2005-04-16 18:20:36 -04:00
|
|
|
/*
|
|
|
|
* linux/fs/nfs/pagelist.c
|
|
|
|
*
|
|
|
|
* A set of helper functions for managing NFS read and write requests.
|
|
|
|
* The main purpose of these routines is to provide support for the
|
|
|
|
* coalescing of several requests into a single RPC call.
|
|
|
|
*
|
|
|
|
* Copyright 2000, 2001 (c) Trond Myklebust <trond.myklebust@fys.uio.no>
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/slab.h>
|
|
|
|
#include <linux/file.h>
|
|
|
|
#include <linux/sunrpc/clnt.h>
|
|
|
|
#include <linux/nfs3.h>
|
|
|
|
#include <linux/nfs4.h>
|
|
|
|
#include <linux/nfs_page.h>
|
|
|
|
#include <linux/nfs_fs.h>
|
|
|
|
#include <linux/nfs_mount.h>
|
2006-09-17 14:46:44 -04:00
|
|
|
#include <linux/writeback.h>
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
#define NFS_PARANOIA 1
|
|
|
|
|
2006-12-06 23:33:20 -05:00
|
|
|
static struct kmem_cache *nfs_page_cachep;
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
static inline struct nfs_page *
|
|
|
|
nfs_page_alloc(void)
|
|
|
|
{
|
|
|
|
struct nfs_page *p;
|
2006-12-06 23:33:17 -05:00
|
|
|
p = kmem_cache_alloc(nfs_page_cachep, GFP_KERNEL);
|
2005-04-16 18:20:36 -04:00
|
|
|
if (p) {
|
|
|
|
memset(p, 0, sizeof(*p));
|
|
|
|
INIT_LIST_HEAD(&p->wb_list);
|
|
|
|
}
|
|
|
|
return p;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void
|
|
|
|
nfs_page_free(struct nfs_page *p)
|
|
|
|
{
|
|
|
|
kmem_cache_free(nfs_page_cachep, p);
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* 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);
|
2006-03-20 13:44:04 -05:00
|
|
|
BUG_ON(PagePrivate(page));
|
|
|
|
BUG_ON(!PageLocked(page));
|
|
|
|
BUG_ON(page->mapping->host != inode);
|
2005-04-16 18:20:36 -04:00
|
|
|
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;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* nfs_unlock_request - Unlock request and wake up sleepers.
|
|
|
|
* @req:
|
|
|
|
*/
|
|
|
|
void nfs_unlock_request(struct nfs_page *req)
|
|
|
|
{
|
|
|
|
if (!NFS_WBACK_BUSY(req)) {
|
|
|
|
printk(KERN_ERR "NFS: Invalid unlock attempted\n");
|
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
smp_mb__before_clear_bit();
|
|
|
|
clear_bit(PG_BUSY, &req->wb_flags);
|
|
|
|
smp_mb__after_clear_bit();
|
2005-06-22 13:16:21 -04:00
|
|
|
wake_up_bit(&req->wb_flags, PG_BUSY);
|
2005-04-16 18:20:36 -04:00
|
|
|
nfs_release_request(req);
|
|
|
|
}
|
|
|
|
|
2005-06-22 13:16:30 -04:00
|
|
|
/**
|
|
|
|
* nfs_set_page_writeback_locked - Lock a request for writeback
|
|
|
|
* @req:
|
|
|
|
*/
|
|
|
|
int nfs_set_page_writeback_locked(struct nfs_page *req)
|
|
|
|
{
|
|
|
|
struct nfs_inode *nfsi = NFS_I(req->wb_context->dentry->d_inode);
|
|
|
|
|
|
|
|
if (!nfs_lock_request(req))
|
|
|
|
return 0;
|
|
|
|
radix_tree_tag_set(&nfsi->nfs_page_tree, req->wb_index, NFS_PAGE_TAG_WRITEBACK);
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* nfs_clear_page_writeback - Unlock request and wake up sleepers
|
|
|
|
*/
|
|
|
|
void nfs_clear_page_writeback(struct nfs_page *req)
|
|
|
|
{
|
|
|
|
struct nfs_inode *nfsi = NFS_I(req->wb_context->dentry->d_inode);
|
|
|
|
|
2006-03-20 13:44:50 -05:00
|
|
|
if (req->wb_page != NULL) {
|
|
|
|
spin_lock(&nfsi->req_lock);
|
|
|
|
radix_tree_tag_clear(&nfsi->nfs_page_tree, req->wb_index, NFS_PAGE_TAG_WRITEBACK);
|
|
|
|
spin_unlock(&nfsi->req_lock);
|
|
|
|
}
|
2005-06-22 13:16:30 -04:00
|
|
|
nfs_unlock_request(req);
|
|
|
|
}
|
|
|
|
|
2005-04-16 18:20:36 -04:00
|
|
|
/**
|
|
|
|
* nfs_clear_request - Free up all resources allocated to the request
|
|
|
|
* @req:
|
|
|
|
*
|
|
|
|
* Release page resources associated with a write request after it
|
|
|
|
* has completed.
|
|
|
|
*/
|
|
|
|
void nfs_clear_request(struct nfs_page *req)
|
|
|
|
{
|
2006-03-20 13:44:04 -05:00
|
|
|
struct page *page = req->wb_page;
|
|
|
|
if (page != NULL) {
|
|
|
|
page_cache_release(page);
|
2005-04-16 18:20:36 -04:00
|
|
|
req->wb_page = NULL;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
* nfs_release_request - Release the count on an NFS read/write request
|
|
|
|
* @req: request to release
|
|
|
|
*
|
|
|
|
* Note: Should never be called with the spinlock held!
|
|
|
|
*/
|
|
|
|
void
|
|
|
|
nfs_release_request(struct nfs_page *req)
|
|
|
|
{
|
|
|
|
if (!atomic_dec_and_test(&req->wb_count))
|
|
|
|
return;
|
|
|
|
|
|
|
|
#ifdef NFS_PARANOIA
|
|
|
|
BUG_ON (!list_empty(&req->wb_list));
|
|
|
|
BUG_ON (NFS_WBACK_BUSY(req));
|
|
|
|
#endif
|
|
|
|
|
|
|
|
/* Release struct file or cached credential */
|
|
|
|
nfs_clear_request(req);
|
|
|
|
put_nfs_open_context(req->wb_context);
|
|
|
|
nfs_page_free(req);
|
|
|
|
}
|
|
|
|
|
2005-06-22 13:16:21 -04:00
|
|
|
static int nfs_wait_bit_interruptible(void *word)
|
|
|
|
{
|
|
|
|
int ret = 0;
|
|
|
|
|
|
|
|
if (signal_pending(current))
|
|
|
|
ret = -ERESTARTSYS;
|
|
|
|
else
|
|
|
|
schedule();
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2005-04-16 18:20:36 -04:00
|
|
|
/**
|
|
|
|
* nfs_wait_on_request - Wait for a request to complete.
|
|
|
|
* @req: request to wait upon.
|
|
|
|
*
|
|
|
|
* Interruptible by signals only if mounted with intr flag.
|
|
|
|
* The user is responsible for holding a count on the request.
|
|
|
|
*/
|
|
|
|
int
|
|
|
|
nfs_wait_on_request(struct nfs_page *req)
|
|
|
|
{
|
2005-06-22 13:16:21 -04:00
|
|
|
struct rpc_clnt *clnt = NFS_CLIENT(req->wb_context->dentry->d_inode);
|
|
|
|
sigset_t oldmask;
|
|
|
|
int ret = 0;
|
|
|
|
|
|
|
|
if (!test_bit(PG_BUSY, &req->wb_flags))
|
|
|
|
goto out;
|
|
|
|
/*
|
|
|
|
* Note: the call to rpc_clnt_sigmask() suffices to ensure that we
|
|
|
|
* are not interrupted if intr flag is not set
|
|
|
|
*/
|
|
|
|
rpc_clnt_sigmask(clnt, &oldmask);
|
|
|
|
ret = out_of_line_wait_on_bit(&req->wb_flags, PG_BUSY,
|
|
|
|
nfs_wait_bit_interruptible, TASK_INTERRUPTIBLE);
|
|
|
|
rpc_clnt_sigunmask(clnt, &oldmask);
|
|
|
|
out:
|
|
|
|
return ret;
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* nfs_coalesce_requests - Split coalesced requests out from a list.
|
|
|
|
* @head: source list
|
|
|
|
* @dst: destination list
|
|
|
|
* @nmax: maximum number of requests to coalesce
|
|
|
|
*
|
|
|
|
* Moves a maximum of 'nmax' elements from one list to another.
|
|
|
|
* The elements are checked to ensure that they form a contiguous set
|
|
|
|
* of pages, and that the RPC credentials are the same.
|
|
|
|
*/
|
|
|
|
int
|
|
|
|
nfs_coalesce_requests(struct list_head *head, struct list_head *dst,
|
|
|
|
unsigned int nmax)
|
|
|
|
{
|
|
|
|
struct nfs_page *req = NULL;
|
|
|
|
unsigned int npages = 0;
|
|
|
|
|
|
|
|
while (!list_empty(head)) {
|
|
|
|
struct nfs_page *prev = req;
|
|
|
|
|
|
|
|
req = nfs_list_entry(head->next);
|
|
|
|
if (prev) {
|
|
|
|
if (req->wb_context->cred != prev->wb_context->cred)
|
|
|
|
break;
|
|
|
|
if (req->wb_context->lockowner != prev->wb_context->lockowner)
|
|
|
|
break;
|
|
|
|
if (req->wb_context->state != prev->wb_context->state)
|
|
|
|
break;
|
|
|
|
if (req->wb_index != (prev->wb_index + 1))
|
|
|
|
break;
|
|
|
|
|
|
|
|
if (req->wb_pgbase != 0)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
nfs_list_remove_request(req);
|
|
|
|
nfs_list_add_request(req, dst);
|
|
|
|
npages++;
|
|
|
|
if (req->wb_pgbase + req->wb_bytes != PAGE_CACHE_SIZE)
|
|
|
|
break;
|
|
|
|
if (npages >= nmax)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
return npages;
|
|
|
|
}
|
|
|
|
|
2005-06-22 13:16:31 -04:00
|
|
|
#define NFS_SCAN_MAXENTRIES 16
|
|
|
|
/**
|
2006-09-17 14:46:44 -04:00
|
|
|
* nfs_scan_dirty - Scan the radix tree for dirty requests
|
|
|
|
* @mapping: pointer to address space
|
|
|
|
* @wbc: writeback_control structure
|
2005-06-22 13:16:31 -04:00
|
|
|
* @dst: Destination list
|
|
|
|
*
|
|
|
|
* Moves elements from one of the inode request lists.
|
|
|
|
* If the number of requests is set to 0, the entire address_space
|
|
|
|
* starting at index idx_start, is scanned.
|
|
|
|
* The requests are *not* checked to ensure that they form a contiguous set.
|
|
|
|
* You must be holding the inode's req_lock when calling this function
|
|
|
|
*/
|
2006-09-17 14:46:44 -04:00
|
|
|
long nfs_scan_dirty(struct address_space *mapping,
|
|
|
|
struct writeback_control *wbc,
|
|
|
|
struct list_head *dst)
|
2005-06-22 13:16:31 -04:00
|
|
|
{
|
2006-09-17 14:46:44 -04:00
|
|
|
struct nfs_inode *nfsi = NFS_I(mapping->host);
|
2005-06-22 13:16:31 -04:00
|
|
|
struct nfs_page *pgvec[NFS_SCAN_MAXENTRIES];
|
|
|
|
struct nfs_page *req;
|
2006-09-17 14:46:44 -04:00
|
|
|
pgoff_t idx_start, idx_end;
|
|
|
|
long res = 0;
|
2005-06-22 13:16:31 -04:00
|
|
|
int found, i;
|
|
|
|
|
2006-12-05 00:36:56 -05:00
|
|
|
if (nfsi->ndirty == 0)
|
2006-09-17 14:46:44 -04:00
|
|
|
return 0;
|
|
|
|
if (wbc->range_cyclic) {
|
|
|
|
idx_start = 0;
|
|
|
|
idx_end = ULONG_MAX;
|
|
|
|
} else if (wbc->range_end == 0) {
|
|
|
|
idx_start = wbc->range_start >> PAGE_CACHE_SHIFT;
|
|
|
|
idx_end = ULONG_MAX;
|
|
|
|
} else {
|
|
|
|
idx_start = wbc->range_start >> PAGE_CACHE_SHIFT;
|
|
|
|
idx_end = wbc->range_end >> PAGE_CACHE_SHIFT;
|
|
|
|
}
|
2005-06-22 13:16:31 -04:00
|
|
|
|
|
|
|
for (;;) {
|
2006-09-17 14:46:44 -04:00
|
|
|
unsigned int toscan = NFS_SCAN_MAXENTRIES;
|
|
|
|
|
2005-06-22 13:16:31 -04:00
|
|
|
found = radix_tree_gang_lookup_tag(&nfsi->nfs_page_tree,
|
2006-09-17 14:46:44 -04:00
|
|
|
(void **)&pgvec[0], idx_start, toscan,
|
2005-06-22 13:16:31 -04:00
|
|
|
NFS_PAGE_TAG_DIRTY);
|
2006-09-17 14:46:44 -04:00
|
|
|
|
|
|
|
/* Did we make progress? */
|
2005-06-22 13:16:31 -04:00
|
|
|
if (found <= 0)
|
|
|
|
break;
|
2006-09-17 14:46:44 -04:00
|
|
|
|
2005-06-22 13:16:31 -04:00
|
|
|
for (i = 0; i < found; i++) {
|
|
|
|
req = pgvec[i];
|
2006-09-17 14:46:44 -04:00
|
|
|
if (!wbc->range_cyclic && req->wb_index > idx_end)
|
2005-06-22 13:16:31 -04:00
|
|
|
goto out;
|
|
|
|
|
2006-09-17 14:46:44 -04:00
|
|
|
/* Try to lock request and mark it for writeback */
|
|
|
|
if (!nfs_set_page_writeback_locked(req))
|
|
|
|
goto next;
|
|
|
|
radix_tree_tag_clear(&nfsi->nfs_page_tree,
|
|
|
|
req->wb_index, NFS_PAGE_TAG_DIRTY);
|
|
|
|
nfsi->ndirty--;
|
|
|
|
nfs_list_remove_request(req);
|
|
|
|
nfs_list_add_request(req, dst);
|
|
|
|
res++;
|
|
|
|
if (res == LONG_MAX)
|
|
|
|
goto out;
|
|
|
|
next:
|
|
|
|
idx_start = req->wb_index + 1;
|
2005-06-22 13:16:31 -04:00
|
|
|
}
|
|
|
|
}
|
|
|
|
out:
|
2006-09-17 14:46:44 -04:00
|
|
|
WARN_ON ((nfsi->ndirty == 0) != list_empty(&nfsi->dirty));
|
2005-06-22 13:16:31 -04:00
|
|
|
return res;
|
|
|
|
}
|
|
|
|
|
2005-04-16 18:20:36 -04:00
|
|
|
/**
|
|
|
|
* nfs_scan_list - Scan a list for matching requests
|
2006-05-31 01:13:38 -04:00
|
|
|
* @nfsi: NFS inode
|
2005-04-16 18:20:36 -04:00
|
|
|
* @head: One of the NFS inode request lists
|
|
|
|
* @dst: Destination list
|
|
|
|
* @idx_start: lower bound of page->index to scan
|
|
|
|
* @npages: idx_start + npages sets the upper bound to scan.
|
|
|
|
*
|
|
|
|
* Moves elements from one of the inode request lists.
|
|
|
|
* If the number of requests is set to 0, the entire address_space
|
|
|
|
* starting at index idx_start, is scanned.
|
|
|
|
* The requests are *not* checked to ensure that they form a contiguous set.
|
|
|
|
* You must be holding the inode's req_lock when calling this function
|
|
|
|
*/
|
2006-05-31 01:13:38 -04:00
|
|
|
int nfs_scan_list(struct nfs_inode *nfsi, struct list_head *head,
|
|
|
|
struct list_head *dst, unsigned long idx_start,
|
|
|
|
unsigned int npages)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
2006-05-31 01:13:38 -04:00
|
|
|
struct nfs_page *pgvec[NFS_SCAN_MAXENTRIES];
|
|
|
|
struct nfs_page *req;
|
|
|
|
unsigned long idx_end;
|
|
|
|
int found, i;
|
|
|
|
int res;
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
res = 0;
|
|
|
|
if (npages == 0)
|
|
|
|
idx_end = ~0;
|
|
|
|
else
|
|
|
|
idx_end = idx_start + npages - 1;
|
|
|
|
|
2006-05-31 01:13:38 -04:00
|
|
|
for (;;) {
|
|
|
|
found = radix_tree_gang_lookup(&nfsi->nfs_page_tree,
|
|
|
|
(void **)&pgvec[0], idx_start,
|
|
|
|
NFS_SCAN_MAXENTRIES);
|
|
|
|
if (found <= 0)
|
2005-04-16 18:20:36 -04:00
|
|
|
break;
|
2006-05-31 01:13:38 -04:00
|
|
|
for (i = 0; i < found; i++) {
|
|
|
|
req = pgvec[i];
|
|
|
|
if (req->wb_index > idx_end)
|
|
|
|
goto out;
|
|
|
|
idx_start = req->wb_index + 1;
|
|
|
|
if (req->wb_list_head != head)
|
|
|
|
continue;
|
|
|
|
if (nfs_set_page_writeback_locked(req)) {
|
|
|
|
nfs_list_remove_request(req);
|
|
|
|
nfs_list_add_request(req, dst);
|
|
|
|
res++;
|
|
|
|
}
|
|
|
|
}
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
}
|
2006-05-31 01:13:38 -04:00
|
|
|
out:
|
2005-04-16 18:20:36 -04:00
|
|
|
return res;
|
|
|
|
}
|
|
|
|
|
NFS: Split fs/nfs/inode.c
As fs/nfs/inode.c is rather large, heterogenous and unwieldy, the attached
patch splits it up into a number of files:
(*) fs/nfs/inode.c
Strictly inode specific functions.
(*) fs/nfs/super.c
Superblock management functions for NFS and NFS4, normal access, clones
and referrals. The NFS4 superblock functions _could_ move out into a
separate conditionally compiled file, but it's probably not worth it as
there're so many common bits.
(*) fs/nfs/namespace.c
Some namespace-specific functions have been moved here.
(*) fs/nfs/nfs4namespace.c
NFS4-specific namespace functions (this could be merged into the previous
file). This file is conditionally compiled.
(*) fs/nfs/internal.h
Inter-file declarations, plus a few simple utility functions moved from
fs/nfs/inode.c.
Additionally, all the in-.c-file externs have been moved here, and those
files they were moved from now includes this file.
For the most part, the functions have not been changed, only some multiplexor
functions have changed significantly.
I've also:
(*) Added some extra banner comments above some functions.
(*) Rearranged the function order within the files to be more logical and
better grouped (IMO), though someone may prefer a different order.
(*) Reduced the number of #ifdefs in .c files.
(*) Added missing __init and __exit directives.
Signed-Off-By: David Howells <dhowells@redhat.com>
2006-06-09 09:34:33 -04:00
|
|
|
int __init nfs_init_nfspagecache(void)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
|
|
|
nfs_page_cachep = kmem_cache_create("nfs_page",
|
|
|
|
sizeof(struct nfs_page),
|
|
|
|
0, SLAB_HWCACHE_ALIGN,
|
|
|
|
NULL, NULL);
|
|
|
|
if (nfs_page_cachep == NULL)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2006-06-27 15:59:15 -04:00
|
|
|
void nfs_destroy_nfspagecache(void)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
2006-09-27 04:49:40 -04:00
|
|
|
kmem_cache_destroy(nfs_page_cachep);
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|