summaryrefslogtreecommitdiff
path: root/fs/nfs
diff options
context:
space:
mode:
Diffstat (limited to 'fs/nfs')
-rw-r--r--fs/nfs/blocklayout/extent_tree.c10
-rw-r--r--fs/nfs/callback_proc.c52
-rw-r--r--fs/nfs/dir.c34
-rw-r--r--fs/nfs/direct.c60
-rw-r--r--fs/nfs/file.c10
-rw-r--r--fs/nfs/filelayout/filelayout.c20
-rw-r--r--fs/nfs/flexfilelayout/flexfilelayout.c205
-rw-r--r--fs/nfs/flexfilelayout/flexfilelayout.h1
-rw-r--r--fs/nfs/flexfilelayout/flexfilelayoutdev.c117
-rw-r--r--fs/nfs/inode.c60
-rw-r--r--fs/nfs/internal.h41
-rw-r--r--fs/nfs/nfs3acl.c4
-rw-r--r--fs/nfs/nfs42proc.c156
-rw-r--r--fs/nfs/nfs4file.c111
-rw-r--r--fs/nfs/nfs4proc.c143
-rw-r--r--fs/nfs/nfs4sysctl.c2
-rw-r--r--fs/nfs/nfs4trace.c1
-rw-r--r--fs/nfs/nfs4trace.h431
-rw-r--r--fs/nfs/nfstrace.h1
-rw-r--r--fs/nfs/pagelist.c126
-rw-r--r--fs/nfs/pnfs.c345
-rw-r--r--fs/nfs/pnfs.h58
-rw-r--r--fs/nfs/pnfs_nfs.c10
-rw-r--r--fs/nfs/read.c43
-rw-r--r--fs/nfs/symlink.c39
-rw-r--r--fs/nfs/write.c140
26 files changed, 1468 insertions, 752 deletions
diff --git a/fs/nfs/blocklayout/extent_tree.c b/fs/nfs/blocklayout/extent_tree.c
index c59a59c37..35ab51c04 100644
--- a/fs/nfs/blocklayout/extent_tree.c
+++ b/fs/nfs/blocklayout/extent_tree.c
@@ -476,6 +476,7 @@ static void ext_tree_free_commitdata(struct nfs4_layoutcommit_args *arg,
for (i = 0; i < nr_pages; i++)
put_page(arg->layoutupdate_pages[i]);
+ vfree(arg->start_p);
kfree(arg->layoutupdate_pages);
} else {
put_page(arg->layoutupdate_page);
@@ -559,10 +560,15 @@ retry:
if (unlikely(arg->layoutupdate_pages != &arg->layoutupdate_page)) {
void *p = start_p, *end = p + arg->layoutupdate_len;
+ struct page *page = NULL;
int i = 0;
- for ( ; p < end; p += PAGE_SIZE)
- arg->layoutupdate_pages[i++] = vmalloc_to_page(p);
+ arg->start_p = start_p;
+ for ( ; p < end; p += PAGE_SIZE) {
+ page = vmalloc_to_page(p);
+ arg->layoutupdate_pages[i++] = page;
+ get_page(page);
+ }
}
dprintk("%s found %zu ranges\n", __func__, count);
diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c
index 807eb6ef4..f0939d097 100644
--- a/fs/nfs/callback_proc.c
+++ b/fs/nfs/callback_proc.c
@@ -83,8 +83,11 @@ __be32 nfs4_callback_recall(struct cb_recallargs *args, void *dummy,
res = htonl(NFS4ERR_BADHANDLE);
inode = nfs_delegation_find_inode(cps->clp, &args->fh);
- if (inode == NULL)
+ if (inode == NULL) {
+ trace_nfs4_cb_recall(cps->clp, &args->fh, NULL,
+ &args->stateid, -ntohl(res));
goto out;
+ }
/* Set up a helper thread to actually return the delegation */
switch (nfs_async_inode_return_delegation(inode, &args->stateid)) {
case 0:
@@ -96,7 +99,8 @@ __be32 nfs4_callback_recall(struct cb_recallargs *args, void *dummy,
default:
res = htonl(NFS4ERR_RESOURCE);
}
- trace_nfs4_recall_delegation(inode, -ntohl(res));
+ trace_nfs4_cb_recall(cps->clp, &args->fh, inode,
+ &args->stateid, -ntohl(res));
iput(inode);
out:
dprintk("%s: exit with status = %d\n", __func__, ntohl(res));
@@ -160,6 +164,22 @@ static struct pnfs_layout_hdr * get_layout_by_fh(struct nfs_client *clp,
return lo;
}
+/*
+ * Enforce RFC5661 section 12.5.5.2.1. (Layout Recall and Return Sequencing)
+ */
+static bool pnfs_check_stateid_sequence(struct pnfs_layout_hdr *lo,
+ const nfs4_stateid *new)
+{
+ u32 oldseq, newseq;
+
+ oldseq = be32_to_cpu(lo->plh_stateid.seqid);
+ newseq = be32_to_cpu(new->seqid);
+
+ if (newseq > oldseq + 1)
+ return false;
+ return true;
+}
+
static u32 initiate_file_draining(struct nfs_client *clp,
struct cb_layoutrecallargs *args)
{
@@ -169,34 +189,52 @@ static u32 initiate_file_draining(struct nfs_client *clp,
LIST_HEAD(free_me_list);
lo = get_layout_by_fh(clp, &args->cbl_fh, &args->cbl_stateid);
- if (!lo)
+ if (!lo) {
+ trace_nfs4_cb_layoutrecall_file(clp, &args->cbl_fh, NULL,
+ &args->cbl_stateid, -rv);
goto out;
+ }
ino = lo->plh_inode;
spin_lock(&ino->i_lock);
+ if (!pnfs_check_stateid_sequence(lo, &args->cbl_stateid)) {
+ rv = NFS4ERR_DELAY;
+ goto unlock;
+ }
pnfs_set_layout_stateid(lo, &args->cbl_stateid, true);
spin_unlock(&ino->i_lock);
pnfs_layoutcommit_inode(ino, false);
spin_lock(&ino->i_lock);
- if (test_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags) ||
- pnfs_mark_matching_lsegs_invalid(lo, &free_me_list,
- &args->cbl_range)) {
+ /*
+ * Enforce RFC5661 Section 12.5.5.2.1.5 (Bulk Recall and Return)
+ */
+ if (test_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags)) {
rv = NFS4ERR_DELAY;
goto unlock;
}
+ if (pnfs_mark_matching_lsegs_return(lo, &free_me_list,
+ &args->cbl_range)) {
+ rv = NFS4_OK;
+ goto unlock;
+ }
+
if (NFS_SERVER(ino)->pnfs_curr_ld->return_range) {
NFS_SERVER(ino)->pnfs_curr_ld->return_range(lo,
&args->cbl_range);
}
+ pnfs_mark_layout_returned_if_empty(lo);
unlock:
spin_unlock(&ino->i_lock);
pnfs_free_lseg_list(&free_me_list);
+ /* Free all lsegs that are attached to commit buckets */
+ nfs_commit_inode(ino, 0);
pnfs_put_layout_hdr(lo);
- trace_nfs4_cb_layoutrecall_inode(clp, &args->cbl_fh, ino, -rv);
+ trace_nfs4_cb_layoutrecall_file(clp, &args->cbl_fh, ino,
+ &args->cbl_stateid, -rv);
iput(ino);
out:
return rv;
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index ce5a21861..9cce67043 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -940,7 +940,7 @@ static loff_t nfs_llseek_dir(struct file *filp, loff_t offset, int whence)
dfprintk(FILE, "NFS: llseek dir(%pD2, %lld, %d)\n",
filp, offset, whence);
- mutex_lock(&inode->i_mutex);
+ inode_lock(inode);
switch (whence) {
case 1:
offset += filp->f_pos;
@@ -957,7 +957,7 @@ static loff_t nfs_llseek_dir(struct file *filp, loff_t offset, int whence)
dir_ctx->duped = 0;
}
out:
- mutex_unlock(&inode->i_mutex);
+ inode_unlock(inode);
return offset;
}
@@ -972,9 +972,9 @@ static int nfs_fsync_dir(struct file *filp, loff_t start, loff_t end,
dfprintk(FILE, "NFS: fsync dir(%pD2) datasync %d\n", filp, datasync);
- mutex_lock(&inode->i_mutex);
+ inode_lock(inode);
nfs_inc_stats(inode, NFSIOS_VFSFSYNC);
- mutex_unlock(&inode->i_mutex);
+ inode_unlock(inode);
return 0;
}
@@ -1894,15 +1894,14 @@ int nfs_symlink(struct inode *dir, struct dentry *dentry, const char *symname)
attr.ia_mode = S_IFLNK | S_IRWXUGO;
attr.ia_valid = ATTR_MODE;
- page = alloc_page(GFP_HIGHUSER);
+ page = alloc_page(GFP_USER);
if (!page)
return -ENOMEM;
- kaddr = kmap_atomic(page);
+ kaddr = page_address(page);
memcpy(kaddr, symname, pathlen);
if (pathlen < PAGE_SIZE)
memset(kaddr + pathlen, 0, PAGE_SIZE - pathlen);
- kunmap_atomic(kaddr);
trace_nfs_symlink_enter(dir, dentry);
error = NFS_PROTO(dir)->symlink(dir, dentry, page, pathlen, &attr);
@@ -2432,6 +2431,20 @@ int nfs_may_open(struct inode *inode, struct rpc_cred *cred, int openflags)
}
EXPORT_SYMBOL_GPL(nfs_may_open);
+static int nfs_execute_ok(struct inode *inode, int mask)
+{
+ struct nfs_server *server = NFS_SERVER(inode);
+ int ret;
+
+ if (mask & MAY_NOT_BLOCK)
+ ret = nfs_revalidate_inode_rcu(server, inode);
+ else
+ ret = nfs_revalidate_inode(server, inode);
+ if (ret == 0 && !execute_ok(inode))
+ ret = -EACCES;
+ return ret;
+}
+
int nfs_permission(struct inode *inode, int mask)
{
struct rpc_cred *cred;
@@ -2449,6 +2462,9 @@ int nfs_permission(struct inode *inode, int mask)
case S_IFLNK:
goto out;
case S_IFREG:
+ if ((mask & MAY_OPEN) &&
+ nfs_server_capable(inode, NFS_CAP_ATOMIC_OPEN))
+ return 0;
break;
case S_IFDIR:
/*
@@ -2481,8 +2497,8 @@ force_lookup:
res = PTR_ERR(cred);
}
out:
- if (!res && (mask & MAY_EXEC) && !execute_ok(inode))
- res = -EACCES;
+ if (!res && (mask & MAY_EXEC))
+ res = nfs_execute_ok(inode, mask);
dfprintk(VFS, "NFS: permission(%s/%lu), mask=0x%x, res=%d\n",
inode->i_sb->s_id, inode->i_ino, mask, res);
diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c
index 4b1d08f56..7a0cfd326 100644
--- a/fs/nfs/direct.c
+++ b/fs/nfs/direct.c
@@ -117,12 +117,6 @@ static inline int put_dreq(struct nfs_direct_req *dreq)
return atomic_dec_and_test(&dreq->io_count);
}
-void nfs_direct_set_resched_writes(struct nfs_direct_req *dreq)
-{
- dreq->flags = NFS_ODIRECT_RESCHED_WRITES;
-}
-EXPORT_SYMBOL_GPL(nfs_direct_set_resched_writes);
-
static void
nfs_direct_good_bytes(struct nfs_direct_req *dreq, struct nfs_pgio_header *hdr)
{
@@ -586,7 +580,7 @@ ssize_t nfs_file_direct_read(struct kiocb *iocb, struct iov_iter *iter,
if (!count)
goto out;
- mutex_lock(&inode->i_mutex);
+ inode_lock(inode);
result = nfs_sync_mapping(mapping);
if (result)
goto out_unlock;
@@ -614,7 +608,7 @@ ssize_t nfs_file_direct_read(struct kiocb *iocb, struct iov_iter *iter,
NFS_I(inode)->read_io += count;
result = nfs_direct_read_schedule_iovec(dreq, iter, pos);
- mutex_unlock(&inode->i_mutex);
+ inode_unlock(inode);
if (!result) {
result = nfs_direct_wait(dreq);
@@ -628,7 +622,7 @@ ssize_t nfs_file_direct_read(struct kiocb *iocb, struct iov_iter *iter,
out_release:
nfs_direct_req_release(dreq);
out_unlock:
- mutex_unlock(&inode->i_mutex);
+ inode_unlock(inode);
out:
return result;
}
@@ -670,6 +664,10 @@ static void nfs_direct_write_reschedule(struct nfs_direct_req *dreq)
req = nfs_list_entry(reqs.next);
nfs_direct_setup_mirroring(dreq, &desc, req);
+ if (desc.pg_error < 0) {
+ list_splice_init(&reqs, &failed);
+ goto out_failed;
+ }
list_for_each_entry_safe(req, tmp, &reqs, wb_list) {
if (!nfs_pageio_add_request(&desc, req)) {
@@ -677,13 +675,17 @@ static void nfs_direct_write_reschedule(struct nfs_direct_req *dreq)
nfs_list_add_request(req, &failed);
spin_lock(cinfo.lock);
dreq->flags = 0;
- dreq->error = -EIO;
+ if (desc.pg_error < 0)
+ dreq->error = desc.pg_error;
+ else
+ dreq->error = -EIO;
spin_unlock(cinfo.lock);
}
nfs_release_request(req);
}
nfs_pageio_complete(&desc);
+out_failed:
while (!list_empty(&failed)) {
req = nfs_list_entry(failed.next);
nfs_list_remove_request(req);
@@ -727,14 +729,20 @@ static void nfs_direct_commit_complete(struct nfs_commit_data *data)
nfs_direct_write_complete(dreq, data->inode);
}
-static void nfs_direct_error_cleanup(struct nfs_inode *nfsi)
+static void nfs_direct_resched_write(struct nfs_commit_info *cinfo,
+ struct nfs_page *req)
{
- /* There is no lock to clear */
+ struct nfs_direct_req *dreq = cinfo->dreq;
+
+ spin_lock(&dreq->lock);
+ dreq->flags = NFS_ODIRECT_RESCHED_WRITES;
+ spin_unlock(&dreq->lock);
+ nfs_mark_request_commit(req, NULL, cinfo, 0);
}
static const struct nfs_commit_completion_ops nfs_direct_commit_completion_ops = {
.completion = nfs_direct_commit_complete,
- .error_cleanup = nfs_direct_error_cleanup,
+ .resched_write = nfs_direct_resched_write,
};
static void nfs_direct_commit_schedule(struct nfs_direct_req *dreq)
@@ -839,10 +847,25 @@ static void nfs_write_sync_pgio_error(struct list_head *head)
}
}
+static void nfs_direct_write_reschedule_io(struct nfs_pgio_header *hdr)
+{
+ struct nfs_direct_req *dreq = hdr->dreq;
+
+ spin_lock(&dreq->lock);
+ if (dreq->error == 0) {
+ dreq->flags = NFS_ODIRECT_RESCHED_WRITES;
+ /* fake unstable write to let common nfs resend pages */
+ hdr->verf.committed = NFS_UNSTABLE;
+ hdr->good_bytes = hdr->args.count;
+ }
+ spin_unlock(&dreq->lock);
+}
+
static const struct nfs_pgio_completion_ops nfs_direct_write_completion_ops = {
.error_cleanup = nfs_write_sync_pgio_error,
.init_hdr = nfs_direct_pgio_init,
.completion = nfs_direct_write_completion,
+ .reschedule_io = nfs_direct_write_reschedule_io,
};
@@ -900,6 +923,11 @@ static ssize_t nfs_direct_write_schedule_iovec(struct nfs_direct_req *dreq,
}
nfs_direct_setup_mirroring(dreq, &desc, req);
+ if (desc.pg_error < 0) {
+ nfs_free_request(req);
+ result = desc.pg_error;
+ break;
+ }
nfs_lock_request(req);
req->wb_index = pos >> PAGE_SHIFT;
@@ -977,7 +1005,7 @@ ssize_t nfs_file_direct_write(struct kiocb *iocb, struct iov_iter *iter)
pos = iocb->ki_pos;
end = (pos + iov_iter_count(iter) - 1) >> PAGE_CACHE_SHIFT;
- mutex_lock(&inode->i_mutex);
+ inode_lock(inode);
result = nfs_sync_mapping(mapping);
if (result)
@@ -1017,7 +1045,7 @@ ssize_t nfs_file_direct_write(struct kiocb *iocb, struct iov_iter *iter)
pos >> PAGE_CACHE_SHIFT, end);
}
- mutex_unlock(&inode->i_mutex);
+ inode_unlock(inode);
if (!result) {
result = nfs_direct_wait(dreq);
@@ -1038,7 +1066,7 @@ ssize_t nfs_file_direct_write(struct kiocb *iocb, struct iov_iter *iter)
out_release:
nfs_direct_req_release(dreq);
out_unlock:
- mutex_unlock(&inode->i_mutex);
+ inode_unlock(inode);
return result;
}
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 93e236429..748bb813b 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -278,9 +278,9 @@ nfs_file_fsync(struct file *file, loff_t start, loff_t end, int datasync)
ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
if (ret != 0)
break;
- mutex_lock(&inode->i_mutex);
+ inode_lock(inode);
ret = nfs_file_fsync_commit(file, start, end, datasync);
- mutex_unlock(&inode->i_mutex);
+ inode_unlock(inode);
/*
* If nfs_file_fsync_commit detected a server reboot, then
* resend all dirty pages that might have been covered by
@@ -514,7 +514,7 @@ static void nfs_check_dirty_writeback(struct page *page,
* so it will not block due to pages that will shortly be freeable.
*/
nfsi = NFS_I(mapping->host);
- if (test_bit(NFS_INO_COMMIT, &nfsi->flags)) {
+ if (atomic_read(&nfsi->commit_info.rpcs_out)) {
*writeback = true;
return;
}
@@ -545,7 +545,7 @@ static int nfs_launder_page(struct page *page)
inode->i_ino, (long long)page_offset(page));
nfs_fscache_wait_on_page_write(nfsi, page);
- return nfs_wb_page(inode, page);
+ return nfs_wb_launder_page(inode, page);
}
static int nfs_swap_activate(struct swap_info_struct *sis, struct file *file,
@@ -756,7 +756,7 @@ do_unlk(struct file *filp, int cmd, struct file_lock *fl, int is_local)
l_ctx = nfs_get_lock_context(nfs_file_open_context(filp));
if (!IS_ERR(l_ctx)) {
- status = nfs_iocounter_wait(&l_ctx->io_count);
+ status = nfs_iocounter_wait(l_ctx);
nfs_put_lock_context(l_ctx);
if (status < 0)
return status;
diff --git a/fs/nfs/filelayout/filelayout.c b/fs/nfs/filelayout/filelayout.c
index 02ec07973..3384dc8e6 100644
--- a/fs/nfs/filelayout/filelayout.c
+++ b/fs/nfs/filelayout/filelayout.c
@@ -202,6 +202,7 @@ static int filelayout_async_handle_error(struct rpc_task *task,
task->tk_status);
nfs4_mark_deviceid_unavailable(devid);
pnfs_error_mark_layout_for_return(inode, lseg);
+ pnfs_set_lo_fail(lseg);
rpc_wake_up(&tbl->slot_tbl_waitq);
/* fall through */
default:
@@ -883,13 +884,19 @@ static void
filelayout_pg_init_read(struct nfs_pageio_descriptor *pgio,
struct nfs_page *req)
{
- if (!pgio->pg_lseg)
+ if (!pgio->pg_lseg) {
pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode,
req->wb_context,
0,
NFS4_MAX_UINT64,
IOMODE_READ,
GFP_KERNEL);
+ if (IS_ERR(pgio->pg_lseg)) {
+ pgio->pg_error = PTR_ERR(pgio->pg_lseg);
+ pgio->pg_lseg = NULL;
+ return;
+ }
+ }
/* If no lseg, fall back to read through mds */
if (pgio->pg_lseg == NULL)
nfs_pageio_reset_read_mds(pgio);
@@ -902,13 +909,20 @@ filelayout_pg_init_write(struct nfs_pageio_descriptor *pgio,
struct nfs_commit_info cinfo;
int status;
- if (!pgio->pg_lseg)
+ if (!pgio->pg_lseg) {
pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode,
req->wb_context,
0,
NFS4_MAX_UINT64,
IOMODE_RW,
GFP_NOFS);
+ if (IS_ERR(pgio->pg_lseg)) {
+ pgio->pg_error = PTR_ERR(pgio->pg_lseg);
+ pgio->pg_lseg = NULL;
+ return;
+ }
+ }
+
/* If no lseg, fall back to write through mds */
if (pgio->pg_lseg == NULL)
goto out_mds;
@@ -957,7 +971,7 @@ filelayout_mark_request_commit(struct nfs_page *req,
u32 i, j;
if (fl->commit_through_mds) {
- nfs_request_add_commit_list(req, &cinfo->mds->list, cinfo);
+ nfs_request_add_commit_list(req, cinfo);
} else {
/* Note that we are calling nfs4_fl_calc_j_index on each page
* that ends up being committed to a data server. An attractive
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c
index 2a2e2d8dd..0cb1abd53 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.c
+++ b/fs/nfs/flexfilelayout/flexfilelayout.c
@@ -505,9 +505,17 @@ ff_layout_alloc_lseg(struct pnfs_layout_hdr *lh,
}
p = xdr_inline_decode(&stream, 4);
- if (p)
- fls->flags = be32_to_cpup(p);
+ if (!p)
+ goto out_sort_mirrors;
+ fls->flags = be32_to_cpup(p);
+
+ p = xdr_inline_decode(&stream, 4);
+ if (!p)
+ goto out_sort_mirrors;
+ for (i=0; i < fls->mirror_array_cnt; i++)
+ fls->mirror_array[i]->report_interval = be32_to_cpup(p);
+out_sort_mirrors:
ff_layout_sort_mirrors(fls);
rc = ff_layout_check_layout(lgr);
if (rc)
@@ -603,7 +611,9 @@ nfs4_ff_layoutstat_start_io(struct nfs4_ff_layout_mirror *mirror,
mirror->start_time = now;
if (ktime_equal(mirror->last_report_time, notime))
mirror->last_report_time = now;
- if (layoutstats_timer != 0)
+ if (mirror->report_interval != 0)
+ report_interval = (s64)mirror->report_interval * 1000LL;
+ else if (layoutstats_timer != 0)
report_interval = (s64)layoutstats_timer * 1000LL;
if (ktime_to_ms(ktime_sub(now, mirror->last_report_time)) >=
report_interval) {
@@ -785,13 +795,19 @@ ff_layout_pg_init_read(struct nfs_pageio_descriptor *pgio,
int ds_idx;
/* Use full layout for now */
- if (!pgio->pg_lseg)
+ if (!pgio->pg_lseg) {
pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode,
req->wb_context,
0,
NFS4_MAX_UINT64,
IOMODE_READ,
GFP_KERNEL);
+ if (IS_ERR(pgio->pg_lseg)) {
+ pgio->pg_error = PTR_ERR(pgio->pg_lseg);
+ pgio->pg_lseg = NULL;
+ return;
+ }
+ }
/* If no lseg, fall back to read through mds */
if (pgio->pg_lseg == NULL)
goto out_mds;
@@ -825,13 +841,19 @@ ff_layout_pg_init_write(struct nfs_pageio_descriptor *pgio,
int i;
int status;
- if (!pgio->pg_lseg)
+ if (!pgio->pg_lseg) {
pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode,
req->wb_context,
0,
NFS4_MAX_UINT64,
IOMODE_RW,
GFP_NOFS);
+ if (IS_ERR(pgio->pg_lseg)) {
+ pgio->pg_error = PTR_ERR(pgio->pg_lseg);
+ pgio->pg_lseg = NULL;
+ return;
+ }
+ }
/* If no lseg, fall back to write through mds */
if (pgio->pg_lseg == NULL)
goto out_mds;
@@ -867,18 +889,25 @@ static unsigned int
ff_layout_pg_get_mirror_count_write(struct nfs_pageio_descriptor *pgio,
struct nfs_page *req)
{
- if (!pgio->pg_lseg)
+ if (!pgio->pg_lseg) {
pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode,
req->wb_context,
0,
NFS4_MAX_UINT64,
IOMODE_RW,
GFP_NOFS);
+ if (IS_ERR(pgio->pg_lseg)) {
+ pgio->pg_error = PTR_ERR(pgio->pg_lseg);
+ pgio->pg_lseg = NULL;
+ goto out;
+ }
+ }
if (pgio->pg_lseg)
return FF_LAYOUT_MIRROR_COUNT(pgio->pg_lseg);
/* no lseg means that pnfs is not in use, so no mirroring here */
nfs_pageio_reset_write_mds(pgio);
+out:
return 1;
}
@@ -912,18 +941,7 @@ static void ff_layout_reset_write(struct nfs_pgio_header *hdr, bool retry_pnfs)
hdr->args.count,
(unsigned long long)hdr->args.offset);
- if (!hdr->dreq) {
- struct nfs_open_context *ctx;
-
- ctx = nfs_list_entry(hdr->pages.next)->wb_context;
- set_bit(NFS_CONTEXT_RESEND_WRITES, &ctx->flags);
- hdr->completion_ops->error_cleanup(&hdr->pages);
- } else {
- nfs_direct_set_resched_writes(hdr->dreq);
- /* fake unstable write to let common nfs resend pages */
- hdr->verf.committed = NFS_UNSTABLE;
- hdr->good_bytes = hdr->args.count;
- }
+ hdr->completion_ops->reschedule_io(hdr);
return;
}
@@ -1101,7 +1119,7 @@ static int ff_layout_async_handle_error_v3(struct rpc_task *task,
return -NFS4ERR_RESET_TO_PNFS;
out_retry:
task->tk_status = 0;
- rpc_restart_call(task);
+ rpc_restart_call_prepare(task);
rpc_delay(task, NFS_JUKEBOX_RETRY_TIME);
return -EAGAIN;
}
@@ -1159,6 +1177,14 @@ static void ff_layout_io_track_ds_error(struct pnfs_layout_segment *lseg,
}
}
+ switch (status) {
+ case NFS4ERR_DELAY:
+ case NFS4ERR_GRACE:
+ return;
+ default:
+ break;
+ }
+
mirror = FF_LAYOUT_COMP(lseg, idx);
err = ff_layout_track_ds_error(FF_LAYOUT_FROM_HDR(lseg->pls_layout),
mirror, offset, length, status, opnum,
@@ -1189,7 +1215,7 @@ static int ff_layout_read_done_cb(struct rpc_task *task,
hdr->pgio_mirror_idx + 1,
&hdr->pgio_mirror_idx))
goto out_eagain;
- set_bit(NFS_LAYOUT_RETURN_BEFORE_CLOSE,
+ set_bit(NFS_LAYOUT_RETURN_REQUESTED,
&hdr->lseg->pls_layout->plh_flags);
pnfs_read_resend_pnfs(hdr);
return task->tk_status;
@@ -1242,14 +1268,31 @@ ff_layout_reset_to_mds(struct pnfs_layout_segment *lseg, int idx)
return ff_layout_test_devid_unavailable(node);
}
-static int ff_layout_read_prepare_common(struct rpc_task *task,
- struct nfs_pgio_header *hdr)
+static void ff_layout_read_record_layoutstats_start(struct rpc_task *task,
+ struct nfs_pgio_header *hdr)
{
+ if (test_and_set_bit(NFS_IOHDR_STAT, &hdr->flags))
+ return;
nfs4_ff_layout_stat_io_start_read(hdr->inode,
FF_LAYOUT_COMP(hdr->lseg, hdr->pgio_mirror_idx),
hdr->args.count,
task->tk_start);
+}
+
+static void ff_layout_read_record_layoutstats_done(struct rpc_task *task,
+ struct nfs_pgio_header *hdr)
+{
+ if (!test_and_clear_bit(NFS_IOHDR_STAT, &hdr->flags))
+ return;
+ nfs4_ff_layout_stat_io_end_read(task,
+ FF_LAYOUT_COMP(hdr->lseg, hdr->pgio_mirror_idx),
+ hdr->args.count,
+ hdr->res.count);
+}
+static int ff_layout_read_prepare_common(struct rpc_task *task,
+ struct nfs_pgio_header *hdr)
+{
if (unlikely(test_bit(NFS_CONTEXT_BAD, &hdr->args.context->flags))) {
rpc_exit(task, -EIO);
return -EIO;
@@ -1265,6 +1308,7 @@ static int ff_layout_read_prepare_common(struct rpc_task *task,
}
hdr->pgio_done_cb = ff_layout_read_done_cb;
+ ff_layout_read_record_layoutstats_start(task, hdr);
return 0;
}
@@ -1323,10 +1367,6 @@ static void ff_layout_read_call_done(struct rpc_task *task, void *data)
dprintk("--> %s task->tk_status %d\n", __func__, task->tk_status);
- nfs4_ff_layout_stat_io_end_read(task,
- FF_LAYOUT_COMP(hdr->lseg, hdr->pgio_mirror_idx),
- hdr->args.count, hdr->res.count);
-
if (test_bit(NFS_IOHDR_REDO, &hdr->flags) &&
task->tk_status == 0) {
nfs4_sequence_done(task, &hdr->res.seq_res);
@@ -1341,10 +1381,20 @@ static void ff_layout_read_count_stats(struct rpc_task *task, void *data)
{
struct nfs_pgio_header *hdr = data;
+ ff_layout_read_record_layoutstats_done(task, hdr);
rpc_count_iostats_metrics(task,
&NFS_CLIENT(hdr->inode)->cl_metrics[NFSPROC4_CLNT_READ]);
}
+static void ff_layout_read_release(void *data)
+{
+ struct nfs_pgio_header *hdr = data;
+
+ ff_layout_read_record_layoutstats_done(&hdr->task, hdr);
+ pnfs_generic_rw_release(data);
+}
+
+
static int ff_layout_write_done_cb(struct rpc_task *task,
struct nfs_pgio_header *hdr)
{
@@ -1362,15 +1412,12 @@ static int ff_layout_write_done_cb(struct rpc_task *task,
switch (err) {
case -NFS4ERR_RESET_TO_PNFS:
- pnfs_set_retry_layoutget(hdr->lseg->pls_layout);
ff_layout_reset_write(hdr, true);
return task->tk_status;
case -NFS4ERR_RESET_TO_MDS:
- pnfs_clear_retry_layoutget(hdr->lseg->pls_layout);
ff_layout_reset_write(hdr, false);
return task->tk_status;
case -EAGAIN:
- rpc_restart_call_prepare(task);
return -EAGAIN;
}
@@ -1402,11 +1449,9 @@ static int ff_layout_commit_done_cb(struct rpc_task *task,
switch (err) {
case -NFS4ERR_RESET_TO_PNFS:
- pnfs_set_retry_layoutget(data->lseg->pls_layout);
pnfs_generic_prepare_to_resend_writes(data);
return -EAGAIN;
case -NFS4ERR_RESET_TO_MDS:
- pnfs_clear_retry_layoutget(data->lseg->pls_layout);
pnfs_generic_prepare_to_resend_writes(data);
return -EAGAIN;
case -EAGAIN:
@@ -1421,14 +1466,31 @@ static int ff_layout_commit_done_cb(struct rpc_task *task,
return 0;
}
-static int ff_layout_write_prepare_common(struct rpc_task *task,
- struct nfs_pgio_header *hdr)
+static void ff_layout_write_record_layoutstats_start(struct rpc_task *task,
+ struct nfs_pgio_header *hdr)
{
+ if (test_and_set_bit(NFS_IOHDR_STAT, &hdr->flags))
+ return;
nfs4_ff_layout_stat_io_start_write(hdr->inode,
FF_LAYOUT_COMP(hdr->lseg, hdr->pgio_mirror_idx),
hdr->args.count,
task->tk_start);
+}
+
+static void ff_layout_write_record_layoutstats_done(struct rpc_task *task,
+ struct nfs_pgio_header *hdr)
+{
+ if (!test_and_clear_bit(NFS_IOHDR_STAT, &hdr->flags))
+ return;
+ nfs4_ff_layout_stat_io_end_write(task,
+ FF_LAYOUT_COMP(hdr->lseg, hdr->pgio_mirror_idx),
+ hdr->args.count, hdr->res.count,
+ hdr->res.verf->committed);
+}
+static int ff_layout_write_prepare_common(struct rpc_task *task,
+ struct nfs_pgio_header *hdr)
+{
if (unlikely(test_bit(NFS_CONTEXT_BAD, &hdr->args.context->flags))) {
rpc_exit(task, -EIO);
return -EIO;
@@ -1445,6 +1507,7 @@ static int ff_layout_write_prepare_common(struct rpc_task *task,
return -EAGAIN;
}
+ ff_layout_write_record_layoutstats_start(task, hdr);
return 0;
}
@@ -1480,11 +1543,6 @@ static void ff_layout_write_call_done(struct rpc_task *task, void *data)
{
struct nfs_pgio_header *hdr = data;
- nfs4_ff_layout_stat_io_end_write(task,
- FF_LAYOUT_COMP(hdr->lseg, hdr->pgio_mirror_idx),
- hdr->args.count, hdr->res.count,
- hdr->res.verf->committed);
-
if (test_bit(NFS_IOHDR_REDO, &hdr->flags) &&
task->tk_status == 0) {
nfs4_sequence_done(task, &hdr->res.seq_res);
@@ -1499,18 +1557,53 @@ static void ff_layout_write_count_stats(struct rpc_task *task, void *data)
{
struct nfs_pgio_header *hdr = data;
+ ff_layout_write_record_layoutstats_done(task, hdr);
rpc_count_iostats_metrics(task,
&NFS_CLIENT(hdr->inode)->cl_metrics[NFSPROC4_CLNT_WRITE]);
}
-static void ff_layout_commit_prepare_common(struct rpc_task *task,
+static void ff_layout_write_release(void *data)
+{
+ struct nfs_pgio_header *hdr = data;
+
+ ff_layout_write_record_layoutstats_done(&hdr->task, hdr);
+ pnfs_generic_rw_release(data);
+}
+
+static void ff_layout_commit_record_layoutstats_start(struct rpc_task *task,
struct nfs_commit_data *cdata)
{
+ if (test_and_set_bit(NFS_IOHDR_STAT, &cdata->flags))
+ return;
nfs4_ff_layout_stat_io_start_write(cdata->inode,
FF_LAYOUT_COMP(cdata->lseg, cdata->ds_commit_index),
0, task->tk_start);
}
+static void ff_layout_commit_record_layoutstats_done(struct rpc_task *task,
+ struct nfs_commit_data *cdata)
+{
+ struct nfs_page *req;
+ __u64 count = 0;
+
+ if (!test_and_clear_bit(NFS_IOHDR_STAT, &cdata->flags))
+ return;
+
+ if (task->tk_status == 0) {
+ list_for_each_entry(req, &cdata->pages, wb_list)
+ count += req->wb_bytes;
+ }
+ nfs4_ff_layout_stat_io_end_write(task,
+ FF_LAYOUT_COMP(cdata->lseg, cdata->ds_commit_index),
+ count, count, NFS_FILE_SYNC);
+}
+
+static void ff_layout_commit_prepare_common(struct rpc_task *task,
+ struct nfs_commit_data *cdata)
+{
+ ff_layout_commit_record_layoutstats_start(task, cdata);
+}
+
static void ff_layout_commit_prepare_v3(struct rpc_task *task, void *data)
{
ff_layout_commit_prepare_common(task, data);
@@ -1531,19 +1624,6 @@ static void ff_layout_commit_prepare_v4(struct rpc_task *task, void *data)
static void ff_layout_commit_done(struct rpc_task *task, void *data)
{
- struct nfs_commit_data *cdata = data;
- struct nfs_page *req;
- __u64 count = 0;
-
- if (task->tk_status == 0) {
- list_for_each_entry(req, &cdata->pages, wb_list)
- count += req->wb_bytes;
- }
-
- nfs4_ff_layout_stat_io_end_write(task,
- FF_LAYOUT_COMP(cdata->lseg, cdata->ds_commit_index),
- count, count, NFS_FILE_SYNC);
-
pnfs_generic_write_commit_done(task, data);
}
@@ -1551,50 +1631,59 @@ static void ff_layout_commit_count_stats(struct rpc_task *task, void *data)
{
struct nfs_commit_data *cdata = data;
+ ff_layout_commit_record_layoutstats_done(task, cdata);
rpc_count_iostats_metrics(task,
&NFS_CLIENT(cdata->inode)->cl_metrics[NFSPROC4_CLNT_COMMIT]);
}
+static void ff_layout_commit_release(void *data)
+{
+ struct nfs_commit_data *cdata = data;
+
+ ff_layout_commit_record_layoutstats_done(&cdata->task, cdata);
+ pnfs_generic_commit_release(data);
+}
+
static const struct rpc_call_ops ff_layout_read_call_ops_v3 = {
.rpc_call_prepare = ff_layout_read_prepare_v3,
.rpc_call_done = ff_layout_read_call_done,
.rpc_count_stats = ff_layout_read_count_stats,
- .rpc_release = pnfs_generic_rw_release,
+ .rpc_release = ff_layout_read_release,
};
static const struct rpc_call_ops ff_layout_read_call_ops_v4 = {
.rpc_call_prepare = ff_layout_read_prepare_v4,
.rpc_call_done = ff_layout_read_call_done,
.rpc_count_stats = ff_layout_read_count_stats,
- .rpc_release = pnfs_generic_rw_release,
+ .rpc_release = ff_layout_read_release,
};
static const struct rpc_call_ops ff_layout_write_call_ops_v3 = {
.rpc_call_prepare = ff_layout_write_prepare_v3,
.rpc_call_done = ff_layout_write_call_done,
.rpc_count_stats = ff_layout_write_count_stats,
- .rpc_release = pnfs_generic_rw_release,
+ .rpc_release = ff_layout_write_release,
};
static const struct rpc_call_ops ff_layout_write_call_ops_v4 = {
.rpc_call_prepare = ff_layout_write_prepare_v4,
.rpc_call_done = ff_layout_write_call_done,
.rpc_count_stats = ff_layout_write_count_stats,
- .rpc_release = pnfs_generic_rw_release,
+ .rpc_release = ff_layout_write_release,
};
static const struct rpc_call_ops ff_layout_commit_call_ops_v3 = {
.rpc_call_prepare = ff_layout_commit_prepare_v3,
.rpc_call_done = ff_layout_commit_done,
.rpc_count_stats = ff_layout_commit_count_stats,
- .rpc_release = pnfs_generic_commit_release,
+ .rpc_release = ff_layout_commit_release,
};
static const struct rpc_call_ops ff_layout_commit_call_ops_v4 = {
.rpc_call_prepare = ff_layout_commit_prepare_v4,
.rpc_call_done = ff_layout_commit_done,
.rpc_count_stats = ff_layout_commit_count_stats,
- .rpc_release = pnfs_generic_commit_release,
+ .rpc_release = ff_layout_commit_release,
};
static enum pnfs_try_status
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.h b/fs/nfs/flexfilelayout/flexfilelayout.h
index 2bb08bc6a..dd353bb7d 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.h
+++ b/fs/nfs/flexfilelayout/flexfilelayout.h
@@ -85,6 +85,7 @@ struct nfs4_ff_layout_mirror {
struct nfs4_ff_layoutstat write_stat;
ktime_t start_time;
ktime_t last_report_time;
+ u32 report_interval;
};
struct nfs4_ff_layout_segment {
diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c
index e125e55de..eb370460c 100644
--- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c
+++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c
@@ -218,63 +218,55 @@ static void extend_ds_error(struct nfs4_ff_layout_ds_err *err,
err->length = end - err->offset;
}
-static bool ds_error_can_merge(struct nfs4_ff_layout_ds_err *err, u64 offset,
- u64 length, int status, enum nfs_opnum4 opnum,
- nfs4_stateid *stateid,
- struct nfs4_deviceid *deviceid)
+static int
+ff_ds_error_match(const struct nfs4_ff_layout_ds_err *e1,
+ const struct nfs4_ff_layout_ds_err *e2)
{
- return err->status == status && err->opnum == opnum &&
- nfs4_stateid_match(&err->stateid, stateid) &&
- !memcmp(&err->deviceid, deviceid, sizeof(*deviceid)) &&
- end_offset(err->offset, err->length) >= offset &&
- err->offset <= end_offset(offset, length);
-}
-
-static bool merge_ds_error(struct nfs4_ff_layout_ds_err *old,
- struct nfs4_ff_layout_ds_err *new)
-{
- if (!ds_error_can_merge(old, new->offset, new->length, new->status,
- new->opnum, &new->stateid, &new->deviceid))
- return false;
-
- extend_ds_error(old, new->offset, new->length);
- return true;
+ int ret;
+
+ if (e1->opnum != e2->opnum)
+ return e1->opnum < e2->opnum ? -1 : 1;
+ if (e1->status != e2->status)
+ return e1->status < e2->status ? -1 : 1;
+ ret = memcmp(&e1->stateid, &e2->stateid, sizeof(e1->stateid));
+ if (ret != 0)
+ return ret;
+ ret = memcmp(&e1->deviceid, &e2->deviceid, sizeof(e1->deviceid));
+ if (ret != 0)
+ return ret;
+ if (end_offset(e1->offset, e1->length) < e2->offset)
+ return -1;
+ if (e1->offset > end_offset(e2->offset, e2->length))
+ return 1;
+ /* If ranges overlap or are contiguous, they are the same */
+ return 0;
}
-static bool
+static void
ff_layout_add_ds_error_locked(struct nfs4_flexfile_layout *flo,
struct nfs4_ff_layout_ds_err *dserr)
{
- struct nfs4_ff_layout_ds_err *err;
-
- list_for_each_entry(err, &flo->error_list, list) {
- if (merge_ds_error(err, dserr)) {
- return true;
- }
- }
-
- list_add(&dserr->list, &flo->error_list);
- return false;
-}
-
-static bool
-ff_layout_update_ds_error(struct nfs4_flexfile_layout *flo, u64 offset,
- u64 length, int status, enum nfs_opnum4 opnum,
- nfs4_stateid *stateid, struct nfs4_deviceid *deviceid)
-{
- bool found = false;
- struct nfs4_ff_layout_ds_err *err;
-
- list_for_each_entry(err, &flo->error_list, list) {
- if (ds_error_can_merge(err, offset, length, status, opnum,
- stateid, deviceid)) {
- found = true;
- extend_ds_error(err, offset, length);
+ struct nfs4_ff_layout_ds_err *err, *tmp;
+ struct list_head *head = &flo->error_list;
+ int match;
+
+ /* Do insertion sort w/ merges */
+ list_for_each_entry_safe(err, tmp, &flo->error_list, list) {
+ match = ff_ds_error_match(err, dserr);
+ if (match < 0)
+ continue;
+ if (match > 0) {
+ /* Add entry "dserr" _before_ entry "err" */
+ head = &err->list;
break;
}
+ /* Entries match, so merge "err" into "dserr" */
+ extend_ds_error(dserr, err->offset, err->length);
+ list_del(&err->list);
+ kfree(err);
}
- return found;
+ list_add_tail(&dserr->list, head);
}
int ff_layout_track_ds_error(struct nfs4_flexfile_layout *flo,
@@ -283,7 +275,6 @@ int ff_layout_track_ds_error(struct nfs4_flexfile_layout *flo,
gfp_t gfp_flags)
{
struct nfs4_ff_layout_ds_err *dserr;
- bool needfree;
if (status == 0)
return 0;
@@ -291,14 +282,6 @@ int ff_layout_track_ds_error(struct nfs4_flexfile_layout *flo,
if (mirror->mirror_ds == NULL)
return -EINVAL;
- spin_lock(&flo->generic_hdr.plh_inode->i_lock);
- if (ff_layout_update_ds_error(flo, offset, length, status, opnum,
- &mirror->stateid,
- &mirror->mirror_ds->id_node.deviceid)) {
- spin_unlock(&flo->generic_hdr.plh_inode->i_lock);
- return 0;
- }
- spin_unlock(&flo->generic_hdr.plh_inode->i_lock);
dserr = kmalloc(sizeof(*dserr), gfp_flags);
if (!dserr)
return -ENOMEM;
@@ -313,10 +296,8 @@ int ff_layout_track_ds_error(struct nfs4_flexfile_layout *flo,
NFS4_DEVICEID4_SIZE);
spin_lock(&flo->generic_hdr.plh_inode->i_lock);
- needfree = ff_layout_add_ds_error_locked(flo, dserr);
+ ff_layout_add_ds_error_locked(flo, dserr);
spin_unlock(&flo->generic_hdr.plh_inode->i_lock);
- if (needfree)
- kfree(dserr);
return 0;
}
@@ -429,22 +410,14 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, u32 ds_idx,
mirror, lseg->pls_range.offset,
lseg->pls_range.length, NFS4ERR_NXIO,
OP_ILLEGAL, GFP_NOIO);
- if (fail_return) {
- pnfs_error_mark_layout_for_return(ino, lseg);
+ if (!fail_return) {
if (ff_layout_has_available_ds(lseg))
- pnfs_set_retry_layoutget(lseg->pls_layout);
- else
- pnfs_clear_retry_layoutget(lseg->pls_layout);
-
- } else {
- if (ff_layout_has_available_ds(lseg))
- set_bit(NFS_LAYOUT_RETURN_BEFORE_CLOSE,
+ set_bit(NFS_LAYOUT_RETURN_REQUESTED,
&lseg->pls_layout->plh_flags);
- else {
+ else
pnfs_error_mark_layout_for_return(ino, lseg);
- pnfs_clear_retry_layoutget(lseg->pls_layout);
- }
- }
+ } else
+ pnfs_error_mark_layout_for_return(ino, lseg);
}
out_update_creds:
if (ff_layout_update_mirror_cred(mirror, ds))
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 3e2071a17..86faecf8f 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -71,19 +71,25 @@ nfs_fattr_to_ino_t(struct nfs_fattr *fattr)
return nfs_fileid_to_ino_t(fattr->fileid);
}
-/**
- * nfs_wait_bit_killable - helper for functions that are sleeping on bit locks
- * @word: long word containing the bit lock
- */
-int nfs_wait_bit_killable(struct wait_bit_key *key, int mode)
+static int nfs_wait_killable(int mode)
{
freezable_schedule_unsafe();
if (signal_pending_state(mode, current))
return -ERESTARTSYS;
return 0;
}
+
+int nfs_wait_bit_killable(struct wait_bit_key *key, int mode)
+{
+ return nfs_wait_killable(mode);
+}
EXPORT_SYMBOL_GPL(nfs_wait_bit_killable);
+int nfs_wait_atomic_killable(atomic_t *p)
+{
+ return nfs_wait_killable(TASK_KILLABLE);
+}
+
/**
* nfs_compat_user_ino64 - returns the user-visible inode number
* @fileid: 64-bit fileid
@@ -408,9 +414,10 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr, st
inode->i_fop = NULL;
inode->i_flags |= S_AUTOMOUNT;
}
- } else if (S_ISLNK(inode->i_mode))
+ } else if (S_ISLNK(inode->i_mode)) {
inode->i_op = &nfs_symlink_inode_operations;
- else
+ inode_nohighmem(inode);
+ } else
init_special_inode(inode, inode->i_mode, fattr->rdev);
memset(&inode->i_atime, 0, sizeof(inode->i_atime));
@@ -654,9 +661,9 @@ int nfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
trace_nfs_getattr_enter(inode);
/* Flush out writes to the server in order to update c/mtime. */
if (S_ISREG(inode->i_mode)) {
- mutex_lock(&inode->i_mutex);
+ inode_lock(inode);
err = nfs_sync_inode(inode);
- mutex_unlock(&inode->i_mutex);
+ inode_unlock(inode);
if (err)
goto out;
}
@@ -699,7 +706,7 @@ static void nfs_init_lock_context(struct nfs_lock_context *l_ctx)
l_ctx->lockowner.l_owner = current->files;
l_ctx->lockowner.l_pid = current->tgid;
INIT_LIST_HEAD(&l_ctx->list);
- nfs_iocounter_init(&l_ctx->io_count);
+ atomic_set(&l_ctx->io_count, 0);
}
static struct nfs_lock_context *__nfs_find_lock_context(struct nfs_open_context *ctx)
@@ -912,6 +919,12 @@ void nfs_file_clear_open_context(struct file *filp)
if (ctx) {
struct inode *inode = d_inode(ctx->dentry);
+ /*
+ * We fatal error on write before. Try to writeback
+ * every page again.
+ */
+ if (ctx->error < 0)
+ invalidate_inode_pages2(inode->i_mapping);
filp->private_data = NULL;
spin_lock(&inode->i_lock);
list_move_tail(&ctx->list, &NFS_I(inode)->open_files);
@@ -1086,6 +1099,27 @@ static bool nfs_mapping_need_revalidate_inode(struct inode *inode)
|| NFS_STALE(inode);
}
+int nfs_revalidate_mapping_rcu(struct inode *inode)
+{
+ struct nfs_inode *nfsi = NFS_I(inode);
+ unsigned long *bitlock = &nfsi->flags;
+ int ret = 0;
+
+ if (IS_SWAPFILE(inode))
+ goto out;
+ if (nfs_mapping_need_revalidate_inode(inode)) {
+ ret = -ECHILD;
+ goto out;
+ }
+ spin_lock(&inode->i_lock);
+ if (test_bit(NFS_INO_INVALIDATING, bitlock) ||
+ (nfsi->cache_validity & NFS_INO_INVALID_DATA))
+ ret = -ECHILD;
+ spin_unlock(&inode->i_lock);
+out:
+ return ret;
+}
+
/**
* __nfs_revalidate_mapping - Revalidate the pagecache
* @inode - pointer to host inode
@@ -1144,9 +1178,9 @@ static int __nfs_revalidate_mapping(struct inode *inode,
spin_unlock(&inode->i_lock);
trace_nfs_invalidate_mapping_enter(inode);
if (may_lock) {
- mutex_lock(&inode->i_mutex);
+ inode_lock(inode);
ret = nfs_invalidate_mapping(inode, mapping);
- mutex_unlock(&inode->i_mutex);
+ inode_unlock(inode);
} else
ret = nfs_invalidate_mapping(inode, mapping);
trace_nfs_invalidate_mapping_exit(inode, ret);
@@ -1935,7 +1969,7 @@ static int __init nfs_init_inodecache(void)
nfs_inode_cachep = kmem_cache_create("nfs_inode_cache",
sizeof(struct nfs_inode),
0, (SLAB_RECLAIM_ACCOUNT|
- SLAB_MEM_SPREAD),
+ SLAB_MEM_SPREAD|SLAB_ACCOUNT),
init_once);
if (nfs_inode_cachep == NULL)
return -ENOMEM;
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index 9dea85f7f..9a547aa3e 100644
--- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -238,7 +238,7 @@ extern void nfs_pgheader_init(struct nfs_pageio_descriptor *desc,
struct nfs_pgio_header *hdr,
void (*release)(struct nfs_pgio_header *hdr));
void nfs_set_pgio_error(struct nfs_pgio_header *hdr, int error, loff_t pos);
-int nfs_iocounter_wait(struct nfs_io_counter *c);
+int nfs_iocounter_wait(struct nfs_lock_context *l_ctx);
extern const struct nfs_pageio_ops nfs_pgio_rw_ops;
struct nfs_pgio_header *nfs_pgio_header_alloc(const struct nfs_rw_ops *);
@@ -252,18 +252,18 @@ void nfs_free_request(struct nfs_page *req);
struct nfs_pgio_mirror *
nfs_pgio_current_mirror(struct nfs_pageio_descriptor *desc);
-static inline void nfs_iocounter_init(struct nfs_io_counter *c)
-{
- c->flags = 0;
- atomic_set(&c->io_count, 0);
-}
-
static inline bool nfs_pgio_has_mirroring(struct nfs_pageio_descriptor *desc)
{
WARN_ON_ONCE(desc->pg_mirror_count < 1);
return desc->pg_mirror_count > 1;
}
+static inline bool nfs_match_open_context(const struct nfs_open_context *ctx1,
+ const struct nfs_open_context *ctx2)
+{
+ return ctx1->cred == ctx2->cred && ctx1->state == ctx2->state;
+}
+
/* nfs2xdr.c */
extern struct rpc_procinfo nfs_procedures[];
extern int nfs2_decode_dirent(struct xdr_stream *,
@@ -380,6 +380,7 @@ extern void nfs_clear_inode(struct inode *);
extern void nfs_evict_inode(struct inode *);
void nfs_zap_acl_cache(struct inode *inode);
extern int nfs_wait_bit_killable(struct wait_bit_key *key, int mode);
+extern int nfs_wait_atomic_killable(atomic_t *p);
/* super.c */
extern const struct super_operations nfs_sops;
@@ -483,7 +484,7 @@ void nfs_retry_commit(struct list_head *page_list,
struct nfs_commit_info *cinfo,
u32 ds_commit_idx);
void nfs_commitdata_release(struct nfs_commit_data *data);
-void nfs_request_add_commit_list(struct nfs_page *req, struct list_head *dst,
+void nfs_request_add_commit_list(struct nfs_page *req,
struct nfs_commit_info *cinfo);
void nfs_request_add_commit_list_locked(struct nfs_page *req,
struct list_head *dst,
@@ -519,7 +520,6 @@ static inline void nfs_inode_dio_wait(struct inode *inode)
inode_dio_wait(inode);
}
extern ssize_t nfs_dreq_bytes_left(struct nfs_direct_req *dreq);
-extern void nfs_direct_set_resched_writes(struct nfs_direct_req *dreq);
/* nfs4proc.c */
extern void __nfs4_read_done_cb(struct nfs_pgio_header *);
@@ -696,9 +696,32 @@ static inline u32 nfs_fhandle_hash(const struct nfs_fh *fh)
{
return ~crc32_le(0xFFFFFFFF, &fh->data[0], fh->size);
}
+static inline u32 nfs_stateid_hash(const nfs4_stateid *stateid)
+{
+ return ~crc32_le(0xFFFFFFFF, &stateid->other[0],
+ NFS4_STATEID_OTHER_SIZE);
+}
#else
static inline u32 nfs_fhandle_hash(const struct nfs_fh *fh)
{
return 0;
}
+static inline u32 nfs_stateid_hash(nfs4_stateid *stateid)
+{
+ return 0;
+}
#endif
+
+static inline bool nfs_error_is_fatal(int err)
+{
+ switch (err) {
+ case -ERESTARTSYS:
+ case -EIO:
+ case -ENOSPC:
+ case -EROFS:
+ case -E2BIG:
+ return true;
+ default:
+ return false;
+ }
+}
diff --git a/fs/nfs/nfs3acl.c b/fs/nfs/nfs3acl.c
index 1ebe2fc7c..17c0fa1ec 100644
--- a/fs/nfs/nfs3acl.c
+++ b/fs/nfs/nfs3acl.c
@@ -284,12 +284,12 @@ nfs3_listxattr(struct dentry *dentry, char *data, size_t size)
int error;
error = nfs3_list_one_acl(inode, ACL_TYPE_ACCESS,
- POSIX_ACL_XATTR_ACCESS, data, size, &result);
+ XATTR_NAME_POSIX_ACL_ACCESS, data, size, &result);
if (error)
return error;
error = nfs3_list_one_acl(inode, ACL_TYPE_DEFAULT,
- POSIX_ACL_XATTR_DEFAULT, data, size, &result);
+ XATTR_NAME_POSIX_ACL_DEFAULT, data, size, &result);
if (error)
return error;
return result;
diff --git a/fs/nfs/nfs42proc.c b/fs/nfs/nfs42proc.c
index 6b1ce9825..dff83460e 100644
--- a/fs/nfs/nfs42proc.c
+++ b/fs/nfs/nfs42proc.c
@@ -16,29 +16,8 @@
#define NFSDBG_FACILITY NFSDBG_PROC
-static int nfs42_set_rw_stateid(nfs4_stateid *dst, struct file *file,
- fmode_t fmode)
-{
- struct nfs_open_context *open;
- struct nfs_lock_context *lock;
- int ret;
-
- open = get_nfs_open_context(nfs_file_open_context(file));
- lock = nfs_get_lock_context(open);
- if (IS_ERR(lock)) {
- put_nfs_open_context(open);
- return PTR_ERR(lock);
- }
-
- ret = nfs4_set_rw_stateid(dst, open, lock, fmode);
-
- nfs_put_lock_context(lock);
- put_nfs_open_context(open);
- return ret;
-}
-
static int _nfs42_proc_fallocate(struct rpc_message *msg, struct file *filep,
- loff_t offset, loff_t len)
+ struct nfs_lock_context *lock, loff_t offset, loff_t len)
{
struct inode *inode = file_inode(filep);
struct nfs_server *server = NFS_SERVER(inode);
@@ -56,7 +35,8 @@ static int _nfs42_proc_fallocate(struct rpc_message *msg, struct file *filep,
msg->rpc_argp = &args;
msg->rpc_resp = &res;
- status = nfs42_set_rw_stateid(&args.falloc_stateid, filep, FMODE_WRITE);
+ status = nfs4_set_rw_stateid(&args.falloc_stateid, lock->open_context,
+ lock, FMODE_WRITE);
if (status)
return status;
@@ -78,15 +58,26 @@ static int nfs42_proc_fallocate(struct rpc_message *msg, struct file *filep,
{
struct nfs_server *server = NFS_SERVER(file_inode(filep));
struct nfs4_exception exception = { };
+ struct nfs_lock_context *lock;
int err;
+ lock = nfs_get_lock_context(nfs_file_open_context(filep));
+ if (IS_ERR(lock))
+ return PTR_ERR(lock);
+
+ exception.inode = file_inode(filep);
+ exception.state = lock->open_context->state;
+
do {
- err = _nfs42_proc_fallocate(msg, filep, offset, len);
- if (err == -ENOTSUPP)
- return -EOPNOTSUPP;
+ err = _nfs42_proc_fallocate(msg, filep, lock, offset, len);
+ if (err == -ENOTSUPP) {
+ err = -EOPNOTSUPP;
+ break;
+ }
err = nfs4_handle_exception(server, err, &exception);
} while (exception.retry);
+ nfs_put_lock_context(lock);
return err;
}
@@ -101,13 +92,13 @@ int nfs42_proc_allocate(struct file *filep, loff_t offset, loff_t len)
if (!nfs_server_capable(inode, NFS_CAP_ALLOCATE))
return -EOPNOTSUPP;
- mutex_lock(&inode->i_mutex);
+ inode_lock(inode);
err = nfs42_proc_fallocate(&msg, filep, offset, len);
if (err == -EOPNOTSUPP)
NFS_SERVER(inode)->caps &= ~NFS_CAP_ALLOCATE;
- mutex_unlock(&inode->i_mutex);
+ inode_unlock(inode);
return err;
}
@@ -123,7 +114,7 @@ int nfs42_proc_deallocate(struct file *filep, loff_t offset, loff_t len)
return -EOPNOTSUPP;
nfs_wb_all(inode);
- mutex_lock(&inode->i_mutex);
+ inode_lock(inode);
err = nfs42_proc_fallocate(&msg, filep, offset, len);
if (err == 0)
@@ -131,11 +122,12 @@ int nfs42_proc_deallocate(struct file *filep, loff_t offset, loff_t len)
if (err == -EOPNOTSUPP)
NFS_SERVER(inode)->caps &= ~NFS_CAP_DEALLOCATE;
- mutex_unlock(&inode->i_mutex);
+ inode_unlock(inode);
return err;
}
-static loff_t _nfs42_proc_llseek(struct file *filep, loff_t offset, int whence)
+static loff_t _nfs42_proc_llseek(struct file *filep,
+ struct nfs_lock_context *lock, loff_t offset, int whence)
{
struct inode *inode = file_inode(filep);
struct nfs42_seek_args args = {
@@ -156,7 +148,8 @@ static loff_t _nfs42_proc_llseek(struct file *filep, loff_t offset, int whence)
if (!nfs_server_capable(inode, NFS_CAP_SEEK))
return -ENOTSUPP;
- status = nfs42_set_rw_stateid(&args.sa_stateid, filep, FMODE_READ);
+ status = nfs4_set_rw_stateid(&args.sa_stateid, lock->open_context,
+ lock, FMODE_READ);
if (status)
return status;
@@ -175,17 +168,28 @@ loff_t nfs42_proc_llseek(struct file *filep, loff_t offset, int whence)
{
struct nfs_server *server = NFS_SERVER(file_inode(filep));
struct nfs4_exception exception = { };
+ struct nfs_lock_context *lock;
loff_t err;
+ lock = nfs_get_lock_context(nfs_file_open_context(filep));
+ if (IS_ERR(lock))
+ return PTR_ERR(lock);
+
+ exception.inode = file_inode(filep);
+ exception.state = lock->open_context->state;
+
do {
- err = _nfs42_proc_llseek(filep, offset, whence);
+ err = _nfs42_proc_llseek(filep, lock, offset, whence);
if (err >= 0)
break;
- if (err == -ENOTSUPP)
- return -EOPNOTSUPP;
+ if (err == -ENOTSUPP) {
+ err = -EOPNOTSUPP;
+ break;
+ }
err = nfs4_handle_exception(server, err, &exception);
} while (exception.retry);
+ nfs_put_lock_context(lock);
return err;
}
@@ -204,6 +208,8 @@ static void
nfs42_layoutstat_done(struct rpc_task *task, void *calldata)
{
struct nfs42_layoutstat_data *data = calldata;
+ struct inode *inode = data->inode;
+ struct pnfs_layout_hdr *lo;
if (!nfs4_sequence_done(task, &data->res.seq_res))
return;
@@ -211,12 +217,35 @@ nfs42_layoutstat_done(struct rpc_task *task, void *calldata)
switch (task->tk_status) {
case 0:
break;
+ case -NFS4ERR_EXPIRED:
+ case -NFS4ERR_STALE_STATEID:
+ case -NFS4ERR_OLD_STATEID:
+ case -NFS4ERR_BAD_STATEID:
+ spin_lock(&inode->i_lock);
+ lo = NFS_I(inode)->layout;
+ if (lo && nfs4_stateid_match(&data->args.stateid,
+ &lo->plh_stateid)) {
+ LIST_HEAD(head);
+
+ /*
+ * Mark the bad layout state as invalid, then retry
+ * with the current stateid.
+ */
+ set_bit(NFS_LAYOUT_INVALID_STID, &lo->plh_flags);
+ pnfs_mark_matching_lsegs_invalid(lo, &head, NULL);
+ spin_unlock(&inode->i_lock);
+ pnfs_free_lseg_list(&head);
+ } else
+ spin_unlock(&inode->i_lock);
+ break;
case -ENOTSUPP:
case -EOPNOTSUPP:
- NFS_SERVER(data->inode)->caps &= ~NFS_CAP_LAYOUTSTATS;
+ NFS_SERVER(inode)->caps &= ~NFS_CAP_LAYOUTSTATS;
default:
- dprintk("%s server returns %d\n", __func__, task->tk_status);
+ break;
}
+
+ dprintk("%s server returns %d\n", __func__, task->tk_status);
}
static void
@@ -273,8 +302,9 @@ int nfs42_proc_layoutstats_generic(struct nfs_server *server,
}
static int _nfs42_proc_clone(struct rpc_message *msg, struct file *src_f,
- struct file *dst_f, loff_t src_offset,
- loff_t dst_offset, loff_t count)
+ struct file *dst_f, struct nfs_lock_context *src_lock,
+ struct nfs_lock_context *dst_lock, loff_t src_offset,
+ loff_t dst_offset, loff_t count)
{
struct inode *src_inode = file_inode(src_f);
struct inode *dst_inode = file_inode(dst_f);
@@ -295,11 +325,13 @@ static int _nfs42_proc_clone(struct rpc_message *msg, struct file *src_f,
msg->rpc_argp = &args;
msg->rpc_resp = &res;
- status = nfs42_set_rw_stateid(&args.src_stateid, src_f, FMODE_READ);
+ status = nfs4_set_rw_stateid(&args.src_stateid, src_lock->open_context,
+ src_lock, FMODE_READ);
if (status)
return status;
- status = nfs42_set_rw_stateid(&args.dst_stateid, dst_f, FMODE_WRITE);
+ status = nfs4_set_rw_stateid(&args.dst_stateid, dst_lock->open_context,
+ dst_lock, FMODE_WRITE);
if (status)
return status;
@@ -324,22 +356,48 @@ int nfs42_proc_clone(struct file *src_f, struct file *dst_f,
};
struct inode *inode = file_inode(src_f);
struct nfs_server *server = NFS_SERVER(file_inode(src_f));
- struct nfs4_exception exception = { };
- int err;
+ struct nfs_lock_context *src_lock;
+ struct nfs_lock_context *dst_lock;
+ struct nfs4_exception src_exception = { };
+ struct nfs4_exception dst_exception = { };
+ int err, err2;
if (!nfs_server_capable(inode, NFS_CAP_CLONE))
return -EOPNOTSUPP;
+ src_lock = nfs_get_lock_context(nfs_file_open_context(src_f));
+ if (IS_ERR(src_lock))
+ return PTR_ERR(src_lock);
+
+ src_exception.inode = file_inode(src_f);
+ src_exception.state = src_lock->open_context->state;
+
+ dst_lock = nfs_get_lock_context(nfs_file_open_context(dst_f));
+ if (IS_ERR(dst_lock)) {
+ err = PTR_ERR(dst_lock);
+ goto out_put_src_lock;
+ }
+
+ dst_exception.inode = file_inode(dst_f);
+ dst_exception.state = dst_lock->open_context->state;
+
do {
- err = _nfs42_proc_clone(&msg, src_f, dst_f, src_offset,
- dst_offset, count);
+ err = _nfs42_proc_clone(&msg, src_f, dst_f, src_lock, dst_lock,
+ src_offset, dst_offset, count);
if (err == -ENOTSUPP || err == -EOPNOTSUPP) {
NFS_SERVER(inode)->caps &= ~NFS_CAP_CLONE;
- return -EOPNOTSUPP;
+ err = -EOPNOTSUPP;
+ break;
}
- err = nfs4_handle_exception(server, err, &exception);
- } while (exception.retry);
- return err;
+ err2 = nfs4_handle_exception(server, err, &src_exception);
+ err = nfs4_handle_exception(server, err, &dst_exception);
+ if (!err)
+ err = err2;
+ } while (src_exception.retry || dst_exception.retry);
+ nfs_put_lock_context(dst_lock);
+out_put_src_lock:
+ nfs_put_lock_context(src_lock);
+ return err;
}
diff --git a/fs/nfs/nfs4file.c b/fs/nfs/nfs4file.c
index db9b5fea5..57ca1c803 100644
--- a/fs/nfs/nfs4file.c
+++ b/fs/nfs/nfs4file.c
@@ -141,11 +141,11 @@ nfs4_file_fsync(struct file *file, loff_t start, loff_t end, int datasync)
ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
if (ret != 0)
break;
- mutex_lock(&inode->i_mutex);
+ inode_lock(inode);
ret = nfs_file_fsync_commit(file, start, end, datasync);
if (!ret)
ret = pnfs_sync_inode(inode, !!datasync);
- mutex_unlock(&inode->i_mutex);
+ inode_unlock(inode);
/*
* If nfs_file_fsync_commit detected a server reboot, then
* resend all dirty pages that might have been covered by
@@ -195,75 +195,37 @@ static long nfs42_fallocate(struct file *filep, int mode, loff_t offset, loff_t
return nfs42_proc_allocate(filep, offset, len);
}
-static noinline long
-nfs42_ioctl_clone(struct file *dst_file, unsigned long srcfd,
- u64 src_off, u64 dst_off, u64 count)
+static int nfs42_clone_file_range(struct file *src_file, loff_t src_off,
+ struct file *dst_file, loff_t dst_off, u64 count)
{
struct inode *dst_inode = file_inode(dst_file);
struct nfs_server *server = NFS_SERVER(dst_inode);
- struct fd src_file;
- struct inode *src_inode;
+ struct inode *src_inode = file_inode(src_file);
unsigned int bs = server->clone_blksize;
bool same_inode = false;
int ret;
- /* dst file must be opened for writing */
- if (!(dst_file->f_mode & FMODE_WRITE))
- return -EINVAL;
-
- ret = mnt_want_write_file(dst_file);
- if (ret)
- return ret;
-
- src_file = fdget(srcfd);
- if (!src_file.file) {
- ret = -EBADF;
- goto out_drop_write;
- }
-
- src_inode = file_inode(src_file.file);
-
- if (src_inode == dst_inode)
- same_inode = true;
-
- /* src file must be opened for reading */
- if (!(src_file.file->f_mode & FMODE_READ))
- goto out_fput;
-
- /* src and dst must be regular files */
- ret = -EISDIR;
- if (!S_ISREG(src_inode->i_mode) || !S_ISREG(dst_inode->i_mode))
- goto out_fput;
-
- ret = -EXDEV;
- if (src_file.file->f_path.mnt != dst_file->f_path.mnt ||
- src_inode->i_sb != dst_inode->i_sb)
- goto out_fput;
-
/* check alignment w.r.t. clone_blksize */
ret = -EINVAL;
if (bs) {
if (!IS_ALIGNED(src_off, bs) || !IS_ALIGNED(dst_off, bs))
- goto out_fput;
+ goto out;
if (!IS_ALIGNED(count, bs) && i_size_read(src_inode) != (src_off + count))
- goto out_fput;
+ goto out;
}
- /* verify if ranges are overlapped within the same file */
- if (same_inode) {
- if (dst_off + count > src_off && dst_off < src_off + count)
- goto out_fput;
- }
+ if (src_inode == dst_inode)
+ same_inode = true;
/* XXX: do we lock at all? what if server needs CB_RECALL_LAYOUT? */
if (same_inode) {
- mutex_lock(&src_inode->i_mutex);
+ inode_lock(src_inode);
} else if (dst_inode < src_inode) {
- mutex_lock_nested(&dst_inode->i_mutex, I_MUTEX_PARENT);
- mutex_lock_nested(&src_inode->i_mutex, I_MUTEX_CHILD);
+ inode_lock_nested(dst_inode, I_MUTEX_PARENT);
+ inode_lock_nested(src_inode, I_MUTEX_CHILD);
} else {
- mutex_lock_nested(&src_inode->i_mutex, I_MUTEX_PARENT);
- mutex_lock_nested(&dst_inode->i_mutex, I_MUTEX_CHILD);
+ inode_lock_nested(src_inode, I_MUTEX_PARENT);
+ inode_lock_nested(dst_inode, I_MUTEX_CHILD);
}
/* flush all pending writes on both src and dst so that server
@@ -275,7 +237,7 @@ nfs42_ioctl_clone(struct file *dst_file, unsigned long srcfd,
if (ret)
goto out_unlock;
- ret = nfs42_proc_clone(src_file.file, dst_file, src_off, dst_off, count);
+ ret = nfs42_proc_clone(src_file, dst_file, src_off, dst_off, count);
/* truncate inode page cache of the dst range so that future reads can fetch
* new data from server */
@@ -284,45 +246,17 @@ nfs42_ioctl_clone(struct file *dst_file, unsigned long srcfd,
out_unlock:
if (same_inode) {
- mutex_unlock(&src_inode->i_mutex);
+ inode_unlock(src_inode);
} else if (dst_inode < src_inode) {
- mutex_unlock(&src_inode->i_mutex);
- mutex_unlock(&dst_inode->i_mutex);
+ inode_unlock(src_inode);
+ inode_unlock(dst_inode);
} else {
- mutex_unlock(&dst_inode->i_mutex);
- mutex_unlock(&src_inode->i_mutex);
+ inode_unlock(dst_inode);
+ inode_unlock(src_inode);
}
-out_fput:
- fdput(src_file);
-out_drop_write:
- mnt_drop_write_file(dst_file);
+out:
return ret;
}
-
-static long nfs42_ioctl_clone_range(struct file *dst_file, void __user *argp)
-{
- struct btrfs_ioctl_clone_range_args args;
-
- if (copy_from_user(&args, argp, sizeof(args)))
- return -EFAULT;
-
- return nfs42_ioctl_clone(dst_file, args.src_fd, args.src_offset,
- args.dest_offset, args.src_length);
-}
-
-long nfs4_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
-{
- void __user *argp = (void __user *)arg;
-
- switch (cmd) {
- case BTRFS_IOC_CLONE:
- return nfs42_ioctl_clone(file, arg, 0, 0, 0);
- case BTRFS_IOC_CLONE_RANGE:
- return nfs42_ioctl_clone_range(file, argp);
- }
-
- return -ENOTTY;
-}
#endif /* CONFIG_NFS_V4_2 */
const struct file_operations nfs4_file_operations = {
@@ -342,8 +276,7 @@ const struct file_operations nfs4_file_operations = {
#ifdef CONFIG_NFS_V4_2
.llseek = nfs4_file_llseek,
.fallocate = nfs42_fallocate,
- .unlocked_ioctl = nfs4_ioctl,
- .compat_ioctl = nfs4_ioctl,
+ .clone_file_range = nfs42_clone_file_range,
#else
.llseek = nfs_file_llseek,
#endif
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 98a441573..14881594d 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -208,6 +208,9 @@ static const u32 nfs4_pnfs_open_bitmap[3] = {
| FATTR4_WORD1_TIME_METADATA
| FATTR4_WORD1_TIME_MODIFY,
FATTR4_WORD2_MDSTHRESHOLD
+#ifdef CONFIG_NFS_V4_SECURITY_LABEL
+ | FATTR4_WORD2_SECURITY_LABEL
+#endif
};
static const u32 nfs4_open_noattr_bitmap[3] = {
@@ -1598,6 +1601,7 @@ _nfs4_opendata_to_nfs4_state(struct nfs4_opendata *data)
if (!data->rpc_done) {
state = nfs4_try_open_cached(data);
+ trace_nfs4_cached_open(data->state);
goto out;
}
@@ -2015,6 +2019,7 @@ static void nfs4_open_prepare(struct rpc_task *task, void *calldata)
}
return;
unlock_no_action:
+ trace_nfs4_cached_open(data->state);
rcu_read_unlock();
out_no_action:
task->tk_action = NULL;
@@ -2703,6 +2708,7 @@ static int _nfs4_do_setattr(struct inode *inode, struct rpc_cred *cred,
status = nfs4_call_sync(server->client, server, &msg, &arg.seq_args, &res.seq_res, 1);
if (status == 0 && state != NULL)
renew_lease(server, timestamp);
+ trace_nfs4_setattr(inode, &arg.stateid, status);
return status;
}
@@ -2719,7 +2725,6 @@ static int nfs4_do_setattr(struct inode *inode, struct rpc_cred *cred,
int err;
do {
err = _nfs4_do_setattr(inode, cred, fattr, sattr, state, ilabel, olabel);
- trace_nfs4_setattr(inode, err);
switch (err) {
case -NFS4ERR_OPENMODE:
if (!(sattr->ia_valid & ATTR_SIZE)) {
@@ -5048,7 +5053,6 @@ static void nfs4_init_boot_verifier(const struct nfs_client *clp,
static int
nfs4_init_nonuniform_client_string(struct nfs_client *clp)
{
- int result;
size_t len;
char *str;
@@ -5076,7 +5080,7 @@ nfs4_init_nonuniform_client_string(struct nfs_client *clp)
return -ENOMEM;
rcu_read_lock();
- result = scnprintf(str, len, "Linux NFSv4.0 %s/%s %s",
+ scnprintf(str, len, "Linux NFSv4.0 %s/%s %s",
clp->cl_ipaddr,
rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_ADDR),
rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_PROTO));
@@ -5089,7 +5093,6 @@ nfs4_init_nonuniform_client_string(struct nfs_client *clp)
static int
nfs4_init_uniquifier_client_string(struct nfs_client *clp)
{
- int result;
size_t len;
char *str;
@@ -5109,7 +5112,7 @@ nfs4_init_uniquifier_client_string(struct nfs_client *clp)
if (!str)
return -ENOMEM;
- result = scnprintf(str, len, "Linux NFSv%u.%u %s/%s",
+ scnprintf(str, len, "Linux NFSv%u.%u %s/%s",
clp->rpc_ops->version, clp->cl_minorversion,
nfs4_client_id_uniquifier,
clp->cl_rpcclient->cl_nodename);
@@ -5120,7 +5123,6 @@ nfs4_init_uniquifier_client_string(struct nfs_client *clp)
static int
nfs4_init_uniform_client_string(struct nfs_client *clp)
{
- int result;
size_t len;
char *str;
@@ -5145,7 +5147,7 @@ nfs4_init_uniform_client_string(struct nfs_client *clp)
if (!str)
return -ENOMEM;
- result = scnprintf(str, len, "Linux NFSv%u.%u %s",
+ scnprintf(str, len, "Linux NFSv%u.%u %s",
clp->rpc_ops->version, clp->cl_minorversion,
clp->cl_rpcclient->cl_nodename);
clp->cl_owner_id = str;
@@ -5384,6 +5386,11 @@ static int _nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, co
if (data == NULL)
return -ENOMEM;
nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 1);
+
+ nfs4_state_protect(server->nfs_client,
+ NFS_SP4_MACH_CRED_CLEANUP,
+ &task_setup_data.rpc_client, &msg);
+
data->args.fhandle = &data->fh;
data->args.stateid = &data->stateid;
data->args.bitmask = server->cache_consistency_bitmask;
@@ -5426,7 +5433,7 @@ int nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, const nfs4
int err;
do {
err = _nfs4_proc_delegreturn(inode, cred, stateid, issync);
- trace_nfs4_delegreturn(inode, err);
+ trace_nfs4_delegreturn(inode, stateid, err);
switch (err) {
case -NFS4ERR_STALE_STATEID:
case -NFS4ERR_EXPIRED:
@@ -5936,6 +5943,7 @@ static int _nfs4_do_setlk(struct nfs4_state *state, int cmd, struct file_lock *f
data->cancelled = 1;
rpc_put_task(task);
dprintk("%s: done, ret = %d!\n", __func__, ret);
+ trace_nfs4_set_lock(fl, state, &data->res.stateid, cmd, ret);
return ret;
}
@@ -5952,7 +5960,6 @@ static int nfs4_lock_reclaim(struct nfs4_state *state, struct file_lock *request
if (test_bit(NFS_DELEGATED_STATE, &state->flags) != 0)
return 0;
err = _nfs4_do_setlk(state, F_SETLK, request, NFS_LOCK_RECLAIM);
- trace_nfs4_lock_reclaim(request, state, F_SETLK, err);
if (err != -NFS4ERR_DELAY)
break;
nfs4_handle_exception(server, err, &exception);
@@ -5979,7 +5986,6 @@ static int nfs4_lock_expired(struct nfs4_state *state, struct file_lock *request
if (test_bit(NFS_DELEGATED_STATE, &state->flags) != 0)
return 0;
err = _nfs4_do_setlk(state, F_SETLK, request, NFS_LOCK_EXPIRED);
- trace_nfs4_lock_expired(request, state, F_SETLK, err);
switch (err) {
default:
goto out;
@@ -6087,7 +6093,6 @@ static int nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock *
do {
err = _nfs4_proc_setlk(state, cmd, request);
- trace_nfs4_set_lock(request, state, cmd, err);
if (err == -NFS4ERR_DENIED)
err = -EAGAIN;
err = nfs4_handle_exception(NFS_SERVER(state->inode),
@@ -6253,9 +6258,6 @@ static int nfs4_xattr_set_nfs4_acl(const struct xattr_handler *handler,
const void *buf, size_t buflen,
int flags)
{
- if (strcmp(key, "") != 0)
- return -EINVAL;
-
return nfs4_proc_set_acl(d_inode(dentry), buf, buflen);
}
@@ -6263,32 +6265,15 @@ static int nfs4_xattr_get_nfs4_acl(const struct xattr_handler *handler,
struct dentry *dentry, const char *key,
void *buf, size_t buflen)
{
- if (strcmp(key, "") != 0)
- return -EINVAL;
-
return nfs4_proc_get_acl(d_inode(dentry), buf, buflen);
}
-static size_t nfs4_xattr_list_nfs4_acl(const struct xattr_handler *handler,
- struct dentry *dentry, char *list,
- size_t list_len, const char *name,
- size_t name_len)
+static bool nfs4_xattr_list_nfs4_acl(struct dentry *dentry)
{
- size_t len = sizeof(XATTR_NAME_NFSV4_ACL);
-
- if (!nfs4_server_supports_acls(NFS_SERVER(d_inode(dentry))))
- return 0;
-
- if (list && len <= list_len)
- memcpy(list, XATTR_NAME_NFSV4_ACL, len);
- return len;
+ return nfs4_server_supports_acls(NFS_SERVER(d_inode(dentry)));
}
#ifdef CONFIG_NFS_V4_SECURITY_LABEL
-static inline int nfs4_server_supports_labels(struct nfs_server *server)
-{
- return server->caps & NFS_CAP_SECURITY_LABEL;
-}
static int nfs4_xattr_set_nfs4_label(const struct xattr_handler *handler,
struct dentry *dentry, const char *key,
@@ -6310,29 +6295,34 @@ static int nfs4_xattr_get_nfs4_label(const struct xattr_handler *handler,
return -EOPNOTSUPP;
}
-static size_t nfs4_xattr_list_nfs4_label(const struct xattr_handler *handler,
- struct dentry *dentry, char *list,
- size_t list_len, const char *name,
- size_t name_len)
+static ssize_t
+nfs4_listxattr_nfs4_label(struct inode *inode, char *list, size_t list_len)
{
- size_t len = 0;
+ int len = 0;
- if (nfs_server_capable(d_inode(dentry), NFS_CAP_SECURITY_LABEL)) {
- len = security_inode_listsecurity(d_inode(dentry), NULL, 0);
- if (list && len <= list_len)
- security_inode_listsecurity(d_inode(dentry), list, len);
+ if (nfs_server_capable(inode, NFS_CAP_SECURITY_LABEL)) {
+ len = security_inode_listsecurity(inode, list, list_len);
+ if (list_len && len > list_len)
+ return -ERANGE;
}
return len;
}
static const struct xattr_handler nfs4_xattr_nfs4_label_handler = {
.prefix = XATTR_SECURITY_PREFIX,
- .list = nfs4_xattr_list_nfs4_label,
.get = nfs4_xattr_get_nfs4_label,
.set = nfs4_xattr_set_nfs4_label,
};
-#endif
+#else
+
+static ssize_t
+nfs4_listxattr_nfs4_label(struct inode *inode, char *list, size_t list_len)
+{
+ return 0;
+}
+
+#endif
/*
* nfs_fhget will use either the mounted_on_fileid or the fileid
@@ -6862,10 +6852,13 @@ static const struct nfs41_state_protection nfs4_sp4_mach_cred_request = {
},
.allow.u.words = {
[0] = 1 << (OP_CLOSE) |
+ 1 << (OP_OPEN_DOWNGRADE) |
1 << (OP_LOCKU) |
+ 1 << (OP_DELEGRETURN) |
1 << (OP_COMMIT),
[1] = 1 << (OP_SECINFO - 32) |
1 << (OP_SECINFO_NO_NAME - 32) |
+ 1 << (OP_LAYOUTRETURN - 32) |
1 << (OP_TEST_STATEID - 32) |
1 << (OP_FREE_STATEID - 32) |
1 << (OP_WRITE - 32)
@@ -6930,11 +6923,19 @@ static int nfs4_sp4_select_mode(struct nfs_client *clp,
}
if (test_bit(OP_CLOSE, sp->allow.u.longs) &&
+ test_bit(OP_OPEN_DOWNGRADE, sp->allow.u.longs) &&
+ test_bit(OP_DELEGRETURN, sp->allow.u.longs) &&
test_bit(OP_LOCKU, sp->allow.u.longs)) {
dfprintk(MOUNT, " cleanup mode enabled\n");
set_bit(NFS_SP4_MACH_CRED_CLEANUP, &clp->cl_sp4_flags);
}
+ if (test_bit(OP_LAYOUTRETURN, sp->allow.u.longs)) {
+ dfprintk(MOUNT, " pnfs cleanup mode enabled\n");
+ set_bit(NFS_SP4_MACH_CRED_PNFS_CLEANUP,
+ &clp->cl_sp4_flags);
+ }
+
if (test_bit(OP_SECINFO, sp->allow.u.longs) &&
test_bit(OP_SECINFO_NO_NAME, sp->allow.u.longs)) {
dfprintk(MOUNT, " secinfo mode enabled\n");
@@ -7763,6 +7764,7 @@ nfs4_layoutget_prepare(struct rpc_task *task, void *calldata)
struct nfs4_layoutget *lgp = calldata;
struct nfs_server *server = NFS_SERVER(lgp->args.inode);
struct nfs4_session *session = nfs4_get_session(server);
+ int ret;
dprintk("--> %s\n", __func__);
/* Note the is a race here, where a CB_LAYOUTRECALL can come in
@@ -7773,12 +7775,12 @@ nfs4_layoutget_prepare(struct rpc_task *task, void *calldata)
if (nfs41_setup_sequence(session, &lgp->args.seq_args,
&lgp->res.seq_res, task))
return;
- if (pnfs_choose_layoutget_stateid(&lgp->args.stateid,
+ ret = pnfs_choose_layoutget_stateid(&lgp->args.stateid,
NFS_I(lgp->args.inode)->layout,
&lgp->args.range,
- lgp->args.ctx->state)) {
- rpc_exit(task, NFS4_OK);
- }
+ lgp->args.ctx->state);
+ if (ret < 0)
+ rpc_exit(task, ret);
}
static void nfs4_layoutget_done(struct rpc_task *task, void *calldata)
@@ -7798,6 +7800,15 @@ static void nfs4_layoutget_done(struct rpc_task *task, void *calldata)
switch (task->tk_status) {
case 0:
goto out;
+
+ /*
+ * NFS4ERR_LAYOUTUNAVAILABLE means we are not supposed to use pnfs
+ * on the file. set tk_status to -ENODATA to tell upper layer to
+ * retry go inband.
+ */
+ case -NFS4ERR_LAYOUTUNAVAILABLE:
+ task->tk_status = -ENODATA;
+ goto out;
/*
* NFS4ERR_BADLAYOUT means the MDS cannot return a layout of
* length lgp->args.minlength != 0 (see RFC5661 section 18.43.3).
@@ -7994,6 +8005,7 @@ nfs4_proc_layoutget(struct nfs4_layoutget *lgp, gfp_t gfp_flags)
trace_nfs4_layoutget(lgp->args.ctx,
&lgp->args.range,
&lgp->res.range,
+ &lgp->res.stateid,
status);
/* if layoutp->len is 0, nfs4_layoutget_prepare called rpc_exit */
if (status == 0 && lgp->res.layoutp->len)
@@ -8050,9 +8062,10 @@ static void nfs4_layoutreturn_release(void *calldata)
dprintk("--> %s\n", __func__);
spin_lock(&lo->plh_inode->i_lock);
+ pnfs_mark_matching_lsegs_invalid(lo, &freeme, &lrp->args.range);
+ pnfs_mark_layout_returned_if_empty(lo);
if (lrp->res.lrs_present)
pnfs_set_layout_stateid(lo, &lrp->res.stateid, true);
- pnfs_mark_matching_lsegs_invalid(lo, &freeme, &lrp->args.range);
pnfs_clear_layoutreturn_waitbit(lo);
spin_unlock(&lo->plh_inode->i_lock);
pnfs_free_lseg_list(&freeme);
@@ -8085,6 +8098,10 @@ int nfs4_proc_layoutreturn(struct nfs4_layoutreturn *lrp, bool sync)
};
int status = 0;
+ nfs4_state_protect(NFS_SERVER(lrp->args.inode)->nfs_client,
+ NFS_SP4_MACH_CRED_PNFS_CLEANUP,
+ &task_setup_data.rpc_client, &msg);
+
dprintk("--> %s\n", __func__);
if (!sync) {
lrp->inode = nfs_igrab_and_active(lrp->args.inode);
@@ -8100,7 +8117,7 @@ int nfs4_proc_layoutreturn(struct nfs4_layoutreturn *lrp, bool sync)
return PTR_ERR(task);
if (sync)
status = task->tk_status;
- trace_nfs4_layoutreturn(lrp->args.inode, status);
+ trace_nfs4_layoutreturn(lrp->args.inode, &lrp->args.stateid, status);
dprintk("<-- %s status=%d\n", __func__, status);
rpc_put_task(task);
return status;
@@ -8248,7 +8265,7 @@ nfs4_proc_layoutcommit(struct nfs4_layoutcommit_data *data, bool sync)
return PTR_ERR(task);
if (sync)
status = task->tk_status;
- trace_nfs4_layoutcommit(data->args.inode, status);
+ trace_nfs4_layoutcommit(data->args.inode, &data->args.stateid, status);
dprintk("%s: status %d\n", __func__, status);
rpc_put_task(task);
return status;
@@ -8748,6 +8765,24 @@ const struct nfs4_minor_version_ops *nfs_v4_minor_ops[] = {
#endif
};
+ssize_t nfs4_listxattr(struct dentry *dentry, char *list, size_t size)
+{
+ ssize_t error, error2;
+
+ error = generic_listxattr(dentry, list, size);
+ if (error < 0)
+ return error;
+ if (list) {
+ list += error;
+ size -= error;
+ }
+
+ error2 = nfs4_listxattr_nfs4_label(d_inode(dentry), list, size);
+ if (error2 < 0)
+ return error2;
+ return error + error2;
+}
+
static const struct inode_operations nfs4_dir_inode_operations = {
.create = nfs_create,
.lookup = nfs_lookup,
@@ -8764,7 +8799,7 @@ static const struct inode_operations nfs4_dir_inode_operations = {
.setattr = nfs_setattr,
.getxattr = generic_getxattr,
.setxattr = generic_setxattr,
- .listxattr = generic_listxattr,
+ .listxattr = nfs4_listxattr,
.removexattr = generic_removexattr,
};
@@ -8774,7 +8809,7 @@ static const struct inode_operations nfs4_file_inode_operations = {
.setattr = nfs_setattr,
.getxattr = generic_getxattr,
.setxattr = generic_setxattr,
- .listxattr = generic_listxattr,
+ .listxattr = nfs4_listxattr,
.removexattr = generic_removexattr,
};
@@ -8833,7 +8868,7 @@ const struct nfs_rpc_ops nfs_v4_clientops = {
};
static const struct xattr_handler nfs4_xattr_nfs4_acl_handler = {
- .prefix = XATTR_NAME_NFSV4_ACL,
+ .name = XATTR_NAME_NFSV4_ACL,
.list = nfs4_xattr_list_nfs4_acl,
.get = nfs4_xattr_get_nfs4_acl,
.set = nfs4_xattr_set_nfs4_acl,
diff --git a/fs/nfs/nfs4sysctl.c b/fs/nfs/nfs4sysctl.c
index 0fbd3ab1b..8693d77c4 100644
--- a/fs/nfs/nfs4sysctl.c
+++ b/fs/nfs/nfs4sysctl.c
@@ -12,7 +12,7 @@
#include "nfs4idmap.h"
#include "callback.h"
-static const int nfs_set_port_min = 0;
+static const int nfs_set_port_min;
static const int nfs_set_port_max = 65535;
static struct ctl_table_header *nfs4_callback_sysctl_table;
diff --git a/fs/nfs/nfs4trace.c b/fs/nfs/nfs4trace.c
index d774335cc..2850bce19 100644
--- a/fs/nfs/nfs4trace.c
+++ b/fs/nfs/nfs4trace.c
@@ -6,6 +6,7 @@
#include "internal.h"
#include "nfs4session.h"
#include "callback.h"
+#include "pnfs.h"
#define CREATE_TRACE_POINTS
#include "nfs4trace.h"
diff --git a/fs/nfs/nfs4trace.h b/fs/nfs/nfs4trace.h
index 671cf68fe..2c8d05dae 100644
--- a/fs/nfs/nfs4trace.h
+++ b/fs/nfs/nfs4trace.h
@@ -321,6 +321,7 @@ TRACE_EVENT(nfs4_sequence_done,
__entry->highest_slotid = res->sr_highest_slotid;
__entry->target_highest_slotid =
res->sr_target_highest_slotid;
+ __entry->status_flags = res->sr_status_flags;
__entry->error = res->sr_status;
),
TP_printk(
@@ -399,6 +400,10 @@ DECLARE_EVENT_CLASS(nfs4_open_event,
__field(u64, fileid)
__field(u64, dir)
__string(name, ctx->dentry->d_name.name)
+ __field(int, stateid_seq)
+ __field(u32, stateid_hash)
+ __field(int, openstateid_seq)
+ __field(u32, openstateid_hash)
),
TP_fast_assign(
@@ -409,8 +414,22 @@ DECLARE_EVENT_CLASS(nfs4_open_event,
__entry->flags = flags;
__entry->fmode = (__force unsigned int)ctx->mode;
__entry->dev = ctx->dentry->d_sb->s_dev;
- if (!IS_ERR_OR_NULL(state))
+ if (!IS_ERR_OR_NULL(state)) {
inode = state->inode;
+ __entry->stateid_seq =
+ be32_to_cpu(state->stateid.seqid);
+ __entry->stateid_hash =
+ nfs_stateid_hash(&state->stateid);
+ __entry->openstateid_seq =
+ be32_to_cpu(state->open_stateid.seqid);
+ __entry->openstateid_hash =
+ nfs_stateid_hash(&state->open_stateid);
+ } else {
+ __entry->stateid_seq = 0;
+ __entry->stateid_hash = 0;
+ __entry->openstateid_seq = 0;
+ __entry->openstateid_hash = 0;
+ }
if (inode != NULL) {
__entry->fileid = NFS_FILEID(inode);
__entry->fhandle = nfs_fhandle_hash(NFS_FH(inode));
@@ -425,7 +444,8 @@ DECLARE_EVENT_CLASS(nfs4_open_event,
TP_printk(
"error=%d (%s) flags=%d (%s) fmode=%s "
"fileid=%02x:%02x:%llu fhandle=0x%08x "
- "name=%02x:%02x:%llu/%s",
+ "name=%02x:%02x:%llu/%s stateid=%d:0x%08x "
+ "openstateid=%d:0x%08x",
__entry->error,
show_nfsv4_errors(__entry->error),
__entry->flags,
@@ -436,7 +456,9 @@ DECLARE_EVENT_CLASS(nfs4_open_event,
__entry->fhandle,
MAJOR(__entry->dev), MINOR(__entry->dev),
(unsigned long long)__entry->dir,
- __get_str(name)
+ __get_str(name),
+ __entry->stateid_seq, __entry->stateid_hash,
+ __entry->openstateid_seq, __entry->openstateid_hash
)
);
@@ -452,6 +474,45 @@ DEFINE_NFS4_OPEN_EVENT(nfs4_open_reclaim);
DEFINE_NFS4_OPEN_EVENT(nfs4_open_expired);
DEFINE_NFS4_OPEN_EVENT(nfs4_open_file);
+TRACE_EVENT(nfs4_cached_open,
+ TP_PROTO(
+ const struct nfs4_state *state
+ ),
+ TP_ARGS(state),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(u32, fhandle)
+ __field(u64, fileid)
+ __field(unsigned int, fmode)
+ __field(int, stateid_seq)
+ __field(u32, stateid_hash)
+ ),
+
+ TP_fast_assign(
+ const struct inode *inode = state->inode;
+
+ __entry->dev = inode->i_sb->s_dev;
+ __entry->fileid = NFS_FILEID(inode);
+ __entry->fhandle = nfs_fhandle_hash(NFS_FH(inode));
+ __entry->fmode = (__force unsigned int)state->state;
+ __entry->stateid_seq =
+ be32_to_cpu(state->stateid.seqid);
+ __entry->stateid_hash =
+ nfs_stateid_hash(&state->stateid);
+ ),
+
+ TP_printk(
+ "fmode=%s fileid=%02x:%02x:%llu "
+ "fhandle=0x%08x stateid=%d:0x%08x",
+ __entry->fmode ? show_fmode_flags(__entry->fmode) :
+ "closed",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ (unsigned long long)__entry->fileid,
+ __entry->fhandle,
+ __entry->stateid_seq, __entry->stateid_hash
+ )
+);
+
TRACE_EVENT(nfs4_close,
TP_PROTO(
const struct nfs4_state *state,
@@ -468,6 +529,8 @@ TRACE_EVENT(nfs4_close,
__field(u64, fileid)
__field(unsigned int, fmode)
__field(int, error)
+ __field(int, stateid_seq)
+ __field(u32, stateid_hash)
),
TP_fast_assign(
@@ -478,18 +541,23 @@ TRACE_EVENT(nfs4_close,
__entry->fhandle = nfs_fhandle_hash(NFS_FH(inode));
__entry->fmode = (__force unsigned int)state->state;
__entry->error = error;
+ __entry->stateid_seq =
+ be32_to_cpu(args->stateid.seqid);
+ __entry->stateid_hash =
+ nfs_stateid_hash(&args->stateid);
),
TP_printk(
"error=%d (%s) fmode=%s fileid=%02x:%02x:%llu "
- "fhandle=0x%08x",
+ "fhandle=0x%08x openstateid=%d:0x%08x",
__entry->error,
show_nfsv4_errors(__entry->error),
__entry->fmode ? show_fmode_flags(__entry->fmode) :
"closed",
MAJOR(__entry->dev), MINOR(__entry->dev),
(unsigned long long)__entry->fileid,
- __entry->fhandle
+ __entry->fhandle,
+ __entry->stateid_seq, __entry->stateid_hash
)
);
@@ -523,6 +591,8 @@ DECLARE_EVENT_CLASS(nfs4_lock_event,
__field(dev_t, dev)
__field(u32, fhandle)
__field(u64, fileid)
+ __field(int, stateid_seq)
+ __field(u32, stateid_hash)
),
TP_fast_assign(
@@ -536,11 +606,16 @@ DECLARE_EVENT_CLASS(nfs4_lock_event,
__entry->dev = inode->i_sb->s_dev;
__entry->fileid = NFS_FILEID(inode);
__entry->fhandle = nfs_fhandle_hash(NFS_FH(inode));
+ __entry->stateid_seq =
+ be32_to_cpu(state->stateid.seqid);
+ __entry->stateid_hash =
+ nfs_stateid_hash(&state->stateid);
),
TP_printk(
"error=%d (%s) cmd=%s:%s range=%lld:%lld "
- "fileid=%02x:%02x:%llu fhandle=0x%08x",
+ "fileid=%02x:%02x:%llu fhandle=0x%08x "
+ "stateid=%d:0x%08x",
__entry->error,
show_nfsv4_errors(__entry->error),
show_lock_cmd(__entry->cmd),
@@ -549,7 +624,8 @@ DECLARE_EVENT_CLASS(nfs4_lock_event,
(long long)__entry->end,
MAJOR(__entry->dev), MINOR(__entry->dev),
(unsigned long long)__entry->fileid,
- __entry->fhandle
+ __entry->fhandle,
+ __entry->stateid_seq, __entry->stateid_hash
)
);
@@ -563,11 +639,73 @@ DECLARE_EVENT_CLASS(nfs4_lock_event,
), \
TP_ARGS(request, state, cmd, error))
DEFINE_NFS4_LOCK_EVENT(nfs4_get_lock);
-DEFINE_NFS4_LOCK_EVENT(nfs4_set_lock);
-DEFINE_NFS4_LOCK_EVENT(nfs4_lock_reclaim);
-DEFINE_NFS4_LOCK_EVENT(nfs4_lock_expired);
DEFINE_NFS4_LOCK_EVENT(nfs4_unlock);
+TRACE_EVENT(nfs4_set_lock,
+ TP_PROTO(
+ const struct file_lock *request,
+ const struct nfs4_state *state,
+ const nfs4_stateid *lockstateid,
+ int cmd,
+ int error
+ ),
+
+ TP_ARGS(request, state, lockstateid, cmd, error),
+
+ TP_STRUCT__entry(
+ __field(int, error)
+ __field(int, cmd)
+ __field(char, type)
+ __field(loff_t, start)
+ __field(loff_t, end)
+ __field(dev_t, dev)
+ __field(u32, fhandle)
+ __field(u64, fileid)
+ __field(int, stateid_seq)
+ __field(u32, stateid_hash)
+ __field(int, lockstateid_seq)
+ __field(u32, lockstateid_hash)
+ ),
+
+ TP_fast_assign(
+ const struct inode *inode = state->inode;
+
+ __entry->error = error;
+ __entry->cmd = cmd;
+ __entry->type = request->fl_type;
+ __entry->start = request->fl_start;
+ __entry->end = request->fl_end;
+ __entry->dev = inode->i_sb->s_dev;
+ __entry->fileid = NFS_FILEID(inode);
+ __entry->fhandle = nfs_fhandle_hash(NFS_FH(inode));
+ __entry->stateid_seq =
+ be32_to_cpu(state->stateid.seqid);
+ __entry->stateid_hash =
+ nfs_stateid_hash(&state->stateid);
+ __entry->lockstateid_seq =
+ be32_to_cpu(lockstateid->seqid);
+ __entry->lockstateid_hash =
+ nfs_stateid_hash(lockstateid);
+ ),
+
+ TP_printk(
+ "error=%d (%s) cmd=%s:%s range=%lld:%lld "
+ "fileid=%02x:%02x:%llu fhandle=0x%08x "
+ "stateid=%d:0x%08x lockstateid=%d:0x%08x",
+ __entry->error,
+ show_nfsv4_errors(__entry->error),
+ show_lock_cmd(__entry->cmd),
+ show_lock_type(__entry->type),
+ (long long)__entry->start,
+ (long long)__entry->end,
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ (unsigned long long)__entry->fileid,
+ __entry->fhandle,
+ __entry->stateid_seq, __entry->stateid_hash,
+ __entry->lockstateid_seq, __entry->lockstateid_hash
+ )
+);
+
DECLARE_EVENT_CLASS(nfs4_set_delegation_event,
TP_PROTO(
const struct inode *inode,
@@ -621,20 +759,28 @@ TRACE_EVENT(nfs4_delegreturn_exit,
__field(dev_t, dev)
__field(u32, fhandle)
__field(int, error)
+ __field(int, stateid_seq)
+ __field(u32, stateid_hash)
),
TP_fast_assign(
__entry->dev = res->server->s_dev;
__entry->fhandle = nfs_fhandle_hash(args->fhandle);
__entry->error = error;
+ __entry->stateid_seq =
+ be32_to_cpu(args->stateid->seqid);
+ __entry->stateid_hash =
+ nfs_stateid_hash(args->stateid);
),
TP_printk(
- "error=%d (%s) dev=%02x:%02x fhandle=0x%08x",
+ "error=%d (%s) dev=%02x:%02x fhandle=0x%08x "
+ "stateid=%d:0x%08x",
__entry->error,
show_nfsv4_errors(__entry->error),
MAJOR(__entry->dev), MINOR(__entry->dev),
- __entry->fhandle
+ __entry->fhandle,
+ __entry->stateid_seq, __entry->stateid_hash
)
);
@@ -653,6 +799,8 @@ DECLARE_EVENT_CLASS(nfs4_test_stateid_event,
__field(dev_t, dev)
__field(u32, fhandle)
__field(u64, fileid)
+ __field(int, stateid_seq)
+ __field(u32, stateid_hash)
),
TP_fast_assign(
@@ -662,15 +810,21 @@ DECLARE_EVENT_CLASS(nfs4_test_stateid_event,
__entry->dev = inode->i_sb->s_dev;
__entry->fileid = NFS_FILEID(inode);
__entry->fhandle = nfs_fhandle_hash(NFS_FH(inode));
+ __entry->stateid_seq =
+ be32_to_cpu(state->stateid.seqid);
+ __entry->stateid_hash =
+ nfs_stateid_hash(&state->stateid);
),
TP_printk(
- "error=%d (%s) fileid=%02x:%02x:%llu fhandle=0x%08x",
+ "error=%d (%s) fileid=%02x:%02x:%llu fhandle=0x%08x "
+ "stateid=%d:0x%08x",
__entry->error,
show_nfsv4_errors(__entry->error),
MAJOR(__entry->dev), MINOR(__entry->dev),
(unsigned long long)__entry->fileid,
- __entry->fhandle
+ __entry->fhandle,
+ __entry->stateid_seq, __entry->stateid_hash
)
);
@@ -820,7 +974,6 @@ DECLARE_EVENT_CLASS(nfs4_inode_event,
), \
TP_ARGS(inode, error))
-DEFINE_NFS4_INODE_EVENT(nfs4_setattr);
DEFINE_NFS4_INODE_EVENT(nfs4_access);
DEFINE_NFS4_INODE_EVENT(nfs4_readlink);
DEFINE_NFS4_INODE_EVENT(nfs4_readdir);
@@ -830,8 +983,59 @@ DEFINE_NFS4_INODE_EVENT(nfs4_set_acl);
DEFINE_NFS4_INODE_EVENT(nfs4_get_security_label);
DEFINE_NFS4_INODE_EVENT(nfs4_set_security_label);
#endif /* CONFIG_NFS_V4_SECURITY_LABEL */
-DEFINE_NFS4_INODE_EVENT(nfs4_recall_delegation);
-DEFINE_NFS4_INODE_EVENT(nfs4_delegreturn);
+
+DECLARE_EVENT_CLASS(nfs4_inode_stateid_event,
+ TP_PROTO(
+ const struct inode *inode,
+ const nfs4_stateid *stateid,
+ int error
+ ),
+
+ TP_ARGS(inode, stateid, error),
+
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(u32, fhandle)
+ __field(u64, fileid)
+ __field(int, error)
+ __field(int, stateid_seq)
+ __field(u32, stateid_hash)
+ ),
+
+ TP_fast_assign(
+ __entry->dev = inode->i_sb->s_dev;
+ __entry->fileid = NFS_FILEID(inode);
+ __entry->fhandle = nfs_fhandle_hash(NFS_FH(inode));
+ __entry->error = error;
+ __entry->stateid_seq =
+ be32_to_cpu(stateid->seqid);
+ __entry->stateid_hash =
+ nfs_stateid_hash(stateid);
+ ),
+
+ TP_printk(
+ "error=%d (%s) fileid=%02x:%02x:%llu fhandle=0x%08x "
+ "stateid=%d:0x%08x",
+ __entry->error,
+ show_nfsv4_errors(__entry->error),
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ (unsigned long long)__entry->fileid,
+ __entry->fhandle,
+ __entry->stateid_seq, __entry->stateid_hash
+ )
+);
+
+#define DEFINE_NFS4_INODE_STATEID_EVENT(name) \
+ DEFINE_EVENT(nfs4_inode_stateid_event, name, \
+ TP_PROTO( \
+ const struct inode *inode, \
+ const nfs4_stateid *stateid, \
+ int error \
+ ), \
+ TP_ARGS(inode, stateid, error))
+
+DEFINE_NFS4_INODE_STATEID_EVENT(nfs4_setattr);
+DEFINE_NFS4_INODE_STATEID_EVENT(nfs4_delegreturn);
DECLARE_EVENT_CLASS(nfs4_getattr_event,
TP_PROTO(
@@ -941,8 +1145,74 @@ DECLARE_EVENT_CLASS(nfs4_inode_callback_event,
), \
TP_ARGS(clp, fhandle, inode, error))
DEFINE_NFS4_INODE_CALLBACK_EVENT(nfs4_cb_getattr);
-DEFINE_NFS4_INODE_CALLBACK_EVENT(nfs4_cb_layoutrecall_inode);
+DECLARE_EVENT_CLASS(nfs4_inode_stateid_callback_event,
+ TP_PROTO(
+ const struct nfs_client *clp,
+ const struct nfs_fh *fhandle,
+ const struct inode *inode,
+ const nfs4_stateid *stateid,
+ int error
+ ),
+
+ TP_ARGS(clp, fhandle, inode, stateid, error),
+
+ TP_STRUCT__entry(
+ __field(int, error)
+ __field(dev_t, dev)
+ __field(u32, fhandle)
+ __field(u64, fileid)
+ __string(dstaddr, clp ?
+ rpc_peeraddr2str(clp->cl_rpcclient,
+ RPC_DISPLAY_ADDR) : "unknown")
+ __field(int, stateid_seq)
+ __field(u32, stateid_hash)
+ ),
+
+ TP_fast_assign(
+ __entry->error = error;
+ __entry->fhandle = nfs_fhandle_hash(fhandle);
+ if (inode != NULL) {
+ __entry->fileid = NFS_FILEID(inode);
+ __entry->dev = inode->i_sb->s_dev;
+ } else {
+ __entry->fileid = 0;
+ __entry->dev = 0;
+ }
+ __assign_str(dstaddr, clp ?
+ rpc_peeraddr2str(clp->cl_rpcclient,
+ RPC_DISPLAY_ADDR) : "unknown")
+ __entry->stateid_seq =
+ be32_to_cpu(stateid->seqid);
+ __entry->stateid_hash =
+ nfs_stateid_hash(stateid);
+ ),
+
+ TP_printk(
+ "error=%d (%s) fileid=%02x:%02x:%llu fhandle=0x%08x "
+ "stateid=%d:0x%08x dstaddr=%s",
+ __entry->error,
+ show_nfsv4_errors(__entry->error),
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ (unsigned long long)__entry->fileid,
+ __entry->fhandle,
+ __entry->stateid_seq, __entry->stateid_hash,
+ __get_str(dstaddr)
+ )
+);
+
+#define DEFINE_NFS4_INODE_STATEID_CALLBACK_EVENT(name) \
+ DEFINE_EVENT(nfs4_inode_stateid_callback_event, name, \
+ TP_PROTO( \
+ const struct nfs_client *clp, \
+ const struct nfs_fh *fhandle, \
+ const struct inode *inode, \
+ const nfs4_stateid *stateid, \
+ int error \
+ ), \
+ TP_ARGS(clp, fhandle, inode, stateid, error))
+DEFINE_NFS4_INODE_STATEID_CALLBACK_EVENT(nfs4_cb_recall);
+DEFINE_NFS4_INODE_STATEID_CALLBACK_EVENT(nfs4_cb_layoutrecall_file);
DECLARE_EVENT_CLASS(nfs4_idmap_event,
TP_PROTO(
@@ -1005,28 +1275,37 @@ DECLARE_EVENT_CLASS(nfs4_read_event,
__field(loff_t, offset)
__field(size_t, count)
__field(int, error)
+ __field(int, stateid_seq)
+ __field(u32, stateid_hash)
),
TP_fast_assign(
const struct inode *inode = hdr->inode;
+ const struct nfs4_state *state =
+ hdr->args.context->state;
__entry->dev = inode->i_sb->s_dev;
__entry->fileid = NFS_FILEID(inode);
__entry->fhandle = nfs_fhandle_hash(NFS_FH(inode));
__entry->offset = hdr->args.offset;
__entry->count = hdr->args.count;
__entry->error = error;
+ __entry->stateid_seq =
+ be32_to_cpu(state->stateid.seqid);
+ __entry->stateid_hash =
+ nfs_stateid_hash(&state->stateid);
),
TP_printk(
"error=%d (%s) fileid=%02x:%02x:%llu fhandle=0x%08x "
- "offset=%lld count=%zu",
+ "offset=%lld count=%zu stateid=%d:0x%08x",
__entry->error,
show_nfsv4_errors(__entry->error),
MAJOR(__entry->dev), MINOR(__entry->dev),
(unsigned long long)__entry->fileid,
__entry->fhandle,
(long long)__entry->offset,
- __entry->count
+ __entry->count,
+ __entry->stateid_seq, __entry->stateid_hash
)
);
#define DEFINE_NFS4_READ_EVENT(name) \
@@ -1056,28 +1335,37 @@ DECLARE_EVENT_CLASS(nfs4_write_event,
__field(loff_t, offset)
__field(size_t, count)
__field(int, error)
+ __field(int, stateid_seq)
+ __field(u32, stateid_hash)
),
TP_fast_assign(
const struct inode *inode = hdr->inode;
+ const struct nfs4_state *state =
+ hdr->args.context->state;
__entry->dev = inode->i_sb->s_dev;
__entry->fileid = NFS_FILEID(inode);
__entry->fhandle = nfs_fhandle_hash(NFS_FH(inode));
__entry->offset = hdr->args.offset;
__entry->count = hdr->args.count;
__entry->error = error;
+ __entry->stateid_seq =
+ be32_to_cpu(state->stateid.seqid);
+ __entry->stateid_hash =
+ nfs_stateid_hash(&state->stateid);
),
TP_printk(
"error=%d (%s) fileid=%02x:%02x:%llu fhandle=0x%08x "
- "offset=%lld count=%zu",
+ "offset=%lld count=%zu stateid=%d:0x%08x",
__entry->error,
show_nfsv4_errors(__entry->error),
MAJOR(__entry->dev), MINOR(__entry->dev),
(unsigned long long)__entry->fileid,
__entry->fhandle,
(long long)__entry->offset,
- __entry->count
+ __entry->count,
+ __entry->stateid_seq, __entry->stateid_hash
)
);
@@ -1154,10 +1442,11 @@ TRACE_EVENT(nfs4_layoutget,
const struct nfs_open_context *ctx,
const struct pnfs_layout_range *args,
const struct pnfs_layout_range *res,
+ const nfs4_stateid *layout_stateid,
int error
),
- TP_ARGS(ctx, args, res, error),
+ TP_ARGS(ctx, args, res, layout_stateid, error),
TP_STRUCT__entry(
__field(dev_t, dev)
@@ -1167,10 +1456,15 @@ TRACE_EVENT(nfs4_layoutget,
__field(u64, offset)
__field(u64, count)
__field(int, error)
+ __field(int, stateid_seq)
+ __field(u32, stateid_hash)
+ __field(int, layoutstateid_seq)
+ __field(u32, layoutstateid_hash)
),
TP_fast_assign(
const struct inode *inode = d_inode(ctx->dentry);
+ const struct nfs4_state *state = ctx->state;
__entry->dev = inode->i_sb->s_dev;
__entry->fileid = NFS_FILEID(inode);
__entry->fhandle = nfs_fhandle_hash(NFS_FH(inode));
@@ -1178,11 +1472,25 @@ TRACE_EVENT(nfs4_layoutget,
__entry->offset = args->offset;
__entry->count = args->length;
__entry->error = error;
+ __entry->stateid_seq =
+ be32_to_cpu(state->stateid.seqid);
+ __entry->stateid_hash =
+ nfs_stateid_hash(&state->stateid);
+ if (!error) {
+ __entry->layoutstateid_seq =
+ be32_to_cpu(layout_stateid->seqid);
+ __entry->layoutstateid_hash =
+ nfs_stateid_hash(layout_stateid);
+ } else {
+ __entry->layoutstateid_seq = 0;
+ __entry->layoutstateid_hash = 0;
+ }
),
TP_printk(
"error=%d (%s) fileid=%02x:%02x:%llu fhandle=0x%08x "
- "iomode=%s offset=%llu count=%llu",
+ "iomode=%s offset=%llu count=%llu stateid=%d:0x%08x "
+ "layoutstateid=%d:0x%08x",
__entry->error,
show_nfsv4_errors(__entry->error),
MAJOR(__entry->dev), MINOR(__entry->dev),
@@ -1190,14 +1498,83 @@ TRACE_EVENT(nfs4_layoutget,
__entry->fhandle,
show_pnfs_iomode(__entry->iomode),
(unsigned long long)__entry->offset,
- (unsigned long long)__entry->count
+ (unsigned long long)__entry->count,
+ __entry->stateid_seq, __entry->stateid_hash,
+ __entry->layoutstateid_seq, __entry->layoutstateid_hash
)
);
-DEFINE_NFS4_INODE_EVENT(nfs4_layoutcommit);
-DEFINE_NFS4_INODE_EVENT(nfs4_layoutreturn);
+DEFINE_NFS4_INODE_STATEID_EVENT(nfs4_layoutcommit);
+DEFINE_NFS4_INODE_STATEID_EVENT(nfs4_layoutreturn);
DEFINE_NFS4_INODE_EVENT(nfs4_layoutreturn_on_close);
+#define show_pnfs_update_layout_reason(reason) \
+ __print_symbolic(reason, \
+ { PNFS_UPDATE_LAYOUT_UNKNOWN, "unknown" }, \
+ { PNFS_UPDATE_LAYOUT_NO_PNFS, "no pnfs" }, \
+ { PNFS_UPDATE_LAYOUT_RD_ZEROLEN, "read+zerolen" }, \
+ { PNFS_UPDATE_LAYOUT_MDSTHRESH, "mdsthresh" }, \
+ { PNFS_UPDATE_LAYOUT_NOMEM, "nomem" }, \
+ { PNFS_UPDATE_LAYOUT_BULK_RECALL, "bulk recall" }, \
+ { PNFS_UPDATE_LAYOUT_IO_TEST_FAIL, "io test fail" }, \
+ { PNFS_UPDATE_LAYOUT_FOUND_CACHED, "found cached" }, \
+ { PNFS_UPDATE_LAYOUT_RETURN, "layoutreturn" }, \
+ { PNFS_UPDATE_LAYOUT_BLOCKED, "layouts blocked" }, \
+ { PNFS_UPDATE_LAYOUT_SEND_LAYOUTGET, "sent layoutget" })
+
+TRACE_EVENT(pnfs_update_layout,
+ TP_PROTO(struct inode *inode,
+ loff_t pos,
+ u64 count,
+ enum pnfs_iomode iomode,
+ struct pnfs_layout_hdr *lo,
+ enum pnfs_update_layout_reason reason
+ ),
+ TP_ARGS(inode, pos, count, iomode, lo, reason),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(u64, fileid)
+ __field(u32, fhandle)
+ __field(loff_t, pos)
+ __field(u64, count)
+ __field(enum pnfs_iomode, iomode)
+ __field(int, layoutstateid_seq)
+ __field(u32, layoutstateid_hash)
+ __field(enum pnfs_update_layout_reason, reason)
+ ),
+ TP_fast_assign(
+ __entry->dev = inode->i_sb->s_dev;
+ __entry->fileid = NFS_FILEID(inode);
+ __entry->fhandle = nfs_fhandle_hash(NFS_FH(inode));
+ __entry->pos = pos;
+ __entry->count = count;
+ __entry->iomode = iomode;
+ __entry->reason = reason;
+ if (lo != NULL) {
+ __entry->layoutstateid_seq =
+ be32_to_cpu(lo->plh_stateid.seqid);
+ __entry->layoutstateid_hash =
+ nfs_stateid_hash(&lo->plh_stateid);
+ } else {
+ __entry->layoutstateid_seq = 0;
+ __entry->layoutstateid_hash = 0;
+ }
+ ),
+ TP_printk(
+ "fileid=%02x:%02x:%llu fhandle=0x%08x "
+ "iomode=%s pos=%llu count=%llu "
+ "layoutstateid=%d:0x%08x (%s)",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ (unsigned long long)__entry->fileid,
+ __entry->fhandle,
+ show_pnfs_iomode(__entry->iomode),
+ (unsigned long long)__entry->pos,
+ (unsigned long long)__entry->count,
+ __entry->layoutstateid_seq, __entry->layoutstateid_hash,
+ show_pnfs_update_layout_reason(__entry->reason)
+ )
+);
+
#endif /* CONFIG_NFS_V4_1 */
#endif /* _TRACE_NFS4_H */
diff --git a/fs/nfs/nfstrace.h b/fs/nfs/nfstrace.h
index 59f838cdc..9f80a086b 100644
--- a/fs/nfs/nfstrace.h
+++ b/fs/nfs/nfstrace.h
@@ -39,7 +39,6 @@
{ 1 << NFS_INO_INVALIDATING, "INVALIDATING" }, \
{ 1 << NFS_INO_FLUSHING, "FLUSHING" }, \
{ 1 << NFS_INO_FSCACHE, "FSCACHE" }, \
- { 1 << NFS_INO_COMMIT, "COMMIT" }, \
{ 1 << NFS_INO_LAYOUTCOMMIT, "NEED_LAYOUTCOMMIT" }, \
{ 1 << NFS_INO_LAYOUTCOMMITTING, "LAYOUTCOMMIT" })
diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c
index 452a011ba..8ce4f61cb 100644
--- a/fs/nfs/pagelist.c
+++ b/fs/nfs/pagelist.c
@@ -101,53 +101,18 @@ nfs_page_free(struct nfs_page *p)
kmem_cache_free(nfs_page_cachep, p);
}
-static void
-nfs_iocounter_inc(struct nfs_io_counter *c)
-{
- atomic_inc(&c->io_count);
-}
-
-static void
-nfs_iocounter_dec(struct nfs_io_counter *c)
-{
- if (atomic_dec_and_test(&c->io_count)) {
- clear_bit(NFS_IO_INPROGRESS, &c->flags);
- smp_mb__after_atomic();
- wake_up_bit(&c->flags, NFS_IO_INPROGRESS);
- }
-}
-
-static int
-__nfs_iocounter_wait(struct nfs_io_counter *c)
-{
- wait_queue_head_t *wq = bit_waitqueue(&c->flags, NFS_IO_INPROGRESS);
- DEFINE_WAIT_BIT(q, &c->flags, NFS_IO_INPROGRESS);
- int ret = 0;
-
- do {
- prepare_to_wait(wq, &q.wait, TASK_KILLABLE);
- set_bit(NFS_IO_INPROGRESS, &c->flags);
- if (atomic_read(&c->io_count) == 0)
- break;
- ret = nfs_wait_bit_killable(&q.key, TASK_KILLABLE);
- } while (atomic_read(&c->io_count) != 0 && !ret);
- finish_wait(wq, &q.wait);
- return ret;
-}
-
/**
* nfs_iocounter_wait - wait for i/o to complete
- * @c: nfs_io_counter to use
+ * @l_ctx: nfs_lock_context with io_counter to use
*
* returns -ERESTARTSYS if interrupted by a fatal signal.
* Otherwise returns 0 once the io_count hits 0.
*/
int
-nfs_iocounter_wait(struct nfs_io_counter *c)
+nfs_iocounter_wait(struct nfs_lock_context *l_ctx)
{
- if (atomic_read(&c->io_count) == 0)
- return 0;
- return __nfs_iocounter_wait(c);
+ return wait_on_atomic_t(&l_ctx->io_count, nfs_wait_atomic_killable,
+ TASK_KILLABLE);
}
/*
@@ -370,7 +335,7 @@ nfs_create_request(struct nfs_open_context *ctx, struct page *page,
return ERR_CAST(l_ctx);
}
req->wb_lock_context = l_ctx;
- nfs_iocounter_inc(&l_ctx->io_count);
+ atomic_inc(&l_ctx->io_count);
/* Initialize the request struct. Initially, we assume a
* long write-back delay. This will be adjusted in
@@ -431,7 +396,8 @@ static void nfs_clear_request(struct nfs_page *req)
req->wb_page = NULL;
}
if (l_ctx != NULL) {
- nfs_iocounter_dec(&l_ctx->io_count);
+ if (atomic_dec_and_test(&l_ctx->io_count))
+ wake_up_atomic_t(&l_ctx->io_count);
nfs_put_lock_context(l_ctx);
req->wb_lock_context = NULL;
}
@@ -664,22 +630,11 @@ EXPORT_SYMBOL_GPL(nfs_initiate_pgio);
* @desc: IO descriptor
* @hdr: pageio header
*/
-static int nfs_pgio_error(struct nfs_pageio_descriptor *desc,
- struct nfs_pgio_header *hdr)
+static void nfs_pgio_error(struct nfs_pgio_header *hdr)
{
- struct nfs_pgio_mirror *mirror;
- u32 midx;
-
set_bit(NFS_IOHDR_REDO, &hdr->flags);
nfs_pgio_data_destroy(hdr);
hdr->completion_ops->completion(hdr);
- /* TODO: Make sure it's right to clean up all mirrors here
- * and not just hdr->pgio_mirror_idx */
- for (midx = 0; midx < desc->pg_mirror_count; midx++) {
- mirror = &desc->pg_mirrors[midx];
- desc->pg_completion_ops->error_cleanup(&mirror->pg_list);
- }
- return -ENOMEM;
}
/**
@@ -800,8 +755,11 @@ int nfs_generic_pgio(struct nfs_pageio_descriptor *desc,
unsigned int pagecount, pageused;
pagecount = nfs_page_array_len(mirror->pg_base, mirror->pg_count);
- if (!nfs_pgarray_set(&hdr->page_array, pagecount))
- return nfs_pgio_error(desc, hdr);
+ if (!nfs_pgarray_set(&hdr->page_array, pagecount)) {
+ nfs_pgio_error(hdr);
+ desc->pg_error = -ENOMEM;
+ return desc->pg_error;
+ }
nfs_init_cinfo(&cinfo, desc->pg_inode, desc->pg_dreq);
pages = hdr->page_array.pagevec;
@@ -819,8 +777,11 @@ int nfs_generic_pgio(struct nfs_pageio_descriptor *desc,
*pages++ = last_page = req->wb_page;
}
}
- if (WARN_ON_ONCE(pageused != pagecount))
- return nfs_pgio_error(desc, hdr);
+ if (WARN_ON_ONCE(pageused != pagecount)) {
+ nfs_pgio_error(hdr);
+ desc->pg_error = -EINVAL;
+ return desc->pg_error;
+ }
if ((desc->pg_ioflags & FLUSH_COND_STABLE) &&
(desc->pg_moreio || nfs_reqs_to_commit(&cinfo)))
@@ -835,18 +796,13 @@ EXPORT_SYMBOL_GPL(nfs_generic_pgio);
static int nfs_generic_pg_pgios(struct nfs_pageio_descriptor *desc)
{
- struct nfs_pgio_mirror *mirror;
struct nfs_pgio_header *hdr;
int ret;
- mirror = nfs_pgio_current_mirror(desc);
-
hdr = nfs_pgio_header_alloc(desc->pg_rw_ops);
if (!hdr) {
- /* TODO: make sure this is right with mirroring - or
- * should it back out all mirrors? */
- desc->pg_completion_ops->error_cleanup(&mirror->pg_list);
- return -ENOMEM;
+ desc->pg_error = -ENOMEM;
+ return desc->pg_error;
}
nfs_pgheader_init(desc, hdr, nfs_pgio_header_free);
ret = nfs_generic_pgio(desc, hdr);
@@ -874,6 +830,9 @@ static int nfs_pageio_setup_mirroring(struct nfs_pageio_descriptor *pgio,
mirror_count = pgio->pg_ops->pg_get_mirror_count(pgio, req);
+ if (pgio->pg_error < 0)
+ return pgio->pg_error;
+
if (!mirror_count || mirror_count > NFS_PAGEIO_DESCRIPTOR_MIRROR_MAX)
return -EINVAL;
@@ -903,12 +862,6 @@ static void nfs_pageio_cleanup_mirroring(struct nfs_pageio_descriptor *pgio)
pgio->pg_mirrors_dynamic = NULL;
}
-static bool nfs_match_open_context(const struct nfs_open_context *ctx1,
- const struct nfs_open_context *ctx2)
-{
- return ctx1->cred == ctx2->cred && ctx1->state == ctx2->state;
-}
-
static bool nfs_match_lock_context(const struct nfs_lock_context *l1,
const struct nfs_lock_context *l2)
{
@@ -982,6 +935,8 @@ static int nfs_pageio_do_add_request(struct nfs_pageio_descriptor *desc,
} else {
if (desc->pg_ops->pg_init)
desc->pg_ops->pg_init(desc, req);
+ if (desc->pg_error < 0)
+ return 0;
mirror->pg_base = req->wb_pgbase;
}
if (!nfs_can_coalesce_requests(prev, req, desc))
@@ -1147,6 +1102,8 @@ int nfs_pageio_add_request(struct nfs_pageio_descriptor *desc,
bytes = req->wb_bytes;
nfs_pageio_setup_mirroring(desc, req);
+ if (desc->pg_error < 0)
+ goto out_failed;
for (midx = 0; midx < desc->pg_mirror_count; midx++) {
if (midx) {
@@ -1163,7 +1120,8 @@ int nfs_pageio_add_request(struct nfs_pageio_descriptor *desc,
if (IS_ERR(dupreq)) {
nfs_page_group_unlock(req);
- return 0;
+ desc->pg_error = PTR_ERR(dupreq);
+ goto out_failed;
}
nfs_lock_request(dupreq);
@@ -1176,10 +1134,32 @@ int nfs_pageio_add_request(struct nfs_pageio_descriptor *desc,
if (nfs_pgio_has_mirroring(desc))
desc->pg_mirror_idx = midx;
if (!nfs_pageio_add_request_mirror(desc, dupreq))
- return 0;
+ goto out_failed;
}
return 1;
+
+out_failed:
+ /*
+ * We might have failed before sending any reqs over wire.
+ * Clean up rest of the reqs in mirror pg_list.
+ */
+ if (desc->pg_error) {
+ struct nfs_pgio_mirror *mirror;
+ void (*func)(struct list_head *);
+
+ /* remember fatal errors */
+ if (nfs_error_is_fatal(desc->pg_error))
+ mapping_set_error(desc->pg_inode->i_mapping,
+ desc->pg_error);
+
+ func = desc->pg_completion_ops->error_cleanup;
+ for (midx = 0; midx < desc->pg_mirror_count; midx++) {
+ mirror = &desc->pg_mirrors[midx];
+ func(&mirror->pg_list);
+ }
+ }
+ return 0;
}
/*
@@ -1232,7 +1212,7 @@ int nfs_pageio_resend(struct nfs_pageio_descriptor *desc,
nfs_pageio_complete(desc);
if (!list_empty(&failed)) {
list_move(&failed, &hdr->pages);
- return -EIO;
+ return desc->pg_error < 0 ? desc->pg_error : -EIO;
}
return 0;
}
diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
index bec038449..2fa483e6d 100644
--- a/fs/nfs/pnfs.c
+++ b/fs/nfs/pnfs.c
@@ -52,9 +52,7 @@ static DEFINE_SPINLOCK(pnfs_spinlock);
*/
static LIST_HEAD(pnfs_modules_tbl);
-static int
-pnfs_send_layoutreturn(struct pnfs_layout_hdr *lo, nfs4_stateid stateid,
- enum pnfs_iomode iomode, bool sync);
+static void pnfs_layoutreturn_before_put_layout_hdr(struct pnfs_layout_hdr *lo);
/* Return the registered pnfs layout driver module matching given id */
static struct pnfs_layoutdriver_type *
@@ -243,6 +241,8 @@ pnfs_put_layout_hdr(struct pnfs_layout_hdr *lo)
{
struct inode *inode = lo->plh_inode;
+ pnfs_layoutreturn_before_put_layout_hdr(lo);
+
if (atomic_dec_and_lock(&lo->plh_refcount, &inode->i_lock)) {
if (!list_empty(&lo->plh_segs))
WARN_ONCE(1, "NFS: BUG unfreed layout segments.\n");
@@ -252,6 +252,27 @@ pnfs_put_layout_hdr(struct pnfs_layout_hdr *lo)
}
}
+/*
+ * Mark a pnfs_layout_hdr and all associated layout segments as invalid
+ *
+ * In order to continue using the pnfs_layout_hdr, a full recovery
+ * is required.
+ * Note that caller must hold inode->i_lock.
+ */
+static int
+pnfs_mark_layout_stateid_invalid(struct pnfs_layout_hdr *lo,
+ struct list_head *lseg_list)
+{
+ struct pnfs_layout_range range = {
+ .iomode = IOMODE_ANY,
+ .offset = 0,
+ .length = NFS4_MAX_UINT64,
+ };
+
+ set_bit(NFS_LAYOUT_INVALID_STID, &lo->plh_flags);
+ return pnfs_mark_matching_lsegs_invalid(lo, lseg_list, &range);
+}
+
static int
pnfs_iomode_to_fail_bit(u32 iomode)
{
@@ -345,58 +366,6 @@ pnfs_layout_remove_lseg(struct pnfs_layout_hdr *lo,
rpc_wake_up(&NFS_SERVER(inode)->roc_rpcwaitq);
}
-/* Return true if layoutreturn is needed */
-static bool
-pnfs_layout_need_return(struct pnfs_layout_hdr *lo,
- struct pnfs_layout_segment *lseg)
-{
- struct pnfs_layout_segment *s;
-
- if (!test_and_clear_bit(NFS_LSEG_LAYOUTRETURN, &lseg->pls_flags))
- return false;
-
- list_for_each_entry(s, &lo->plh_segs, pls_list)
- if (s != lseg && test_bit(NFS_LSEG_LAYOUTRETURN, &s->pls_flags))
- return false;
-
- return true;
-}
-
-static bool
-pnfs_prepare_layoutreturn(struct pnfs_layout_hdr *lo)
-{
- if (test_and_set_bit(NFS_LAYOUT_RETURN, &lo->plh_flags))
- return false;
- lo->plh_return_iomode = 0;
- pnfs_get_layout_hdr(lo);
- clear_bit(NFS_LAYOUT_RETURN_BEFORE_CLOSE, &lo->plh_flags);
- return true;
-}
-
-static void pnfs_layoutreturn_before_put_lseg(struct pnfs_layout_segment *lseg,
- struct pnfs_layout_hdr *lo, struct inode *inode)
-{
- lo = lseg->pls_layout;
- inode = lo->plh_inode;
-
- spin_lock(&inode->i_lock);
- if (pnfs_layout_need_return(lo, lseg)) {
- nfs4_stateid stateid;
- enum pnfs_iomode iomode;
- bool send;
-
- stateid = lo->plh_stateid;
- iomode = lo->plh_return_iomode;
- send = pnfs_prepare_layoutreturn(lo);
- spin_unlock(&inode->i_lock);
- if (send) {
- /* Send an async layoutreturn so we dont deadlock */
- pnfs_send_layoutreturn(lo, stateid, iomode, false);
- }
- } else
- spin_unlock(&inode->i_lock);
-}
-
void
pnfs_put_lseg(struct pnfs_layout_segment *lseg)
{
@@ -410,15 +379,8 @@ pnfs_put_lseg(struct pnfs_layout_segment *lseg)
atomic_read(&lseg->pls_refcount),
test_bit(NFS_LSEG_VALID, &lseg->pls_flags));
- /* Handle the case where refcount != 1 */
- if (atomic_add_unless(&lseg->pls_refcount, -1, 1))
- return;
-
lo = lseg->pls_layout;
inode = lo->plh_inode;
- /* Do we need a layoutreturn? */
- if (test_bit(NFS_LSEG_LAYOUTRETURN, &lseg->pls_flags))
- pnfs_layoutreturn_before_put_lseg(lseg, lo, inode);
if (atomic_dec_and_lock(&lseg->pls_refcount, &inode->i_lock)) {
if (test_bit(NFS_LSEG_VALID, &lseg->pls_flags)) {
@@ -566,10 +528,10 @@ static int mark_lseg_invalid(struct pnfs_layout_segment *lseg,
int
pnfs_mark_matching_lsegs_invalid(struct pnfs_layout_hdr *lo,
struct list_head *tmp_list,
- struct pnfs_layout_range *recall_range)
+ const struct pnfs_layout_range *recall_range)
{
struct pnfs_layout_segment *lseg, *next;
- int invalid = 0, removed = 0;
+ int remaining = 0;
dprintk("%s:Begin lo %p\n", __func__, lo);
@@ -582,11 +544,11 @@ pnfs_mark_matching_lsegs_invalid(struct pnfs_layout_hdr *lo,
"offset %llu length %llu\n", __func__,
lseg, lseg->pls_range.iomode, lseg->pls_range.offset,
lseg->pls_range.length);
- invalid++;
- removed += mark_lseg_invalid(lseg, tmp_list);
+ if (!mark_lseg_invalid(lseg, tmp_list))
+ remaining++;
}
- dprintk("%s:Return %i\n", __func__, invalid - removed);
- return invalid - removed;
+ dprintk("%s:Return %i\n", __func__, remaining);
+ return remaining;
}
/* note free_me must contain lsegs from a single layout_hdr */
@@ -613,12 +575,10 @@ pnfs_destroy_layout(struct nfs_inode *nfsi)
spin_lock(&nfsi->vfs_inode.i_lock);
lo = nfsi->layout;
if (lo) {
- lo->plh_block_lgets++; /* permanently block new LAYOUTGETs */
- pnfs_mark_matching_lsegs_invalid(lo, &tmp_list, NULL);
pnfs_get_layout_hdr(lo);
+ pnfs_mark_layout_stateid_invalid(lo, &tmp_list);
pnfs_layout_clear_fail_bit(lo, NFS_LAYOUT_RO_FAILED);
pnfs_layout_clear_fail_bit(lo, NFS_LAYOUT_RW_FAILED);
- pnfs_clear_retry_layoutget(lo);
spin_unlock(&nfsi->vfs_inode.i_lock);
pnfs_free_lseg_list(&tmp_list);
pnfs_put_layout_hdr(lo);
@@ -677,11 +637,6 @@ pnfs_layout_free_bulk_destroy_list(struct list_head *layout_list,
{
struct pnfs_layout_hdr *lo;
struct inode *inode;
- struct pnfs_layout_range range = {
- .iomode = IOMODE_ANY,
- .offset = 0,
- .length = NFS4_MAX_UINT64,
- };
LIST_HEAD(lseg_list);
int ret = 0;
@@ -696,13 +651,15 @@ pnfs_layout_free_bulk_destroy_list(struct list_head *layout_list,
spin_lock(&inode->i_lock);
list_del_init(&lo->plh_bulk_destroy);
- lo->plh_block_lgets++; /* permanently block new LAYOUTGETs */
- if (is_bulk_recall)
- set_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags);
- if (pnfs_mark_matching_lsegs_invalid(lo, &lseg_list, &range))
+ if (pnfs_mark_layout_stateid_invalid(lo, &lseg_list)) {
+ if (is_bulk_recall)
+ set_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags);
ret = -EAGAIN;
+ }
spin_unlock(&inode->i_lock);
pnfs_free_lseg_list(&lseg_list);
+ /* Free all lsegs that are attached to commit buckets */
+ nfs_commit_inode(inode, 0);
pnfs_put_layout_hdr(lo);
iput(inode);
}
@@ -826,7 +783,7 @@ pnfs_layoutgets_blocked(const struct pnfs_layout_hdr *lo)
int
pnfs_choose_layoutget_stateid(nfs4_stateid *dst, struct pnfs_layout_hdr *lo,
- struct pnfs_layout_range *range,
+ const struct pnfs_layout_range *range,
struct nfs4_state *open_state)
{
int status = 0;
@@ -861,7 +818,7 @@ pnfs_choose_layoutget_stateid(nfs4_stateid *dst, struct pnfs_layout_hdr *lo,
static struct pnfs_layout_segment *
send_layoutget(struct pnfs_layout_hdr *lo,
struct nfs_open_context *ctx,
- struct pnfs_layout_range *range,
+ const struct pnfs_layout_range *range,
gfp_t gfp_flags)
{
struct inode *ino = lo->plh_inode;
@@ -894,7 +851,7 @@ send_layoutget(struct pnfs_layout_hdr *lo,
lgp->args.minlength = i_size - range->offset;
}
lgp->args.maxcount = PNFS_LAYOUT_MAXSIZE;
- lgp->args.range = *range;
+ pnfs_copy_range(&lgp->args.range, range);
lgp->args.type = server->pnfs_curr_ld->id;
lgp->args.inode = ino;
lgp->args.ctx = get_nfs_open_context(ctx);
@@ -904,17 +861,9 @@ send_layoutget(struct pnfs_layout_hdr *lo,
lseg = nfs4_proc_layoutget(lgp, gfp_flags);
} while (lseg == ERR_PTR(-EAGAIN));
- if (IS_ERR(lseg)) {
- switch (PTR_ERR(lseg)) {
- case -ENOMEM:
- case -ERESTARTSYS:
- break;
- default:
- /* remember that LAYOUTGET failed and suspend trying */
- pnfs_layout_io_set_failed(lo, range->iomode);
- }
- return NULL;
- } else
+ if (IS_ERR(lseg) && !nfs_error_is_fatal(PTR_ERR(lseg)))
+ lseg = NULL;
+ else
pnfs_layout_clear_fail_bit(lo,
pnfs_iomode_to_fail_bit(range->iomode));
@@ -944,8 +893,19 @@ void pnfs_clear_layoutreturn_waitbit(struct pnfs_layout_hdr *lo)
rpc_wake_up(&NFS_SERVER(lo->plh_inode)->roc_rpcwaitq);
}
+static bool
+pnfs_prepare_layoutreturn(struct pnfs_layout_hdr *lo)
+{
+ if (test_and_set_bit(NFS_LAYOUT_RETURN, &lo->plh_flags))
+ return false;
+ lo->plh_return_iomode = 0;
+ pnfs_get_layout_hdr(lo);
+ clear_bit(NFS_LAYOUT_RETURN_REQUESTED, &lo->plh_flags);
+ return true;
+}
+
static int
-pnfs_send_layoutreturn(struct pnfs_layout_hdr *lo, nfs4_stateid stateid,
+pnfs_send_layoutreturn(struct pnfs_layout_hdr *lo, const nfs4_stateid *stateid,
enum pnfs_iomode iomode, bool sync)
{
struct inode *ino = lo->plh_inode;
@@ -962,7 +922,7 @@ pnfs_send_layoutreturn(struct pnfs_layout_hdr *lo, nfs4_stateid stateid,
goto out;
}
- lrp->args.stateid = stateid;
+ nfs4_stateid_copy(&lrp->args.stateid, stateid);
lrp->args.layout_type = NFS_SERVER(ino)->pnfs_curr_ld->id;
lrp->args.inode = ino;
lrp->args.range.iomode = iomode;
@@ -978,6 +938,48 @@ out:
return status;
}
+/* Return true if layoutreturn is needed */
+static bool
+pnfs_layout_need_return(struct pnfs_layout_hdr *lo)
+{
+ struct pnfs_layout_segment *s;
+
+ if (!test_bit(NFS_LAYOUT_RETURN_REQUESTED, &lo->plh_flags))
+ return false;
+
+ /* Defer layoutreturn until all lsegs are done */
+ list_for_each_entry(s, &lo->plh_segs, pls_list) {
+ if (test_bit(NFS_LSEG_LAYOUTRETURN, &s->pls_flags))
+ return false;
+ }
+
+ return true;
+}
+
+static void pnfs_layoutreturn_before_put_layout_hdr(struct pnfs_layout_hdr *lo)
+{
+ struct inode *inode= lo->plh_inode;
+
+ if (!test_bit(NFS_LAYOUT_RETURN_REQUESTED, &lo->plh_flags))
+ return;
+ spin_lock(&inode->i_lock);
+ if (pnfs_layout_need_return(lo)) {
+ nfs4_stateid stateid;
+ enum pnfs_iomode iomode;
+ bool send;
+
+ nfs4_stateid_copy(&stateid, &lo->plh_stateid);
+ iomode = lo->plh_return_iomode;
+ send = pnfs_prepare_layoutreturn(lo);
+ spin_unlock(&inode->i_lock);
+ if (send) {
+ /* Send an async layoutreturn so we dont deadlock */
+ pnfs_send_layoutreturn(lo, &stateid, iomode, false);
+ }
+ } else
+ spin_unlock(&inode->i_lock);
+}
+
/*
* Initiates a LAYOUTRETURN(FILE), and removes the pnfs_layout_hdr
* when the layout segment list is empty.
@@ -1005,7 +1007,7 @@ _pnfs_return_layout(struct inode *ino)
dprintk("NFS: %s no layout to return\n", __func__);
goto out;
}
- stateid = nfsi->layout->plh_stateid;
+ nfs4_stateid_copy(&stateid, &nfsi->layout->plh_stateid);
/* Reference matched in nfs4_layoutreturn_release */
pnfs_get_layout_hdr(lo);
empty = list_empty(&lo->plh_segs);
@@ -1033,7 +1035,7 @@ _pnfs_return_layout(struct inode *ino)
spin_unlock(&ino->i_lock);
pnfs_free_lseg_list(&tmp_list);
if (send)
- status = pnfs_send_layoutreturn(lo, stateid, IOMODE_ANY, true);
+ status = pnfs_send_layoutreturn(lo, &stateid, IOMODE_ANY, true);
out_put_layout_hdr:
pnfs_put_layout_hdr(lo);
out:
@@ -1096,13 +1098,12 @@ bool pnfs_roc(struct inode *ino)
goto out_noroc;
}
- stateid = lo->plh_stateid;
+ nfs4_stateid_copy(&stateid, &lo->plh_stateid);
/* always send layoutreturn if being marked so */
- if (test_and_clear_bit(NFS_LAYOUT_RETURN_BEFORE_CLOSE,
+ if (test_and_clear_bit(NFS_LAYOUT_RETURN_REQUESTED,
&lo->plh_flags))
layoutreturn = pnfs_prepare_layoutreturn(lo);
- pnfs_clear_retry_layoutget(lo);
list_for_each_entry_safe(lseg, tmp, &lo->plh_segs, pls_list)
/* If we are sending layoutreturn, invalidate all valid lsegs */
if (layoutreturn || test_bit(NFS_LSEG_ROC, &lseg->pls_flags)) {
@@ -1124,7 +1125,7 @@ out_noroc:
pnfs_free_lseg_list(&tmp_list);
pnfs_layoutcommit_inode(ino, true);
if (layoutreturn)
- pnfs_send_layoutreturn(lo, stateid, IOMODE_ANY, true);
+ pnfs_send_layoutreturn(lo, &stateid, IOMODE_ANY, true);
return roc;
}
@@ -1149,6 +1150,7 @@ void pnfs_roc_set_barrier(struct inode *ino, u32 barrier)
spin_lock(&ino->i_lock);
lo = NFS_I(ino)->layout;
+ pnfs_mark_layout_returned_if_empty(lo);
if (pnfs_seqid_is_newer(barrier, lo->plh_barrier))
lo->plh_barrier = barrier;
spin_unlock(&ino->i_lock);
@@ -1465,25 +1467,15 @@ static bool pnfs_within_mdsthreshold(struct nfs_open_context *ctx,
return ret;
}
-/* stop waiting if someone clears NFS_LAYOUT_RETRY_LAYOUTGET bit. */
-static int pnfs_layoutget_retry_bit_wait(struct wait_bit_key *key, int mode)
-{
- if (!test_bit(NFS_LAYOUT_RETRY_LAYOUTGET, key->flags))
- return 1;
- return nfs_wait_bit_killable(key, mode);
-}
-
static bool pnfs_prepare_to_retry_layoutget(struct pnfs_layout_hdr *lo)
{
- if (!pnfs_should_retry_layoutget(lo))
- return false;
/*
* send layoutcommit as it can hold up layoutreturn due to lseg
* reference
*/
pnfs_layoutcommit_inode(lo->plh_inode, false);
return !wait_on_bit_action(&lo->plh_flags, NFS_LAYOUT_RETURN,
- pnfs_layoutget_retry_bit_wait,
+ nfs_wait_bit_killable,
TASK_UNINTERRUPTIBLE);
}
@@ -1520,14 +1512,23 @@ pnfs_update_layout(struct inode *ino,
struct pnfs_layout_segment *lseg = NULL;
bool first;
- if (!pnfs_enabled_sb(NFS_SERVER(ino)))
+ if (!pnfs_enabled_sb(NFS_SERVER(ino))) {
+ trace_pnfs_update_layout(ino, pos, count, iomode, NULL,
+ PNFS_UPDATE_LAYOUT_NO_PNFS);
goto out;
+ }
- if (iomode == IOMODE_READ && i_size_read(ino) == 0)
+ if (iomode == IOMODE_READ && i_size_read(ino) == 0) {
+ trace_pnfs_update_layout(ino, pos, count, iomode, NULL,
+ PNFS_UPDATE_LAYOUT_RD_ZEROLEN);
goto out;
+ }
- if (pnfs_within_mdsthreshold(ctx, ino, iomode))
+ if (pnfs_within_mdsthreshold(ctx, ino, iomode)) {
+ trace_pnfs_update_layout(ino, pos, count, iomode, NULL,
+ PNFS_UPDATE_LAYOUT_MDSTHRESH);
goto out;
+ }
lookup_again:
first = false;
@@ -1535,19 +1536,25 @@ lookup_again:
lo = pnfs_find_alloc_layout(ino, ctx, gfp_flags);
if (lo == NULL) {
spin_unlock(&ino->i_lock);
+ trace_pnfs_update_layout(ino, pos, count, iomode, NULL,
+ PNFS_UPDATE_LAYOUT_NOMEM);
goto out;
}
/* Do we even need to bother with this? */
if (test_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags)) {
+ trace_pnfs_update_layout(ino, pos, count, iomode, lo,
+ PNFS_UPDATE_LAYOUT_BULK_RECALL);
dprintk("%s matches recall, use MDS\n", __func__);
goto out_unlock;
}
/* if LAYOUTGET already failed once we don't try again */
- if (pnfs_layout_io_test_failed(lo, iomode) &&
- !pnfs_should_retry_layoutget(lo))
+ if (pnfs_layout_io_test_failed(lo, iomode)) {
+ trace_pnfs_update_layout(ino, pos, count, iomode, lo,
+ PNFS_UPDATE_LAYOUT_IO_TEST_FAIL);
goto out_unlock;
+ }
first = list_empty(&lo->plh_segs);
if (first) {
@@ -1567,8 +1574,11 @@ lookup_again:
* already exists
*/
lseg = pnfs_find_lseg(lo, &arg);
- if (lseg)
+ if (lseg) {
+ trace_pnfs_update_layout(ino, pos, count, iomode, lo,
+ PNFS_UPDATE_LAYOUT_FOUND_CACHED);
goto out_unlock;
+ }
}
/*
@@ -1585,11 +1595,16 @@ lookup_again:
dprintk("%s retrying\n", __func__);
goto lookup_again;
}
+ trace_pnfs_update_layout(ino, pos, count, iomode, lo,
+ PNFS_UPDATE_LAYOUT_RETURN);
goto out_put_layout_hdr;
}
- if (pnfs_layoutgets_blocked(lo))
+ if (pnfs_layoutgets_blocked(lo)) {
+ trace_pnfs_update_layout(ino, pos, count, iomode, lo,
+ PNFS_UPDATE_LAYOUT_BLOCKED);
goto out_unlock;
+ }
atomic_inc(&lo->plh_outstanding);
spin_unlock(&ino->i_lock);
@@ -1612,8 +1627,9 @@ lookup_again:
arg.length = PAGE_CACHE_ALIGN(arg.length);
lseg = send_layoutget(lo, ctx, &arg, gfp_flags);
- pnfs_clear_retry_layoutget(lo);
atomic_dec(&lo->plh_outstanding);
+ trace_pnfs_update_layout(ino, pos, count, iomode, lo,
+ PNFS_UPDATE_LAYOUT_SEND_LAYOUTGET);
out_put_layout_hdr:
if (first)
pnfs_clear_first_layoutget(lo);
@@ -1623,7 +1639,7 @@ out:
"(%s, offset: %llu, length: %llu)\n",
__func__, ino->i_sb->s_id,
(unsigned long long)NFS_FILEID(ino),
- lseg == NULL ? "not found" : "found",
+ IS_ERR_OR_NULL(lseg) ? "not found" : "found",
iomode==IOMODE_RW ? "read/write" : "read-only",
(unsigned long long)pos,
(unsigned long long)count);
@@ -1730,16 +1746,40 @@ out_forget_reply:
}
static void
+pnfs_set_plh_return_iomode(struct pnfs_layout_hdr *lo, enum pnfs_iomode iomode)
+{
+ if (lo->plh_return_iomode == iomode)
+ return;
+ if (lo->plh_return_iomode != 0)
+ iomode = IOMODE_ANY;
+ lo->plh_return_iomode = iomode;
+ set_bit(NFS_LAYOUT_RETURN_REQUESTED, &lo->plh_flags);
+}
+
+/**
+ * pnfs_mark_matching_lsegs_return - Free or return matching layout segments
+ * @lo: pointer to layout header
+ * @tmp_list: list header to be used with pnfs_free_lseg_list()
+ * @return_range: describe layout segment ranges to be returned
+ *
+ * This function is mainly intended for use by layoutrecall. It attempts
+ * to free the layout segment immediately, or else to mark it for return
+ * as soon as its reference count drops to zero.
+ */
+int
pnfs_mark_matching_lsegs_return(struct pnfs_layout_hdr *lo,
struct list_head *tmp_list,
- struct pnfs_layout_range *return_range)
+ const struct pnfs_layout_range *return_range)
{
struct pnfs_layout_segment *lseg, *next;
+ int remaining = 0;
dprintk("%s:Begin lo %p\n", __func__, lo);
if (list_empty(&lo->plh_segs))
- return;
+ return 0;
+
+ assert_spin_locked(&lo->plh_inode->i_lock);
list_for_each_entry_safe(lseg, next, &lo->plh_segs, pls_list)
if (should_free_lseg(&lseg->pls_range, return_range)) {
@@ -1748,39 +1788,47 @@ pnfs_mark_matching_lsegs_return(struct pnfs_layout_hdr *lo,
lseg, lseg->pls_range.iomode,
lseg->pls_range.offset,
lseg->pls_range.length);
+ if (mark_lseg_invalid(lseg, tmp_list))
+ continue;
+ remaining++;
set_bit(NFS_LSEG_LAYOUTRETURN, &lseg->pls_flags);
- mark_lseg_invalid(lseg, tmp_list);
- set_bit(NFS_LAYOUT_RETURN_BEFORE_CLOSE,
- &lo->plh_flags);
+ pnfs_set_plh_return_iomode(lo, return_range->iomode);
}
+ return remaining;
}
void pnfs_error_mark_layout_for_return(struct inode *inode,
struct pnfs_layout_segment *lseg)
{
struct pnfs_layout_hdr *lo = NFS_I(inode)->layout;
- int iomode = pnfs_iomode_to_fail_bit(lseg->pls_range.iomode);
struct pnfs_layout_range range = {
.iomode = lseg->pls_range.iomode,
.offset = 0,
.length = NFS4_MAX_UINT64,
};
LIST_HEAD(free_me);
+ bool return_now = false;
spin_lock(&inode->i_lock);
- /* set failure bit so that pnfs path will be retried later */
- pnfs_layout_set_fail_bit(lo, iomode);
- if (lo->plh_return_iomode == 0)
- lo->plh_return_iomode = range.iomode;
- else if (lo->plh_return_iomode != range.iomode)
- lo->plh_return_iomode = IOMODE_ANY;
+ pnfs_set_plh_return_iomode(lo, range.iomode);
/*
* mark all matching lsegs so that we are sure to have no live
* segments at hand when sending layoutreturn. See pnfs_put_lseg()
* for how it works.
*/
- pnfs_mark_matching_lsegs_return(lo, &free_me, &range);
- spin_unlock(&inode->i_lock);
+ if (!pnfs_mark_matching_lsegs_return(lo, &free_me, &range)) {
+ nfs4_stateid stateid;
+ enum pnfs_iomode iomode = lo->plh_return_iomode;
+
+ nfs4_stateid_copy(&stateid, &lo->plh_stateid);
+ return_now = pnfs_prepare_layoutreturn(lo);
+ spin_unlock(&inode->i_lock);
+ if (return_now)
+ pnfs_send_layoutreturn(lo, &stateid, iomode, false);
+ } else {
+ spin_unlock(&inode->i_lock);
+ nfs_commit_inode(inode, 0);
+ }
pnfs_free_lseg_list(&free_me);
}
EXPORT_SYMBOL_GPL(pnfs_error_mark_layout_for_return);
@@ -1802,6 +1850,11 @@ pnfs_generic_pg_init_read(struct nfs_pageio_descriptor *pgio, struct nfs_page *r
rd_size,
IOMODE_READ,
GFP_KERNEL);
+ if (IS_ERR(pgio->pg_lseg)) {
+ pgio->pg_error = PTR_ERR(pgio->pg_lseg);
+ pgio->pg_lseg = NULL;
+ return;
+ }
}
/* If no lseg, fall back to read through mds */
if (pgio->pg_lseg == NULL)
@@ -1814,13 +1867,19 @@ void
pnfs_generic_pg_init_write(struct nfs_pageio_descriptor *pgio,
struct nfs_page *req, u64 wb_size)
{
- if (pgio->pg_lseg == NULL)
+ if (pgio->pg_lseg == NULL) {
pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode,
req->wb_context,
req_offset(req),
wb_size,
IOMODE_RW,
GFP_NOFS);
+ if (IS_ERR(pgio->pg_lseg)) {
+ pgio->pg_error = PTR_ERR(pgio->pg_lseg);
+ pgio->pg_lseg = NULL;
+ return;
+ }
+ }
/* If no lseg, fall back to write through mds */
if (pgio->pg_lseg == NULL)
nfs_pageio_reset_write_mds(pgio);
@@ -1988,15 +2047,13 @@ static void pnfs_writehdr_free(struct nfs_pgio_header *hdr)
int
pnfs_generic_pg_writepages(struct nfs_pageio_descriptor *desc)
{
- struct nfs_pgio_mirror *mirror = nfs_pgio_current_mirror(desc);
-
struct nfs_pgio_header *hdr;
int ret;
hdr = nfs_pgio_header_alloc(desc->pg_rw_ops);
if (!hdr) {
- desc->pg_completion_ops->error_cleanup(&mirror->pg_list);
- return -ENOMEM;
+ desc->pg_error = -ENOMEM;
+ return desc->pg_error;
}
nfs_pgheader_init(desc, hdr, pnfs_writehdr_free);
@@ -2119,15 +2176,13 @@ static void pnfs_readhdr_free(struct nfs_pgio_header *hdr)
int
pnfs_generic_pg_readpages(struct nfs_pageio_descriptor *desc)
{
- struct nfs_pgio_mirror *mirror = nfs_pgio_current_mirror(desc);
-
struct nfs_pgio_header *hdr;
int ret;
hdr = nfs_pgio_header_alloc(desc->pg_rw_ops);
if (!hdr) {
- desc->pg_completion_ops->error_cleanup(&mirror->pg_list);
- return -ENOMEM;
+ desc->pg_error = -ENOMEM;
+ return desc->pg_error;
}
nfs_pgheader_init(desc, hdr, pnfs_readhdr_free);
hdr->lseg = pnfs_get_lseg(desc->pg_lseg);
diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h
index d1990e90e..1ac1db5f6 100644
--- a/fs/nfs/pnfs.h
+++ b/fs/nfs/pnfs.h
@@ -94,11 +94,10 @@ enum {
NFS_LAYOUT_RO_FAILED = 0, /* get ro layout failed stop trying */
NFS_LAYOUT_RW_FAILED, /* get rw layout failed stop trying */
NFS_LAYOUT_BULK_RECALL, /* bulk recall affecting layout */
- NFS_LAYOUT_RETURN, /* Return this layout ASAP */
- NFS_LAYOUT_RETURN_BEFORE_CLOSE, /* Return this layout before close */
+ NFS_LAYOUT_RETURN, /* layoutreturn in progress */
+ NFS_LAYOUT_RETURN_REQUESTED, /* Return this layout ASAP */
NFS_LAYOUT_INVALID_STID, /* layout stateid id is invalid */
NFS_LAYOUT_FIRST_LAYOUTGET, /* Serialize first layoutget */
- NFS_LAYOUT_RETRY_LAYOUTGET, /* Retry layoutget */
};
enum layoutdriver_policy_flags {
@@ -261,11 +260,14 @@ void pnfs_set_layout_stateid(struct pnfs_layout_hdr *lo,
bool update_barrier);
int pnfs_choose_layoutget_stateid(nfs4_stateid *dst,
struct pnfs_layout_hdr *lo,
- struct pnfs_layout_range *range,
+ const struct pnfs_layout_range *range,
struct nfs4_state *open_state);
int pnfs_mark_matching_lsegs_invalid(struct pnfs_layout_hdr *lo,
struct list_head *tmp_list,
- struct pnfs_layout_range *recall_range);
+ const struct pnfs_layout_range *recall_range);
+int pnfs_mark_matching_lsegs_return(struct pnfs_layout_hdr *lo,
+ struct list_head *tmp_list,
+ const struct pnfs_layout_range *recall_range);
bool pnfs_roc(struct inode *ino);
void pnfs_roc_release(struct inode *ino);
void pnfs_roc_set_barrier(struct inode *ino, u32 barrier);
@@ -379,26 +381,6 @@ nfs4_get_deviceid(struct nfs4_deviceid_node *d)
return d;
}
-static inline void pnfs_set_retry_layoutget(struct pnfs_layout_hdr *lo)
-{
- if (!test_and_set_bit(NFS_LAYOUT_RETRY_LAYOUTGET, &lo->plh_flags))
- atomic_inc(&lo->plh_refcount);
-}
-
-static inline void pnfs_clear_retry_layoutget(struct pnfs_layout_hdr *lo)
-{
- if (test_and_clear_bit(NFS_LAYOUT_RETRY_LAYOUTGET, &lo->plh_flags)) {
- atomic_dec(&lo->plh_refcount);
- /* wake up waiters for LAYOUTRETURN as that is not needed */
- wake_up_bit(&lo->plh_flags, NFS_LAYOUT_RETURN);
- }
-}
-
-static inline bool pnfs_should_retry_layoutget(struct pnfs_layout_hdr *lo)
-{
- return test_bit(NFS_LAYOUT_RETRY_LAYOUTGET, &lo->plh_flags);
-}
-
static inline struct pnfs_layout_segment *
pnfs_get_lseg(struct pnfs_layout_segment *lseg)
{
@@ -409,6 +391,12 @@ pnfs_get_lseg(struct pnfs_layout_segment *lseg)
return lseg;
}
+static inline bool
+pnfs_is_valid_lseg(struct pnfs_layout_segment *lseg)
+{
+ return test_bit(NFS_LSEG_VALID, &lseg->pls_flags) != 0;
+}
+
/* Return true if a layout driver is being used for this mountpoint */
static inline int pnfs_enabled_sb(struct nfs_server *nfss)
{
@@ -556,6 +544,26 @@ pnfs_calc_offset_length(u64 offset, u64 end)
return 1 + end - offset;
}
+/**
+ * pnfs_mark_layout_returned_if_empty - marks the layout as returned
+ * @lo: layout header
+ *
+ * Note: Caller must hold inode->i_lock
+ */
+static inline void
+pnfs_mark_layout_returned_if_empty(struct pnfs_layout_hdr *lo)
+{
+ if (list_empty(&lo->plh_segs))
+ set_bit(NFS_LAYOUT_INVALID_STID, &lo->plh_flags);
+}
+
+static inline void
+pnfs_copy_range(struct pnfs_layout_range *dst,
+ const struct pnfs_layout_range *src)
+{
+ memcpy(dst, src, sizeof(*dst));
+}
+
extern unsigned int layoutstats_timer;
#ifdef NFS_DEBUG
diff --git a/fs/nfs/pnfs_nfs.c b/fs/nfs/pnfs_nfs.c
index 24655b807..81ac6480f 100644
--- a/fs/nfs/pnfs_nfs.c
+++ b/fs/nfs/pnfs_nfs.c
@@ -266,17 +266,14 @@ pnfs_generic_commit_pagelist(struct inode *inode, struct list_head *mds_pages,
} else {
nfs_retry_commit(mds_pages, NULL, cinfo, 0);
pnfs_generic_retry_commit(cinfo, 0);
- cinfo->completion_ops->error_cleanup(NFS_I(inode));
return -ENOMEM;
}
}
nreq += pnfs_generic_alloc_ds_commits(cinfo, &list);
- if (nreq == 0) {
- cinfo->completion_ops->error_cleanup(NFS_I(inode));
+ if (nreq == 0)
goto out;
- }
atomic_add(nreq, &cinfo->mds->rpcs_out);
@@ -871,6 +868,11 @@ pnfs_layout_mark_request_commit(struct nfs_page *req,
buckets = cinfo->ds->buckets;
list = &buckets[ds_commit_idx].written;
if (list_empty(list)) {
+ if (!pnfs_is_valid_lseg(lseg)) {
+ spin_unlock(cinfo->lock);
+ cinfo->completion_ops->resched_write(cinfo, req);
+ return;
+ }
/* Non-empty buckets hold a reference on the lseg. That ref
* is normally transferred to the COMMIT call and released
* there. It could also be released if the last req is pulled
diff --git a/fs/nfs/read.c b/fs/nfs/read.c
index 0a5e33f33..eb31e23e7 100644
--- a/fs/nfs/read.c
+++ b/fs/nfs/read.c
@@ -85,6 +85,23 @@ void nfs_pageio_reset_read_mds(struct nfs_pageio_descriptor *pgio)
}
EXPORT_SYMBOL_GPL(nfs_pageio_reset_read_mds);
+static void nfs_readpage_release(struct nfs_page *req)
+{
+ struct inode *inode = d_inode(req->wb_context->dentry);
+
+ dprintk("NFS: read done (%s/%llu %d@%lld)\n", inode->i_sb->s_id,
+ (unsigned long long)NFS_FILEID(inode), req->wb_bytes,
+ (long long)req_offset(req));
+
+ if (nfs_page_group_sync_on_bit(req, PG_UNLOCKPAGE)) {
+ if (PageUptodate(req->wb_page))
+ nfs_readpage_to_fscache(inode, req->wb_page, 0);
+
+ unlock_page(req->wb_page);
+ }
+ nfs_release_request(req);
+}
+
int nfs_readpage_async(struct nfs_open_context *ctx, struct inode *inode,
struct page *page)
{
@@ -106,7 +123,10 @@ int nfs_readpage_async(struct nfs_open_context *ctx, struct inode *inode,
nfs_pageio_init_read(&pgio, inode, false,
&nfs_async_read_completion_ops);
- nfs_pageio_add_request(&pgio, new);
+ if (!nfs_pageio_add_request(&pgio, new)) {
+ nfs_list_remove_request(new);
+ nfs_readpage_release(new);
+ }
nfs_pageio_complete(&pgio);
/* It doesn't make sense to do mirrored reads! */
@@ -115,24 +135,7 @@ int nfs_readpage_async(struct nfs_open_context *ctx, struct inode *inode,
pgm = &pgio.pg_mirrors[0];
NFS_I(inode)->read_io += pgm->pg_bytes_written;
- return 0;
-}
-
-static void nfs_readpage_release(struct nfs_page *req)
-{
- struct inode *inode = d_inode(req->wb_context->dentry);
-
- dprintk("NFS: read done (%s/%llu %d@%lld)\n", inode->i_sb->s_id,
- (unsigned long long)NFS_FILEID(inode), req->wb_bytes,
- (long long)req_offset(req));
-
- if (nfs_page_group_sync_on_bit(req, PG_UNLOCKPAGE)) {
- if (PageUptodate(req->wb_page))
- nfs_readpage_to_fscache(inode, req->wb_page, 0);
-
- unlock_page(req->wb_page);
- }
- nfs_release_request(req);
+ return pgio.pg_error < 0 ? pgio.pg_error : 0;
}
static void nfs_page_group_set_uptodate(struct nfs_page *req)
@@ -361,6 +364,8 @@ readpage_async_filler(void *data, struct page *page)
if (len < PAGE_CACHE_SIZE)
zero_user_segment(page, len, PAGE_CACHE_SIZE);
if (!nfs_pageio_add_request(desc->pgio, new)) {
+ nfs_list_remove_request(new);
+ nfs_readpage_release(new);
error = desc->pgio->pg_error;
goto out_unlock;
}
diff --git a/fs/nfs/symlink.c b/fs/nfs/symlink.c
index b6de433da..4fe3eead3 100644
--- a/fs/nfs/symlink.c
+++ b/fs/nfs/symlink.c
@@ -42,21 +42,35 @@ error:
return -EIO;
}
-static const char *nfs_follow_link(struct dentry *dentry, void **cookie)
+static const char *nfs_get_link(struct dentry *dentry,
+ struct inode *inode,
+ struct delayed_call *done)
{
- struct inode *inode = d_inode(dentry);
struct page *page;
void *err;
- err = ERR_PTR(nfs_revalidate_mapping(inode, inode->i_mapping));
- if (err)
- return err;
- page = read_cache_page(&inode->i_data, 0,
- (filler_t *)nfs_symlink_filler, inode);
- if (IS_ERR(page))
- return ERR_CAST(page);
- *cookie = page;
- return kmap(page);
+ if (!dentry) {
+ err = ERR_PTR(nfs_revalidate_mapping_rcu(inode));
+ if (err)
+ return err;
+ page = find_get_page(inode->i_mapping, 0);
+ if (!page)
+ return ERR_PTR(-ECHILD);
+ if (!PageUptodate(page)) {
+ put_page(page);
+ return ERR_PTR(-ECHILD);
+ }
+ } else {
+ err = ERR_PTR(nfs_revalidate_mapping(inode, inode->i_mapping));
+ if (err)
+ return err;
+ page = read_cache_page(&inode->i_data, 0,
+ (filler_t *)nfs_symlink_filler, inode);
+ if (IS_ERR(page))
+ return ERR_CAST(page);
+ }
+ set_delayed_call(done, page_put_link, page);
+ return page_address(page);
}
/*
@@ -64,8 +78,7 @@ static const char *nfs_follow_link(struct dentry *dentry, void **cookie)
*/
const struct inode_operations nfs_symlink_inode_operations = {
.readlink = generic_readlink,
- .follow_link = nfs_follow_link,
- .put_link = page_put_link,
+ .get_link = nfs_get_link,
.getattr = nfs_getattr,
.setattr = nfs_setattr,
};
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index 7b9316406..5754835a2 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -21,6 +21,8 @@
#include <linux/nfs_page.h>
#include <linux/backing-dev.h>
#include <linux/export.h>
+#include <linux/freezer.h>
+#include <linux/wait.h>
#include <asm/uaccess.h>
@@ -244,11 +246,9 @@ static int wb_priority(struct writeback_control *wbc)
{
int ret = 0;
if (wbc->for_reclaim)
- return FLUSH_HIGHPRI | FLUSH_STABLE;
+ return FLUSH_HIGHPRI | FLUSH_COND_STABLE;
if (wbc->sync_mode == WB_SYNC_ALL)
ret = FLUSH_COND_STABLE;
- if (wbc->for_kupdate || wbc->for_background)
- ret |= FLUSH_LOWPRI;
return ret;
}
@@ -545,12 +545,22 @@ try_again:
return head;
}
+static void nfs_write_error_remove_page(struct nfs_page *req)
+{
+ nfs_unlock_request(req);
+ nfs_end_page_writeback(req);
+ nfs_release_request(req);
+ generic_error_remove_page(page_file_mapping(req->wb_page),
+ req->wb_page);
+}
+
/*
* Find an associated nfs write request, and prepare to flush it out
* May return an error if the user signalled nfs_wait_on_request().
*/
static int nfs_page_async_flush(struct nfs_pageio_descriptor *pgio,
- struct page *page, bool nonblock)
+ struct page *page, bool nonblock,
+ bool launder)
{
struct nfs_page *req;
int ret = 0;
@@ -567,8 +577,21 @@ static int nfs_page_async_flush(struct nfs_pageio_descriptor *pgio,
ret = 0;
if (!nfs_pageio_add_request(pgio, req)) {
- nfs_redirty_request(req);
ret = pgio->pg_error;
+ /*
+ * Remove the problematic req upon fatal errors
+ * in launder case, while other dirty pages can
+ * still be around until they get flushed.
+ */
+ if (nfs_error_is_fatal(ret)) {
+ nfs_context_set_write_error(req->wb_context, ret);
+ if (launder) {
+ nfs_write_error_remove_page(req);
+ goto out;
+ }
+ }
+ nfs_redirty_request(req);
+ ret = -EAGAIN;
} else
nfs_add_stats(page_file_mapping(page)->host,
NFSIOS_WRITEPAGES, 1);
@@ -576,12 +599,14 @@ out:
return ret;
}
-static int nfs_do_writepage(struct page *page, struct writeback_control *wbc, struct nfs_pageio_descriptor *pgio)
+static int nfs_do_writepage(struct page *page, struct writeback_control *wbc,
+ struct nfs_pageio_descriptor *pgio, bool launder)
{
int ret;
nfs_pageio_cond_complete(pgio, page_file_index(page));
- ret = nfs_page_async_flush(pgio, page, wbc->sync_mode == WB_SYNC_NONE);
+ ret = nfs_page_async_flush(pgio, page, wbc->sync_mode == WB_SYNC_NONE,
+ launder);
if (ret == -EAGAIN) {
redirty_page_for_writepage(wbc, page);
ret = 0;
@@ -592,7 +617,9 @@ static int nfs_do_writepage(struct page *page, struct writeback_control *wbc, st
/*
* Write an mmapped page to the server.
*/
-static int nfs_writepage_locked(struct page *page, struct writeback_control *wbc)
+static int nfs_writepage_locked(struct page *page,
+ struct writeback_control *wbc,
+ bool launder)
{
struct nfs_pageio_descriptor pgio;
struct inode *inode = page_file_mapping(page)->host;
@@ -601,7 +628,7 @@ static int nfs_writepage_locked(struct page *page, struct writeback_control *wbc
nfs_inc_stats(inode, NFSIOS_VFSWRITEPAGE);
nfs_pageio_init_write(&pgio, inode, wb_priority(wbc),
false, &nfs_async_write_completion_ops);
- err = nfs_do_writepage(page, wbc, &pgio);
+ err = nfs_do_writepage(page, wbc, &pgio, launder);
nfs_pageio_complete(&pgio);
if (err < 0)
return err;
@@ -614,7 +641,7 @@ int nfs_writepage(struct page *page, struct writeback_control *wbc)
{
int ret;
- ret = nfs_writepage_locked(page, wbc);
+ ret = nfs_writepage_locked(page, wbc, false);
unlock_page(page);
return ret;
}
@@ -623,7 +650,7 @@ static int nfs_writepages_callback(struct page *page, struct writeback_control *
{
int ret;
- ret = nfs_do_writepage(page, wbc, data);
+ ret = nfs_do_writepage(page, wbc, data, false);
unlock_page(page);
return ret;
}
@@ -803,11 +830,10 @@ EXPORT_SYMBOL_GPL(nfs_request_add_commit_list_locked);
* holding the nfs_page lock.
*/
void
-nfs_request_add_commit_list(struct nfs_page *req, struct list_head *dst,
- struct nfs_commit_info *cinfo)
+nfs_request_add_commit_list(struct nfs_page *req, struct nfs_commit_info *cinfo)
{
spin_lock(cinfo->lock);
- nfs_request_add_commit_list_locked(req, dst, cinfo);
+ nfs_request_add_commit_list_locked(req, &cinfo->mds->list, cinfo);
spin_unlock(cinfo->lock);
nfs_mark_page_unstable(req->wb_page, cinfo);
}
@@ -865,7 +891,7 @@ nfs_mark_request_commit(struct nfs_page *req, struct pnfs_layout_segment *lseg,
{
if (pnfs_mark_request_commit(req, lseg, cinfo, ds_commit_idx))
return;
- nfs_request_add_commit_list(req, &cinfo->mds->list, cinfo);
+ nfs_request_add_commit_list(req, cinfo);
}
static void
@@ -1128,7 +1154,8 @@ int nfs_flush_incompatible(struct file *file, struct page *page)
if (req == NULL)
return 0;
l_ctx = req->wb_lock_context;
- do_flush = req->wb_page != page || req->wb_context != ctx;
+ do_flush = req->wb_page != page ||
+ !nfs_match_open_context(req->wb_context, ctx);
/* for now, flush if more than 1 request in page_group */
do_flush |= req->wb_this_page != req;
if (l_ctx && flctx &&
@@ -1326,9 +1353,15 @@ static void nfs_async_write_error(struct list_head *head)
}
}
+static void nfs_async_write_reschedule_io(struct nfs_pgio_header *hdr)
+{
+ nfs_async_write_error(&hdr->pages);
+}
+
static const struct nfs_pgio_completion_ops nfs_async_write_completion_ops = {
.error_cleanup = nfs_async_write_error,
.completion = nfs_write_completion,
+ .reschedule_io = nfs_async_write_reschedule_io,
};
void nfs_pageio_init_write(struct nfs_pageio_descriptor *pgio,
@@ -1529,27 +1562,21 @@ static void nfs_writeback_result(struct rpc_task *task,
}
}
-
-static int nfs_commit_set_lock(struct nfs_inode *nfsi, int may_wait)
+static int wait_on_commit(struct nfs_mds_commit_info *cinfo)
{
- int ret;
+ return wait_on_atomic_t(&cinfo->rpcs_out,
+ nfs_wait_atomic_killable, TASK_KILLABLE);
+}
- if (!test_and_set_bit(NFS_INO_COMMIT, &nfsi->flags))
- return 1;
- if (!may_wait)
- return 0;
- ret = out_of_line_wait_on_bit_lock(&nfsi->flags,
- NFS_INO_COMMIT,
- nfs_wait_bit_killable,
- TASK_KILLABLE);
- return (ret < 0) ? ret : 1;
+static void nfs_commit_begin(struct nfs_mds_commit_info *cinfo)
+{
+ atomic_inc(&cinfo->rpcs_out);
}
-static void nfs_commit_clear_lock(struct nfs_inode *nfsi)
+static void nfs_commit_end(struct nfs_mds_commit_info *cinfo)
{
- clear_bit(NFS_INO_COMMIT, &nfsi->flags);
- smp_mb__after_atomic();
- wake_up_bit(&nfsi->flags, NFS_INO_COMMIT);
+ if (atomic_dec_and_test(&cinfo->rpcs_out))
+ wake_up_atomic_t(&cinfo->rpcs_out);
}
void nfs_commitdata_release(struct nfs_commit_data *data)
@@ -1666,6 +1693,13 @@ void nfs_retry_commit(struct list_head *page_list,
}
EXPORT_SYMBOL_GPL(nfs_retry_commit);
+static void
+nfs_commit_resched_write(struct nfs_commit_info *cinfo,
+ struct nfs_page *req)
+{
+ __set_page_dirty_nobuffers(req->wb_page);
+}
+
/*
* Commit dirty pages
*/
@@ -1687,7 +1721,6 @@ nfs_commit_list(struct inode *inode, struct list_head *head, int how,
data->mds_ops, how, 0);
out_bad:
nfs_retry_commit(head, NULL, cinfo, 0);
- cinfo->completion_ops->error_cleanup(NFS_I(inode));
return -ENOMEM;
}
@@ -1749,8 +1782,7 @@ static void nfs_commit_release_pages(struct nfs_commit_data *data)
clear_bdi_congested(&nfss->backing_dev_info, BLK_RW_ASYNC);
nfs_init_cinfo(&cinfo, data->inode, data->dreq);
- if (atomic_dec_and_test(&cinfo.mds->rpcs_out))
- nfs_commit_clear_lock(NFS_I(data->inode));
+ nfs_commit_end(cinfo.mds);
}
static void nfs_commit_release(void *calldata)
@@ -1769,7 +1801,7 @@ static const struct rpc_call_ops nfs_commit_ops = {
static const struct nfs_commit_completion_ops nfs_commit_completion_ops = {
.completion = nfs_commit_release_pages,
- .error_cleanup = nfs_commit_clear_lock,
+ .resched_write = nfs_commit_resched_write,
};
int nfs_generic_commit_list(struct inode *inode, struct list_head *head,
@@ -1788,30 +1820,25 @@ int nfs_commit_inode(struct inode *inode, int how)
LIST_HEAD(head);
struct nfs_commit_info cinfo;
int may_wait = how & FLUSH_SYNC;
+ int error = 0;
int res;
- res = nfs_commit_set_lock(NFS_I(inode), may_wait);
- if (res <= 0)
- goto out_mark_dirty;
nfs_init_cinfo_from_inode(&cinfo, inode);
+ nfs_commit_begin(cinfo.mds);
res = nfs_scan_commit(inode, &head, &cinfo);
- if (res) {
- int error;
-
+ if (res)
error = nfs_generic_commit_list(inode, &head, how, &cinfo);
- if (error < 0)
- return error;
- if (!may_wait)
- goto out_mark_dirty;
- error = wait_on_bit_action(&NFS_I(inode)->flags,
- NFS_INO_COMMIT,
- nfs_wait_bit_killable,
- TASK_KILLABLE);
- if (error < 0)
- return error;
- } else
- nfs_commit_clear_lock(NFS_I(inode));
+ nfs_commit_end(cinfo.mds);
+ if (error < 0)
+ goto out_error;
+ if (!may_wait)
+ goto out_mark_dirty;
+ error = wait_on_commit(cinfo.mds);
+ if (error < 0)
+ return error;
return res;
+out_error:
+ res = error;
/* Note: If we exit without ensuring that the commit is complete,
* we must mark the inode as dirty. Otherwise, future calls to
* sync_inode() with the WB_SYNC_ALL flag set will fail to ensure
@@ -1821,6 +1848,7 @@ out_mark_dirty:
__mark_inode_dirty(inode, I_DIRTY_DATASYNC);
return res;
}
+EXPORT_SYMBOL_GPL(nfs_commit_inode);
int nfs_write_inode(struct inode *inode, struct writeback_control *wbc)
{
@@ -1911,7 +1939,7 @@ int nfs_wb_page_cancel(struct inode *inode, struct page *page)
/*
* Write back all requests on one page - we do this before reading it.
*/
-int nfs_wb_page(struct inode *inode, struct page *page)
+int nfs_wb_single_page(struct inode *inode, struct page *page, bool launder)
{
loff_t range_start = page_file_offset(page);
loff_t range_end = range_start + (loff_t)(PAGE_CACHE_SIZE - 1);
@@ -1928,7 +1956,7 @@ int nfs_wb_page(struct inode *inode, struct page *page)
for (;;) {
wait_on_page_writeback(page);
if (clear_page_dirty_for_io(page)) {
- ret = nfs_writepage_locked(page, &wbc);
+ ret = nfs_writepage_locked(page, &wbc, launder);
if (ret < 0)
goto out_error;
continue;