/**
* nfs_create_request - Create an NFS read/write request.
* @ctx: open context to use
* @inode: inode to which the request is attached
* @page: page to write
* @offset: starting offset within the page for the write
* @count: number of bytes to read/write
*
* The page must be locked by the caller. This makes sure we never
* create two different requests for the same page.
* User should ensure it is safe to sleep in this function.
*/
struct nfs_page *
nfs_create_request(struct nfs_open_context *ctx, struct inode *inode,
struct page *page,
unsigned int offset, unsigned int count)
{
struct nfs_page *req;
struct nfs_lock_context *l_ctx;
/* try to allocate the request struct */
req = nfs_page_alloc();
if (req == NULL)
return ERR_PTR(-ENOMEM);
/* get lock context early so we can deal with alloc failures */
l_ctx = nfs_get_lock_context(ctx);
if (IS_ERR(l_ctx)) {
nfs_page_free(req);
return ERR_CAST(l_ctx);
}
req->wb_lock_context = l_ctx;
/* Initialize the request struct. Initially, we assume a
* long write-back delay. This will be adjusted in
* update_nfs_request below if the region is not locked. */
req->wb_page = page;
req->wb_index = page_file_index(page);
page_cache_get(page);
req->wb_offset = offset;
req->wb_pgbase = offset;
req->wb_bytes = count;
req->wb_context = get_nfs_open_context(ctx);
kref_init(&req->wb_kref);
return req;
}
/**
* nfs_create_request - Create an NFS read/write request.
* @file: file descriptor to use
* @inode: inode to which the request is attached
* @page: page to write
* @offset: starting offset within the page for the write
* @count: number of bytes to read/write
*
* The page must be locked by the caller. This makes sure we never
* create two different requests for the same page.
* User should ensure it is safe to sleep in this function.
*/
struct nfs_page *
nfs_create_request(struct nfs_open_context *ctx, struct inode *inode,
struct page *page,
unsigned int offset, unsigned int count)
{
struct nfs_page *req;
/* try to allocate the request struct */
req = nfs_page_alloc();
if (req == NULL)
return ERR_PTR(-ENOMEM);
/* Initialize the request struct. Initially, we assume a
* long write-back delay. This will be adjusted in
* update_nfs_request below if the region is not locked. */
req->wb_page = page;
atomic_set(&req->wb_complete, 0);
req->wb_index = page->index;
page_cache_get(page);
BUG_ON(PagePrivate(page));
BUG_ON(!PageLocked(page));
BUG_ON(page->mapping->host != inode);
req->wb_offset = offset;
req->wb_pgbase = offset;
req->wb_bytes = count;
req->wb_context = get_nfs_open_context(ctx);
kref_init(&req->wb_kref);
return req;
}
/**
* nfs_pgio_rpcsetup - Set up arguments for a pageio call
* @hdr: The pageio hdr
* @count: Number of bytes to read
* @offset: Initial offset
* @how: How to commit data (writes only)
* @cinfo: Commit information for the call (writes only)
*/
static void nfs_pgio_rpcsetup(struct nfs_pgio_header *hdr,
unsigned int count, unsigned int offset,
int how, struct nfs_commit_info *cinfo)
{
struct nfs_page *req = hdr->req;
/* Set up the RPC argument and reply structs
* NB: take care not to mess about with hdr->commit et al. */
hdr->args.fh = NFS_FH(hdr->inode);
hdr->args.offset = req_offset(req) + offset;
/* pnfs_set_layoutcommit needs this */
hdr->mds_offset = hdr->args.offset;
hdr->args.pgbase = req->wb_pgbase + offset;
hdr->args.pages = hdr->page_array.pagevec;
hdr->args.count = count;
hdr->args.context = get_nfs_open_context(req->wb_context);
hdr->args.lock_context = req->wb_lock_context;
hdr->args.stable = NFS_UNSTABLE;
switch (how & (FLUSH_STABLE | FLUSH_COND_STABLE)) {
case 0:
break;
case FLUSH_COND_STABLE:
if (nfs_reqs_to_commit(cinfo))
break;
default:
hdr->args.stable = NFS_FILE_SYNC;
}
hdr->res.fattr = &hdr->fattr;
hdr->res.count = count;
hdr->res.eof = 0;
hdr->res.verf = &hdr->verf;
nfs_fattr_init(&hdr->fattr);
}
static void nfs3_nlm_alloc_call(void *data)
{
struct nfs_lock_context *l_ctx = data;
if (l_ctx && test_bit(NFS_CONTEXT_UNLOCK, &l_ctx->open_context->flags)) {
get_nfs_open_context(l_ctx->open_context);
nfs_get_lock_context(l_ctx->open_context);
}
}
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;
}
/**
* nfs_create_request - Create an NFS read/write request.
* @file: file descriptor to use
* @inode: inode to which the request is attached
* @page: page to write
* @offset: starting offset within the page for the write
* @count: number of bytes to read/write
*
* The page must be locked by the caller. This makes sure we never
* create two different requests for the same page, and avoids
* a possible deadlock when we reach the hard limit on the number
* of dirty pages.
* User should ensure it is safe to sleep in this function.
*/
struct nfs_page *
nfs_create_request(struct nfs_open_context *ctx, struct inode *inode,
struct page *page,
unsigned int offset, unsigned int count)
{
struct nfs_server *server = NFS_SERVER(inode);
struct nfs_page *req;
/* Deal with hard limits. */
for (;;) {
/* try to allocate the request struct */
req = nfs_page_alloc();
if (req != NULL)
break;
/* Try to free up at least one request in order to stay
* below the hard limit
*/
if (signalled() && (server->flags & NFS_MOUNT_INTR))
return ERR_PTR(-ERESTARTSYS);
yield();
}
/* Initialize the request struct. Initially, we assume a
* long write-back delay. This will be adjusted in
* update_nfs_request below if the region is not locked. */
req->wb_page = page;
atomic_set(&req->wb_complete, 0);
req->wb_index = page->index;
page_cache_get(page);
BUG_ON(PagePrivate(page));
BUG_ON(!PageLocked(page));
BUG_ON(page->mapping->host != inode);
req->wb_offset = offset;
req->wb_pgbase = offset;
req->wb_bytes = count;
atomic_set(&req->wb_count, 1);
req->wb_context = get_nfs_open_context(ctx);
return req;
}
/*
* Set up the argument/result storage required for the RPC call.
*/
static int nfs_write_rpcsetup(struct nfs_page *req,
struct nfs_write_data *data,
const struct rpc_call_ops *call_ops,
unsigned int count, unsigned int offset,
int how)
{
struct inode *inode = req->wb_context->path.dentry->d_inode;
int flags = (how & FLUSH_SYNC) ? 0 : RPC_TASK_ASYNC;
int priority = flush_task_priority(how);
struct rpc_task *task;
struct rpc_message msg = {
.rpc_argp = &data->args,
.rpc_resp = &data->res,
.rpc_cred = req->wb_context->cred,
};
struct rpc_task_setup task_setup_data = {
.rpc_client = NFS_CLIENT(inode),
.task = &data->task,
.rpc_message = &msg,
.callback_ops = call_ops,
.callback_data = data,
.workqueue = nfsiod_workqueue,
.flags = flags,
.priority = priority,
};
/* Set up the RPC argument and reply structs
* NB: take care not to mess about with data->commit et al. */
data->req = req;
data->inode = inode = req->wb_context->path.dentry->d_inode;
data->cred = msg.rpc_cred;
data->args.fh = NFS_FH(inode);
data->args.offset = req_offset(req) + offset;
data->args.pgbase = req->wb_pgbase + offset;
data->args.pages = data->pagevec;
data->args.count = count;
data->args.context = get_nfs_open_context(req->wb_context);
data->args.stable = NFS_UNSTABLE;
if (how & FLUSH_STABLE) {
data->args.stable = NFS_DATA_SYNC;
if (!nfs_need_commit(NFS_I(inode)))
data->args.stable = NFS_FILE_SYNC;
}
data->res.fattr = &data->fattr;
data->res.count = count;
data->res.verf = &data->verf;
nfs_fattr_init(&data->fattr);
/* Set up the initial task struct. */
NFS_PROTO(inode)->write_setup(data, &msg);
dprintk("NFS: %5u initiated write call "
"(req %s/%lld, %u bytes @ offset %llu)\n",
data->task.tk_pid,
inode->i_sb->s_id,
(long long)NFS_FILEID(inode),
count,
(unsigned long long)data->args.offset);
task = rpc_run_task(&task_setup_data);
if (IS_ERR(task))
return PTR_ERR(task);
rpc_put_task(task);
return 0;
}
/* If a nfs_flush_* function fails, it should remove reqs from @head and
* call this on each, which will prepare them to be retried on next
* writeback using standard nfs.
*/
static void nfs_redirty_request(struct nfs_page *req)
{
nfs_mark_request_dirty(req);
nfs_end_page_writeback(req->wb_page);
nfs_clear_page_tag_locked(req);
}
/*
* Generate multiple small requests to write out a single
* contiguous dirty area on one page.
*/
static int nfs_flush_multi(struct inode *inode, struct list_head *head, unsigned int npages, size_t count, int how)
{
struct nfs_page *req = nfs_list_entry(head->next);
struct page *page = req->wb_page;
struct nfs_write_data *data;
size_t wsize = NFS_SERVER(inode)->wsize, nbytes;
unsigned int offset;
int requests = 0;
int ret = 0;
LIST_HEAD(list);
nfs_list_remove_request(req);
nbytes = count;
do {
size_t len = min(nbytes, wsize);
data = nfs_writedata_alloc(1);
if (!data)
goto out_bad;
list_add(&data->pages, &list);
requests++;
nbytes -= len;
} while (nbytes != 0);
atomic_set(&req->wb_complete, requests);
ClearPageError(page);
offset = 0;
nbytes = count;
do {
int ret2;
data = list_entry(list.next, struct nfs_write_data, pages);
list_del_init(&data->pages);
data->pagevec[0] = page;
if (nbytes < wsize)
wsize = nbytes;
ret2 = nfs_write_rpcsetup(req, data, &nfs_write_partial_ops,
wsize, offset, how);
if (ret == 0)
ret = ret2;
offset += wsize;
nbytes -= wsize;
} while (nbytes != 0);
return ret;
out_bad:
while (!list_empty(&list)) {
data = list_entry(list.next, struct nfs_write_data, pages);
list_del(&data->pages);
nfs_writedata_release(data);
}
nfs_redirty_request(req);
return -ENOMEM;
}
/*
* Create an RPC task for the given write request and kick it.
* The page must have been locked by the caller.
*
* It may happen that the page we're passed is not marked dirty.
* This is the case if nfs_updatepage detects a conflicting request
* that has been written but not committed.
*/
static int nfs_flush_one(struct inode *inode, struct list_head *head, unsigned int npages, size_t count, int how)
{
struct nfs_page *req;
struct page **pages;
struct nfs_write_data *data;
data = nfs_writedata_alloc(npages);
if (!data)
goto out_bad;
pages = data->pagevec;
while (!list_empty(head)) {
req = nfs_list_entry(head->next);
nfs_list_remove_request(req);
nfs_list_add_request(req, &data->pages);
ClearPageError(req->wb_page);
*pages++ = req->wb_page;
}
req = nfs_list_entry(data->pages.next);
/* Set up the argument struct */
return nfs_write_rpcsetup(req, data, &nfs_write_full_ops, count, 0, how);
out_bad:
while (!list_empty(head)) {
req = nfs_list_entry(head->next);
nfs_list_remove_request(req);
nfs_redirty_request(req);
}
return -ENOMEM;
}
static void nfs_pageio_init_write(struct nfs_pageio_descriptor *pgio,
struct inode *inode, int ioflags)
{
size_t wsize = NFS_SERVER(inode)->wsize;
if (wsize < PAGE_CACHE_SIZE)
nfs_pageio_init(pgio, inode, nfs_flush_multi, wsize, ioflags);
else
nfs_pageio_init(pgio, inode, nfs_flush_one, wsize, ioflags);
}
/*
* Set up the argument/result storage required for the RPC call.
*/
static int nfs_commit_rpcsetup(struct list_head *head,
struct nfs_write_data *data,
int how)
{
struct nfs_page *first = nfs_list_entry(head->next);
struct inode *inode = first->wb_context->path.dentry->d_inode;
int flags = (how & FLUSH_SYNC) ? 0 : RPC_TASK_ASYNC;
int priority = flush_task_priority(how);
struct rpc_task *task;
struct rpc_message msg = {
.rpc_argp = &data->args,
.rpc_resp = &data->res,
.rpc_cred = first->wb_context->cred,
};
struct rpc_task_setup task_setup_data = {
.task = &data->task,
.rpc_client = NFS_CLIENT(inode),
.rpc_message = &msg,
.callback_ops = &nfs_commit_ops,
.callback_data = data,
.workqueue = nfsiod_workqueue,
.flags = flags,
.priority = priority,
};
/* Set up the RPC argument and reply structs
* NB: take care not to mess about with data->commit et al. */
list_splice_init(head, &data->pages);
data->inode = inode;
data->cred = msg.rpc_cred;
data->args.fh = NFS_FH(data->inode);
/* Note: we always request a commit of the entire inode */
data->args.offset = 0;
data->args.count = 0;
data->args.context = get_nfs_open_context(first->wb_context);
data->res.count = 0;
data->res.fattr = &data->fattr;
data->res.verf = &data->verf;
nfs_fattr_init(&data->fattr);
/* Set up the initial task struct. */
NFS_PROTO(inode)->commit_setup(data, &msg);
dprintk("NFS: %5u initiated commit call\n", data->task.tk_pid);
task = rpc_run_task(&task_setup_data);
if (IS_ERR(task))
return PTR_ERR(task);
rpc_put_task(task);
return 0;
}
/*
* Commit dirty pages
*/
static int
nfs_commit_list(struct inode *inode, struct list_head *head, int how)
{
struct nfs_write_data *data;
struct nfs_page *req;
data = nfs_commitdata_alloc();
if (!data)
goto out_bad;
/* Set up the argument struct */
return nfs_commit_rpcsetup(head, data, how);
out_bad:
while (!list_empty(head)) {
req = nfs_list_entry(head->next);
nfs_list_remove_request(req);
nfs_mark_request_commit(req);
dec_zone_page_state(req->wb_page, NR_UNSTABLE_NFS);
dec_bdi_stat(req->wb_page->mapping->backing_dev_info,
BDI_RECLAIMABLE);
nfs_clear_page_tag_locked(req);
}
return -ENOMEM;
}
/*
* COMMIT call returned
*/
static void nfs_commit_done(struct rpc_task *task, void *calldata)
{
struct nfs_write_data *data = calldata;
dprintk("NFS: %5u nfs_commit_done (status %d)\n",
task->tk_pid, task->tk_status);
/* Call the NFS version-specific code */
if (NFS_PROTO(data->inode)->commit_done(task, data) != 0)
return;
}
static void nfs_commit_release(void *calldata)
{
struct nfs_write_data *data = calldata;
struct nfs_page *req;
int status = data->task.tk_status;
while (!list_empty(&data->pages)) {
req = nfs_list_entry(data->pages.next);
nfs_list_remove_request(req);
nfs_clear_request_commit(req);
dprintk("NFS: commit (%s/%lld %d@%lld)",
req->wb_context->path.dentry->d_inode->i_sb->s_id,
(long long)NFS_FILEID(req->wb_context->path.dentry->d_inode),
req->wb_bytes,
(long long)req_offset(req));
if (status < 0) {
nfs_context_set_write_error(req->wb_context, status);
nfs_inode_remove_request(req);
dprintk(", error = %d\n", status);
goto next;
}
/* Okay, COMMIT succeeded, apparently. Check the verifier
* returned by the server against all stored verfs. */
if (!memcmp(req->wb_verf.verifier, data->verf.verifier, sizeof(data->verf.verifier))) {
/* We have a match */
nfs_inode_remove_request(req);
dprintk(" OK\n");
goto next;
}
/* We have a mismatch. Write the page again */
dprintk(" mismatch\n");
nfs_mark_request_dirty(req);
next:
nfs_clear_page_tag_locked(req);
}
nfs_commitdata_release(calldata);
}
static const struct rpc_call_ops nfs_commit_ops = {
#if defined(CONFIG_NFS_V4_1)
.rpc_call_prepare = nfs_write_prepare,
#endif /* CONFIG_NFS_V4_1 */
.rpc_call_done = nfs_commit_done,
.rpc_release = nfs_commit_release,
};
int nfs_commit_inode(struct inode *inode, int how)
{
LIST_HEAD(head);
int res;
spin_lock(&inode->i_lock);
res = nfs_scan_commit(inode, &head, 0, 0);
spin_unlock(&inode->i_lock);
if (res) {
int error = nfs_commit_list(inode, &head, how);
if (error < 0)
return error;
}
return res;
}
#else
static inline int nfs_commit_list(struct inode *inode, struct list_head *head, int how)
{
return 0;
}
#endif
long nfs_sync_mapping_wait(struct address_space *mapping, struct writeback_control *wbc, int how)
{
struct inode *inode = mapping->host;
pgoff_t idx_start, idx_end;
unsigned int npages = 0;
LIST_HEAD(head);
int nocommit = how & FLUSH_NOCOMMIT;
long pages, ret;
/* FIXME */
if (wbc->range_cyclic)
idx_start = 0;
else {
idx_start = wbc->range_start >> PAGE_CACHE_SHIFT;
idx_end = wbc->range_end >> PAGE_CACHE_SHIFT;
if (idx_end > idx_start) {
pgoff_t l_npages = 1 + idx_end - idx_start;
npages = l_npages;
if (sizeof(npages) != sizeof(l_npages) &&
(pgoff_t)npages != l_npages)
npages = 0;
}
}
how &= ~FLUSH_NOCOMMIT;
spin_lock(&inode->i_lock);
do {
ret = nfs_wait_on_requests_locked(inode, idx_start, npages);
if (ret != 0)
continue;
if (nocommit)
break;
pages = nfs_scan_commit(inode, &head, idx_start, npages);
if (pages == 0)
break;
if (how & FLUSH_INVALIDATE) {
spin_unlock(&inode->i_lock);
nfs_cancel_commit_list(&head);
ret = pages;
spin_lock(&inode->i_lock);
continue;
}
pages += nfs_scan_commit(inode, &head, 0, 0);
spin_unlock(&inode->i_lock);
ret = nfs_commit_list(inode, &head, how);
spin_lock(&inode->i_lock);
} while (ret >= 0);
spin_unlock(&inode->i_lock);
return ret;
}
static int __nfs_write_mapping(struct address_space *mapping, struct writeback_control *wbc, int how)
{
int ret;
ret = nfs_writepages(mapping, wbc);
if (ret < 0)
goto out;
ret = nfs_sync_mapping_wait(mapping, wbc, how);
if (ret < 0)
goto out;
return 0;
out:
__mark_inode_dirty(mapping->host, I_DIRTY_PAGES);
return ret;
}
/*
* Get an extra reference on a read context.
* - This function can be absent if the completion function doesn't require a
* context.
* - The read context is passed back to NFS in the event that a data read on the
* cache fails with EIO - in which case the server must be contacted to
* retrieve the data, which requires the read context for security.
*/
static void nfs_fh_get_context(void *cookie_netfs_data, void *context)
{
get_nfs_open_context(context);
}
