linux/fs/nfs/pagelist.c
Trond Myklebust 72cb77f4a5 NFS: Throttle page dirtying while we're flushing to disk
The following patch is a combination of a patch by myself and Peter
Staubach.

Trond: If we allow other processes to dirty pages while a process is doing
a consistency sync to disk, we can end up never making progress.

Peter: Attached is a patch which addresses a continuing problem with
the NFS client generating out of order WRITE requests.  While
this is compliant with all of the current protocol
specifications, there are servers in the market which can not
handle out of order WRITE requests very well.  Also, this may
lead to sub-optimal block allocations in the underlying file
system on the server.  This may cause the read throughputs to
be reduced when reading the file from the server.

Peter: There has been a lot of work recently done to address out of
order issues on a systemic level.  However, the NFS client is
still susceptible to the problem.  Out of order WRITE
requests can occur when pdflush is in the middle of writing
out pages while the process dirtying the pages calls
generic_file_buffered_write which calls
generic_perform_write which calls
balance_dirty_pages_rate_limited which ends up calling
writeback_inodes which ends up calling back into the NFS
client to writes out dirty pages for the same file that
pdflush happens to be working with.

Signed-off-by: Peter Staubach <staubach@redhat.com>
[modification by Trond to merge the two similar patches]
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2009-03-11 14:10:30 -04:00

439 lines
11 KiB
C

/*
* 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/sched.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>
#include "internal.h"
static struct kmem_cache *nfs_page_cachep;
static inline struct nfs_page *
nfs_page_alloc(void)
{
struct nfs_page *p;
p = kmem_cache_alloc(nfs_page_cachep, GFP_KERNEL);
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.
* 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;
for (;;) {
/* try to allocate the request struct */
req = nfs_page_alloc();
if (req != NULL)
break;
if (fatal_signal_pending(current))
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;
req->wb_context = get_nfs_open_context(ctx);
kref_init(&req->wb_kref);
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();
wake_up_bit(&req->wb_flags, PG_BUSY);
nfs_release_request(req);
}
/**
* nfs_set_page_tag_locked - Tag a request as locked
* @req:
*/
int nfs_set_page_tag_locked(struct nfs_page *req)
{
struct nfs_inode *nfsi = NFS_I(req->wb_context->path.dentry->d_inode);
if (!nfs_lock_request_dontget(req))
return 0;
if (req->wb_page != NULL)
radix_tree_tag_set(&nfsi->nfs_page_tree, req->wb_index, NFS_PAGE_TAG_LOCKED);
return 1;
}
/**
* nfs_clear_page_tag_locked - Clear request tag and wake up sleepers
*/
void nfs_clear_page_tag_locked(struct nfs_page *req)
{
struct inode *inode = req->wb_context->path.dentry->d_inode;
struct nfs_inode *nfsi = NFS_I(inode);
if (req->wb_page != NULL) {
spin_lock(&inode->i_lock);
radix_tree_tag_clear(&nfsi->nfs_page_tree, req->wb_index, NFS_PAGE_TAG_LOCKED);
nfs_unlock_request(req);
spin_unlock(&inode->i_lock);
} else
nfs_unlock_request(req);
}
/**
* 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)
{
struct page *page = req->wb_page;
if (page != NULL) {
page_cache_release(page);
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!
*/
static void nfs_free_request(struct kref *kref)
{
struct nfs_page *req = container_of(kref, struct nfs_page, wb_kref);
/* Release struct file or cached credential */
nfs_clear_request(req);
put_nfs_open_context(req->wb_context);
nfs_page_free(req);
}
void nfs_release_request(struct nfs_page *req)
{
kref_put(&req->wb_kref, nfs_free_request);
}
/**
* nfs_wait_on_request - Wait for a request to complete.
* @req: request to wait upon.
*
* Interruptible by fatal signals only.
* The user is responsible for holding a count on the request.
*/
int
nfs_wait_on_request(struct nfs_page *req)
{
int ret = 0;
if (!test_bit(PG_BUSY, &req->wb_flags))
goto out;
ret = out_of_line_wait_on_bit(&req->wb_flags, PG_BUSY,
nfs_wait_bit_killable, TASK_KILLABLE);
out:
return ret;
}
/**
* nfs_pageio_init - initialise a page io descriptor
* @desc: pointer to descriptor
* @inode: pointer to inode
* @doio: pointer to io function
* @bsize: io block size
* @io_flags: extra parameters for the io function
*/
void nfs_pageio_init(struct nfs_pageio_descriptor *desc,
struct inode *inode,
int (*doio)(struct inode *, struct list_head *, unsigned int, size_t, int),
size_t bsize,
int io_flags)
{
INIT_LIST_HEAD(&desc->pg_list);
desc->pg_bytes_written = 0;
desc->pg_count = 0;
desc->pg_bsize = bsize;
desc->pg_base = 0;
desc->pg_inode = inode;
desc->pg_doio = doio;
desc->pg_ioflags = io_flags;
desc->pg_error = 0;
}
/**
* nfs_can_coalesce_requests - test two requests for compatibility
* @prev: pointer to nfs_page
* @req: pointer to nfs_page
*
* The nfs_page structures 'prev' and 'req' are compared to ensure that the
* page data area they describe is contiguous, and that their RPC
* credentials, NFSv4 open state, and lockowners are the same.
*
* Return 'true' if this is the case, else return 'false'.
*/
static int nfs_can_coalesce_requests(struct nfs_page *prev,
struct nfs_page *req)
{
if (req->wb_context->cred != prev->wb_context->cred)
return 0;
if (req->wb_context->lockowner != prev->wb_context->lockowner)
return 0;
if (req->wb_context->state != prev->wb_context->state)
return 0;
if (req->wb_index != (prev->wb_index + 1))
return 0;
if (req->wb_pgbase != 0)
return 0;
if (prev->wb_pgbase + prev->wb_bytes != PAGE_CACHE_SIZE)
return 0;
return 1;
}
/**
* nfs_pageio_do_add_request - Attempt to coalesce a request into a page list.
* @desc: destination io descriptor
* @req: request
*
* Returns true if the request 'req' was successfully coalesced into the
* existing list of pages 'desc'.
*/
static int nfs_pageio_do_add_request(struct nfs_pageio_descriptor *desc,
struct nfs_page *req)
{
size_t newlen = req->wb_bytes;
if (desc->pg_count != 0) {
struct nfs_page *prev;
/*
* FIXME: ideally we should be able to coalesce all requests
* that are not block boundary aligned, but currently this
* is problematic for the case of bsize < PAGE_CACHE_SIZE,
* since nfs_flush_multi and nfs_pagein_multi assume you
* can have only one struct nfs_page.
*/
if (desc->pg_bsize < PAGE_SIZE)
return 0;
newlen += desc->pg_count;
if (newlen > desc->pg_bsize)
return 0;
prev = nfs_list_entry(desc->pg_list.prev);
if (!nfs_can_coalesce_requests(prev, req))
return 0;
} else
desc->pg_base = req->wb_pgbase;
nfs_list_remove_request(req);
nfs_list_add_request(req, &desc->pg_list);
desc->pg_count = newlen;
return 1;
}
/*
* Helper for nfs_pageio_add_request and nfs_pageio_complete
*/
static void nfs_pageio_doio(struct nfs_pageio_descriptor *desc)
{
if (!list_empty(&desc->pg_list)) {
int error = desc->pg_doio(desc->pg_inode,
&desc->pg_list,
nfs_page_array_len(desc->pg_base,
desc->pg_count),
desc->pg_count,
desc->pg_ioflags);
if (error < 0)
desc->pg_error = error;
else
desc->pg_bytes_written += desc->pg_count;
}
if (list_empty(&desc->pg_list)) {
desc->pg_count = 0;
desc->pg_base = 0;
}
}
/**
* nfs_pageio_add_request - Attempt to coalesce a request into a page list.
* @desc: destination io descriptor
* @req: request
*
* Returns true if the request 'req' was successfully coalesced into the
* existing list of pages 'desc'.
*/
int nfs_pageio_add_request(struct nfs_pageio_descriptor *desc,
struct nfs_page *req)
{
while (!nfs_pageio_do_add_request(desc, req)) {
nfs_pageio_doio(desc);
if (desc->pg_error < 0)
return 0;
}
return 1;
}
/**
* nfs_pageio_complete - Complete I/O on an nfs_pageio_descriptor
* @desc: pointer to io descriptor
*/
void nfs_pageio_complete(struct nfs_pageio_descriptor *desc)
{
nfs_pageio_doio(desc);
}
/**
* nfs_pageio_cond_complete - Conditional I/O completion
* @desc: pointer to io descriptor
* @index: page index
*
* It is important to ensure that processes don't try to take locks
* on non-contiguous ranges of pages as that might deadlock. This
* function should be called before attempting to wait on a locked
* nfs_page. It will complete the I/O if the page index 'index'
* is not contiguous with the existing list of pages in 'desc'.
*/
void nfs_pageio_cond_complete(struct nfs_pageio_descriptor *desc, pgoff_t index)
{
if (!list_empty(&desc->pg_list)) {
struct nfs_page *prev = nfs_list_entry(desc->pg_list.prev);
if (index != prev->wb_index + 1)
nfs_pageio_doio(desc);
}
}
#define NFS_SCAN_MAXENTRIES 16
/**
* nfs_scan_list - Scan a list for matching requests
* @nfsi: NFS inode
* @dst: Destination list
* @idx_start: lower bound of page->index to scan
* @npages: idx_start + npages sets the upper bound to scan.
* @tag: tag to scan for
*
* 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 i_lock when calling this function
*/
int nfs_scan_list(struct nfs_inode *nfsi,
struct list_head *dst, pgoff_t idx_start,
unsigned int npages, int tag)
{
struct nfs_page *pgvec[NFS_SCAN_MAXENTRIES];
struct nfs_page *req;
pgoff_t idx_end;
int found, i;
int res;
res = 0;
if (npages == 0)
idx_end = ~0;
else
idx_end = idx_start + npages - 1;
for (;;) {
found = radix_tree_gang_lookup_tag(&nfsi->nfs_page_tree,
(void **)&pgvec[0], idx_start,
NFS_SCAN_MAXENTRIES, tag);
if (found <= 0)
break;
for (i = 0; i < found; i++) {
req = pgvec[i];
if (req->wb_index > idx_end)
goto out;
idx_start = req->wb_index + 1;
if (nfs_set_page_tag_locked(req)) {
kref_get(&req->wb_kref);
nfs_list_remove_request(req);
radix_tree_tag_clear(&nfsi->nfs_page_tree,
req->wb_index, tag);
nfs_list_add_request(req, dst);
res++;
if (res == INT_MAX)
goto out;
}
}
/* for latency reduction */
cond_resched_lock(&nfsi->vfs_inode.i_lock);
}
out:
return res;
}
int __init nfs_init_nfspagecache(void)
{
nfs_page_cachep = kmem_cache_create("nfs_page",
sizeof(struct nfs_page),
0, SLAB_HWCACHE_ALIGN,
NULL);
if (nfs_page_cachep == NULL)
return -ENOMEM;
return 0;
}
void nfs_destroy_nfspagecache(void)
{
kmem_cache_destroy(nfs_page_cachep);
}