-/* -*- 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.
/*
* Copyright (c) 2002, 2010, Oracle and/or its affiliates. All rights reserved.
* Use is subject to license terms.
- */
-/*
- * Copyright (c) 2011 Whamcloud, Inc.
+ *
+ * Copyright (c) 2011, 2012, Whamcloud, Inc.
*/
/*
* 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.
*
*/
/* returns the page unlocked, but with a reference */
-static int ll_dir_readpage(struct file *file, struct page *page0)
+static int ll_dir_filler(void *_hash, struct page *page0)
{
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 obd_capa *oc;
- __u64 hash;
+ struct md_op_data *op_data;
+ __u64 hash = *((__u64 *)_hash);
struct page **page_pool;
struct page *page;
#ifndef HAVE_ADD_TO_PAGE_CACHE_LRU
int rc;
ENTRY;
- if (file) {
- struct ll_file_data *fd = LUSTRE_FPRIVATE(file);
-
- hash = fd->fd_dir.lfd_next;
- } else {
- hash = ll_i2info(inode)->lli_sa_pos;
- }
CDEBUG(D_VFSTRACE, "VFS Op:inode=%lu/%u(%p) hash "LPU64"\n",
inode->i_ino, inode->i_generation, inode, hash);
page_pool[npages] = page;
}
- oc = ll_mdscapa_get(inode);
- rc = md_readpage(exp, ll_inode2fid(inode), oc, hash, page_pool, npages,
- &request);
- capa_put(oc);
+ op_data = ll_prep_md_op_data(NULL, inode, NULL, NULL, 0, 0,
+ LUSTRE_OPC_ANY, NULL);
+ 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() */
GFP_KERNEL);
if (ret == 0) {
unlock_page(page);
- page_cache_get(page);
if (ll_pagevec_add(&lru_pvec, page) == 0)
ll_pagevec_lru_add_file(&lru_pvec);
} else {
return rc;
}
-#ifndef MS_HAS_NEW_AOPS
-struct address_space_operations ll_dir_aops = {
- .readpage = ll_dir_readpage,
-};
-#else
-struct address_space_operations_ext ll_dir_aops = {
- .orig_aops.readpage = ll_dir_readpage,
-};
-#endif
-
static void ll_check_page(struct inode *dir, struct page *page)
{
/* XXX: check page format later */
if (remove) {
lock_page(page);
if (likely(page->mapping != NULL))
- ll_truncate_complete_page(page);
+ truncate_complete_page(page->mapping, page);
unlock_page(page);
}
page_cache_release(page);
* 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)) {
return page;
}
-struct page *ll_get_dir_page(struct file *filp, struct inode *dir, __u64 hash,
+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} };
}
ldlm_lock_dump_handle(D_OTHER, &lockh);
- cfs_down(&lli->lli_readdir_sem);
+ cfs_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",
}
page = read_cache_page(mapping, hash_x_index(hash, hash64),
- (filler_t*)mapping->a_ops->readpage, filp);
+ 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 fail;
}
out_unlock:
- cfs_up(&lli->lli_readdir_sem);
+ cfs_mutex_unlock(&lli->lli_readdir_mutex);
ldlm_lock_decref(&lockh, mode);
return page;
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);
struct ll_sb_info *sbi = ll_i2sbi(inode);
- struct ll_file_data *fd = LUSTRE_FPRIVATE(filp);
- __u64 pos = fd->fd_dir.lfd_pos;
+ __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 done;
- int shift;
- int rc;
+ int done = 0;
+ int rc = 0;
ENTRY;
- CDEBUG(D_VFSTRACE, "VFS Op:inode=%lu/%u(%p) pos %lu/%llu 32bit_api %d\n",
- inode->i_ino, inode->i_generation, inode,
- (unsigned long)pos, i_size_read(inode), api32);
-
- if (pos == MDS_DIR_END_OFF)
- /*
- * end-of-file.
- */
- RETURN(0);
-
- rc = 0;
- done = 0;
- shift = 0;
ll_dir_chain_init(&chain);
- fd->fd_dir.lfd_next = pos;
- page = ll_get_dir_page(filp, inode, pos, &chain);
+ page = ll_get_dir_page(inode, pos, &chain);
while (rc == 0 && !done) {
struct lu_dirpage *dp;
ll_release_page(page,
le32_to_cpu(dp->ldp_flags) &
LDF_COLLIDE);
- fd->fd_dir.lfd_next = pos;
- page = ll_get_dir_page(filp, inode, pos,
+ next = pos;
+ page = ll_get_dir_page(inode, pos,
&chain);
} else {
/*
}
}
- fd->fd_dir.lfd_pos = pos;
+ *_pos = pos;
+ ll_dir_chain_fini(&chain);
+ RETURN(rc);
+}
+
+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 = lfd->lfd_pos;
+ int hash64 = sbi->ll_flags & LL_SBI_64BIT_HASH;
+ int api32 = ll_need_32bit_api(sbi);
+ int rc;
+ ENTRY;
+
+ CDEBUG(D_VFSTRACE, "VFS Op:inode=%lu/%u(%p) pos %lu/%llu "
+ " 32bit_api %d\n", inode->i_ino, inode->i_generation,
+ 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);
+ lfd->lfd_pos = pos;
if (pos == MDS_DIR_END_OFF) {
if (api32)
filp->f_pos = LL_DIR_END_OFF_32BIT;
filp->f_version = inode->i_version;
touch_atime(filp->f_vfsmnt, filp->f_dentry);
- ll_dir_chain_fini(&chain);
+out:
+ if (!rc)
+ ll_stats_ops_tally(sbi, LPROC_LL_READDIR, 1);
RETURN(rc);
}
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)
-{
- 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;
-}
-
int ll_dir_setstripe(struct inode *inode, struct lov_user_md *lump,
int set_default)
{
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;
if (lump != 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 ? le32_to_cpu(lump->lmm_stripe_size) : 0);
- 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 ? le16_to_cpu(lump->lmm_stripe_count) : 0);
- 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 ? 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);
- 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,
op_data = ll_prep_md_op_data(NULL, inode, NULL, NULL,
0, lmmsize, LUSTRE_OPC_ANY,
NULL);
- if (op_data == NULL)
- RETURN(-ENOMEM);
+ 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);
op_data = ll_prep_md_op_data(NULL, inode, NULL, NULL, 0,
0, LUSTRE_OPC_ANY, NULL);
- if (op_data == NULL)
- RETURN(-ENOMEM);
+ if (IS_ERR(op_data))
+ RETURN(PTR_ERR(op_data));
- op_data->op_valid |= OBD_MD_MDTIDX;
+ 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);
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 int ll_dir_ioctl(struct inode *inode, struct file *file,
- unsigned int cmd, unsigned long arg)
+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;
op_data = ll_prep_md_op_data(NULL, inode, NULL, filename, namelen,
0, LUSTRE_OPC_ANY, NULL);
- if (op_data == NULL)
- GOTO(out_free, rc = -ENOMEM);
+ 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);
return rc;
}
case OBD_IOC_LLOG_CATINFO: {
- struct ptlrpc_request *req = NULL;
- char *buf = NULL;
- char *str;
- int len = 0;
-
- 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 (cfs_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;
OBD_FREE_PTR(check);
RETURN(rc);
}
-#if LUSTRE_VERSION_CODE < OBD_OCD_VERSION(2,7,50,0)
+#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. */
}
#else
#warning "remove old LL_IOC_QUOTACTL_18 compatibility code"
-#endif /* LUSTRE_VERSION_CODE < OBD_OCD_VERSION(2,7,50,0) */
+#endif /* LUSTRE_VERSION_CODE < OBD_OCD_VERSION(2, 7, 50, 0) */
case LL_IOC_QUOTACTL: {
struct if_quotactl *qctl;
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 (cfs_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
/* 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(exp, sizeof(KEY_TGT_COUNT), KEY_TGT_COUNT,
- &vallen, &count, NULL);
+ 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);
sizeof(struct ioc_changelog));
RETURN(rc);
case OBD_IOC_FID2PATH:
- RETURN(ll_fid2path(ll_i2mdexp(inode), (void *)arg));
+ RETURN(ll_fid2path(inode, (void *)arg));
case LL_IOC_HSM_CT_START:
rc = copy_and_ioctl(cmd, sbi->ll_md_exp, (void *)arg,
sizeof(struct lustre_kernelcomm));
if (offset != file->f_pos) {
if ((api32 && offset == LL_DIR_END_OFF_32BIT) ||
(!api32 && offset == LL_DIR_END_OFF))
- fd->fd_dir.lfd_pos = MDS_DIR_END_OFF;
+ fd->lfd_pos = MDS_DIR_END_OFF;
else if (api32 && sbi->ll_flags & LL_SBI_64BIT_HASH)
- fd->fd_dir.lfd_pos = offset << 32;
+ fd->lfd_pos = offset << 32;
else
- fd->fd_dir.lfd_pos = offset;
+ fd->lfd_pos = offset;
file->f_pos = offset;
file->f_version = 0;
}
.release = ll_dir_release,
.read = generic_read_dir,
.readdir = ll_readdir,
- .ioctl = ll_dir_ioctl,
- .fsync = ll_fsync
+ .unlocked_ioctl = ll_dir_ioctl,
+ .fsync = ll_fsync,
};
git://git.whamcloud.com / fs / lustre-release.git / blobdiff