 0 for + * DoM files to avoid divide-by-zero for older userspace that + * calls this ioctl, e.g. lustre ADIO driver. + */ + if (lmm->lmm_stripe_count == 0) + lmm->lmm_stripe_count = 1; + if (lmm->lmm_stripe_size == 0) { + /* Since the first component of the file data is placed + * on the MDT for faster access, the stripe_size of the + * second one is always that applications which are + * doing large IOs. + */ + if (lmm->lmm_pattern == LOV_PATTERN_MDT) + i = comp_v1->lcm_entry_count > 1 ? 1 : 0; + else + i = comp_v1->lcm_entry_count > 1 ? + comp_v1->lcm_entry_count - 1 : 0; + ent = &comp_v1->lcm_entries[i]; + off = ent->lcme_offset; + v1 = (struct lov_user_md_v1 *)((char *)lmm + off); + lmm->lmm_stripe_size = v1->lmm_stripe_size; + } + } out: *lmmp = lmm; *lmm_size = lmmsize; *request = req; - return rc; + RETURN(rc); } static int ll_lov_setea(struct inode *inode, struct file *file, @@ -2186,7 +2482,7 @@ static int ll_lov_setea(struct inode *inode, struct file *file, int rc; ENTRY; - if (!cfs_capable(CFS_CAP_SYS_ADMIN)) + if (!capable(CAP_SYS_ADMIN)) RETURN(-EPERM); OBD_ALLOC_LARGE(lump, lum_size); @@ -2249,6 +2545,13 @@ static int ll_lov_setstripe(struct inode *inode, struct file *file, GOTO(out, rc); rc = ll_file_getstripe(inode, arg, lum_size); + if (S_ISREG(inode->i_mode) && IS_ENCRYPTED(inode) && + ll_i2info(inode)->lli_clob) { + struct iattr attr = { 0 }; + + rc = cl_setattr_ost(ll_i2info(inode)->lli_clob, &attr, + OP_XVALID_FLAGS, LUSTRE_ENCRYPT_FL); + } } cl_lov_delay_create_clear(&file->f_flags); @@ -2317,12 +2620,12 @@ retry: GOTO(out, rc = PTR_ERR(env)); rc = cl_object_layout_get(env, obj, &cl); - if (!rc && cl.cl_is_composite) + if (rc >= 0 && cl.cl_is_composite) rc = ll_layout_write_intent(inode, LAYOUT_INTENT_WRITE, &ext); cl_env_put(env, &refcheck); - if (rc) + if (rc < 0) GOTO(out, rc); } @@ -2398,26 +2701,26 @@ out: */ int ll_release_openhandle(struct dentry *dentry, struct lookup_intent *it) { - struct inode *inode = dentry->d_inode; - struct obd_client_handle *och; - int rc; - ENTRY; + struct inode *inode = dentry->d_inode; + struct obd_client_handle *och; + int rc; + ENTRY; - LASSERT(inode); + LASSERT(inode); - /* Root ? Do nothing. */ - if (dentry->d_inode->i_sb->s_root == dentry) - RETURN(0); + /* Root ? Do nothing. */ + if (is_root_inode(inode)) + RETURN(0); - /* No open handle to close? Move away */ - if (!it_disposition(it, DISP_OPEN_OPEN)) - RETURN(0); + /* No open handle to close? Move away */ + if (!it_disposition(it, DISP_OPEN_OPEN)) + RETURN(0); - LASSERT(it_open_error(DISP_OPEN_OPEN, it) == 0); + LASSERT(it_open_error(DISP_OPEN_OPEN, it) == 0); - OBD_ALLOC(och, sizeof(*och)); - if (!och) - GOTO(out, rc = -ENOMEM); + OBD_ALLOC(och, sizeof(*och)); + if (!och) + GOTO(out, rc = -ENOMEM); rc = ll_och_fill(ll_i2sbi(inode)->ll_md_exp, it, och); if (rc) @@ -2471,7 +2774,7 @@ static int ll_do_fiemap(struct inode *inode, struct fiemap *fiemap, GOTO(out, rc); } - fmkey.lfik_oa.o_valid = OBD_MD_FLID | OBD_MD_FLGROUP; + fmkey.lfik_oa.o_valid = OBD_MD_FLID | OBD_MD_FLGROUP | OBD_MD_FLPROJID; obdo_from_inode(&fmkey.lfik_oa, inode, OBD_MD_FLSIZE); obdo_set_parent_fid(&fmkey.lfik_oa, &ll_i2info(inode)->lli_fid); @@ -2501,8 +2804,8 @@ int ll_fid2path(struct inode *inode, void __user *arg) ENTRY; - if (!cfs_capable(CFS_CAP_DAC_READ_SEARCH) && - !(ll_i2sbi(inode)->ll_flags & LL_SBI_USER_FID2PATH)) + if (!capable(CAP_DAC_READ_SEARCH) && + !test_bit(LL_SBI_USER_FID2PATH, ll_i2sbi(inode)->ll_flags)) RETURN(-EPERM); /* Only need to get the buflen */ @@ -2615,8 +2918,9 @@ int ll_hsm_release(struct inode *inode) struct lu_env *env; struct obd_client_handle *och = NULL; __u64 data_version = 0; - int rc; __u16 refcheck; + int rc; + ENTRY; CDEBUG(D_INODE, "%s: Releasing file "DFID".\n", @@ -2628,7 +2932,8 @@ int ll_hsm_release(struct inode *inode) GOTO(out, rc = PTR_ERR(och)); /* Grab latest data_version and [am]time values */ - rc = ll_data_version(inode, &data_version, LL_DV_WR_FLUSH); + rc = ll_data_version(inode, &data_version, + LL_DV_WR_FLUSH | LL_DV_SZ_UPDATE); if (rc != 0) GOTO(out, rc); @@ -2789,7 +3094,7 @@ int ll_hsm_state_set(struct inode *inode, struct hsm_state_set *hss) /* Non-root users are forbidden to set or clear flags which are * NOT defined in HSM_USER_MASK. */ if (((hss->hss_setmask | hss->hss_clearmask) & ~HSM_USER_MASK) && - !cfs_capable(CFS_CAP_SYS_ADMIN)) + !capable(CAP_SYS_ADMIN)) RETURN(-EPERM); if (!exp_connect_archive_id_array(exp)) { @@ -2993,8 +3298,7 @@ int ll_file_lock_ahead(struct file *file, struct llapi_lu_ladvise *ladvise) descr->cld_mode = cl_mode; /* CEF_MUST is used because we do not want to convert a * lockahead request to a lockless lock */ - descr->cld_enq_flags = CEF_MUST | CEF_LOCK_NO_EXPAND | - CEF_NONBLOCK; + descr->cld_enq_flags = CEF_MUST | CEF_LOCK_NO_EXPAND; if (ladvise->lla_peradvice_flags & LF_ASYNC) descr->cld_enq_flags |= CEF_SPECULATIVE; @@ -3067,7 +3371,7 @@ static int ll_ladvise_sanity(struct inode *inode, ladvise_names[advice], rc); GOTO(out, rc); } - /* fallthrough */ + fallthrough; case LU_LADVISE_WILLREAD: case LU_LADVISE_DONTNEED: default: @@ -3165,7 +3469,7 @@ int ll_ioctl_fsgetxattr(struct inode *inode, unsigned int cmd, RETURN(-EFAULT); fsxattr.fsx_xflags = ll_inode_flags_to_xflags(inode->i_flags); - if (ll_file_test_flag(ll_i2info(inode), LLIF_PROJECT_INHERIT)) + if (test_bit(LLIF_PROJECT_INHERIT, &ll_i2info(inode)->lli_flags)) fsxattr.fsx_xflags |= FS_XFLAG_PROJINHERIT; fsxattr.fsx_projid = ll_i2info(inode)->lli_projid; if (copy_to_user((struct fsxattr __user *)arg, @@ -3175,48 +3479,51 @@ int ll_ioctl_fsgetxattr(struct inode *inode, unsigned int cmd, RETURN(0); } -int ll_ioctl_check_project(struct inode *inode, struct fsxattr *fa) +int ll_ioctl_check_project(struct inode *inode, __u32 xflags, + __u32 projid) { /* * Project Quota ID state is only allowed to change from within the init * namespace. Enforce that restriction only if we are trying to change * the quota ID state. Everything else is allowed in user namespaces. */ - if (current_user_ns() == &init_user_ns) + if (current_user_ns() == &init_user_ns) { + /* + * Caller is allowed to change the project ID. if it is being + * changed, make sure that the new value is valid. + */ + if (ll_i2info(inode)->lli_projid != projid && + !projid_valid(make_kprojid(&init_user_ns, projid))) + return -EINVAL; + return 0; + } - if (ll_i2info(inode)->lli_projid != fa->fsx_projid) + if (ll_i2info(inode)->lli_projid != projid) return -EINVAL; - if (ll_file_test_flag(ll_i2info(inode), LLIF_PROJECT_INHERIT)) { - if (!(fa->fsx_xflags & FS_XFLAG_PROJINHERIT)) + if (test_bit(LLIF_PROJECT_INHERIT, &ll_i2info(inode)->lli_flags)) { + if (!(xflags & FS_XFLAG_PROJINHERIT)) return -EINVAL; } else { - if (fa->fsx_xflags & FS_XFLAG_PROJINHERIT) + if (xflags & FS_XFLAG_PROJINHERIT) return -EINVAL; } return 0; } -int ll_ioctl_fssetxattr(struct inode *inode, unsigned int cmd, - unsigned long arg) +static int ll_set_project(struct inode *inode, __u32 xflags, __u32 projid) { - - struct md_op_data *op_data; struct ptlrpc_request *req = NULL; - int rc = 0; - struct fsxattr fsxattr; + struct md_op_data *op_data; struct cl_object *obj; - struct iattr *attr; - int flags; - - if (copy_from_user(&fsxattr, - (const struct fsxattr __user *)arg, - sizeof(fsxattr))) - RETURN(-EFAULT); + unsigned int inode_flags; + int rc = 0; - rc = ll_ioctl_check_project(inode, &fsxattr); + CDEBUG(D_QUOTA, DFID" xflags=%x projid=%u\n", + PFID(ll_inode2fid(inode)), xflags, projid); + rc = ll_ioctl_check_project(inode, xflags, projid); if (rc) RETURN(rc); @@ -3225,39 +3532,116 @@ int ll_ioctl_fssetxattr(struct inode *inode, unsigned int cmd, if (IS_ERR(op_data)) RETURN(PTR_ERR(op_data)); - flags = ll_xflags_to_inode_flags(fsxattr.fsx_xflags); - op_data->op_attr_flags = ll_inode_to_ext_flags(flags); - if (fsxattr.fsx_xflags & FS_XFLAG_PROJINHERIT) + inode_flags = ll_xflags_to_inode_flags(xflags); + op_data->op_attr_flags = ll_inode_to_ext_flags(inode_flags); + if (xflags & FS_XFLAG_PROJINHERIT) op_data->op_attr_flags |= LUSTRE_PROJINHERIT_FL; - op_data->op_projid = fsxattr.fsx_projid; + + /* pass projid to md_op_data */ + op_data->op_projid = projid; + op_data->op_xvalid |= OP_XVALID_PROJID | OP_XVALID_FLAGS; - rc = md_setattr(ll_i2sbi(inode)->ll_md_exp, op_data, NULL, - 0, &req); + rc = md_setattr(ll_i2sbi(inode)->ll_md_exp, op_data, NULL, 0, &req); ptlrpc_req_finished(req); if (rc) GOTO(out_fsxattr, rc); ll_update_inode_flags(inode, op_data->op_attr_flags); - obj = ll_i2info(inode)->lli_clob; - if (obj == NULL) - GOTO(out_fsxattr, rc); - /* Avoiding OST RPC if this is only project ioctl */ - if (fsxattr.fsx_xflags == 0 || - fsxattr.fsx_xflags == FS_XFLAG_PROJINHERIT) + /* Avoid OST RPC if this is only ioctl setting project inherit flag */ + if (xflags == 0 || xflags == FS_XFLAG_PROJINHERIT) GOTO(out_fsxattr, rc); - OBD_ALLOC_PTR(attr); - if (attr == NULL) - GOTO(out_fsxattr, rc = -ENOMEM); + obj = ll_i2info(inode)->lli_clob; + if (obj) { + struct iattr attr = { 0 }; + + rc = cl_setattr_ost(obj, &attr, OP_XVALID_FLAGS, xflags); + } - rc = cl_setattr_ost(obj, attr, OP_XVALID_FLAGS, - fsxattr.fsx_xflags); - OBD_FREE_PTR(attr); out_fsxattr: ll_finish_md_op_data(op_data); RETURN(rc); } +int ll_ioctl_fssetxattr(struct inode *inode, unsigned int cmd, + unsigned long arg) +{ + struct fsxattr fsxattr; + + ENTRY; + + if (copy_from_user(&fsxattr, + (const struct fsxattr __user *)arg, + sizeof(fsxattr))) + RETURN(-EFAULT); + + RETURN(ll_set_project(inode, fsxattr.fsx_xflags, + fsxattr.fsx_projid)); +} + +int ll_ioctl_project(struct file *file, unsigned int cmd, + unsigned long arg) +{ + struct lu_project lu_project; + struct dentry *dentry = file_dentry(file); + struct inode *inode = file_inode(file); + struct dentry *child_dentry = NULL; + int rc = 0, name_len; + + if (copy_from_user(&lu_project, + (const struct lu_project __user *)arg, + sizeof(lu_project))) + RETURN(-EFAULT); + + /* apply child dentry if name is valid */ + name_len = strnlen(lu_project.project_name, NAME_MAX); + if (name_len > 0 && name_len <= NAME_MAX) { + inode_lock(inode); + child_dentry = lookup_one_len(lu_project.project_name, + dentry, name_len); + inode_unlock(inode); + if (IS_ERR(child_dentry)) { + rc = PTR_ERR(child_dentry); + goto out; + } + inode = child_dentry->d_inode; + if (!inode) { + rc = -ENOENT; + goto out; + } + } else if (name_len > NAME_MAX) { + rc = -EINVAL; + goto out; + } + + switch (lu_project.project_type) { + case LU_PROJECT_SET: + rc = ll_set_project(inode, lu_project.project_xflags, + lu_project.project_id); + break; + case LU_PROJECT_GET: + lu_project.project_xflags = + ll_inode_flags_to_xflags(inode->i_flags); + if (test_bit(LLIF_PROJECT_INHERIT, + &ll_i2info(inode)->lli_flags)) + lu_project.project_xflags |= FS_XFLAG_PROJINHERIT; + lu_project.project_id = ll_i2info(inode)->lli_projid; + if (copy_to_user((struct lu_project __user *)arg, + &lu_project, sizeof(lu_project))) { + rc = -EFAULT; + goto out; + } + break; + default: + rc = -EINVAL; + break; + } +out: + if (!IS_ERR_OR_NULL(child_dentry)) + dput(child_dentry); + RETURN(rc); +} + static long ll_file_unlock_lease(struct file *file, struct ll_ioc_lease *ioc, unsigned long arg) { @@ -3347,6 +3731,7 @@ static long ll_file_unlock_lease(struct file *file, struct ll_ioc_lease *ioc, if (!layout_file) GOTO(out_lease_close, rc = -EBADF); + /* if layout_file == file, it means to destroy the mirror */ sp.sp_inode = file_inode(layout_file); sp.sp_mirror_id = (__u16)mirror_id; data = &sp; @@ -3357,6 +3742,9 @@ static long ll_file_unlock_lease(struct file *file, struct ll_ioc_lease *ioc, if (ioc->lil_count != 1) RETURN(-EINVAL); + if (IS_ENCRYPTED(inode)) + RETURN(-EOPNOTSUPP); + arg += sizeof(*ioc); if (copy_from_user(¶m.pa_archive_id, (void __user *)arg, sizeof(__u32))) @@ -3404,6 +3792,8 @@ out: case LL_LEASE_LAYOUT_SPLIT: if (layout_file) fput(layout_file); + + ll_layout_refresh(inode, &fd->fd_layout_version); break; case LL_LEASE_PCC_ATTACH: if (!rc) @@ -3676,6 +4066,7 @@ out: RETURN(0); } + case OBD_IOC_GETNAME_OLD: case OBD_IOC_GETDTNAME: case OBD_IOC_GETMDNAME: RETURN(ll_get_obd_name(inode, cmd, arg)); @@ -3724,9 +4115,10 @@ out: RETURN(rc); } case LL_IOC_HSM_ACTION: { - struct md_op_data *op_data; - struct hsm_current_action *hca; - int rc; + struct md_op_data *op_data; + struct hsm_current_action *hca; + const char *action; + int rc; OBD_ALLOC_PTR(hca); if (hca == NULL) @@ -3741,10 +4133,26 @@ out: rc = obd_iocontrol(cmd, ll_i2mdexp(inode), sizeof(*op_data), op_data, NULL); + if (rc < 0) + GOTO(skip_copy, rc); + + /* The hsm_current_action retreived from the server could + * contain corrupt information. If it is incorrect data collect + * debug information. We still send the data even if incorrect + * to user land to handle. + */ + action = hsm_user_action2name(hca->hca_action); + if (strcmp(action, "UNKNOWN") == 0 || + hca->hca_state > HPS_DONE) { + CDEBUG(D_HSM, + "HSM current state %s action %s, offset = %llu, length %llu\n", + hsm_progress_state2name(hca->hca_state), action, + hca->hca_location.offset, hca->hca_location.length); + } if (copy_to_user((char __user *)arg, hca, sizeof(*hca))) rc = -EFAULT; - +skip_copy: ll_finish_md_op_data(op_data); OBD_FREE_PTR(hca); RETURN(rc); @@ -3903,6 +4311,8 @@ out_ladvise: RETURN(ll_ioctl_fsgetxattr(inode, cmd, arg)); case FS_IOC_FSSETXATTR: RETURN(ll_ioctl_fssetxattr(inode, cmd, arg)); + case LL_IOC_PROJECT: + RETURN(ll_ioctl_project(file, cmd, arg)); case BLKSSZGET: RETURN(put_user(PAGE_SIZE, (int __user *)arg)); case LL_IOC_HEAT_GET: { @@ -3951,7 +4361,7 @@ out_ladvise: if (!S_ISREG(inode->i_mode)) GOTO(out_detach_free, rc = -EINVAL); - if (!inode_owner_or_capable(inode)) + if (!inode_owner_or_capable(&init_user_ns, inode)) GOTO(out_detach_free, rc = -EPERM); rc = pcc_ioctl_detach(inode, detach->pccd_opt); @@ -3982,34 +4392,138 @@ out_state: OBD_FREE_PTR(state); RETURN(rc); } +#ifdef HAVE_LUSTRE_CRYPTO + case LL_IOC_SET_ENCRYPTION_POLICY: + if (!ll_sbi_has_encrypt(ll_i2sbi(inode))) + return -EOPNOTSUPP; + return llcrypt_ioctl_set_policy(file, (const void __user *)arg); + case LL_IOC_GET_ENCRYPTION_POLICY_EX: + if (!ll_sbi_has_encrypt(ll_i2sbi(inode))) + return -EOPNOTSUPP; + return llcrypt_ioctl_get_policy_ex(file, (void __user *)arg); + case LL_IOC_ADD_ENCRYPTION_KEY: + if (!ll_sbi_has_encrypt(ll_i2sbi(inode))) + return -EOPNOTSUPP; + return llcrypt_ioctl_add_key(file, (void __user *)arg); + case LL_IOC_REMOVE_ENCRYPTION_KEY: + if (!ll_sbi_has_encrypt(ll_i2sbi(inode))) + return -EOPNOTSUPP; + return llcrypt_ioctl_remove_key(file, (void __user *)arg); + case LL_IOC_REMOVE_ENCRYPTION_KEY_ALL_USERS: + if (!ll_sbi_has_encrypt(ll_i2sbi(inode))) + return -EOPNOTSUPP; + return llcrypt_ioctl_remove_key_all_users(file, + (void __user *)arg); + case LL_IOC_GET_ENCRYPTION_KEY_STATUS: + if (!ll_sbi_has_encrypt(ll_i2sbi(inode))) + return -EOPNOTSUPP; + return llcrypt_ioctl_get_key_status(file, (void __user *)arg); +#endif + + case LL_IOC_UNLOCK_FOREIGN: { + struct dentry *dentry = file_dentry(file); + + /* if not a foreign symlink do nothing */ + if (ll_foreign_is_removable(dentry, true)) { + CDEBUG(D_INFO, + "prevent unlink of non-foreign file ("DFID")\n", + PFID(ll_inode2fid(inode))); + RETURN(-EOPNOTSUPP); + } + RETURN(0); + } + default: RETURN(obd_iocontrol(cmd, ll_i2dtexp(inode), 0, NULL, (void __user *)arg)); } } +loff_t ll_lseek(struct file *file, loff_t offset, int whence) +{ + struct inode *inode = file_inode(file); + struct lu_env *env; + struct cl_io *io; + struct cl_lseek_io *lsio; + __u16 refcheck; + int rc; + loff_t retval; + + ENTRY; + + env = cl_env_get(&refcheck); + if (IS_ERR(env)) + RETURN(PTR_ERR(env)); + + io = vvp_env_thread_io(env); + io->ci_obj = ll_i2info(inode)->lli_clob; + ll_io_set_mirror(io, file); + + lsio = &io->u.ci_lseek; + lsio->ls_start = offset; + lsio->ls_whence = whence; + lsio->ls_result = -ENXIO; + + do { + rc = cl_io_init(env, io, CIT_LSEEK, io->ci_obj); + if (!rc) { + struct vvp_io *vio = vvp_env_io(env); + + vio->vui_fd = file->private_data; + rc = cl_io_loop(env, io); + } else { + rc = io->ci_result; + } + retval = rc ? : lsio->ls_result; + cl_io_fini(env, io); + } while (unlikely(io->ci_need_restart)); + + cl_env_put(env, &refcheck); + + /* Without the key, SEEK_HOLE return value has to be + * rounded up to next LUSTRE_ENCRYPTION_UNIT_SIZE. + */ + if (llcrypt_require_key(inode) == -ENOKEY && whence == SEEK_HOLE) + retval = round_up(retval, LUSTRE_ENCRYPTION_UNIT_SIZE); + + RETURN(retval); +} + static loff_t ll_file_seek(struct file *file, loff_t offset, int origin) { struct inode *inode = file_inode(file); - loff_t retval, eof = 0; + loff_t retval = offset, eof = 0; ktime_t kstart = ktime_get(); ENTRY; - retval = offset + ((origin == SEEK_END) ? i_size_read(inode) : - (origin == SEEK_CUR) ? file->f_pos : 0); + CDEBUG(D_VFSTRACE, "VFS Op:inode="DFID"(%p), to=%llu=%#llx(%d)\n", PFID(ll_inode2fid(inode)), inode, retval, retval, origin); - if (origin == SEEK_END || origin == SEEK_HOLE || origin == SEEK_DATA) { + if (origin == SEEK_END) { retval = ll_glimpse_size(inode); if (retval != 0) RETURN(retval); eof = i_size_read(inode); } - retval = generic_file_llseek_size(file, offset, origin, - ll_file_maxbytes(inode), eof); + if (origin == SEEK_HOLE || origin == SEEK_DATA) { + if (offset < 0) + return -ENXIO; + + /* flush local cache first if any */ + cl_sync_file_range(inode, offset, OBD_OBJECT_EOF, + CL_FSYNC_LOCAL, 0); + + retval = ll_lseek(file, offset, origin); + if (retval < 0) + return retval; + retval = vfs_setpos(file, retval, ll_file_maxbytes(inode)); + } else { + retval = generic_file_llseek_size(file, offset, origin, + ll_file_maxbytes(inode), eof); + } if (retval >= 0) ll_stats_ops_tally(ll_i2sbi(inode), LPROC_LL_LLSEEK, ktime_us_delta(ktime_get(), kstart)); @@ -4114,7 +4628,6 @@ int ll_fsync(struct file *file, loff_t start, loff_t end, int datasync) /* fsync's caller has already called _fdata{sync,write}, we want * that IO to finish before calling the osc and mdc sync methods */ rc = filemap_write_and_wait_range(inode->i_mapping, start, end); - inode_lock(inode); /* catch async errors that were recorded back when async writeback * failed for pages in this mapping. */ @@ -4155,8 +4668,6 @@ int ll_fsync(struct file *file, loff_t start, loff_t end, int datasync) fd->fd_write_failed = false; } - inode_unlock(inode); - if (!rc) ll_stats_ops_tally(ll_i2sbi(inode), LPROC_LL_FSYNC, ktime_us_delta(ktime_get(), kstart)); @@ -4314,10 +4825,10 @@ int ll_get_fid_by_name(struct inode *parent, const char *name, int namelen, struct lu_fid *fid, struct inode **inode) { - struct md_op_data *op_data = NULL; - struct mdt_body *body; - struct ptlrpc_request *req; - int rc; + struct md_op_data *op_data = NULL; + struct mdt_body *body; + struct ptlrpc_request *req; + int rc; ENTRY; op_data = ll_prep_md_op_data(NULL, parent, NULL, name, namelen, 0, @@ -4338,14 +4849,14 @@ int ll_get_fid_by_name(struct inode *parent, const char *name, *fid = body->mbo_fid1; if (inode != NULL) - rc = ll_prep_inode(inode, req, parent->i_sb, NULL); + rc = ll_prep_inode(inode, &req->rq_pill, parent->i_sb, NULL); out_req: ptlrpc_req_finished(req); RETURN(rc); } int ll_migrate(struct inode *parent, struct file *file, struct lmv_user_md *lum, - const char *name) + const char *name, __u32 flags) { struct dentry *dchild = NULL; struct inode *child_inode = NULL; @@ -4404,7 +4915,7 @@ int ll_migrate(struct inode *parent, struct file *file, struct lmv_user_md *lum, * by checking the migrate FID against the FID of the * filesystem root. */ - if (child_inode == parent->i_sb->s_root->d_inode) + if (is_root_inode(child_inode)) GOTO(out_iput, rc = -EINVAL); op_data = ll_prep_md_op_data(NULL, parent, NULL, name, namelen, @@ -4425,6 +4936,11 @@ int ll_migrate(struct inode *parent, struct file *file, struct lmv_user_md *lum, op_data->op_data = lum; op_data->op_data_size = lumlen; + /* migrate dirent only for subdirs if MDS_MIGRATE_NSONLY set */ + if (S_ISDIR(child_inode->i_mode) && (flags & MDS_MIGRATE_NSONLY) && + lmv_dir_layout_changing(ll_i2info(parent)->lli_lsm_md)) + op_data->op_bias |= MDS_MIGRATE_NSONLY; + again: if (S_ISREG(child_inode->i_mode)) { och = ll_lease_open(child_inode, NULL, FMODE_WRITE, 0); @@ -4449,8 +4965,9 @@ again: spin_unlock(&och->och_mod->mod_open_req->rq_lock); } - rc = md_rename(ll_i2sbi(parent)->ll_md_exp, op_data, name, namelen, - name, namelen, &request); + rc = md_rename(ll_i2sbi(parent)->ll_md_exp, op_data, + op_data->op_name, op_data->op_namelen, + op_data->op_name, op_data->op_namelen, &request); if (rc == 0) { LASSERT(request != NULL); ll_update_times(request, parent); @@ -4507,7 +5024,7 @@ ll_file_noflock(struct file *file, int cmd, struct file_lock *file_lock) */ if (!(fd->fd_flags & LL_FILE_FLOCK_WARNING)) { fd->fd_flags |= LL_FILE_FLOCK_WARNING; - CDEBUG_LIMIT(D_TTY | D_CONSOLE, + CDEBUG_LIMIT(D_CONSOLE, "flock disabled, mount with '-o [local]flock' to enable\r\n"); } RETURN(-ENOSYS); @@ -4610,6 +5127,7 @@ static int ll_inode_revalidate_fini(struct inode *inode, int rc) static int ll_inode_revalidate(struct dentry *dentry, enum ldlm_intent_flags op) { + struct inode *parent; struct inode *inode = dentry->d_inode; struct obd_export *exp = ll_i2mdexp(inode); struct lookup_intent oit = { @@ -4617,18 +5135,30 @@ static int ll_inode_revalidate(struct dentry *dentry, enum ldlm_intent_flags op) }; struct ptlrpc_request *req = NULL; struct md_op_data *op_data; + const char *name = NULL; + size_t namelen = 0; int rc = 0; ENTRY; CDEBUG(D_VFSTRACE, "VFS Op:inode="DFID"(%p),name=%s\n", PFID(ll_inode2fid(inode)), inode, dentry->d_name.name); - /* Call getattr by fid, so do not provide name at all. */ - op_data = ll_prep_md_op_data(NULL, inode, inode, NULL, 0, 0, + if (exp_connect_flags2(exp) & OBD_CONNECT2_GETATTR_PFID) { + parent = dentry->d_parent->d_inode; + name = dentry->d_name.name; + namelen = dentry->d_name.len; + } else { + parent = inode; + } + + op_data = ll_prep_md_op_data(NULL, parent, inode, name, namelen, 0, LUSTRE_OPC_ANY, NULL); if (IS_ERR(op_data)) RETURN(PTR_ERR(op_data)); + /* Call getattr by fid */ + if (exp_connect_flags2(exp) & OBD_CONNECT2_GETATTR_PFID) + op_data->op_flags = MF_GETATTR_BY_FID; rc = md_intent_lock(exp, op_data, &oit, &req, &ll_md_blocking_ast, 0); ll_finish_md_op_data(op_data); if (rc < 0) { @@ -4646,11 +5176,8 @@ static int ll_inode_revalidate(struct dentry *dentry, enum ldlm_intent_flags op) * do_lookup() -> ll_revalidate_it(). We cannot use d_drop * here to preserve get_cwd functionality on 2.6. * Bug 10503 */ - if (!dentry->d_inode->i_nlink) { - spin_lock(&inode->i_lock); - d_lustre_invalidate(dentry, 0); - spin_unlock(&inode->i_lock); - } + if (!dentry->d_inode->i_nlink) + d_lustre_invalidate(dentry); ll_lookup_finish_locks(&oit, dentry); out: @@ -4665,19 +5192,24 @@ static int ll_merge_md_attr(struct inode *inode) struct cl_attr attr = { 0 }; int rc; - LASSERT(lli->lli_lsm_md != NULL); - - if (!lmv_dir_striped(lli->lli_lsm_md)) + if (!lli->lli_lsm_md) RETURN(0); down_read(&lli->lli_lsm_sem); + if (!lmv_dir_striped(lli->lli_lsm_md)) { + up_read(&lli->lli_lsm_sem); + RETURN(0); + } rc = md_merge_attr(ll_i2mdexp(inode), ll_i2info(inode)->lli_lsm_md, &attr, ll_md_blocking_ast); up_read(&lli->lli_lsm_sem); if (rc != 0) RETURN(rc); + spin_lock(&inode->i_lock); set_nlink(inode, attr.cat_nlink); + spin_unlock(&inode->i_lock); + inode->i_blocks = attr.cat_blocks; i_size_write(inode, attr.cat_size); @@ -4689,7 +5221,7 @@ static int ll_merge_md_attr(struct inode *inode) } int ll_getattr_dentry(struct dentry *de, struct kstat *stat, u32 request_mask, - unsigned int flags) + unsigned int flags, bool foreign) { struct inode *inode = de->d_inode; struct ll_sb_info *sbi = ll_i2sbi(inode); @@ -4715,7 +5247,10 @@ int ll_getattr_dentry(struct dentry *de, struct kstat *stat, u32 request_mask, if (rc < 0) RETURN(rc); - if (S_ISREG(inode->i_mode)) { + /* foreign file/dir are always of zero length, so don't + * need to validate size. + */ + if (S_ISREG(inode->i_mode) && !foreign) { bool cached; if (!need_glimpse) @@ -4755,14 +5290,15 @@ int ll_getattr_dentry(struct dentry *de, struct kstat *stat, u32 request_mask, * restore the MDT holds the layout lock so the glimpse will * block up to the end of restore (getattr will block) */ - if (!ll_file_test_flag(lli, LLIF_FILE_RESTORING)) { + if (!test_bit(LLIF_FILE_RESTORING, &lli->lli_flags)) { rc = ll_glimpse_size(inode); if (rc < 0) RETURN(rc); } } else { /* If object isn't regular a file then don't validate size. */ - if (ll_dir_striped(inode)) { + /* foreign dir is not striped dir */ + if (ll_dir_striped(inode) && !foreign) { rc = ll_merge_md_attr(inode); if (rc < 0) RETURN(rc); @@ -4789,7 +5325,12 @@ fill_attr: stat->rdev = inode->i_rdev; } - stat->mode = inode->i_mode; + /* foreign symlink to be exposed as a real symlink */ + if (!foreign) + stat->mode = inode->i_mode; + else + stat->mode = (inode->i_mode & ~S_IFMT) | S_IFLNK; + stat->uid = inode->i_uid; stat->gid = inode->i_gid; stat->atime = inode->i_atime; @@ -4808,7 +5349,7 @@ fill_attr: stat->size = i_size_read(inode); stat->blocks = inode->i_blocks; -#ifdef HAVE_INODEOPS_ENHANCED_GETATTR +#if defined(HAVE_USER_NAMESPACE_ARG) || defined(HAVE_INODEOPS_ENHANCED_GETATTR) if (flags & AT_STATX_DONT_SYNC) { if (stat->size == 0 && lli->lli_attr_valid & OBD_MD_FLLAZYSIZE) @@ -4834,27 +5375,29 @@ fill_attr: return 0; } -#ifdef HAVE_INODEOPS_ENHANCED_GETATTR -int ll_getattr(const struct path *path, struct kstat *stat, - u32 request_mask, unsigned int flags) +#if defined(HAVE_USER_NAMESPACE_ARG) || defined(HAVE_INODEOPS_ENHANCED_GETATTR) +int ll_getattr(struct user_namespace *mnt_userns, const struct path *path, + struct kstat *stat, u32 request_mask, unsigned int flags) { - return ll_getattr_dentry(path->dentry, stat, request_mask, flags); + return ll_getattr_dentry(path->dentry, stat, request_mask, flags, + false); } #else int ll_getattr(struct vfsmount *mnt, struct dentry *de, struct kstat *stat) { return ll_getattr_dentry(de, stat, STATX_BASIC_STATS, - AT_STATX_SYNC_AS_STAT); + AT_STATX_SYNC_AS_STAT, false); } #endif -int cl_falloc(struct inode *inode, int mode, loff_t offset, loff_t len) +int cl_falloc(struct file *file, struct inode *inode, int mode, loff_t offset, + loff_t len) { + loff_t size = i_size_read(inode); struct lu_env *env; struct cl_io *io; __u16 refcheck; - int rc; loff_t sa_falloc_end; - loff_t size = i_size_read(inode); + int rc; ENTRY; @@ -4864,43 +5407,48 @@ int cl_falloc(struct inode *inode, int mode, loff_t offset, loff_t len) io = vvp_env_thread_io(env); io->ci_obj = ll_i2info(inode)->lli_clob; + ll_io_set_mirror(io, file); + io->ci_verify_layout = 1; io->u.ci_setattr.sa_parent_fid = lu_object_fid(&io->ci_obj->co_lu); io->u.ci_setattr.sa_falloc_mode = mode; io->u.ci_setattr.sa_falloc_offset = offset; - io->u.ci_setattr.sa_falloc_len = len; - io->u.ci_setattr.sa_falloc_end = io->u.ci_setattr.sa_falloc_offset + - io->u.ci_setattr.sa_falloc_len; + io->u.ci_setattr.sa_falloc_end = offset + len; io->u.ci_setattr.sa_subtype = CL_SETATTR_FALLOCATE; - sa_falloc_end = io->u.ci_setattr.sa_falloc_end; - if (sa_falloc_end > size) { + + CDEBUG(D_INODE, "UID %u GID %u PRJID %u\n", + from_kuid(&init_user_ns, inode->i_uid), + from_kgid(&init_user_ns, inode->i_gid), + ll_i2info(inode)->lli_projid); + + io->u.ci_setattr.sa_falloc_uid = from_kuid(&init_user_ns, inode->i_uid); + io->u.ci_setattr.sa_falloc_gid = from_kgid(&init_user_ns, inode->i_gid); + io->u.ci_setattr.sa_falloc_projid = ll_i2info(inode)->lli_projid; + + if (io->u.ci_setattr.sa_falloc_end > size) { + loff_t newsize = io->u.ci_setattr.sa_falloc_end; + /* Check new size against VFS/VM file size limit and rlimit */ - rc = inode_newsize_ok(inode, sa_falloc_end); + rc = inode_newsize_ok(inode, newsize); if (rc) goto out; - if (sa_falloc_end > ll_file_maxbytes(inode)) { + if (newsize > ll_file_maxbytes(inode)) { CDEBUG(D_INODE, "file size too large %llu > %llu\n", - (unsigned long long)(sa_falloc_end), + (unsigned long long)newsize, ll_file_maxbytes(inode)); rc = -EFBIG; goto out; } - io->u.ci_setattr.sa_attr.lvb_size = sa_falloc_end; - if (!(mode & FALLOC_FL_KEEP_SIZE)) - io->u.ci_setattr.sa_avalid |= ATTR_SIZE; - } else { - io->u.ci_setattr.sa_attr.lvb_size = size; } -again: - if (cl_io_init(env, io, CIT_SETATTR, io->ci_obj) == 0) - rc = cl_io_loop(env, io); - else - rc = io->ci_result; - - cl_io_fini(env, io); - if (unlikely(io->ci_need_restart)) - goto again; + do { + rc = cl_io_init(env, io, CIT_SETATTR, io->ci_obj); + if (!rc) + rc = cl_io_loop(env, io); + else + rc = io->ci_result; + cl_io_fini(env, io); + } while (unlikely(io->ci_need_restart)); out: cl_env_put(env, &refcheck); @@ -4909,19 +5457,40 @@ out: long ll_fallocate(struct file *filp, int mode, loff_t offset, loff_t len) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); int rc; + if (offset < 0 || len <= 0) + RETURN(-EINVAL); + /* + * Encrypted inodes can't handle collapse range or zero range or insert + * range since we would need to re-encrypt blocks with a different IV or + * XTS tweak (which are based on the logical block number). + * Similar to what ext4 does. + */ + if (IS_ENCRYPTED(inode) && + (mode & (FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_INSERT_RANGE | + FALLOC_FL_ZERO_RANGE))) + RETURN(-EOPNOTSUPP); + /* - * Only mode == 0 (which is standard prealloc) is supported now. - * Punch is not supported yet. + * mode == 0 (which is standard prealloc) and PUNCH is supported + * Rest of mode options are not supported yet. */ - if (mode & ~FALLOC_FL_KEEP_SIZE) + if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE)) RETURN(-EOPNOTSUPP); ll_stats_ops_tally(ll_i2sbi(inode), LPROC_LL_FALLOCATE, 1); - rc = cl_falloc(inode, mode, offset, len); + rc = cl_falloc(filp, inode, mode, offset, len); + /* + * ENOTSUPP (524) is an NFSv3 specific error code erroneously + * used by Lustre in several places. Retuning it here would + * confuse applications that explicity test for EOPNOTSUPP + * (95) and fall back to ftruncate(). + */ + if (rc == -ENOTSUPP) + rc = -EOPNOTSUPP; RETURN(rc); } @@ -4952,6 +5521,15 @@ static int ll_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, rc = ll_do_fiemap(inode, fiemap, num_bytes); + if (IS_ENCRYPTED(inode)) { + int i; + + for (i = 0; i < fiemap->fm_mapped_extents; i++) + fiemap->fm_extents[i].fe_flags |= + FIEMAP_EXTENT_DATA_ENCRYPTED | + FIEMAP_EXTENT_ENCODED; + } + fieinfo->fi_flags = fiemap->fm_flags; fieinfo->fi_extents_mapped = fiemap->fm_mapped_extents; if (extent_count > 0 && @@ -4964,88 +5542,14 @@ out: return rc; } -struct posix_acl *ll_get_acl(struct inode *inode, int type) -{ - struct ll_inode_info *lli = ll_i2info(inode); - struct posix_acl *acl = NULL; - ENTRY; - - spin_lock(&lli->lli_lock); - /* VFS' acl_permission_check->check_acl will release the refcount */ - acl = posix_acl_dup(lli->lli_posix_acl); - spin_unlock(&lli->lli_lock); - - RETURN(acl); -} - -#ifdef HAVE_IOP_SET_ACL -#ifdef CONFIG_LUSTRE_FS_POSIX_ACL -int ll_set_acl(struct inode *inode, struct posix_acl *acl, int type) -{ - struct ll_sb_info *sbi = ll_i2sbi(inode); - struct ptlrpc_request *req = NULL; - const char *name = NULL; - char *value = NULL; - size_t value_size = 0; - int rc = 0; - ENTRY; - - switch (type) { - case ACL_TYPE_ACCESS: - name = XATTR_NAME_POSIX_ACL_ACCESS; - if (acl) - rc = posix_acl_update_mode(inode, &inode->i_mode, &acl); - break; - - case ACL_TYPE_DEFAULT: - name = XATTR_NAME_POSIX_ACL_DEFAULT; - if (!S_ISDIR(inode->i_mode)) - rc = acl ? -EACCES : 0; - break; - - default: - rc = -EINVAL; - break; - } - if (rc) - return rc; - - if (acl) { - value_size = posix_acl_xattr_size(acl->a_count); - value = kmalloc(value_size, GFP_NOFS); - if (value == NULL) - GOTO(out, rc = -ENOMEM); - - rc = posix_acl_to_xattr(&init_user_ns, acl, value, value_size); - if (rc < 0) - GOTO(out_value, rc); - } - - rc = md_setxattr(sbi->ll_md_exp, ll_inode2fid(inode), - value ? OBD_MD_FLXATTR : OBD_MD_FLXATTRRM, - name, value, value_size, 0, 0, &req); - - ptlrpc_req_finished(req); -out_value: - kfree(value); -out: - if (rc) - forget_cached_acl(inode, type); - else - set_cached_acl(inode, type, acl); - RETURN(rc); -} -#endif /* CONFIG_LUSTRE_FS_POSIX_ACL */ -#endif /* HAVE_IOP_SET_ACL */ - -int ll_inode_permission(struct inode *inode, int mask) +int ll_inode_permission(struct user_namespace *mnt_userns, struct inode *inode, + int mask) { int rc = 0; struct ll_sb_info *sbi; struct root_squash_info *squash; struct cred *cred = NULL; const struct cred *old_cred = NULL; - cfs_cap_t cap; bool squash_id = false; ktime_t kstart = ktime_get(); @@ -5059,7 +5563,7 @@ int ll_inode_permission(struct inode *inode, int mask) * need to do it before permission check. */ - if (inode == inode->i_sb->s_root->d_inode) { + if (is_root_inode(inode)) { rc = ll_inode_revalidate(inode->i_sb->s_root, IT_LOOKUP); if (rc) RETURN(rc); @@ -5073,7 +5577,7 @@ int ll_inode_permission(struct inode *inode, int mask) squash = &sbi->ll_squash; if (unlikely(squash->rsi_uid != 0 && uid_eq(current_fsuid(), GLOBAL_ROOT_UID) && - !(sbi->ll_flags & LL_SBI_NOROOTSQUASH))) { + !test_bit(LL_SBI_NOROOTSQUASH, sbi->ll_flags))) { squash_id = true; } if (squash_id) { @@ -5089,14 +5593,13 @@ int ll_inode_permission(struct inode *inode, int mask) cred->fsuid = make_kuid(&init_user_ns, squash->rsi_uid); cred->fsgid = make_kgid(&init_user_ns, squash->rsi_gid); - for (cap = 0; cap < sizeof(cfs_cap_t) * 8; cap++) { - if (BIT(cap) & CFS_CAP_FS_MASK) - cap_lower(cred->cap_effective, cap); - } + cred->cap_effective = cap_drop_nfsd_set(cred->cap_effective); + cred->cap_effective = cap_drop_fs_set(cred->cap_effective); + old_cred = override_creds(cred); } - rc = generic_permission(inode, mask); + rc = generic_permission(mnt_userns, inode, mask); /* restore current process's credentials and FS capability */ if (squash_id) { revert_creds(old_cred); @@ -5111,7 +5614,7 @@ int ll_inode_permission(struct inode *inode, int mask) } /* -o localflock - only provides locally consistent flock locks */ -struct file_operations ll_file_operations = { +static const struct file_operations ll_file_operations = { #ifdef HAVE_FILE_OPERATIONS_READ_WRITE_ITER # ifdef HAVE_SYNC_READ_WRITE .read = new_sync_read, @@ -5130,13 +5633,17 @@ struct file_operations ll_file_operations = { .release = ll_file_release, .mmap = ll_file_mmap, .llseek = ll_file_seek, - .splice_read = ll_file_splice_read, +#ifndef HAVE_DEFAULT_FILE_SPLICE_READ_EXPORT + .splice_read = generic_file_splice_read, +#else + .splice_read = pcc_file_splice_read, +#endif .fsync = ll_fsync, .flush = ll_flush, .fallocate = ll_fallocate, }; -struct file_operations ll_file_operations_flock = { +static const struct file_operations ll_file_operations_flock = { #ifdef HAVE_FILE_OPERATIONS_READ_WRITE_ITER # ifdef HAVE_SYNC_READ_WRITE .read = new_sync_read, @@ -5155,7 +5662,11 @@ struct file_operations ll_file_operations_flock = { .release = ll_file_release, .mmap = ll_file_mmap, .llseek = ll_file_seek, - .splice_read = ll_file_splice_read, +#ifndef HAVE_DEFAULT_FILE_SPLICE_READ_EXPORT + .splice_read = generic_file_splice_read, +#else + .splice_read = pcc_file_splice_read, +#endif .fsync = ll_fsync, .flush = ll_flush, .flock = ll_file_flock, @@ -5164,7 +5675,7 @@ struct file_operations ll_file_operations_flock = { }; /* These are for -o noflock - to return ENOSYS on flock calls */ -struct file_operations ll_file_operations_noflock = { +static const struct file_operations ll_file_operations_noflock = { #ifdef HAVE_FILE_OPERATIONS_READ_WRITE_ITER # ifdef HAVE_SYNC_READ_WRITE .read = new_sync_read, @@ -5183,7 +5694,11 @@ struct file_operations ll_file_operations_noflock = { .release = ll_file_release, .mmap = ll_file_mmap, .llseek = ll_file_seek, - .splice_read = ll_file_splice_read, +#ifndef HAVE_DEFAULT_FILE_SPLICE_READ_EXPORT + .splice_read = generic_file_splice_read, +#else + .splice_read = pcc_file_splice_read, +#endif .fsync = ll_fsync, .flush = ll_flush, .flock = ll_file_noflock, @@ -5191,7 +5706,7 @@ struct file_operations ll_file_operations_noflock = { .fallocate = ll_fallocate, }; -struct inode_operations ll_file_inode_operations = { +const struct inode_operations ll_file_inode_operations = { .setattr = ll_setattr, .getattr = ll_getattr, .permission = ll_inode_permission, @@ -5208,6 +5723,18 @@ struct inode_operations ll_file_inode_operations = { #endif }; +const struct file_operations *ll_select_file_operations(struct ll_sb_info *sbi) +{ + const struct file_operations *fops = &ll_file_operations_noflock; + + if (test_bit(LL_SBI_FLOCK, sbi->ll_flags)) + fops = &ll_file_operations_flock; + else if (test_bit(LL_SBI_LOCALFLOCK, sbi->ll_flags)) + fops = &ll_file_operations; + + return fops; +} + int ll_layout_conf(struct inode *inode, const struct cl_object_conf *conf) { struct ll_inode_info *lli = ll_i2info(inode); @@ -5257,7 +5784,7 @@ int ll_layout_conf(struct inode *inode, const struct cl_object_conf *conf) out: cl_env_put(env, &refcheck); - RETURN(rc); + RETURN(rc < 0 ? rc : 0); } /* Fetch layout from MDT with getxattr request, if it's not ready yet */ @@ -5350,7 +5877,9 @@ static int ll_layout_lock_set(struct lustre_handle *lockh, enum ldlm_mode mode, lock = ldlm_handle2lock(lockh); LASSERT(lock != NULL); - LASSERT(ldlm_has_layout(lock)); + + if (!ldlm_has_layout(lock)) + GOTO(out, rc = -EAGAIN); LDLM_DEBUG(lock, "file "DFID"(%p) being reconfigured", PFID(&lli->lli_fid), inode); @@ -5488,7 +6017,8 @@ int ll_layout_refresh(struct inode *inode, __u32 *gen) ENTRY; *gen = ll_layout_version_get(lli); - if (!(sbi->ll_flags & LL_SBI_LAYOUT_LOCK) || *gen != CL_LAYOUT_GEN_NONE) + if (!test_bit(LL_SBI_LAYOUT_LOCK, sbi->ll_flags) || + *gen != CL_LAYOUT_GEN_NONE) RETURN(0); /* sanity checks */