-/* -*- mode: c; c-basic-offset: 8; indent-tabs-mode: nil; -*-
- * vim:expandtab:shiftwidth=8:tabstop=8:
- *
+/*
* GPL HEADER START
*
* DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
* GPL HEADER END
*/
/*
- * Copyright 2008 Sun Microsystems, Inc. All rights reserved
+ * Copyright (c) 2002, 2010, Oracle and/or its affiliates. All rights reserved.
* Use is subject to license terms.
+ *
+ * Copyright (c) 2011, 2013, Intel Corporation.
*/
/*
* This file is part of Lustre, http://www.lustre.org/
#include <linux/pagemap.h>
#include <linux/mm.h>
#include <linux/version.h>
-#include <linux/smp_lock.h>
#include <asm/uaccess.h>
#include <linux/buffer_head.h> // for wait_on_buffer
+#include <linux/pagevec.h>
#define DEBUG_SUBSYSTEM S_LLITE
+#include <lustre/lustre_idl.h>
#include <obd_support.h>
#include <obd_class.h>
#include <lustre_lib.h>
#include <lustre_fid.h>
#include "llite_internal.h"
-#ifndef HAVE_PAGE_CHECKED
-#ifdef HAVE_PG_FS_MISC
-#define PageChecked(page) test_bit(PG_fs_misc, &(page)->flags)
-#define SetPageChecked(page) set_bit(PG_fs_misc, &(page)->flags)
-#else
-#error PageChecked or PageFsMisc not defined in kernel
-#endif
-#endif
-
/*
* (new) readdir implementation overview.
*
*
* page format
*
- *
- *
- *
+ * Page in MDS_READPAGE RPC is packed in LU_PAGE_SIZE, and each page contains
+ * a header lu_dirpage which describes the start/end hash, and whether this
+ * page is empty (contains no dir entry) or hash collide with next page.
+ * After client receives reply, several pages will be integrated into dir page
+ * in PAGE_CACHE_SIZE (if PAGE_CACHE_SIZE greater than LU_PAGE_SIZE), and the
+ * lu_dirpage for this integrated page will be adjusted. See
+ * lmv_adjust_dirpages().
*
*/
/* returns the page unlocked, but with a reference */
-static int ll_dir_readpage(struct file *file, struct page *page)
+static int ll_dir_filler(void *_hash, struct page *page0)
{
- struct inode *inode = page->mapping->host;
- struct ptlrpc_request *request;
- struct mdt_body *body;
- struct obd_capa *oc;
- __u64 hash;
- int rc;
- ENTRY;
-
- hash = (__u64)hash_x_index(page->index);
- CDEBUG(D_VFSTRACE, "VFS Op:inode=%lu/%u(%p) off %lu\n",
- inode->i_ino, inode->i_generation, inode, (unsigned long)hash);
-
- oc = ll_mdscapa_get(inode);
- rc = md_readpage(ll_i2sbi(inode)->ll_md_exp, ll_inode2fid(inode),
- oc, hash, page, &request);
- capa_put(oc);
- if (!rc) {
- body = req_capsule_server_get(&request->rq_pill, &RMF_MDT_BODY);
- /* Checked by mdc_readpage() */
- LASSERT(body != NULL);
-
- if (body->valid & OBD_MD_FLSIZE) {
- ll_inode_size_lock(inode, 0);
- i_size_write(inode, body->size);
- ll_inode_size_unlock(inode, 0);
- }
- SetPageUptodate(page);
- }
- ptlrpc_req_finished(request);
-
- unlock_page(page);
- EXIT;
- return rc;
+ struct inode *inode = page0->mapping->host;
+ int hash64 = ll_i2sbi(inode)->ll_flags & LL_SBI_64BIT_HASH;
+ struct obd_export *exp = ll_i2sbi(inode)->ll_md_exp;
+ struct ptlrpc_request *request;
+ struct mdt_body *body;
+ struct md_op_data *op_data;
+ __u64 hash = *((__u64 *)_hash);
+ struct page **page_pool;
+ struct page *page;
+ struct lu_dirpage *dp;
+ int max_pages = ll_i2sbi(inode)->ll_md_brw_size >> PAGE_CACHE_SHIFT;
+ int nrdpgs = 0; /* number of pages read actually */
+ int npages;
+ int i;
+ int rc;
+ ENTRY;
+
+ CDEBUG(D_VFSTRACE, "VFS Op:inode="DFID"(%p) hash "LPU64"\n",
+ PFID(ll_inode2fid(inode)), inode, hash);
+
+ LASSERT(max_pages > 0 && max_pages <= MD_MAX_BRW_PAGES);
+
+ op_data = ll_prep_md_op_data(NULL, inode, NULL, NULL, 0, 0,
+ LUSTRE_OPC_ANY, NULL);
+ if (IS_ERR(op_data))
+ RETURN(PTR_ERR(op_data));
+
+ OBD_ALLOC(page_pool, sizeof(page) * max_pages);
+ if (page_pool != NULL) {
+ page_pool[0] = page0;
+ } else {
+ page_pool = &page0;
+ max_pages = 1;
+ }
+ for (npages = 1; npages < max_pages; npages++) {
+ page = page_cache_alloc_cold(inode->i_mapping);
+ if (!page)
+ break;
+ page_pool[npages] = page;
+ }
+
+ op_data->op_npages = npages;
+ op_data->op_offset = hash;
+ rc = md_readpage(exp, op_data, page_pool, &request);
+ ll_finish_md_op_data(op_data);
+ if (rc == 0) {
+ body = req_capsule_server_get(&request->rq_pill, &RMF_MDT_BODY);
+ /* Checked by mdc_readpage() */
+ LASSERT(body != NULL);
+
+ if (body->valid & OBD_MD_FLSIZE)
+ cl_isize_write(inode, body->size);
+
+ nrdpgs = (request->rq_bulk->bd_nob_transferred +
+ PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
+ SetPageUptodate(page0);
+ }
+ unlock_page(page0);
+ ptlrpc_req_finished(request);
+
+ CDEBUG(D_VFSTRACE, "read %d/%d pages\n", nrdpgs, npages);
+
+ for (i = 1; i < npages; i++) {
+ unsigned long offset;
+ int ret;
+
+ page = page_pool[i];
+
+ if (rc < 0 || i >= nrdpgs) {
+ page_cache_release(page);
+ continue;
+ }
+
+ SetPageUptodate(page);
+
+ dp = kmap(page);
+ hash = le64_to_cpu(dp->ldp_hash_start);
+ kunmap(page);
+
+ offset = hash_x_index(hash, hash64);
+
+ prefetchw(&page->flags);
+ ret = add_to_page_cache_lru(page, inode->i_mapping, offset,
+ GFP_KERNEL);
+ if (ret == 0)
+ unlock_page(page);
+ else
+ CDEBUG(D_VFSTRACE, "page %lu add to page cache failed:"
+ " %d\n", offset, ret);
+ page_cache_release(page);
+ }
+
+ if (page_pool != &page0)
+ OBD_FREE(page_pool, sizeof(struct page *) * max_pages);
+
+ RETURN(rc);
}
-struct address_space_operations ll_dir_aops = {
- .readpage = ll_dir_readpage,
-};
-
static void ll_check_page(struct inode *dir, struct page *page)
{
/* XXX: check page format later */
SetPageChecked(page);
}
-static void ll_release_page(struct page *page, __u64 hash,
- __u64 start, __u64 end)
+void ll_release_page(struct page *page, int remove)
{
kunmap(page);
- lock_page(page);
- if (likely(page->mapping != NULL)) {
- ll_truncate_complete_page(page);
- unlock_page(page);
- } else {
+ if (remove) {
+ lock_page(page);
+ if (likely(page->mapping != NULL))
+ truncate_complete_page(page->mapping, page);
unlock_page(page);
- CWARN("NULL mapping page %p, truncated by others: "
- "hash("LPX64") | start("LPX64") | end("LPX64")\n",
- page, hash, start, end);
}
page_cache_release(page);
}
/*
* Find, kmap and return page that contains given hash.
*/
-static struct page *ll_dir_page_locate(struct inode *dir, __u64 hash,
+static struct page *ll_dir_page_locate(struct inode *dir, __u64 *hash,
__u64 *start, __u64 *end)
{
+ int hash64 = ll_i2sbi(dir)->ll_flags & LL_SBI_64BIT_HASH;
struct address_space *mapping = dir->i_mapping;
/*
* Complement of hash is used as an index so that
* radix_tree_gang_lookup() can be used to find a page with starting
* hash _smaller_ than one we are looking for.
*/
- unsigned long offset = hash_x_index((unsigned long)hash);
+ unsigned long offset = hash_x_index(*hash, hash64);
struct page *page;
int found;
- TREE_READ_LOCK_IRQ(mapping);
+ spin_lock_irq(&mapping->tree_lock);
found = radix_tree_gang_lookup(&mapping->page_tree,
(void **)&page, offset, 1);
if (found > 0) {
struct lu_dirpage *dp;
page_cache_get(page);
- TREE_READ_UNLOCK_IRQ(mapping);
+ spin_unlock_irq(&mapping->tree_lock);
/*
* In contrast to find_lock_page() we are sure that directory
* page cannot be truncated (while DLM lock is held) and,
* hence, can avoid restart.
*
* In fact, page cannot be locked here at all, because
- * ll_dir_readpage() does synchronous io.
+ * ll_dir_filler() does synchronous io.
*/
- wait_on_page(page);
- if (PageUptodate(page)) {
- dp = kmap(page);
- *start = le64_to_cpu(dp->ldp_hash_start);
- *end = le64_to_cpu(dp->ldp_hash_end);
- LASSERT(*start <= hash);
- if (hash > *end || (*end != *start && hash == *end)) {
- ll_release_page(page, hash, *start, *end);
+ wait_on_page_locked(page);
+ if (PageUptodate(page)) {
+ dp = kmap(page);
+ if (BITS_PER_LONG == 32 && hash64) {
+ *start = le64_to_cpu(dp->ldp_hash_start) >> 32;
+ *end = le64_to_cpu(dp->ldp_hash_end) >> 32;
+ *hash = *hash >> 32;
+ } else {
+ *start = le64_to_cpu(dp->ldp_hash_start);
+ *end = le64_to_cpu(dp->ldp_hash_end);
+ }
+ LASSERTF(*start <= *hash, "start = "LPX64",end = "
+ LPX64",hash = "LPX64"\n", *start, *end, *hash);
+ CDEBUG(D_VFSTRACE, "page %lu [%llu %llu], hash "LPU64"\n",
+ offset, *start, *end, *hash);
+ if (*hash > *end) {
+ ll_release_page(page, 0);
+ page = NULL;
+ } else if (*end != *start && *hash == *end) {
+ /*
+ * upon hash collision, remove this page,
+ * otherwise put page reference, and
+ * ll_get_dir_page() will issue RPC to fetch
+ * the page we want.
+ */
+ ll_release_page(page,
+ le32_to_cpu(dp->ldp_flags) & LDF_COLLIDE);
page = NULL;
}
} else {
}
} else {
- TREE_READ_UNLOCK_IRQ(mapping);
+ spin_unlock_irq(&mapping->tree_lock);
page = NULL;
}
return page;
}
-struct page *ll_get_dir_page(struct inode *dir, __u64 hash, int exact,
+struct page *ll_get_dir_page(struct inode *dir, __u64 hash,
struct ll_dir_chain *chain)
{
ldlm_policy_data_t policy = {.l_inodebits = {MDS_INODELOCK_UPDATE} };
int rc;
__u64 start = 0;
__u64 end = 0;
+ __u64 lhash = hash;
+ struct ll_inode_info *lli = ll_i2info(dir);
+ int hash64 = ll_i2sbi(dir)->ll_flags & LL_SBI_64BIT_HASH;
mode = LCK_PR;
rc = md_lock_match(ll_i2sbi(dir)->ll_md_exp, LDLM_FL_BLOCK_GRANTED,
ll_inode2fid(dir), LDLM_IBITS, &policy, mode, &lockh);
- if (!rc) {
- struct ldlm_enqueue_info einfo = { LDLM_IBITS, mode,
- ll_md_blocking_ast, ldlm_completion_ast,
- NULL, NULL, dir };
- struct lookup_intent it = { .it_op = IT_READDIR };
- struct ptlrpc_request *request;
- struct md_op_data *op_data;
-
- op_data = ll_prep_md_op_data(NULL, dir, NULL, NULL, 0, 0,
- LUSTRE_OPC_ANY, NULL);
- if (IS_ERR(op_data))
- return (void *)op_data;
-
- rc = md_enqueue(ll_i2sbi(dir)->ll_md_exp, &einfo, &it,
- op_data, &lockh, NULL, 0, NULL, 0);
-
- ll_finish_md_op_data(op_data);
-
- request = (struct ptlrpc_request *)it.d.lustre.it_data;
- if (request)
- ptlrpc_req_finished(request);
- if (rc < 0) {
- CERROR("lock enqueue: "DFID" at "LPU64": rc %d\n",
- PFID(ll_inode2fid(dir)), hash, rc);
- return ERR_PTR(rc);
- }
+ if (!rc) {
+ struct ldlm_enqueue_info einfo = {
+ .ei_type = LDLM_IBITS,
+ .ei_mode = mode,
+ .ei_cb_bl = ll_md_blocking_ast,
+ .ei_cb_cp = ldlm_completion_ast,
+ };
+ struct lookup_intent it = { .it_op = IT_READDIR };
+ struct ptlrpc_request *request;
+ struct md_op_data *op_data;
+
+ op_data = ll_prep_md_op_data(NULL, dir, dir, NULL, 0, 0,
+ LUSTRE_OPC_ANY, NULL);
+ if (IS_ERR(op_data))
+ return (void *)op_data;
+
+ rc = md_enqueue(ll_i2sbi(dir)->ll_md_exp, &einfo, &it,
+ op_data, &lockh, NULL, 0, NULL, 0);
+
+ ll_finish_md_op_data(op_data);
+
+ request = (struct ptlrpc_request *)it.d.lustre.it_data;
+ if (request)
+ ptlrpc_req_finished(request);
+ if (rc < 0) {
+ CERROR("lock enqueue: "DFID" at "LPU64": rc %d\n",
+ PFID(ll_inode2fid(dir)), hash, rc);
+ return ERR_PTR(rc);
+ }
+
+ CDEBUG(D_INODE, "setting lr_lvb_inode to inode "DFID"(%p)\n",
+ PFID(ll_inode2fid(dir)), dir);
+ md_set_lock_data(ll_i2sbi(dir)->ll_md_exp,
+ &it.d.lustre.it_lock_handle, dir, NULL);
+ } else {
+ /* for cross-ref object, l_ast_data of the lock may not be set,
+ * we reset it here */
+ md_set_lock_data(ll_i2sbi(dir)->ll_md_exp, &lockh.cookie,
+ dir, NULL);
}
- md_set_lock_data(ll_i2sbi(dir)->ll_md_exp, &lockh.cookie, dir, NULL);
ldlm_lock_dump_handle(D_OTHER, &lockh);
- page = ll_dir_page_locate(dir, hash, &start, &end);
+ mutex_lock(&lli->lli_readdir_mutex);
+ page = ll_dir_page_locate(dir, &lhash, &start, &end);
if (IS_ERR(page)) {
CERROR("dir page locate: "DFID" at "LPU64": rc %ld\n",
- PFID(ll_inode2fid(dir)), hash, PTR_ERR(page));
+ PFID(ll_inode2fid(dir)), lhash, PTR_ERR(page));
GOTO(out_unlock, page);
- }
-
- if (page != NULL) {
+ } else if (page != NULL) {
/*
* XXX nikita: not entirely correct handling of a corner case:
* suppose hash chain of entries with hash value HASH crosses
* it as an "overflow" page. 1. invalidate all pages at
* once. 2. use HASH|1 as an index for P1.
*/
- if (exact && hash != start) {
- /*
- * readdir asked for a page starting _exactly_ from
- * given hash, but cache contains stale page, with
- * entries with smaller hash values. Stale page should
- * be invalidated, and new one fetched.
- */
- CDEBUG(D_OTHER, "Stale readpage page %p: "LPX64" != "LPX64"\n",
- page, hash, start);
- ll_release_page(page, hash, start, end);
- } else {
- GOTO(hash_collision, page);
- }
+ GOTO(hash_collision, page);
}
- page = read_cache_page(mapping, hash_x_index((unsigned long)hash),
- (filler_t*)mapping->a_ops->readpage, NULL);
+ page = read_cache_page(mapping, hash_x_index(hash, hash64),
+ ll_dir_filler, &lhash);
if (IS_ERR(page)) {
CERROR("read cache page: "DFID" at "LPU64": rc %ld\n",
PFID(ll_inode2fid(dir)), hash, PTR_ERR(page));
GOTO(out_unlock, page);
}
- wait_on_page(page);
- (void)kmap(page);
- if (!PageUptodate(page)) {
- CERROR("page not updated: "DFID" at "LPU64": rc %d\n",
- PFID(ll_inode2fid(dir)), hash, -5);
- goto fail;
- }
+ wait_on_page_locked(page);
+ (void)kmap(page);
+ if (!PageUptodate(page)) {
+ CERROR("page not updated: "DFID" at "LPU64": rc %d\n",
+ PFID(ll_inode2fid(dir)), hash, -5);
+ goto fail;
+ }
if (!PageChecked(page))
ll_check_page(dir, page);
if (PageError(page)) {
}
hash_collision:
dp = page_address(page);
-
- start = le64_to_cpu(dp->ldp_hash_start);
- end = le64_to_cpu(dp->ldp_hash_end);
+ if (BITS_PER_LONG == 32 && hash64) {
+ start = le64_to_cpu(dp->ldp_hash_start) >> 32;
+ end = le64_to_cpu(dp->ldp_hash_end) >> 32;
+ lhash = hash >> 32;
+ } else {
+ start = le64_to_cpu(dp->ldp_hash_start);
+ end = le64_to_cpu(dp->ldp_hash_end);
+ lhash = hash;
+ }
if (end == start) {
- LASSERT(start == hash);
- CWARN("Page-wide hash collision: %#lx\n", (unsigned long)end);
+ LASSERT(start == lhash);
+ CWARN("Page-wide hash collision: "LPU64"\n", end);
+ if (BITS_PER_LONG == 32 && hash64)
+ CWARN("Real page-wide hash collision at ["LPU64" "LPU64
+ "] with hash "LPU64"\n",
+ le64_to_cpu(dp->ldp_hash_start),
+ le64_to_cpu(dp->ldp_hash_end), hash);
/*
* Fetch whole overflow chain...
*
goto fail;
}
out_unlock:
+ mutex_unlock(&lli->lli_readdir_mutex);
ldlm_lock_decref(&lockh, mode);
return page;
fail:
- ll_put_page(page);
+ ll_release_page(page, 1);
page = ERR_PTR(-EIO);
goto out_unlock;
}
-int ll_readdir(struct file *filp, void *cookie, filldir_t filldir)
+int ll_dir_read(struct inode *inode, __u64 *_pos, void *cookie,
+ filldir_t filldir)
{
- struct inode *inode = filp->f_dentry->d_inode;
- struct ll_inode_info *info = ll_i2info(inode);
- __u64 pos = filp->f_pos;
+ struct ll_inode_info *info = ll_i2info(inode);
+ struct ll_sb_info *sbi = ll_i2sbi(inode);
+ __u64 pos = *_pos;
+ int api32 = ll_need_32bit_api(sbi);
+ int hash64 = sbi->ll_flags & LL_SBI_64BIT_HASH;
struct page *page;
struct ll_dir_chain chain;
- int rc;
- int done;
- int shift;
- __u16 type;
+ int done = 0;
+ int rc = 0;
ENTRY;
- CDEBUG(D_VFSTRACE, "VFS Op:inode=%lu/%u(%p) pos %lu/%llu\n",
- inode->i_ino, inode->i_generation, inode,
- (unsigned long)pos, i_size_read(inode));
-
- if (pos == DIR_END_OFF)
- /*
- * end-of-file.
- */
- RETURN(0);
-
- rc = 0;
- done = 0;
- shift = 0;
ll_dir_chain_init(&chain);
- page = ll_get_dir_page(inode, pos, 0, &chain);
+ page = ll_get_dir_page(inode, pos, &chain);
while (rc == 0 && !done) {
struct lu_dirpage *dp;
* If page is empty (end of directory is reached),
* use this value.
*/
- __u64 hash = DIR_END_OFF;
+ __u64 hash = MDS_DIR_END_OFF;
__u64 next;
dp = page_address(page);
for (ent = lu_dirent_start(dp); ent != NULL && !done;
ent = lu_dirent_next(ent)) {
- char *name;
+ __u16 type;
int namelen;
struct lu_fid fid;
- ino_t ino;
+ __u64 lhash;
+ __u64 ino;
/*
* XXX: implement correct swabbing here.
*/
- hash = le64_to_cpu(ent->lde_hash);
- namelen = le16_to_cpu(ent->lde_namelen);
-
+ hash = le64_to_cpu(ent->lde_hash);
if (hash < pos)
/*
* Skip until we find target hash
*/
continue;
+ namelen = le16_to_cpu(ent->lde_namelen);
if (namelen == 0)
/*
* Skip dummy record.
*/
continue;
- fid = ent->lde_fid;
- name = ent->lde_name;
- fid_le_to_cpu(&fid, &fid);
- ino = cl_fid_build_ino(&fid);
+ if (api32 && hash64)
+ lhash = hash >> 32;
+ else
+ lhash = hash;
+ fid_le_to_cpu(&fid, &ent->lde_fid);
+ ino = cl_fid_build_ino(&fid, api32);
type = ll_dirent_type_get(ent);
- done = filldir(cookie, name, namelen,
- (loff_t)hash, ino, type);
+ /* For 'll_nfs_get_name_filldir()', it will try
+ * to access the 'ent' through its 'lde_name',
+ * so the parameter 'name' for 'filldir()' must
+ * be part of the 'ent'. */
+ done = filldir(cookie, ent->lde_name, namelen,
+ lhash, ino, type);
}
next = le64_to_cpu(dp->ldp_hash_end);
- ll_put_page(page);
if (!done) {
pos = next;
- if (pos == DIR_END_OFF)
+ if (pos == MDS_DIR_END_OFF) {
/*
* End of directory reached.
*/
done = 1;
- else if (1 /* chain is exhausted*/)
+ ll_release_page(page, 0);
+ } else if (1 /* chain is exhausted*/) {
/*
* Normal case: continue to the next
* page.
*/
- page = ll_get_dir_page(inode, pos, 1,
+ ll_release_page(page,
+ le32_to_cpu(dp->ldp_flags) &
+ LDF_COLLIDE);
+ next = pos;
+ page = ll_get_dir_page(inode, pos,
&chain);
- else {
+ } else {
/*
* go into overflow page.
*/
+ LASSERT(le32_to_cpu(dp->ldp_flags) &
+ LDF_COLLIDE);
+ ll_release_page(page, 1);
}
- } else
+ } else {
pos = hash;
+ ll_release_page(page, 0);
+ }
} else {
rc = PTR_ERR(page);
CERROR("error reading dir "DFID" at %lu: rc %d\n",
}
}
- filp->f_pos = (loff_t)pos;
- filp->f_version = inode->i_version;
- touch_atime(filp->f_vfsmnt, filp->f_dentry);
+ *_pos = pos;
+ ll_dir_chain_fini(&chain);
+ RETURN(rc);
+}
- ll_dir_chain_fini(&chain);
+static int ll_readdir(struct file *filp, void *cookie, filldir_t filldir)
+{
+ struct inode *inode = filp->f_dentry->d_inode;
+ struct ll_file_data *lfd = LUSTRE_FPRIVATE(filp);
+ struct ll_sb_info *sbi = ll_i2sbi(inode);
+ __u64 pos;
+ int hash64 = sbi->ll_flags & LL_SBI_64BIT_HASH;
+ int api32 = ll_need_32bit_api(sbi);
+ int rc;
+#ifdef HAVE_TOUCH_ATIME_1ARG
+ struct path path;
+#endif
+ ENTRY;
+
+ if (lfd != NULL)
+ pos = lfd->lfd_pos;
+ else
+ pos = 0;
+
+ CDEBUG(D_VFSTRACE, "VFS Op:inode="DFID"(%p) pos %lu/%llu "
+ " 32bit_api %d\n", PFID(ll_inode2fid(inode)),
+ inode, (unsigned long)pos, i_size_read(inode), api32);
+
+ if (pos == MDS_DIR_END_OFF)
+ /*
+ * end-of-file.
+ */
+ GOTO(out, rc = 0);
+
+ rc = ll_dir_read(inode, &pos, cookie, filldir);
+ if (lfd != NULL)
+ lfd->lfd_pos = pos;
+ if (pos == MDS_DIR_END_OFF) {
+ if (api32)
+ filp->f_pos = LL_DIR_END_OFF_32BIT;
+ else
+ filp->f_pos = LL_DIR_END_OFF;
+ } else {
+ if (api32 && hash64)
+ filp->f_pos = pos >> 32;
+ else
+ filp->f_pos = pos;
+ }
+ filp->f_version = inode->i_version;
+#ifdef HAVE_TOUCH_ATIME_1ARG
+#ifdef HAVE_F_PATH_MNT
+ path.mnt = filp->f_path.mnt;
+#else
+ path.mnt = filp->f_vfsmnt;
+#endif
+ path.dentry = filp->f_dentry;
+ touch_atime(&path);
+#else
+ touch_atime(filp->f_vfsmnt, filp->f_dentry);
+#endif
- RETURN(rc);
+out:
+ if (!rc)
+ ll_stats_ops_tally(sbi, LPROC_LL_READDIR, 1);
+
+ RETURN(rc);
}
int ll_send_mgc_param(struct obd_export *mgc, char *string)
return -ENOMEM;
strncpy(msp->mgs_param, string, MGS_PARAM_MAXLEN);
- rc = obd_set_info_async(mgc, sizeof(KEY_SET_INFO), KEY_SET_INFO,
+ rc = obd_set_info_async(NULL, mgc, sizeof(KEY_SET_INFO), KEY_SET_INFO,
sizeof(struct mgs_send_param), msp, NULL);
if (rc)
CERROR("Failed to set parameter: %d\n", rc);
return rc;
}
-char *ll_get_fsname(struct inode *inode)
+int ll_dir_setdirstripe(struct inode *dir, struct lmv_user_md *lump,
+ char *filename)
{
- struct lustre_sb_info *lsi = s2lsi(inode->i_sb);
- char *ptr, *fsname;
- int len;
-
- OBD_ALLOC(fsname, MGS_PARAM_MAXLEN);
- len = strlen(lsi->lsi_lmd->lmd_profile);
- ptr = strrchr(lsi->lsi_lmd->lmd_profile, '-');
- if (ptr && (strcmp(ptr, "-client") == 0))
- len -= 7;
- strncpy(fsname, lsi->lsi_lmd->lmd_profile, len);
- fsname[len] = '\0';
-
- return fsname;
+ struct ptlrpc_request *request = NULL;
+ struct md_op_data *op_data;
+ struct ll_sb_info *sbi = ll_i2sbi(dir);
+ int mode;
+ int err;
+
+ ENTRY;
+
+ mode = (0755 & (S_IRWXUGO|S_ISVTX) & ~current->fs->umask) | S_IFDIR;
+ op_data = ll_prep_md_op_data(NULL, dir, NULL, filename,
+ strlen(filename), mode, LUSTRE_OPC_MKDIR,
+ lump);
+ if (IS_ERR(op_data))
+ GOTO(err_exit, err = PTR_ERR(op_data));
+
+ op_data->op_cli_flags |= CLI_SET_MEA;
+ err = md_create(sbi->ll_md_exp, op_data, lump, sizeof(*lump), mode,
+ current_fsuid(), current_fsgid(),
+ cfs_curproc_cap_pack(), 0, &request);
+ ll_finish_md_op_data(op_data);
+ if (err)
+ GOTO(err_exit, err);
+err_exit:
+ ptlrpc_req_finished(request);
+ return err;
}
int ll_dir_setstripe(struct inode *inode, struct lov_user_md *lump,
int rc = 0;
struct lustre_sb_info *lsi = s2lsi(inode->i_sb);
struct obd_device *mgc = lsi->lsi_mgc;
- char *fsname = NULL, *param = NULL;
int lum_size;
+ ENTRY;
- /*
- * This is coming from userspace, so should be in
- * local endian. But the MDS would like it in little
- * endian, so we swab it before we send it.
- */
- switch (lump->lmm_magic) {
- case LOV_USER_MAGIC_V1: {
- if (lump->lmm_magic != cpu_to_le32(LOV_USER_MAGIC_V1))
- lustre_swab_lov_user_md_v1(lump);
- lum_size = sizeof(struct lov_user_md_v1);
- break;
+ if (lump != NULL) {
+ /*
+ * This is coming from userspace, so should be in
+ * local endian. But the MDS would like it in little
+ * endian, so we swab it before we send it.
+ */
+ switch (lump->lmm_magic) {
+ case LOV_USER_MAGIC_V1: {
+ if (lump->lmm_magic != cpu_to_le32(LOV_USER_MAGIC_V1))
+ lustre_swab_lov_user_md_v1(lump);
+ lum_size = sizeof(struct lov_user_md_v1);
+ break;
}
- case LOV_USER_MAGIC_V3: {
- if (lump->lmm_magic != cpu_to_le32(LOV_USER_MAGIC_V3))
- lustre_swab_lov_user_md_v3((struct lov_user_md_v3 *)lump);
- lum_size = sizeof(struct lov_user_md_v3);
- break;
+ case LOV_USER_MAGIC_V3: {
+ if (lump->lmm_magic != cpu_to_le32(LOV_USER_MAGIC_V3))
+ lustre_swab_lov_user_md_v3(
+ (struct lov_user_md_v3 *)lump);
+ lum_size = sizeof(struct lov_user_md_v3);
+ break;
}
- default: {
- CDEBUG(D_IOCTL, "bad userland LOV MAGIC:"
- " %#08x != %#08x nor %#08x\n",
- lump->lmm_magic, LOV_USER_MAGIC_V1,
- LOV_USER_MAGIC_V3);
- RETURN(-EINVAL);
+ default: {
+ CDEBUG(D_IOCTL, "bad userland LOV MAGIC:"
+ " %#08x != %#08x nor %#08x\n",
+ lump->lmm_magic, LOV_USER_MAGIC_V1,
+ LOV_USER_MAGIC_V3);
+ RETURN(-EINVAL);
}
+ }
+ } else {
+ lum_size = sizeof(struct lov_user_md_v1);
}
op_data = ll_prep_md_op_data(NULL, inode, NULL, NULL, 0, 0,
if (IS_ERR(op_data))
RETURN(PTR_ERR(op_data));
+ if (lump != NULL && lump->lmm_magic == cpu_to_le32(LMV_USER_MAGIC))
+ op_data->op_cli_flags |= CLI_SET_MEA;
+
/* swabbing is done in lov_setstripe() on server side */
rc = md_setattr(sbi->ll_md_exp, op_data, lump, lum_size,
NULL, 0, &req, NULL);
LOV_USER_MAGIC_V3 have the same initial fields so we do not
need the make the distiction between the 2 versions */
if (set_default && mgc->u.cli.cl_mgc_mgsexp) {
- OBD_ALLOC(param, MGS_PARAM_MAXLEN);
-
- /* Get fsname and assume devname to be -MDT0000. */
- fsname = ll_get_fsname(inode);
- /* Set root stripesize */
- sprintf(param, "%s-MDT0000.lov.stripesize=%u", fsname,
- lump->lmm_stripe_size);
- rc = ll_send_mgc_param(mgc->u.cli.cl_mgc_mgsexp, param);
- if (rc)
- goto end;
-
- /* Set root stripecount */
- sprintf(param, "%s-MDT0000.lov.stripecount=%hd", fsname,
- lump->lmm_stripe_count);
- rc = ll_send_mgc_param(mgc->u.cli.cl_mgc_mgsexp, param);
- if (rc)
- goto end;
+ char *param = NULL;
+ char *buf;
+
+ OBD_ALLOC(param, MGS_PARAM_MAXLEN);
+ if (param == NULL)
+ GOTO(end, rc = -ENOMEM);
+
+ buf = param;
+ /* Get fsname and assume devname to be -MDT0000. */
+ ll_get_fsname(inode->i_sb, buf, MTI_NAME_MAXLEN);
+ strcat(buf, "-MDT0000.lov");
+ buf += strlen(buf);
+
+ /* Set root stripesize */
+ sprintf(buf, ".stripesize=%u",
+ lump ? le32_to_cpu(lump->lmm_stripe_size) : 0);
+ rc = ll_send_mgc_param(mgc->u.cli.cl_mgc_mgsexp, param);
+ if (rc)
+ GOTO(end, rc);
+
+ /* Set root stripecount */
+ sprintf(buf, ".stripecount=%hd",
+ lump ? le16_to_cpu(lump->lmm_stripe_count) : 0);
+ rc = ll_send_mgc_param(mgc->u.cli.cl_mgc_mgsexp, param);
+ if (rc)
+ GOTO(end, rc);
+
+ /* Set root stripeoffset */
+ sprintf(buf, ".stripeoffset=%hd",
+ lump ? le16_to_cpu(lump->lmm_stripe_offset) :
+ (typeof(lump->lmm_stripe_offset))(-1));
+ rc = ll_send_mgc_param(mgc->u.cli.cl_mgc_mgsexp, param);
- /* Set root stripeoffset */
- sprintf(param, "%s-MDT0000.lov.stripeoffset=%hd", fsname,
- lump->lmm_stripe_offset);
- rc = ll_send_mgc_param(mgc->u.cli.cl_mgc_mgsexp, param);
- if (rc)
- goto end;
end:
- if (fsname)
- OBD_FREE(fsname, MGS_PARAM_MAXLEN);
- if (param)
- OBD_FREE(param, MGS_PARAM_MAXLEN);
- }
- return rc;
+ if (param != NULL)
+ OBD_FREE(param, MGS_PARAM_MAXLEN);
+ }
+ RETURN(rc);
}
int ll_dir_getstripe(struct inode *inode, struct lov_mds_md **lmmp,
struct lov_mds_md *lmm = NULL;
struct ptlrpc_request *req = NULL;
int rc, lmmsize;
- struct obd_capa *oc;
+ struct md_op_data *op_data;
rc = ll_get_max_mdsize(sbi, &lmmsize);
if (rc)
RETURN(rc);
- oc = ll_mdscapa_get(inode);
- rc = md_getattr(sbi->ll_md_exp, ll_inode2fid(inode),
- oc, OBD_MD_FLEASIZE | OBD_MD_FLDIREA,
- lmmsize, &req);
- capa_put(oc);
+ op_data = ll_prep_md_op_data(NULL, inode, NULL, NULL,
+ 0, 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(sbi->ll_md_exp, op_data, &req);
+ ll_finish_md_op_data(op_data);
if (rc < 0) {
- CDEBUG(D_INFO, "md_getattr failed on inode "
- "%lu/%u: rc %d\n", inode->i_ino,
- inode->i_generation, rc);
+ CDEBUG(D_INFO, "md_getattr failed on inode "
+ DFID": rc %d\n", PFID(ll_inode2fid(inode)), rc);
GOTO(out, rc);
}
return rc;
}
-static int ll_dir_ioctl(struct inode *inode, struct file *file,
- unsigned int cmd, unsigned long arg)
+/*
+ * Get MDT index for the inode.
+ */
+int ll_get_mdt_idx(struct inode *inode)
{
struct ll_sb_info *sbi = ll_i2sbi(inode);
+ struct md_op_data *op_data;
+ int rc, mdtidx;
+ ENTRY;
+
+ op_data = ll_prep_md_op_data(NULL, inode, NULL, NULL, 0,
+ 0, LUSTRE_OPC_ANY, NULL);
+ if (IS_ERR(op_data))
+ RETURN(PTR_ERR(op_data));
+
+ op_data->op_flags |= MF_GET_MDT_IDX;
+ rc = md_getattr(sbi->ll_md_exp, op_data, NULL);
+ mdtidx = op_data->op_mds;
+ ll_finish_md_op_data(op_data);
+ if (rc < 0) {
+ CDEBUG(D_INFO, "md_getattr_name: %d\n", rc);
+ RETURN(rc);
+ }
+ return mdtidx;
+}
+
+/**
+ * Generic handler to do any pre-copy work.
+ *
+ * It send a first hsm_progress (with extent length == 0) to coordinator as a
+ * first information for it that real work has started.
+ *
+ * Moreover, for a ARCHIVE request, it will sample the file data version and
+ * store it in \a copy.
+ *
+ * \return 0 on success.
+ */
+static int ll_ioc_copy_start(struct super_block *sb, struct hsm_copy *copy)
+{
+ struct ll_sb_info *sbi = ll_s2sbi(sb);
+ struct hsm_progress_kernel hpk;
+ int rc;
+ ENTRY;
+
+ /* Forge a hsm_progress based on data from copy. */
+ hpk.hpk_fid = copy->hc_hai.hai_fid;
+ hpk.hpk_cookie = copy->hc_hai.hai_cookie;
+ hpk.hpk_extent.offset = copy->hc_hai.hai_extent.offset;
+ hpk.hpk_extent.length = 0;
+ hpk.hpk_flags = 0;
+ hpk.hpk_errval = 0;
+ hpk.hpk_data_version = 0;
+
+
+ /* For archive request, we need to read the current file version. */
+ if (copy->hc_hai.hai_action == HSMA_ARCHIVE) {
+ struct inode *inode;
+ __u64 data_version = 0;
+
+ /* Get inode for this fid */
+ inode = search_inode_for_lustre(sb, ©->hc_hai.hai_fid);
+ if (IS_ERR(inode)) {
+ hpk.hpk_flags |= HP_FLAG_RETRY;
+ /* hpk_errval is >= 0 */
+ hpk.hpk_errval = -PTR_ERR(inode);
+ GOTO(progress, rc = PTR_ERR(inode));
+ }
+
+ /* Read current file data version */
+ rc = ll_data_version(inode, &data_version, LL_DV_RD_FLUSH);
+ iput(inode);
+ if (rc != 0) {
+ CDEBUG(D_HSM, "Could not read file data version of "
+ DFID" (rc = %d). Archive request ("
+ LPX64") could not be done.\n",
+ PFID(©->hc_hai.hai_fid), rc,
+ copy->hc_hai.hai_cookie);
+ hpk.hpk_flags |= HP_FLAG_RETRY;
+ /* hpk_errval must be >= 0 */
+ hpk.hpk_errval = -rc;
+ GOTO(progress, rc);
+ }
+
+ /* Store it the hsm_copy for later copytool use.
+ * Always modified even if no lsm. */
+ copy->hc_data_version = data_version;
+ }
+
+progress:
+ /* On error, the request should be considered as completed */
+ if (hpk.hpk_errval > 0)
+ hpk.hpk_flags |= HP_FLAG_COMPLETED;
+ rc = obd_iocontrol(LL_IOC_HSM_PROGRESS, sbi->ll_md_exp, sizeof(hpk),
+ &hpk, NULL);
+
+ RETURN(rc);
+}
+
+/**
+ * Generic handler to do any post-copy work.
+ *
+ * It will send the last hsm_progress update to coordinator to inform it
+ * that copy is finished and whether it was successful or not.
+ *
+ * Moreover,
+ * - for ARCHIVE request, it will sample the file data version and compare it
+ * with the version saved in ll_ioc_copy_start(). If they do not match, copy
+ * will be considered as failed.
+ * - for RESTORE request, it will sample the file data version and send it to
+ * coordinator which is useful if the file was imported as 'released'.
+ *
+ * \return 0 on success.
+ */
+static int ll_ioc_copy_end(struct super_block *sb, struct hsm_copy *copy)
+{
+ struct ll_sb_info *sbi = ll_s2sbi(sb);
+ struct hsm_progress_kernel hpk;
+ int rc;
+ ENTRY;
+
+ /* If you modify the logic here, also check llapi_hsm_copy_end(). */
+ /* Take care: copy->hc_hai.hai_action, len, gid and data are not
+ * initialized if copy_end was called with copy == NULL.
+ */
+
+ /* Forge a hsm_progress based on data from copy. */
+ hpk.hpk_fid = copy->hc_hai.hai_fid;
+ hpk.hpk_cookie = copy->hc_hai.hai_cookie;
+ hpk.hpk_extent = copy->hc_hai.hai_extent;
+ hpk.hpk_flags = copy->hc_flags | HP_FLAG_COMPLETED;
+ hpk.hpk_errval = copy->hc_errval;
+ hpk.hpk_data_version = 0;
+
+ /* For archive request, we need to check the file data was not changed.
+ *
+ * For restore request, we need to send the file data version, this is
+ * useful when the file was created using hsm_import.
+ */
+ if (((copy->hc_hai.hai_action == HSMA_ARCHIVE) ||
+ (copy->hc_hai.hai_action == HSMA_RESTORE)) &&
+ (copy->hc_errval == 0)) {
+ struct inode *inode;
+ __u64 data_version = 0;
+
+ /* Get lsm for this fid */
+ inode = search_inode_for_lustre(sb, ©->hc_hai.hai_fid);
+ if (IS_ERR(inode)) {
+ hpk.hpk_flags |= HP_FLAG_RETRY;
+ /* hpk_errval must be >= 0 */
+ hpk.hpk_errval = -PTR_ERR(inode);
+ GOTO(progress, rc = PTR_ERR(inode));
+ }
+
+ rc = ll_data_version(inode, &data_version, LL_DV_RD_FLUSH);
+ iput(inode);
+ if (rc) {
+ CDEBUG(D_HSM, "Could not read file data version. "
+ "Request could not be confirmed.\n");
+ if (hpk.hpk_errval == 0)
+ hpk.hpk_errval = -rc;
+ GOTO(progress, rc);
+ }
+
+ /* Store it the hsm_copy for later copytool use.
+ * Always modified even if no lsm. */
+ hpk.hpk_data_version = data_version;
+
+ /* File could have been stripped during archiving, so we need
+ * to check anyway. */
+ if ((copy->hc_hai.hai_action == HSMA_ARCHIVE) &&
+ (copy->hc_data_version != data_version)) {
+ CDEBUG(D_HSM, "File data version mismatched. "
+ "File content was changed during archiving. "
+ DFID", start:"LPX64" current:"LPX64"\n",
+ PFID(©->hc_hai.hai_fid),
+ copy->hc_data_version, data_version);
+ /* File was changed, send error to cdt. Do not ask for
+ * retry because if a file is modified frequently,
+ * the cdt will loop on retried archive requests.
+ * The policy engine will ask for a new archive later
+ * when the file will not be modified for some tunable
+ * time */
+ /* we do not notify caller */
+ hpk.hpk_flags &= ~HP_FLAG_RETRY;
+ /* hpk_errval must be >= 0 */
+ hpk.hpk_errval = EBUSY;
+ }
+
+ }
+
+progress:
+ rc = obd_iocontrol(LL_IOC_HSM_PROGRESS, sbi->ll_md_exp, sizeof(hpk),
+ &hpk, NULL);
+
+ RETURN(rc);
+}
+
+
+static int copy_and_ioctl(int cmd, struct obd_export *exp,
+ const void __user *data, size_t size)
+{
+ void *copy;
+ int rc;
+
+ OBD_ALLOC(copy, size);
+ if (copy == NULL)
+ return -ENOMEM;
+
+ if (copy_from_user(copy, data, size)) {
+ rc = -EFAULT;
+ goto out;
+ }
+
+ rc = obd_iocontrol(cmd, exp, size, copy, NULL);
+out:
+ OBD_FREE(copy, size);
+
+ return rc;
+}
+
+static int quotactl_ioctl(struct ll_sb_info *sbi, struct if_quotactl *qctl)
+{
+ int cmd = qctl->qc_cmd;
+ int type = qctl->qc_type;
+ int id = qctl->qc_id;
+ int valid = qctl->qc_valid;
+ int rc = 0;
+ ENTRY;
+
+ switch (cmd) {
+ case LUSTRE_Q_INVALIDATE:
+ case LUSTRE_Q_FINVALIDATE:
+ case Q_QUOTAON:
+ case Q_QUOTAOFF:
+ case Q_SETQUOTA:
+ case Q_SETINFO:
+ if (!cfs_capable(CFS_CAP_SYS_ADMIN) ||
+ sbi->ll_flags & LL_SBI_RMT_CLIENT)
+ RETURN(-EPERM);
+ break;
+ case Q_GETQUOTA:
+ if (((type == USRQUOTA && current_euid() != id) ||
+ (type == GRPQUOTA && !in_egroup_p(id))) &&
+ (!cfs_capable(CFS_CAP_SYS_ADMIN) ||
+ sbi->ll_flags & LL_SBI_RMT_CLIENT))
+ RETURN(-EPERM);
+ break;
+ case Q_GETINFO:
+ break;
+ default:
+ CERROR("unsupported quotactl op: %#x\n", cmd);
+ RETURN(-ENOTTY);
+ }
+
+ if (valid != QC_GENERAL) {
+ if (sbi->ll_flags & LL_SBI_RMT_CLIENT)
+ RETURN(-EOPNOTSUPP);
+
+ if (cmd == Q_GETINFO)
+ qctl->qc_cmd = Q_GETOINFO;
+ else if (cmd == Q_GETQUOTA)
+ qctl->qc_cmd = Q_GETOQUOTA;
+ else
+ RETURN(-EINVAL);
+
+ switch (valid) {
+ case QC_MDTIDX:
+ rc = obd_iocontrol(OBD_IOC_QUOTACTL, sbi->ll_md_exp,
+ sizeof(*qctl), qctl, NULL);
+ break;
+ case QC_OSTIDX:
+ rc = obd_iocontrol(OBD_IOC_QUOTACTL, sbi->ll_dt_exp,
+ sizeof(*qctl), qctl, NULL);
+ break;
+ case QC_UUID:
+ rc = obd_iocontrol(OBD_IOC_QUOTACTL, sbi->ll_md_exp,
+ sizeof(*qctl), qctl, NULL);
+ if (rc == -EAGAIN)
+ rc = obd_iocontrol(OBD_IOC_QUOTACTL,
+ sbi->ll_dt_exp,
+ sizeof(*qctl), qctl, NULL);
+ break;
+ default:
+ rc = -EINVAL;
+ break;
+ }
+
+ if (rc)
+ RETURN(rc);
+
+ qctl->qc_cmd = cmd;
+ } else {
+ struct obd_quotactl *oqctl;
+
+ OBD_ALLOC_PTR(oqctl);
+ if (oqctl == NULL)
+ RETURN(-ENOMEM);
+
+ QCTL_COPY(oqctl, qctl);
+ rc = obd_quotactl(sbi->ll_md_exp, oqctl);
+ if (rc) {
+ if (rc != -EALREADY && cmd == Q_QUOTAON) {
+ oqctl->qc_cmd = Q_QUOTAOFF;
+ obd_quotactl(sbi->ll_md_exp, oqctl);
+ }
+ OBD_FREE_PTR(oqctl);
+ RETURN(rc);
+ }
+ /* If QIF_SPACE is not set, client should collect the
+ * space usage from OSSs by itself */
+ if (cmd == Q_GETQUOTA &&
+ !(oqctl->qc_dqblk.dqb_valid & QIF_SPACE) &&
+ !oqctl->qc_dqblk.dqb_curspace) {
+ struct obd_quotactl *oqctl_tmp;
+
+ OBD_ALLOC_PTR(oqctl_tmp);
+ if (oqctl_tmp == NULL)
+ GOTO(out, rc = -ENOMEM);
+
+ oqctl_tmp->qc_cmd = Q_GETOQUOTA;
+ oqctl_tmp->qc_id = oqctl->qc_id;
+ oqctl_tmp->qc_type = oqctl->qc_type;
+
+ /* collect space usage from OSTs */
+ oqctl_tmp->qc_dqblk.dqb_curspace = 0;
+ rc = obd_quotactl(sbi->ll_dt_exp, oqctl_tmp);
+ if (!rc || rc == -EREMOTEIO) {
+ oqctl->qc_dqblk.dqb_curspace =
+ oqctl_tmp->qc_dqblk.dqb_curspace;
+ oqctl->qc_dqblk.dqb_valid |= QIF_SPACE;
+ }
+
+ /* collect space & inode usage from MDTs */
+ oqctl_tmp->qc_dqblk.dqb_curspace = 0;
+ oqctl_tmp->qc_dqblk.dqb_curinodes = 0;
+ rc = obd_quotactl(sbi->ll_md_exp, oqctl_tmp);
+ if (!rc || rc == -EREMOTEIO) {
+ oqctl->qc_dqblk.dqb_curspace +=
+ oqctl_tmp->qc_dqblk.dqb_curspace;
+ oqctl->qc_dqblk.dqb_curinodes =
+ oqctl_tmp->qc_dqblk.dqb_curinodes;
+ oqctl->qc_dqblk.dqb_valid |= QIF_INODES;
+ } else {
+ oqctl->qc_dqblk.dqb_valid &= ~QIF_SPACE;
+ }
+
+ OBD_FREE_PTR(oqctl_tmp);
+ }
+out:
+ QCTL_COPY(qctl, oqctl);
+ OBD_FREE_PTR(oqctl);
+ }
+
+ RETURN(rc);
+}
+
+static char *
+ll_getname(const char __user *filename)
+{
+ int ret = 0, len;
+ char *tmp = __getname();
+
+ if (!tmp)
+ return ERR_PTR(-ENOMEM);
+
+ len = strncpy_from_user(tmp, filename, PATH_MAX);
+ if (len == 0)
+ ret = -ENOENT;
+ else if (len > PATH_MAX)
+ ret = -ENAMETOOLONG;
+
+ if (ret) {
+ __putname(tmp);
+ tmp = ERR_PTR(ret);
+ }
+ return tmp;
+}
+
+#define ll_putname(filename) __putname(filename)
+
+static long ll_dir_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
+{
+ struct inode *inode = file->f_dentry->d_inode;
+ struct ll_sb_info *sbi = ll_i2sbi(inode);
struct obd_ioctl_data *data;
+ int rc = 0;
ENTRY;
- CDEBUG(D_VFSTRACE, "VFS Op:inode=%lu/%u(%p), cmd=%#x\n",
- inode->i_ino, inode->i_generation, inode, cmd);
+ CDEBUG(D_VFSTRACE, "VFS Op:inode="DFID"(%p), cmd=%#x\n",
+ PFID(ll_inode2fid(inode)), inode, cmd);
/* asm-ppc{,64} declares TCGETS, et. al. as type 't' not 'T' */
if (_IOC_TYPE(cmd) == 'T' || _IOC_TYPE(cmd) == 't') /* tty ioctls */
case FSFILT_IOC_SETVERSION_OLD:
case FSFILT_IOC_SETVERSION:
*/
+ case LL_IOC_GET_MDTIDX: {
+ int mdtidx;
+
+ mdtidx = ll_get_mdt_idx(inode);
+ if (mdtidx < 0)
+ RETURN(mdtidx);
+
+ if (put_user((int)mdtidx, (int*)arg))
+ RETURN(-EFAULT);
+
+ return 0;
+ }
case IOC_MDC_LOOKUP: {
struct ptlrpc_request *request = NULL;
- int namelen, rc, len = 0;
+ int namelen, len = 0;
char *buf = NULL;
char *filename;
- struct obd_capa *oc;
+ struct md_op_data *op_data;
rc = obd_ioctl_getdata(&buf, &len, (void *)arg);
if (rc)
data = (void *)buf;
filename = data->ioc_inlbuf1;
- namelen = data->ioc_inllen1;
+ namelen = strlen(filename);
if (namelen < 1) {
CDEBUG(D_INFO, "IOC_MDC_LOOKUP missing filename\n");
- GOTO(out, rc = -EINVAL);
+ GOTO(out_free, rc = -EINVAL);
}
- oc = ll_mdscapa_get(inode);
- rc = md_getattr_name(sbi->ll_md_exp, ll_inode2fid(inode), oc,
- filename, namelen, OBD_MD_FLID, 0,
- ll_i2suppgid(inode), &request);
- capa_put(oc);
+ op_data = ll_prep_md_op_data(NULL, inode, NULL, filename, namelen,
+ 0, LUSTRE_OPC_ANY, NULL);
+ if (IS_ERR(op_data))
+ GOTO(out_free, rc = PTR_ERR(op_data));
+
+ op_data->op_valid = OBD_MD_FLID;
+ rc = md_getattr_name(sbi->ll_md_exp, op_data, &request);
+ ll_finish_md_op_data(op_data);
if (rc < 0) {
CDEBUG(D_INFO, "md_getattr_name: %d\n", rc);
- GOTO(out, rc);
+ GOTO(out_free, rc);
}
-
ptlrpc_req_finished(request);
-
EXIT;
- out:
+out_free:
obd_ioctl_freedata(buf, len);
return rc;
}
+ case LL_IOC_LMV_SETSTRIPE: {
+ struct lmv_user_md *lum;
+ char *buf = NULL;
+ char *filename;
+ int namelen = 0;
+ int lumlen = 0;
+ int len;
+ int rc;
+
+ rc = obd_ioctl_getdata(&buf, &len, (void *)arg);
+ if (rc)
+ RETURN(rc);
+
+ data = (void *)buf;
+ if (data->ioc_inlbuf1 == NULL || data->ioc_inlbuf2 == NULL ||
+ data->ioc_inllen1 == 0 || data->ioc_inllen2 == 0)
+ GOTO(lmv_out_free, rc = -EINVAL);
+
+ filename = data->ioc_inlbuf1;
+ namelen = data->ioc_inllen1;
+
+ if (namelen < 1) {
+ CDEBUG(D_INFO, "IOC_MDC_LOOKUP missing filename\n");
+ GOTO(lmv_out_free, rc = -EINVAL);
+ }
+ lum = (struct lmv_user_md *)data->ioc_inlbuf2;
+ lumlen = data->ioc_inllen2;
+
+ if (lum->lum_magic != LMV_USER_MAGIC ||
+ lumlen != sizeof(*lum)) {
+ CERROR("%s: wrong lum magic %x or size %d: rc = %d\n",
+ filename, lum->lum_magic, lumlen, -EFAULT);
+ GOTO(lmv_out_free, rc = -EINVAL);
+ }
+
+ /**
+ * ll_dir_setdirstripe will be used to set dir stripe
+ * mdc_create--->mdt_reint_create (with dirstripe)
+ */
+ rc = ll_dir_setdirstripe(inode, lum, filename);
+lmv_out_free:
+ obd_ioctl_freedata(buf, len);
+ RETURN(rc);
+
+ }
case LL_IOC_LOV_SETSTRIPE: {
struct lov_user_md_v3 lumv3;
struct lov_user_md_v1 *lumv1 = (struct lov_user_md_v1 *)&lumv3;
struct lov_user_md_v1 *lumv1p = (struct lov_user_md_v1 *)arg;
struct lov_user_md_v3 *lumv3p = (struct lov_user_md_v3 *)arg;
- int rc = 0;
int set_default = 0;
LASSERT(sizeof(lumv3) == sizeof(*lumv3p));
LASSERT(sizeof(lumv3.lmm_objects[0]) ==
sizeof(lumv3p->lmm_objects[0]));
/* first try with v1 which is smaller than v3 */
- if (copy_from_user(lumv1, lumv1p, sizeof(*lumv1)))
+ if (copy_from_user(lumv1, lumv1p, sizeof(*lumv1)))
RETURN(-EFAULT);
- if (lumv1->lmm_magic == LOV_USER_MAGIC_V3) {
- if (copy_from_user(&lumv3, lumv3p, sizeof(lumv3)))
+ if ((lumv1->lmm_magic == LOV_USER_MAGIC_V3) ) {
+ if (copy_from_user(&lumv3, lumv3p, sizeof(lumv3)))
RETURN(-EFAULT);
}
RETURN(rc);
}
+ case LL_IOC_LMV_GETSTRIPE: {
+ struct lmv_user_md *lump = (struct lmv_user_md *)arg;
+ struct lmv_user_md lum;
+ struct lmv_user_md *tmp;
+ int lum_size;
+ int rc = 0;
+ int mdtindex;
+
+ if (copy_from_user(&lum, lump, sizeof(struct lmv_user_md)))
+ RETURN(-EFAULT);
+
+ if (lum.lum_magic != LMV_MAGIC_V1)
+ RETURN(-EINVAL);
+
+ lum_size = lmv_user_md_size(1, LMV_MAGIC_V1);
+ OBD_ALLOC(tmp, lum_size);
+ if (tmp == NULL)
+ GOTO(free_lmv, rc = -ENOMEM);
+
+ memcpy(tmp, &lum, sizeof(lum));
+ tmp->lum_type = LMV_STRIPE_TYPE;
+ tmp->lum_stripe_count = 1;
+ mdtindex = ll_get_mdt_idx(inode);
+ if (mdtindex < 0)
+ GOTO(free_lmv, rc = -ENOMEM);
+
+ tmp->lum_stripe_offset = mdtindex;
+ tmp->lum_objects[0].lum_mds = mdtindex;
+ memcpy(&tmp->lum_objects[0].lum_fid, ll_inode2fid(inode),
+ sizeof(struct lu_fid));
+ if (copy_to_user((void *)arg, tmp, lum_size))
+ GOTO(free_lmv, rc = -EFAULT);
+free_lmv:
+ if (tmp)
+ OBD_FREE(tmp, lum_size);
+ RETURN(rc);
+ }
+ case LL_IOC_REMOVE_ENTRY: {
+ char *filename = NULL;
+ int namelen = 0;
+ int rc;
+
+ /* Here is a little hack to avoid sending REINT_RMENTRY to
+ * unsupported server, which might crash the server(LU-2730),
+ * Because both LVB_TYPE and REINT_RMENTRY will be supported
+ * on 2.4, we use OBD_CONNECT_LVB_TYPE to detect whether the
+ * server will support REINT_RMENTRY XXX*/
+ if (!(exp_connect_flags(sbi->ll_md_exp) & OBD_CONNECT_LVB_TYPE))
+ RETURN(-ENOTSUPP);
+
+ filename = ll_getname((const char *)arg);
+ if (IS_ERR(filename))
+ RETURN(PTR_ERR(filename));
+
+ namelen = strlen(filename);
+ if (namelen < 1)
+ GOTO(out_rmdir, rc = -EINVAL);
+
+ rc = ll_rmdir_entry(inode, filename, namelen);
+out_rmdir:
+ if (filename)
+ ll_putname(filename);
+ RETURN(rc);
+ }
+ case LL_IOC_LOV_SWAP_LAYOUTS:
+ RETURN(-EPERM);
case LL_IOC_OBD_STATFS:
RETURN(ll_obd_statfs(inode, (void *)arg));
case LL_IOC_LOV_GETSTRIPE:
struct lov_mds_md *lmm = NULL;
struct mdt_body *body;
char *filename = NULL;
- int rc, lmmsize;
+ int lmmsize;
if (cmd == IOC_MDC_GETFILEINFO ||
cmd == IOC_MDC_GETFILESTRIPE) {
- filename = getname((const char *)arg);
+ filename = ll_getname((const char *)arg);
if (IS_ERR(filename))
RETURN(PTR_ERR(filename));
lmdp = (struct lov_user_mds_data *)arg;
lump = &lmdp->lmd_lmm;
}
- if (copy_to_user(lump, lmm, lmmsize))
- GOTO(out_req, rc = -EFAULT);
+ if (copy_to_user(lump, lmm, lmmsize)) {
+ if (copy_to_user(lump, lmm, sizeof(*lump)))
+ GOTO(out_req, rc = -EFAULT);
+ rc = -EOVERFLOW;
+ }
skip_lmm:
if (cmd == IOC_MDC_GETFILEINFO || cmd == LL_IOC_MDC_GETINFO) {
struct lov_user_mds_data *lmdp;
st.st_gid = body->gid;
st.st_rdev = body->rdev;
st.st_size = body->size;
- st.st_blksize = CFS_PAGE_SIZE;
+ st.st_blksize = PAGE_CACHE_SIZE;
st.st_blocks = body->blocks;
st.st_atime = body->atime;
st.st_mtime = body->mtime;
st.st_ino = inode->i_ino;
lmdp = (struct lov_user_mds_data *)arg;
- if (copy_to_user(&lmdp->lmd_st, &st, sizeof(st)))
+ if (copy_to_user(&lmdp->lmd_st, &st, sizeof(st)))
GOTO(out_req, rc = -EFAULT);
}
out_req:
ptlrpc_req_finished(request);
if (filename)
- putname(filename);
+ ll_putname(filename);
return rc;
}
case IOC_LOV_GETINFO: {
struct lov_mds_md *lmm;
int lmmsize;
lstat_t st;
- int rc;
lumd = (struct lov_user_mds_data *)arg;
lum = &lumd->lmd_lmm;
if (rc)
RETURN(rc);
- OBD_ALLOC(lmm, lmmsize);
- if (copy_from_user(lmm, lum, lmmsize))
- GOTO(free_lmm, rc = -EFAULT);
+ OBD_ALLOC_LARGE(lmm, lmmsize);
+ if (lmm == NULL)
+ RETURN(-ENOMEM);
+
+ if (copy_from_user(lmm, lum, lmmsize))
+ GOTO(free_lmm, rc = -EFAULT);
switch (lmm->lmm_magic) {
case LOV_USER_MAGIC_V1:
if (rc)
GOTO(free_lsm, rc);
- if (copy_to_user(&lumd->lmd_st, &st, sizeof(st)))
+ if (copy_to_user(&lumd->lmd_st, &st, sizeof(st)))
GOTO(free_lsm, rc = -EFAULT);
EXIT;
free_lsm:
obd_free_memmd(sbi->ll_dt_exp, &lsm);
free_lmm:
- OBD_FREE(lmm, lmmsize);
+ OBD_FREE_LARGE(lmm, lmmsize);
return rc;
}
case OBD_IOC_LLOG_CATINFO: {
- struct ptlrpc_request *req = NULL;
- char *buf = NULL;
- char *str;
- int len = 0;
- int rc;
-
- rc = obd_ioctl_getdata(&buf, &len, (void *)arg);
- if (rc)
- RETURN(rc);
- data = (void *)buf;
-
- if (!data->ioc_inlbuf1) {
- obd_ioctl_freedata(buf, len);
- RETURN(-EINVAL);
- }
-
- req = ptlrpc_request_alloc(sbi2mdc(sbi)->cl_import,
- &RQF_LLOG_CATINFO);
- if (req == NULL)
- GOTO(out_catinfo, rc = -ENOMEM);
-
- req_capsule_set_size(&req->rq_pill, &RMF_NAME, RCL_CLIENT,
- data->ioc_inllen1);
- req_capsule_set_size(&req->rq_pill, &RMF_STRING, RCL_CLIENT,
- data->ioc_inllen2);
-
- rc = ptlrpc_request_pack(req, LUSTRE_LOG_VERSION, LLOG_CATINFO);
- if (rc) {
- ptlrpc_request_free(req);
- GOTO(out_catinfo, rc);
- }
-
- str = req_capsule_client_get(&req->rq_pill, &RMF_NAME);
- memcpy(str, data->ioc_inlbuf1, data->ioc_inllen1);
- if (data->ioc_inllen2) {
- str = req_capsule_client_get(&req->rq_pill,
- &RMF_STRING);
- memcpy(str, data->ioc_inlbuf2, data->ioc_inllen2);
- }
-
- req_capsule_set_size(&req->rq_pill, &RMF_STRING, RCL_SERVER,
- data->ioc_plen1);
- ptlrpc_request_set_replen(req);
-
- rc = ptlrpc_queue_wait(req);
- if (!rc) {
- str = req_capsule_server_get(&req->rq_pill,
- &RMF_STRING);
- if (copy_to_user(data->ioc_pbuf1, str, data->ioc_plen1))
- rc = -EFAULT;
- }
- ptlrpc_req_finished(req);
- out_catinfo:
- obd_ioctl_freedata(buf, len);
- RETURN(rc);
+ RETURN(-EOPNOTSUPP);
}
case OBD_IOC_QUOTACHECK: {
struct obd_quotactl *oqctl;
- int rc, error = 0;
+ int error = 0;
if (!cfs_capable(CFS_CAP_SYS_ADMIN) ||
sbi->ll_flags & LL_SBI_RMT_CLIENT)
}
case OBD_IOC_POLL_QUOTACHECK: {
struct if_quotacheck *check;
- int rc;
if (!cfs_capable(CFS_CAP_SYS_ADMIN) ||
sbi->ll_flags & LL_SBI_RMT_CLIENT)
NULL);
if (rc) {
CDEBUG(D_QUOTA, "mdc ioctl %d failed: %d\n", cmd, rc);
- if (copy_to_user((void *)arg, check, sizeof(*check)))
- rc = -EFAULT;
+ if (copy_to_user((void *)arg, check,
+ sizeof(*check)))
+ CDEBUG(D_QUOTA, "copy_to_user failed\n");
GOTO(out_poll, rc);
}
NULL);
if (rc) {
CDEBUG(D_QUOTA, "osc ioctl %d failed: %d\n", cmd, rc);
- if (copy_to_user((void *)arg, check, sizeof(*check)))
- rc = -EFAULT;
+ if (copy_to_user((void *)arg, check,
+ sizeof(*check)))
+ CDEBUG(D_QUOTA, "copy_to_user failed\n");
GOTO(out_poll, rc);
}
out_poll:
OBD_FREE_PTR(check);
RETURN(rc);
}
- case OBD_IOC_QUOTACTL: {
- struct if_quotactl *qctl;
- int cmd, type, id, valid, rc = 0;
-
- OBD_ALLOC_PTR(qctl);
- if (!qctl)
+#if LUSTRE_VERSION_CODE < OBD_OCD_VERSION(2, 7, 50, 0)
+ case LL_IOC_QUOTACTL_18: {
+ /* copy the old 1.x quota struct for internal use, then copy
+ * back into old format struct. For 1.8 compatibility. */
+ struct if_quotactl_18 *qctl_18;
+ struct if_quotactl *qctl_20;
+
+ OBD_ALLOC_PTR(qctl_18);
+ if (!qctl_18)
RETURN(-ENOMEM);
- if (copy_from_user(qctl, (void *)arg, sizeof(*qctl)))
- GOTO(out_quotactl, rc = -EFAULT);
+ OBD_ALLOC_PTR(qctl_20);
+ if (!qctl_20)
+ GOTO(out_quotactl_18, rc = -ENOMEM);
+
+ if (copy_from_user(qctl_18, (void *)arg, sizeof(*qctl_18)))
+ GOTO(out_quotactl_20, rc = -ENOMEM);
+
+ QCTL_COPY(qctl_20, qctl_18);
+ qctl_20->qc_idx = 0;
+
+ /* XXX: dqb_valid was borrowed as a flag to mark that
+ * only mds quota is wanted */
+ if (qctl_18->qc_cmd == Q_GETQUOTA &&
+ qctl_18->qc_dqblk.dqb_valid) {
+ qctl_20->qc_valid = QC_MDTIDX;
+ qctl_20->qc_dqblk.dqb_valid = 0;
+ } else if (qctl_18->obd_uuid.uuid[0] != '\0') {
+ qctl_20->qc_valid = QC_UUID;
+ qctl_20->obd_uuid = qctl_18->obd_uuid;
+ } else {
+ qctl_20->qc_valid = QC_GENERAL;
+ }
- cmd = qctl->qc_cmd;
- type = qctl->qc_type;
- id = qctl->qc_id;
- valid = qctl->qc_valid;
-
- switch (cmd) {
- case LUSTRE_Q_INVALIDATE:
- case LUSTRE_Q_FINVALIDATE:
- case Q_QUOTAON:
- case Q_QUOTAOFF:
- case Q_SETQUOTA:
- case Q_SETINFO:
- if (!cfs_capable(CFS_CAP_SYS_ADMIN) ||
- sbi->ll_flags & LL_SBI_RMT_CLIENT)
- GOTO(out_quotactl, rc = -EPERM);
- break;
- case Q_GETQUOTA:
- if (((type == USRQUOTA && current->euid != id) ||
- (type == GRPQUOTA && !in_egroup_p(id))) &&
- (!cfs_capable(CFS_CAP_SYS_ADMIN) ||
- sbi->ll_flags & LL_SBI_RMT_CLIENT))
- GOTO(out_quotactl, rc = -EPERM);
- break;
- case Q_GETINFO:
- break;
- default:
- CERROR("unsupported quotactl op: %#x\n", cmd);
- GOTO(out_quotactl, rc = -ENOTTY);
+ rc = quotactl_ioctl(sbi, qctl_20);
+
+ if (rc == 0) {
+ QCTL_COPY(qctl_18, qctl_20);
+ qctl_18->obd_uuid = qctl_20->obd_uuid;
+
+ if (copy_to_user((void *)arg, qctl_18,
+ sizeof(*qctl_18)))
+ rc = -EFAULT;
}
- if (valid != QC_GENERAL) {
- if (sbi->ll_flags & LL_SBI_RMT_CLIENT)
- GOTO(out_quotactl, rc = -EOPNOTSUPP);
+ out_quotactl_20:
+ OBD_FREE_PTR(qctl_20);
+ out_quotactl_18:
+ OBD_FREE_PTR(qctl_18);
+ RETURN(rc);
+ }
+#else
+#warning "remove old LL_IOC_QUOTACTL_18 compatibility code"
+#endif /* LUSTRE_VERSION_CODE < OBD_OCD_VERSION(2, 7, 50, 0) */
+ case LL_IOC_QUOTACTL: {
+ struct if_quotactl *qctl;
- if (cmd == Q_GETINFO)
- qctl->qc_cmd = Q_GETOINFO;
- else if (cmd == Q_GETQUOTA)
- qctl->qc_cmd = Q_GETOQUOTA;
- else
- GOTO(out_quotactl, rc = -EINVAL);
+ OBD_ALLOC_PTR(qctl);
+ if (!qctl)
+ RETURN(-ENOMEM);
- switch (valid) {
- case QC_MDTIDX:
- rc = obd_iocontrol(OBD_IOC_QUOTACTL,
- sbi->ll_md_exp,
- sizeof(*qctl), qctl, NULL);
- break;
- case QC_OSTIDX:
- rc = obd_iocontrol(OBD_IOC_QUOTACTL,
- sbi->ll_dt_exp,
- sizeof(*qctl), qctl, NULL);
- break;
- case QC_UUID:
- rc = obd_iocontrol(OBD_IOC_QUOTACTL,
- sbi->ll_md_exp,
- sizeof(*qctl), qctl, NULL);
- if (rc == -EAGAIN)
- rc = obd_iocontrol(OBD_IOC_QUOTACTL,
- sbi->ll_dt_exp,
- sizeof(*qctl), qctl,
- NULL);
- break;
- default:
- rc = -EINVAL;
- break;
- }
+ if (copy_from_user(qctl, (void *)arg, sizeof(*qctl)))
+ GOTO(out_quotactl, rc = -EFAULT);
- if (rc)
- GOTO(out_quotactl, rc);
- else
- qctl->qc_cmd = cmd;
- } else {
- struct obd_quotactl *oqctl;
-
- OBD_ALLOC_PTR(oqctl);
- if (!oqctl)
- GOTO(out_quotactl, rc = -ENOMEM);
-
- QCTL_COPY(oqctl, qctl);
- rc = obd_quotactl(sbi->ll_md_exp, oqctl);
- if (rc) {
- if (rc != -EALREADY && cmd == Q_QUOTAON) {
- oqctl->qc_cmd = Q_QUOTAOFF;
- obd_quotactl(sbi->ll_md_exp, oqctl);
- }
- OBD_FREE_PTR(oqctl);
- GOTO(out_quotactl, rc);
- } else {
- QCTL_COPY(qctl, oqctl);
- OBD_FREE_PTR(oqctl);
- }
- }
+ rc = quotactl_ioctl(sbi, qctl);
- if (copy_to_user((void *)arg, qctl, sizeof(*qctl)))
+ if (rc == 0 && copy_to_user((void *)arg, qctl, sizeof(*qctl)))
rc = -EFAULT;
out_quotactl:
OBD_FREE_PTR(qctl);
RETURN(rc);
}
- case OBD_IOC_GETNAME: {
- struct obd_device *obd = class_exp2obd(sbi->ll_dt_exp);
- if (!obd)
- RETURN(-EFAULT);
- if (copy_to_user((void *)arg, obd->obd_name,
- strlen(obd->obd_name) + 1))
- RETURN (-EFAULT);
- RETURN(0);
- }
+ case OBD_IOC_GETDTNAME:
+ case OBD_IOC_GETMDNAME:
+ RETURN(ll_get_obd_name(inode, cmd, arg));
case LL_IOC_FLUSHCTX:
RETURN(ll_flush_ctx(inode));
#ifdef CONFIG_FS_POSIX_ACL
if (sbi->ll_flags & LL_SBI_RMT_CLIENT &&
inode == inode->i_sb->s_root->d_inode) {
struct ll_file_data *fd = LUSTRE_FPRIVATE(file);
- int rc;
- LASSERT(fd != NULL);
- rc = rct_add(&sbi->ll_rct, cfs_curproc_pid(), arg);
- if (!rc)
- fd->fd_flags |= LL_FILE_RMTACL;
- RETURN(rc);
+ LASSERT(fd != NULL);
+ rc = rct_add(&sbi->ll_rct, current_pid(), arg);
+ if (!rc)
+ fd->fd_flags |= LL_FILE_RMTACL;
+ RETURN(rc);
} else
RETURN(0);
}
#endif
case LL_IOC_GETOBDCOUNT: {
- int count;
+ int count, vallen;
+ struct obd_export *exp;
- if (copy_from_user(&count, (int *)arg, sizeof(int)))
+ if (copy_from_user(&count, (int *)arg, sizeof(int)))
RETURN(-EFAULT);
- if (!count) {
- /* get ost count */
- struct lov_obd *lov = &sbi->ll_dt_exp->exp_obd->u.lov;
- count = lov->desc.ld_tgt_count;
- } else {
- /* get mdt count */
- struct lmv_obd *lmv = &sbi->ll_md_exp->exp_obd->u.lmv;
- count = lmv->desc.ld_tgt_count;
+ /* get ost count when count is zero, get mdt count otherwise */
+ exp = count ? sbi->ll_md_exp : sbi->ll_dt_exp;
+ vallen = sizeof(count);
+ rc = obd_get_info(NULL, exp, sizeof(KEY_TGT_COUNT),
+ KEY_TGT_COUNT, &vallen, &count, NULL);
+ if (rc) {
+ CERROR("get target count failed: %d\n", rc);
+ RETURN(rc);
}
- if (copy_to_user((int *)arg, &count, sizeof(int)))
+ if (copy_to_user((int *)arg, &count, sizeof(int)))
RETURN(-EFAULT);
RETURN(0);
}
case LL_IOC_PATH2FID:
- if (copy_to_user((void *)arg, ll_inode2fid(inode),
- sizeof(struct lu_fid)))
+ if (copy_to_user((void *)arg, ll_inode2fid(inode),
+ sizeof(struct lu_fid)))
RETURN(-EFAULT);
RETURN(0);
- case OBD_IOC_CHANGELOG_CLEAR: {
- struct ioc_changelog_clear *icc;
- int rc;
-
- OBD_ALLOC_PTR(icc);
- if (icc == NULL)
- RETURN(-ENOMEM);
- if (copy_from_user(icc, (void *)arg, sizeof(*icc)))
- GOTO(icc_free, rc = -EFAULT);
+ case LL_IOC_GET_CONNECT_FLAGS: {
+ RETURN(obd_iocontrol(cmd, sbi->ll_md_exp, 0, NULL, (void*)arg));
+ }
+ case OBD_IOC_CHANGELOG_SEND:
+ case OBD_IOC_CHANGELOG_CLEAR:
+ rc = copy_and_ioctl(cmd, sbi->ll_md_exp, (void *)arg,
+ sizeof(struct ioc_changelog));
+ RETURN(rc);
+ case OBD_IOC_FID2PATH:
+ RETURN(ll_fid2path(inode, (void *)arg));
+ case LL_IOC_HSM_REQUEST: {
+ struct hsm_user_request *hur;
+ int totalsize;
+
+ OBD_ALLOC_PTR(hur);
+ if (hur == NULL)
+ RETURN(-ENOMEM);
+
+ /* We don't know the true size yet; copy the fixed-size part */
+ if (copy_from_user(hur, (void *)arg, sizeof(*hur))) {
+ OBD_FREE_PTR(hur);
+ RETURN(-EFAULT);
+ }
+
+ /* Compute the whole struct size */
+ totalsize = hur_len(hur);
+ OBD_FREE_PTR(hur);
+
+ /* Make sure the size is reasonable */
+ if (totalsize >= MDS_MAXREQSIZE)
+ RETURN(-E2BIG);
+
+ OBD_ALLOC_LARGE(hur, totalsize);
+ if (hur == NULL)
+ RETURN(-ENOMEM);
+
+ /* Copy the whole struct */
+ if (copy_from_user(hur, (void *)arg, totalsize)) {
+ OBD_FREE_LARGE(hur, totalsize);
+ RETURN(-EFAULT);
+ }
+
+ if (hur->hur_request.hr_action == HUA_RELEASE) {
+ const struct lu_fid *fid;
+ struct inode *f;
+ int i;
+
+ for (i = 0; i < hur->hur_request.hr_itemcount; i++) {
+ fid = &hur->hur_user_item[i].hui_fid;
+ f = search_inode_for_lustre(inode->i_sb, fid);
+ if (IS_ERR(f)) {
+ rc = PTR_ERR(f);
+ break;
+ }
+
+ rc = ll_hsm_release(f);
+ iput(f);
+ if (rc != 0)
+ break;
+ }
+ } else {
+ rc = obd_iocontrol(cmd, ll_i2mdexp(inode), totalsize,
+ hur, NULL);
+ }
+
+ OBD_FREE_LARGE(hur, totalsize);
+
+ RETURN(rc);
+ }
+ case LL_IOC_HSM_PROGRESS: {
+ struct hsm_progress_kernel hpk;
+ struct hsm_progress hp;
+
+ if (copy_from_user(&hp, (void *)arg, sizeof(hp)))
+ RETURN(-EFAULT);
+
+ hpk.hpk_fid = hp.hp_fid;
+ hpk.hpk_cookie = hp.hp_cookie;
+ hpk.hpk_extent = hp.hp_extent;
+ hpk.hpk_flags = hp.hp_flags;
+ hpk.hpk_errval = hp.hp_errval;
+ hpk.hpk_data_version = 0;
+
+ /* File may not exist in Lustre; all progress
+ * reported to Lustre root */
+ rc = obd_iocontrol(cmd, sbi->ll_md_exp, sizeof(hpk), &hpk,
+ NULL);
+ RETURN(rc);
+ }
+ case LL_IOC_HSM_CT_START:
+ if (!cfs_capable(CFS_CAP_SYS_ADMIN))
+ RETURN(-EPERM);
+
+ rc = copy_and_ioctl(cmd, sbi->ll_md_exp, (void *)arg,
+ sizeof(struct lustre_kernelcomm));
+ RETURN(rc);
+
+ case LL_IOC_HSM_COPY_START: {
+ struct hsm_copy *copy;
+ int rc;
+
+ OBD_ALLOC_PTR(copy);
+ if (copy == NULL)
+ RETURN(-ENOMEM);
+ if (copy_from_user(copy, (char *)arg, sizeof(*copy))) {
+ OBD_FREE_PTR(copy);
+ RETURN(-EFAULT);
+ }
+
+ rc = ll_ioc_copy_start(inode->i_sb, copy);
+ if (copy_to_user((char *)arg, copy, sizeof(*copy)))
+ rc = -EFAULT;
+
+ OBD_FREE_PTR(copy);
+ RETURN(rc);
+ }
+ case LL_IOC_HSM_COPY_END: {
+ struct hsm_copy *copy;
+ int rc;
+
+ OBD_ALLOC_PTR(copy);
+ if (copy == NULL)
+ RETURN(-ENOMEM);
+ if (copy_from_user(copy, (char *)arg, sizeof(*copy))) {
+ OBD_FREE_PTR(copy);
+ RETURN(-EFAULT);
+ }
+
+ rc = ll_ioc_copy_end(inode->i_sb, copy);
+ if (copy_to_user((char *)arg, copy, sizeof(*copy)))
+ rc = -EFAULT;
+
+ OBD_FREE_PTR(copy);
+ RETURN(rc);
+ }
+ default:
+ RETURN(obd_iocontrol(cmd, sbi->ll_dt_exp, 0, NULL,
+ (void *)arg));
+ }
+}
- rc = obd_iocontrol(cmd, sbi->ll_md_exp, sizeof(*icc), icc,NULL);
+static loff_t ll_dir_seek(struct file *file, loff_t offset, int origin)
+{
+ struct inode *inode = file->f_mapping->host;
+ struct ll_file_data *fd = LUSTRE_FPRIVATE(file);
+ struct ll_sb_info *sbi = ll_i2sbi(inode);
+ int api32 = ll_need_32bit_api(sbi);
+ loff_t ret = -EINVAL;
+ ENTRY;
-icc_free:
- OBD_FREE_PTR(icc);
- RETURN(rc);
+ mutex_lock(&inode->i_mutex);
+ switch (origin) {
+ case SEEK_SET:
+ break;
+ case SEEK_CUR:
+ offset += file->f_pos;
+ break;
+ case SEEK_END:
+ if (offset > 0)
+ GOTO(out, ret);
+ if (api32)
+ offset += LL_DIR_END_OFF_32BIT;
+ else
+ offset += LL_DIR_END_OFF;
+ break;
+ default:
+ GOTO(out, ret);
}
- case OBD_IOC_FID2PATH:
- RETURN(ll_fid2path(ll_i2mdexp(inode), (void *)arg));
- default:
- RETURN(obd_iocontrol(cmd, sbi->ll_dt_exp,0,NULL,(void *)arg));
+ if (offset >= 0 &&
+ ((api32 && offset <= LL_DIR_END_OFF_32BIT) ||
+ (!api32 && offset <= LL_DIR_END_OFF))) {
+ if (offset != file->f_pos) {
+ if ((api32 && offset == LL_DIR_END_OFF_32BIT) ||
+ (!api32 && offset == LL_DIR_END_OFF))
+ fd->lfd_pos = MDS_DIR_END_OFF;
+ else if (api32 && sbi->ll_flags & LL_SBI_64BIT_HASH)
+ fd->lfd_pos = offset << 32;
+ else
+ fd->lfd_pos = offset;
+ file->f_pos = offset;
+ file->f_version = 0;
+ }
+ ret = offset;
}
+ GOTO(out, ret);
+
+out:
+ mutex_unlock(&inode->i_mutex);
+ return ret;
}
int ll_dir_open(struct inode *inode, struct file *file)
}
struct file_operations ll_dir_operations = {
+ .llseek = ll_dir_seek,
.open = ll_dir_open,
.release = ll_dir_release,
.read = generic_read_dir,
.readdir = ll_readdir,
- .ioctl = ll_dir_ioctl
+ .unlocked_ioctl = ll_dir_ioctl,
+ .fsync = ll_fsync,
};