2015-02-17 00:58:56 +01:00
|
|
|
/*
|
|
|
|
* fs/dax.c - Direct Access filesystem code
|
|
|
|
* Copyright (c) 2013-2014 Intel Corporation
|
|
|
|
* Author: Matthew Wilcox <matthew.r.wilcox@intel.com>
|
|
|
|
* Author: Ross Zwisler <ross.zwisler@linux.intel.com>
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or modify it
|
|
|
|
* under the terms and conditions of the GNU General Public License,
|
|
|
|
* version 2, as published by the Free Software Foundation.
|
|
|
|
*
|
|
|
|
* This program is distributed in the hope it will be useful, but WITHOUT
|
|
|
|
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
|
|
|
|
* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
|
|
|
|
* more details.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/atomic.h>
|
|
|
|
#include <linux/blkdev.h>
|
|
|
|
#include <linux/buffer_head.h>
|
2015-09-09 18:29:40 +02:00
|
|
|
#include <linux/dax.h>
|
2015-02-17 00:58:56 +01:00
|
|
|
#include <linux/fs.h>
|
|
|
|
#include <linux/genhd.h>
|
2015-02-17 00:59:02 +01:00
|
|
|
#include <linux/highmem.h>
|
|
|
|
#include <linux/memcontrol.h>
|
|
|
|
#include <linux/mm.h>
|
2015-02-17 00:58:56 +01:00
|
|
|
#include <linux/mutex.h>
|
2015-08-18 21:55:40 +02:00
|
|
|
#include <linux/pmem.h>
|
2015-02-17 00:58:59 +01:00
|
|
|
#include <linux/sched.h>
|
2015-02-17 00:58:56 +01:00
|
|
|
#include <linux/uio.h>
|
2015-02-17 00:59:02 +01:00
|
|
|
#include <linux/vmstat.h>
|
2016-01-16 01:56:14 +01:00
|
|
|
#include <linux/pfn_t.h>
|
2016-01-16 01:55:53 +01:00
|
|
|
#include <linux/sizes.h>
|
2015-02-17 00:58:56 +01:00
|
|
|
|
2016-01-16 01:55:59 +01:00
|
|
|
static long dax_map_atomic(struct block_device *bdev, struct blk_dax_ctl *dax)
|
|
|
|
{
|
|
|
|
struct request_queue *q = bdev->bd_queue;
|
|
|
|
long rc = -EIO;
|
|
|
|
|
|
|
|
dax->addr = (void __pmem *) ERR_PTR(-EIO);
|
|
|
|
if (blk_queue_enter(q, true) != 0)
|
|
|
|
return rc;
|
|
|
|
|
|
|
|
rc = bdev_direct_access(bdev, dax);
|
|
|
|
if (rc < 0) {
|
|
|
|
dax->addr = (void __pmem *) ERR_PTR(rc);
|
|
|
|
blk_queue_exit(q);
|
|
|
|
return rc;
|
|
|
|
}
|
|
|
|
return rc;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void dax_unmap_atomic(struct block_device *bdev,
|
|
|
|
const struct blk_dax_ctl *dax)
|
|
|
|
{
|
|
|
|
if (IS_ERR(dax->addr))
|
|
|
|
return;
|
|
|
|
blk_queue_exit(bdev->bd_queue);
|
|
|
|
}
|
|
|
|
|
xfs: Don't use unwritten extents for DAX
DAX has a page fault serialisation problem with block allocation.
Because it allows concurrent page faults and does not have a page
lock to serialise faults to the same page, it can get two concurrent
faults to the page that race.
When two read faults race, this isn't a huge problem as the data
underlying the page is not changing and so "detect and drop" works
just fine. The issues are to do with write faults.
When two write faults occur, we serialise block allocation in
get_blocks() so only one faul will allocate the extent. It will,
however, be marked as an unwritten extent, and that is where the
problem lies - the DAX fault code cannot differentiate between a
block that was just allocated and a block that was preallocated and
needs zeroing. The result is that both write faults end up zeroing
the block and attempting to convert it back to written.
The problem is that the first fault can zero and convert before the
second fault starts zeroing, resulting in the zeroing for the second
fault overwriting the data that the first fault wrote with zeros.
The second fault then attempts to convert the unwritten extent,
which is then a no-op because it's already written. Data loss occurs
as a result of this race.
Because there is no sane locking construct in the page fault code
that we can use for serialisation across the page faults, we need to
ensure block allocation and zeroing occurs atomically in the
filesystem. This means we can still take concurrent page faults and
the only time they will serialise is in the filesystem
mapping/allocation callback. The page fault code will always see
written, initialised extents, so we will be able to remove the
unwritten extent handling from the DAX code when all filesystems are
converted.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2015-11-03 02:37:00 +01:00
|
|
|
/*
|
|
|
|
* dax_clear_blocks() is called from within transaction context from XFS,
|
|
|
|
* and hence this means the stack from this point must follow GFP_NOFS
|
|
|
|
* semantics for all operations.
|
|
|
|
*/
|
2016-01-16 01:55:59 +01:00
|
|
|
int dax_clear_blocks(struct inode *inode, sector_t block, long _size)
|
2015-02-17 00:58:59 +01:00
|
|
|
{
|
|
|
|
struct block_device *bdev = inode->i_sb->s_bdev;
|
2016-01-16 01:55:59 +01:00
|
|
|
struct blk_dax_ctl dax = {
|
|
|
|
.sector = block << (inode->i_blkbits - 9),
|
|
|
|
.size = _size,
|
|
|
|
};
|
2015-02-17 00:58:59 +01:00
|
|
|
|
|
|
|
might_sleep();
|
|
|
|
do {
|
2016-01-16 01:55:53 +01:00
|
|
|
long count, sz;
|
2015-02-17 00:58:59 +01:00
|
|
|
|
2016-01-16 01:55:59 +01:00
|
|
|
count = dax_map_atomic(bdev, &dax);
|
2015-02-17 00:58:59 +01:00
|
|
|
if (count < 0)
|
|
|
|
return count;
|
2016-01-16 01:55:53 +01:00
|
|
|
sz = min_t(long, count, SZ_128K);
|
2016-01-16 01:55:59 +01:00
|
|
|
clear_pmem(dax.addr, sz);
|
|
|
|
dax.size -= sz;
|
|
|
|
dax.sector += sz / 512;
|
|
|
|
dax_unmap_atomic(bdev, &dax);
|
2016-01-16 01:55:53 +01:00
|
|
|
cond_resched();
|
2016-01-16 01:55:59 +01:00
|
|
|
} while (dax.size);
|
2015-02-17 00:58:59 +01:00
|
|
|
|
2015-08-18 21:55:40 +02:00
|
|
|
wmb_pmem();
|
2015-02-17 00:58:59 +01:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dax_clear_blocks);
|
|
|
|
|
2015-08-18 21:55:40 +02:00
|
|
|
/* the clear_pmem() calls are ordered by a wmb_pmem() in the caller */
|
2015-08-18 21:55:41 +02:00
|
|
|
static void dax_new_buf(void __pmem *addr, unsigned size, unsigned first,
|
|
|
|
loff_t pos, loff_t end)
|
2015-02-17 00:58:56 +01:00
|
|
|
{
|
|
|
|
loff_t final = end - pos + first; /* The final byte of the buffer */
|
|
|
|
|
|
|
|
if (first > 0)
|
2015-08-18 21:55:41 +02:00
|
|
|
clear_pmem(addr, first);
|
2015-02-17 00:58:56 +01:00
|
|
|
if (final < size)
|
2015-08-18 21:55:41 +02:00
|
|
|
clear_pmem(addr + final, size - final);
|
2015-02-17 00:58:56 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
static bool buffer_written(struct buffer_head *bh)
|
|
|
|
{
|
|
|
|
return buffer_mapped(bh) && !buffer_unwritten(bh);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* When ext4 encounters a hole, it returns without modifying the buffer_head
|
|
|
|
* which means that we can't trust b_size. To cope with this, we set b_state
|
|
|
|
* to 0 before calling get_block and, if any bit is set, we know we can trust
|
|
|
|
* b_size. Unfortunate, really, since ext4 knows precisely how long a hole is
|
|
|
|
* and would save us time calling get_block repeatedly.
|
|
|
|
*/
|
|
|
|
static bool buffer_size_valid(struct buffer_head *bh)
|
|
|
|
{
|
|
|
|
return bh->b_state != 0;
|
|
|
|
}
|
|
|
|
|
2016-01-16 01:55:59 +01:00
|
|
|
|
|
|
|
static sector_t to_sector(const struct buffer_head *bh,
|
|
|
|
const struct inode *inode)
|
|
|
|
{
|
|
|
|
sector_t sector = bh->b_blocknr << (inode->i_blkbits - 9);
|
|
|
|
|
|
|
|
return sector;
|
|
|
|
}
|
|
|
|
|
2015-03-16 12:33:51 +01:00
|
|
|
static ssize_t dax_io(struct inode *inode, struct iov_iter *iter,
|
|
|
|
loff_t start, loff_t end, get_block_t get_block,
|
|
|
|
struct buffer_head *bh)
|
2015-02-17 00:58:56 +01:00
|
|
|
{
|
2016-01-16 01:55:59 +01:00
|
|
|
loff_t pos = start, max = start, bh_max = start;
|
|
|
|
bool hole = false, need_wmb = false;
|
|
|
|
struct block_device *bdev = NULL;
|
|
|
|
int rw = iov_iter_rw(iter), rc;
|
|
|
|
long map_len = 0;
|
|
|
|
struct blk_dax_ctl dax = {
|
|
|
|
.addr = (void __pmem *) ERR_PTR(-EIO),
|
|
|
|
};
|
|
|
|
|
|
|
|
if (rw == READ)
|
2015-02-17 00:58:56 +01:00
|
|
|
end = min(end, i_size_read(inode));
|
|
|
|
|
|
|
|
while (pos < end) {
|
2015-08-18 21:55:40 +02:00
|
|
|
size_t len;
|
2015-02-17 00:58:56 +01:00
|
|
|
if (pos == max) {
|
|
|
|
unsigned blkbits = inode->i_blkbits;
|
2015-08-14 22:15:31 +02:00
|
|
|
long page = pos >> PAGE_SHIFT;
|
|
|
|
sector_t block = page << (PAGE_SHIFT - blkbits);
|
2015-02-17 00:58:56 +01:00
|
|
|
unsigned first = pos - (block << blkbits);
|
|
|
|
long size;
|
|
|
|
|
|
|
|
if (pos == bh_max) {
|
|
|
|
bh->b_size = PAGE_ALIGN(end - pos);
|
|
|
|
bh->b_state = 0;
|
2016-01-16 01:55:59 +01:00
|
|
|
rc = get_block(inode, block, bh, rw == WRITE);
|
|
|
|
if (rc)
|
2015-02-17 00:58:56 +01:00
|
|
|
break;
|
|
|
|
if (!buffer_size_valid(bh))
|
|
|
|
bh->b_size = 1 << blkbits;
|
|
|
|
bh_max = pos - first + bh->b_size;
|
2016-01-16 01:55:59 +01:00
|
|
|
bdev = bh->b_bdev;
|
2015-02-17 00:58:56 +01:00
|
|
|
} else {
|
|
|
|
unsigned done = bh->b_size -
|
|
|
|
(bh_max - (pos - first));
|
|
|
|
bh->b_blocknr += done >> blkbits;
|
|
|
|
bh->b_size -= done;
|
|
|
|
}
|
|
|
|
|
2016-01-16 01:55:59 +01:00
|
|
|
hole = rw == READ && !buffer_written(bh);
|
2015-02-17 00:58:56 +01:00
|
|
|
if (hole) {
|
|
|
|
size = bh->b_size - first;
|
|
|
|
} else {
|
2016-01-16 01:55:59 +01:00
|
|
|
dax_unmap_atomic(bdev, &dax);
|
|
|
|
dax.sector = to_sector(bh, inode);
|
|
|
|
dax.size = bh->b_size;
|
|
|
|
map_len = dax_map_atomic(bdev, &dax);
|
|
|
|
if (map_len < 0) {
|
|
|
|
rc = map_len;
|
2015-02-17 00:58:56 +01:00
|
|
|
break;
|
2016-01-16 01:55:59 +01:00
|
|
|
}
|
2015-08-18 21:55:40 +02:00
|
|
|
if (buffer_unwritten(bh) || buffer_new(bh)) {
|
2016-01-16 01:55:59 +01:00
|
|
|
dax_new_buf(dax.addr, map_len, first,
|
|
|
|
pos, end);
|
2015-08-18 21:55:40 +02:00
|
|
|
need_wmb = true;
|
|
|
|
}
|
2016-01-16 01:55:59 +01:00
|
|
|
dax.addr += first;
|
|
|
|
size = map_len - first;
|
2015-02-17 00:58:56 +01:00
|
|
|
}
|
|
|
|
max = min(pos + size, end);
|
|
|
|
}
|
|
|
|
|
2015-08-18 21:55:40 +02:00
|
|
|
if (iov_iter_rw(iter) == WRITE) {
|
2016-01-16 01:55:59 +01:00
|
|
|
len = copy_from_iter_pmem(dax.addr, max - pos, iter);
|
2015-08-18 21:55:40 +02:00
|
|
|
need_wmb = true;
|
|
|
|
} else if (!hole)
|
2016-01-16 01:55:59 +01:00
|
|
|
len = copy_to_iter((void __force *) dax.addr, max - pos,
|
2015-08-18 21:55:41 +02:00
|
|
|
iter);
|
2015-02-17 00:58:56 +01:00
|
|
|
else
|
|
|
|
len = iov_iter_zero(max - pos, iter);
|
|
|
|
|
2015-11-11 03:42:49 +01:00
|
|
|
if (!len) {
|
2016-01-16 01:55:59 +01:00
|
|
|
rc = -EFAULT;
|
2015-02-17 00:58:56 +01:00
|
|
|
break;
|
2015-11-11 03:42:49 +01:00
|
|
|
}
|
2015-02-17 00:58:56 +01:00
|
|
|
|
|
|
|
pos += len;
|
2016-01-16 01:55:59 +01:00
|
|
|
if (!IS_ERR(dax.addr))
|
|
|
|
dax.addr += len;
|
2015-02-17 00:58:56 +01:00
|
|
|
}
|
|
|
|
|
2015-08-18 21:55:40 +02:00
|
|
|
if (need_wmb)
|
|
|
|
wmb_pmem();
|
2016-01-16 01:55:59 +01:00
|
|
|
dax_unmap_atomic(bdev, &dax);
|
2015-08-18 21:55:40 +02:00
|
|
|
|
2016-01-16 01:55:59 +01:00
|
|
|
return (pos == start) ? rc : pos - start;
|
2015-02-17 00:58:56 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* dax_do_io - Perform I/O to a DAX file
|
|
|
|
* @iocb: The control block for this I/O
|
|
|
|
* @inode: The file which the I/O is directed at
|
|
|
|
* @iter: The addresses to do I/O from or to
|
|
|
|
* @pos: The file offset where the I/O starts
|
|
|
|
* @get_block: The filesystem method used to translate file offsets to blocks
|
|
|
|
* @end_io: A filesystem callback for I/O completion
|
|
|
|
* @flags: See below
|
|
|
|
*
|
|
|
|
* This function uses the same locking scheme as do_blockdev_direct_IO:
|
|
|
|
* If @flags has DIO_LOCKING set, we assume that the i_mutex is held by the
|
|
|
|
* caller for writes. For reads, we take and release the i_mutex ourselves.
|
|
|
|
* If DIO_LOCKING is not set, the filesystem takes care of its own locking.
|
|
|
|
* As with do_blockdev_direct_IO(), we increment i_dio_count while the I/O
|
|
|
|
* is in progress.
|
|
|
|
*/
|
2015-03-16 12:33:51 +01:00
|
|
|
ssize_t dax_do_io(struct kiocb *iocb, struct inode *inode,
|
|
|
|
struct iov_iter *iter, loff_t pos, get_block_t get_block,
|
|
|
|
dio_iodone_t end_io, int flags)
|
2015-02-17 00:58:56 +01:00
|
|
|
{
|
|
|
|
struct buffer_head bh;
|
|
|
|
ssize_t retval = -EINVAL;
|
|
|
|
loff_t end = pos + iov_iter_count(iter);
|
|
|
|
|
|
|
|
memset(&bh, 0, sizeof(bh));
|
|
|
|
|
2015-03-16 12:33:51 +01:00
|
|
|
if ((flags & DIO_LOCKING) && iov_iter_rw(iter) == READ) {
|
2015-02-17 00:58:56 +01:00
|
|
|
struct address_space *mapping = inode->i_mapping;
|
|
|
|
mutex_lock(&inode->i_mutex);
|
|
|
|
retval = filemap_write_and_wait_range(mapping, pos, end - 1);
|
|
|
|
if (retval) {
|
|
|
|
mutex_unlock(&inode->i_mutex);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Protects against truncate */
|
2015-07-03 16:40:42 +02:00
|
|
|
if (!(flags & DIO_SKIP_DIO_COUNT))
|
|
|
|
inode_dio_begin(inode);
|
2015-02-17 00:58:56 +01:00
|
|
|
|
2015-03-16 12:33:51 +01:00
|
|
|
retval = dax_io(inode, iter, pos, end, get_block, &bh);
|
2015-02-17 00:58:56 +01:00
|
|
|
|
2015-03-16 12:33:51 +01:00
|
|
|
if ((flags & DIO_LOCKING) && iov_iter_rw(iter) == READ)
|
2015-02-17 00:58:56 +01:00
|
|
|
mutex_unlock(&inode->i_mutex);
|
|
|
|
|
|
|
|
if ((retval > 0) && end_io)
|
|
|
|
end_io(iocb, pos, retval, bh.b_private);
|
|
|
|
|
2015-07-03 16:40:42 +02:00
|
|
|
if (!(flags & DIO_SKIP_DIO_COUNT))
|
|
|
|
inode_dio_end(inode);
|
2015-02-17 00:58:56 +01:00
|
|
|
out:
|
|
|
|
return retval;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dax_do_io);
|
2015-02-17 00:59:02 +01:00
|
|
|
|
|
|
|
/*
|
|
|
|
* The user has performed a load from a hole in the file. Allocating
|
|
|
|
* a new page in the file would cause excessive storage usage for
|
|
|
|
* workloads with sparse files. We allocate a page cache page instead.
|
|
|
|
* We'll kick it out of the page cache if it's ever written to,
|
|
|
|
* otherwise it will simply fall out of the page cache under memory
|
|
|
|
* pressure without ever having been dirtied.
|
|
|
|
*/
|
|
|
|
static int dax_load_hole(struct address_space *mapping, struct page *page,
|
|
|
|
struct vm_fault *vmf)
|
|
|
|
{
|
|
|
|
unsigned long size;
|
|
|
|
struct inode *inode = mapping->host;
|
|
|
|
if (!page)
|
|
|
|
page = find_or_create_page(mapping, vmf->pgoff,
|
|
|
|
GFP_KERNEL | __GFP_ZERO);
|
|
|
|
if (!page)
|
|
|
|
return VM_FAULT_OOM;
|
|
|
|
/* Recheck i_size under page lock to avoid truncate race */
|
|
|
|
size = (i_size_read(inode) + PAGE_SIZE - 1) >> PAGE_SHIFT;
|
|
|
|
if (vmf->pgoff >= size) {
|
|
|
|
unlock_page(page);
|
|
|
|
page_cache_release(page);
|
|
|
|
return VM_FAULT_SIGBUS;
|
|
|
|
}
|
|
|
|
|
|
|
|
vmf->page = page;
|
|
|
|
return VM_FAULT_LOCKED;
|
|
|
|
}
|
|
|
|
|
2016-01-16 01:55:59 +01:00
|
|
|
static int copy_user_bh(struct page *to, struct inode *inode,
|
|
|
|
struct buffer_head *bh, unsigned long vaddr)
|
2015-02-17 00:59:02 +01:00
|
|
|
{
|
2016-01-16 01:55:59 +01:00
|
|
|
struct blk_dax_ctl dax = {
|
|
|
|
.sector = to_sector(bh, inode),
|
|
|
|
.size = bh->b_size,
|
|
|
|
};
|
|
|
|
struct block_device *bdev = bh->b_bdev;
|
2015-08-18 21:55:41 +02:00
|
|
|
void *vto;
|
|
|
|
|
2016-01-16 01:55:59 +01:00
|
|
|
if (dax_map_atomic(bdev, &dax) < 0)
|
|
|
|
return PTR_ERR(dax.addr);
|
2015-02-17 00:59:02 +01:00
|
|
|
vto = kmap_atomic(to);
|
2016-01-16 01:55:59 +01:00
|
|
|
copy_user_page(vto, (void __force *)dax.addr, vaddr, to);
|
2015-02-17 00:59:02 +01:00
|
|
|
kunmap_atomic(vto);
|
2016-01-16 01:55:59 +01:00
|
|
|
dax_unmap_atomic(bdev, &dax);
|
2015-02-17 00:59:02 +01:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int dax_insert_mapping(struct inode *inode, struct buffer_head *bh,
|
|
|
|
struct vm_area_struct *vma, struct vm_fault *vmf)
|
|
|
|
{
|
|
|
|
unsigned long vaddr = (unsigned long)vmf->virtual_address;
|
2016-01-16 01:55:59 +01:00
|
|
|
struct address_space *mapping = inode->i_mapping;
|
|
|
|
struct block_device *bdev = bh->b_bdev;
|
|
|
|
struct blk_dax_ctl dax = {
|
|
|
|
.sector = to_sector(bh, inode),
|
|
|
|
.size = bh->b_size,
|
|
|
|
};
|
2015-02-17 00:59:02 +01:00
|
|
|
pgoff_t size;
|
|
|
|
int error;
|
|
|
|
|
2015-10-16 00:28:32 +02:00
|
|
|
i_mmap_lock_read(mapping);
|
|
|
|
|
2015-02-17 00:59:02 +01:00
|
|
|
/*
|
|
|
|
* Check truncate didn't happen while we were allocating a block.
|
|
|
|
* If it did, this block may or may not be still allocated to the
|
|
|
|
* file. We can't tell the filesystem to free it because we can't
|
|
|
|
* take i_mutex here. In the worst case, the file still has blocks
|
|
|
|
* allocated past the end of the file.
|
|
|
|
*/
|
|
|
|
size = (i_size_read(inode) + PAGE_SIZE - 1) >> PAGE_SHIFT;
|
|
|
|
if (unlikely(vmf->pgoff >= size)) {
|
|
|
|
error = -EIO;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2016-01-16 01:55:59 +01:00
|
|
|
if (dax_map_atomic(bdev, &dax) < 0) {
|
|
|
|
error = PTR_ERR(dax.addr);
|
2015-02-17 00:59:02 +01:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2015-08-18 21:55:40 +02:00
|
|
|
if (buffer_unwritten(bh) || buffer_new(bh)) {
|
2016-01-16 01:55:59 +01:00
|
|
|
clear_pmem(dax.addr, PAGE_SIZE);
|
2015-08-18 21:55:40 +02:00
|
|
|
wmb_pmem();
|
|
|
|
}
|
2016-01-16 01:55:59 +01:00
|
|
|
dax_unmap_atomic(bdev, &dax);
|
2015-02-17 00:59:02 +01:00
|
|
|
|
2016-01-16 01:56:40 +01:00
|
|
|
error = vm_insert_mixed(vma, vaddr, dax.pfn);
|
2015-02-17 00:59:02 +01:00
|
|
|
|
|
|
|
out:
|
2015-10-16 00:28:32 +02:00
|
|
|
i_mmap_unlock_read(mapping);
|
|
|
|
|
2015-02-17 00:59:02 +01:00
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
2015-06-04 01:18:18 +02:00
|
|
|
/**
|
|
|
|
* __dax_fault - handle a page fault on a DAX file
|
|
|
|
* @vma: The virtual memory area where the fault occurred
|
|
|
|
* @vmf: The description of the fault
|
|
|
|
* @get_block: The filesystem method used to translate file offsets to blocks
|
2015-07-29 03:48:00 +02:00
|
|
|
* @complete_unwritten: The filesystem method used to convert unwritten blocks
|
|
|
|
* to written so the data written to them is exposed. This is required for
|
|
|
|
* required by write faults for filesystems that will return unwritten
|
|
|
|
* extent mappings from @get_block, but it is optional for reads as
|
|
|
|
* dax_insert_mapping() will always zero unwritten blocks. If the fs does
|
|
|
|
* not support unwritten extents, the it should pass NULL.
|
2015-06-04 01:18:18 +02:00
|
|
|
*
|
|
|
|
* When a page fault occurs, filesystems may call this helper in their
|
|
|
|
* fault handler for DAX files. __dax_fault() assumes the caller has done all
|
|
|
|
* the necessary locking for the page fault to proceed successfully.
|
|
|
|
*/
|
|
|
|
int __dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf,
|
2015-06-04 01:18:18 +02:00
|
|
|
get_block_t get_block, dax_iodone_t complete_unwritten)
|
2015-02-17 00:59:02 +01:00
|
|
|
{
|
|
|
|
struct file *file = vma->vm_file;
|
|
|
|
struct address_space *mapping = file->f_mapping;
|
|
|
|
struct inode *inode = mapping->host;
|
|
|
|
struct page *page;
|
|
|
|
struct buffer_head bh;
|
|
|
|
unsigned long vaddr = (unsigned long)vmf->virtual_address;
|
|
|
|
unsigned blkbits = inode->i_blkbits;
|
|
|
|
sector_t block;
|
|
|
|
pgoff_t size;
|
|
|
|
int error;
|
|
|
|
int major = 0;
|
|
|
|
|
|
|
|
size = (i_size_read(inode) + PAGE_SIZE - 1) >> PAGE_SHIFT;
|
|
|
|
if (vmf->pgoff >= size)
|
|
|
|
return VM_FAULT_SIGBUS;
|
|
|
|
|
|
|
|
memset(&bh, 0, sizeof(bh));
|
|
|
|
block = (sector_t)vmf->pgoff << (PAGE_SHIFT - blkbits);
|
|
|
|
bh.b_size = PAGE_SIZE;
|
|
|
|
|
|
|
|
repeat:
|
|
|
|
page = find_get_page(mapping, vmf->pgoff);
|
|
|
|
if (page) {
|
|
|
|
if (!lock_page_or_retry(page, vma->vm_mm, vmf->flags)) {
|
|
|
|
page_cache_release(page);
|
|
|
|
return VM_FAULT_RETRY;
|
|
|
|
}
|
|
|
|
if (unlikely(page->mapping != mapping)) {
|
|
|
|
unlock_page(page);
|
|
|
|
page_cache_release(page);
|
|
|
|
goto repeat;
|
|
|
|
}
|
|
|
|
size = (i_size_read(inode) + PAGE_SIZE - 1) >> PAGE_SHIFT;
|
|
|
|
if (unlikely(vmf->pgoff >= size)) {
|
|
|
|
/*
|
|
|
|
* We have a struct page covering a hole in the file
|
|
|
|
* from a read fault and we've raced with a truncate
|
|
|
|
*/
|
|
|
|
error = -EIO;
|
2015-10-16 00:28:32 +02:00
|
|
|
goto unlock_page;
|
2015-02-17 00:59:02 +01:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
error = get_block(inode, block, &bh, 0);
|
|
|
|
if (!error && (bh.b_size < PAGE_SIZE))
|
|
|
|
error = -EIO; /* fs corruption? */
|
|
|
|
if (error)
|
2015-10-16 00:28:32 +02:00
|
|
|
goto unlock_page;
|
2015-02-17 00:59:02 +01:00
|
|
|
|
|
|
|
if (!buffer_mapped(&bh) && !buffer_unwritten(&bh) && !vmf->cow_page) {
|
|
|
|
if (vmf->flags & FAULT_FLAG_WRITE) {
|
|
|
|
error = get_block(inode, block, &bh, 1);
|
|
|
|
count_vm_event(PGMAJFAULT);
|
|
|
|
mem_cgroup_count_vm_event(vma->vm_mm, PGMAJFAULT);
|
|
|
|
major = VM_FAULT_MAJOR;
|
|
|
|
if (!error && (bh.b_size < PAGE_SIZE))
|
|
|
|
error = -EIO;
|
|
|
|
if (error)
|
2015-10-16 00:28:32 +02:00
|
|
|
goto unlock_page;
|
2015-02-17 00:59:02 +01:00
|
|
|
} else {
|
|
|
|
return dax_load_hole(mapping, page, vmf);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (vmf->cow_page) {
|
|
|
|
struct page *new_page = vmf->cow_page;
|
|
|
|
if (buffer_written(&bh))
|
2016-01-16 01:55:59 +01:00
|
|
|
error = copy_user_bh(new_page, inode, &bh, vaddr);
|
2015-02-17 00:59:02 +01:00
|
|
|
else
|
|
|
|
clear_user_highpage(new_page, vaddr);
|
|
|
|
if (error)
|
2015-10-16 00:28:32 +02:00
|
|
|
goto unlock_page;
|
2015-02-17 00:59:02 +01:00
|
|
|
vmf->page = page;
|
|
|
|
if (!page) {
|
2015-10-16 00:28:32 +02:00
|
|
|
i_mmap_lock_read(mapping);
|
2015-02-17 00:59:02 +01:00
|
|
|
/* Check we didn't race with truncate */
|
|
|
|
size = (i_size_read(inode) + PAGE_SIZE - 1) >>
|
|
|
|
PAGE_SHIFT;
|
|
|
|
if (vmf->pgoff >= size) {
|
2015-10-16 00:28:32 +02:00
|
|
|
i_mmap_unlock_read(mapping);
|
2015-02-17 00:59:02 +01:00
|
|
|
error = -EIO;
|
2015-10-16 00:28:32 +02:00
|
|
|
goto out;
|
2015-02-17 00:59:02 +01:00
|
|
|
}
|
|
|
|
}
|
|
|
|
return VM_FAULT_LOCKED;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Check we didn't race with a read fault installing a new page */
|
|
|
|
if (!page && major)
|
|
|
|
page = find_lock_page(mapping, vmf->pgoff);
|
|
|
|
|
|
|
|
if (page) {
|
|
|
|
unmap_mapping_range(mapping, vmf->pgoff << PAGE_SHIFT,
|
|
|
|
PAGE_CACHE_SIZE, 0);
|
|
|
|
delete_from_page_cache(page);
|
|
|
|
unlock_page(page);
|
|
|
|
page_cache_release(page);
|
|
|
|
}
|
|
|
|
|
2015-06-04 01:18:18 +02:00
|
|
|
/*
|
|
|
|
* If we successfully insert the new mapping over an unwritten extent,
|
|
|
|
* we need to ensure we convert the unwritten extent. If there is an
|
|
|
|
* error inserting the mapping, the filesystem needs to leave it as
|
|
|
|
* unwritten to prevent exposure of the stale underlying data to
|
|
|
|
* userspace, but we still need to call the completion function so
|
|
|
|
* the private resources on the mapping buffer can be released. We
|
|
|
|
* indicate what the callback should do via the uptodate variable, same
|
|
|
|
* as for normal BH based IO completions.
|
|
|
|
*/
|
2015-02-17 00:59:02 +01:00
|
|
|
error = dax_insert_mapping(inode, &bh, vma, vmf);
|
2015-07-29 03:48:00 +02:00
|
|
|
if (buffer_unwritten(&bh)) {
|
|
|
|
if (complete_unwritten)
|
|
|
|
complete_unwritten(&bh, !error);
|
|
|
|
else
|
|
|
|
WARN_ON_ONCE(!(vmf->flags & FAULT_FLAG_WRITE));
|
|
|
|
}
|
2015-02-17 00:59:02 +01:00
|
|
|
|
|
|
|
out:
|
|
|
|
if (error == -ENOMEM)
|
|
|
|
return VM_FAULT_OOM | major;
|
|
|
|
/* -EBUSY is fine, somebody else faulted on the same PTE */
|
|
|
|
if ((error < 0) && (error != -EBUSY))
|
|
|
|
return VM_FAULT_SIGBUS | major;
|
|
|
|
return VM_FAULT_NOPAGE | major;
|
|
|
|
|
2015-10-16 00:28:32 +02:00
|
|
|
unlock_page:
|
2015-02-17 00:59:02 +01:00
|
|
|
if (page) {
|
|
|
|
unlock_page(page);
|
|
|
|
page_cache_release(page);
|
|
|
|
}
|
|
|
|
goto out;
|
|
|
|
}
|
2015-06-04 01:18:18 +02:00
|
|
|
EXPORT_SYMBOL(__dax_fault);
|
2015-02-17 00:59:02 +01:00
|
|
|
|
|
|
|
/**
|
|
|
|
* dax_fault - handle a page fault on a DAX file
|
|
|
|
* @vma: The virtual memory area where the fault occurred
|
|
|
|
* @vmf: The description of the fault
|
|
|
|
* @get_block: The filesystem method used to translate file offsets to blocks
|
|
|
|
*
|
|
|
|
* When a page fault occurs, filesystems may call this helper in their
|
|
|
|
* fault handler for DAX files.
|
|
|
|
*/
|
|
|
|
int dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf,
|
2015-06-04 01:18:18 +02:00
|
|
|
get_block_t get_block, dax_iodone_t complete_unwritten)
|
2015-02-17 00:59:02 +01:00
|
|
|
{
|
|
|
|
int result;
|
|
|
|
struct super_block *sb = file_inode(vma->vm_file)->i_sb;
|
|
|
|
|
|
|
|
if (vmf->flags & FAULT_FLAG_WRITE) {
|
|
|
|
sb_start_pagefault(sb);
|
|
|
|
file_update_time(vma->vm_file);
|
|
|
|
}
|
2015-06-04 01:18:18 +02:00
|
|
|
result = __dax_fault(vma, vmf, get_block, complete_unwritten);
|
2015-02-17 00:59:02 +01:00
|
|
|
if (vmf->flags & FAULT_FLAG_WRITE)
|
|
|
|
sb_end_pagefault(sb);
|
|
|
|
|
|
|
|
return result;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dax_fault);
|
2015-02-17 00:59:06 +01:00
|
|
|
|
2015-09-08 23:58:57 +02:00
|
|
|
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
|
|
|
|
/*
|
|
|
|
* The 'colour' (ie low bits) within a PMD of a page offset. This comes up
|
|
|
|
* more often than one might expect in the below function.
|
|
|
|
*/
|
|
|
|
#define PG_PMD_COLOUR ((PMD_SIZE >> PAGE_SHIFT) - 1)
|
|
|
|
|
2016-01-16 01:56:58 +01:00
|
|
|
static void __dax_dbg(struct buffer_head *bh, unsigned long address,
|
|
|
|
const char *reason, const char *fn)
|
|
|
|
{
|
|
|
|
if (bh) {
|
|
|
|
char bname[BDEVNAME_SIZE];
|
|
|
|
bdevname(bh->b_bdev, bname);
|
|
|
|
pr_debug("%s: %s addr: %lx dev %s state %lx start %lld "
|
|
|
|
"length %zd fallback: %s\n", fn, current->comm,
|
|
|
|
address, bname, bh->b_state, (u64)bh->b_blocknr,
|
|
|
|
bh->b_size, reason);
|
|
|
|
} else {
|
|
|
|
pr_debug("%s: %s addr: %lx fallback: %s\n", fn,
|
|
|
|
current->comm, address, reason);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
#define dax_pmd_dbg(bh, address, reason) __dax_dbg(bh, address, reason, "dax_pmd")
|
|
|
|
|
2015-09-08 23:58:57 +02:00
|
|
|
int __dax_pmd_fault(struct vm_area_struct *vma, unsigned long address,
|
|
|
|
pmd_t *pmd, unsigned int flags, get_block_t get_block,
|
|
|
|
dax_iodone_t complete_unwritten)
|
|
|
|
{
|
|
|
|
struct file *file = vma->vm_file;
|
|
|
|
struct address_space *mapping = file->f_mapping;
|
|
|
|
struct inode *inode = mapping->host;
|
|
|
|
struct buffer_head bh;
|
|
|
|
unsigned blkbits = inode->i_blkbits;
|
|
|
|
unsigned long pmd_addr = address & PMD_MASK;
|
|
|
|
bool write = flags & FAULT_FLAG_WRITE;
|
2016-01-16 01:55:59 +01:00
|
|
|
struct block_device *bdev;
|
2015-09-08 23:58:57 +02:00
|
|
|
pgoff_t size, pgoff;
|
2016-01-16 01:55:59 +01:00
|
|
|
sector_t block;
|
2015-09-08 23:58:57 +02:00
|
|
|
int result = 0;
|
|
|
|
|
2016-01-16 01:57:01 +01:00
|
|
|
/* dax pmd mappings require pfn_t_devmap() */
|
2015-11-16 01:06:32 +01:00
|
|
|
if (!IS_ENABLED(CONFIG_FS_DAX_PMD))
|
|
|
|
return VM_FAULT_FALLBACK;
|
|
|
|
|
2015-09-08 23:58:57 +02:00
|
|
|
/* Fall back to PTEs if we're going to COW */
|
2016-01-16 01:56:05 +01:00
|
|
|
if (write && !(vma->vm_flags & VM_SHARED)) {
|
|
|
|
split_huge_pmd(vma, pmd, address);
|
2016-01-16 01:56:58 +01:00
|
|
|
dax_pmd_dbg(NULL, address, "cow write");
|
2015-09-08 23:58:57 +02:00
|
|
|
return VM_FAULT_FALLBACK;
|
2016-01-16 01:56:05 +01:00
|
|
|
}
|
2015-09-08 23:58:57 +02:00
|
|
|
/* If the PMD would extend outside the VMA */
|
2016-01-16 01:56:58 +01:00
|
|
|
if (pmd_addr < vma->vm_start) {
|
|
|
|
dax_pmd_dbg(NULL, address, "vma start unaligned");
|
2015-09-08 23:58:57 +02:00
|
|
|
return VM_FAULT_FALLBACK;
|
2016-01-16 01:56:58 +01:00
|
|
|
}
|
|
|
|
if ((pmd_addr + PMD_SIZE) > vma->vm_end) {
|
|
|
|
dax_pmd_dbg(NULL, address, "vma end unaligned");
|
2015-09-08 23:58:57 +02:00
|
|
|
return VM_FAULT_FALLBACK;
|
2016-01-16 01:56:58 +01:00
|
|
|
}
|
2015-09-08 23:58:57 +02:00
|
|
|
|
2015-09-08 23:59:39 +02:00
|
|
|
pgoff = linear_page_index(vma, pmd_addr);
|
2015-09-08 23:58:57 +02:00
|
|
|
size = (i_size_read(inode) + PAGE_SIZE - 1) >> PAGE_SHIFT;
|
|
|
|
if (pgoff >= size)
|
|
|
|
return VM_FAULT_SIGBUS;
|
|
|
|
/* If the PMD would cover blocks out of the file */
|
2016-01-16 01:56:58 +01:00
|
|
|
if ((pgoff | PG_PMD_COLOUR) >= size) {
|
|
|
|
dax_pmd_dbg(NULL, address,
|
|
|
|
"offset + huge page size > file size");
|
2015-09-08 23:58:57 +02:00
|
|
|
return VM_FAULT_FALLBACK;
|
2016-01-16 01:56:58 +01:00
|
|
|
}
|
2015-09-08 23:58:57 +02:00
|
|
|
|
|
|
|
memset(&bh, 0, sizeof(bh));
|
2016-01-23 00:10:31 +01:00
|
|
|
bh.b_bdev = inode->i_sb->s_bdev;
|
2015-09-08 23:58:57 +02:00
|
|
|
block = (sector_t)pgoff << (PAGE_SHIFT - blkbits);
|
|
|
|
|
|
|
|
bh.b_size = PMD_SIZE;
|
2016-01-16 01:55:59 +01:00
|
|
|
if (get_block(inode, block, &bh, write) != 0)
|
2015-09-08 23:58:57 +02:00
|
|
|
return VM_FAULT_SIGBUS;
|
2016-01-16 01:55:59 +01:00
|
|
|
bdev = bh.b_bdev;
|
2015-10-16 00:28:32 +02:00
|
|
|
i_mmap_lock_read(mapping);
|
2015-09-08 23:58:57 +02:00
|
|
|
|
|
|
|
/*
|
|
|
|
* If the filesystem isn't willing to tell us the length of a hole,
|
|
|
|
* just fall back to PTEs. Calling get_block 512 times in a loop
|
|
|
|
* would be silly.
|
|
|
|
*/
|
2016-01-16 01:56:58 +01:00
|
|
|
if (!buffer_size_valid(&bh) || bh.b_size < PMD_SIZE) {
|
|
|
|
dax_pmd_dbg(&bh, address, "allocated block too small");
|
2015-09-08 23:58:57 +02:00
|
|
|
goto fallback;
|
2016-01-16 01:56:58 +01:00
|
|
|
}
|
2015-09-08 23:58:57 +02:00
|
|
|
|
2015-09-08 23:59:42 +02:00
|
|
|
/*
|
|
|
|
* If we allocated new storage, make sure no process has any
|
|
|
|
* zero pages covering this hole
|
|
|
|
*/
|
|
|
|
if (buffer_new(&bh)) {
|
2015-10-16 00:28:32 +02:00
|
|
|
i_mmap_unlock_read(mapping);
|
2015-09-08 23:59:42 +02:00
|
|
|
unmap_mapping_range(mapping, pgoff << PAGE_SHIFT, PMD_SIZE, 0);
|
2015-10-16 00:28:32 +02:00
|
|
|
i_mmap_lock_read(mapping);
|
2015-09-08 23:59:42 +02:00
|
|
|
}
|
|
|
|
|
2015-09-08 23:59:17 +02:00
|
|
|
/*
|
|
|
|
* If a truncate happened while we were allocating blocks, we may
|
|
|
|
* leave blocks allocated to the file that are beyond EOF. We can't
|
|
|
|
* take i_mutex here, so just leave them hanging; they'll be freed
|
|
|
|
* when the file is deleted.
|
|
|
|
*/
|
2015-09-08 23:58:57 +02:00
|
|
|
size = (i_size_read(inode) + PAGE_SIZE - 1) >> PAGE_SHIFT;
|
|
|
|
if (pgoff >= size) {
|
|
|
|
result = VM_FAULT_SIGBUS;
|
|
|
|
goto out;
|
|
|
|
}
|
2016-01-16 01:56:58 +01:00
|
|
|
if ((pgoff | PG_PMD_COLOUR) >= size) {
|
|
|
|
dax_pmd_dbg(&bh, address, "pgoff unaligned");
|
2015-09-08 23:58:57 +02:00
|
|
|
goto fallback;
|
2016-01-16 01:56:58 +01:00
|
|
|
}
|
2015-09-08 23:58:57 +02:00
|
|
|
|
|
|
|
if (!write && !buffer_mapped(&bh) && buffer_uptodate(&bh)) {
|
|
|
|
spinlock_t *ptl;
|
2015-09-08 23:59:34 +02:00
|
|
|
pmd_t entry;
|
2015-09-08 23:58:57 +02:00
|
|
|
struct page *zero_page = get_huge_zero_page();
|
2015-09-08 23:59:34 +02:00
|
|
|
|
2016-01-16 01:56:58 +01:00
|
|
|
if (unlikely(!zero_page)) {
|
|
|
|
dax_pmd_dbg(&bh, address, "no zero page");
|
2015-09-08 23:58:57 +02:00
|
|
|
goto fallback;
|
2016-01-16 01:56:58 +01:00
|
|
|
}
|
2015-09-08 23:58:57 +02:00
|
|
|
|
2015-09-08 23:59:34 +02:00
|
|
|
ptl = pmd_lock(vma->vm_mm, pmd);
|
|
|
|
if (!pmd_none(*pmd)) {
|
|
|
|
spin_unlock(ptl);
|
2016-01-16 01:56:58 +01:00
|
|
|
dax_pmd_dbg(&bh, address, "pmd already present");
|
2015-09-08 23:59:34 +02:00
|
|
|
goto fallback;
|
|
|
|
}
|
|
|
|
|
2016-01-16 01:56:58 +01:00
|
|
|
dev_dbg(part_to_dev(bdev->bd_part),
|
|
|
|
"%s: %s addr: %lx pfn: <zero> sect: %llx\n",
|
|
|
|
__func__, current->comm, address,
|
|
|
|
(unsigned long long) to_sector(&bh, inode));
|
|
|
|
|
2015-09-08 23:59:34 +02:00
|
|
|
entry = mk_pmd(zero_page, vma->vm_page_prot);
|
|
|
|
entry = pmd_mkhuge(entry);
|
|
|
|
set_pmd_at(vma->vm_mm, pmd_addr, pmd, entry);
|
2015-09-08 23:58:57 +02:00
|
|
|
result = VM_FAULT_NOPAGE;
|
2015-09-08 23:59:34 +02:00
|
|
|
spin_unlock(ptl);
|
2015-09-08 23:58:57 +02:00
|
|
|
} else {
|
2016-01-16 01:55:59 +01:00
|
|
|
struct blk_dax_ctl dax = {
|
|
|
|
.sector = to_sector(&bh, inode),
|
|
|
|
.size = PMD_SIZE,
|
|
|
|
};
|
|
|
|
long length = dax_map_atomic(bdev, &dax);
|
|
|
|
|
2015-09-08 23:58:57 +02:00
|
|
|
if (length < 0) {
|
|
|
|
result = VM_FAULT_SIGBUS;
|
|
|
|
goto out;
|
|
|
|
}
|
2016-01-16 01:56:58 +01:00
|
|
|
if (length < PMD_SIZE) {
|
|
|
|
dax_pmd_dbg(&bh, address, "dax-length too small");
|
|
|
|
dax_unmap_atomic(bdev, &dax);
|
|
|
|
goto fallback;
|
|
|
|
}
|
|
|
|
if (pfn_t_to_pfn(dax.pfn) & PG_PMD_COLOUR) {
|
|
|
|
dax_pmd_dbg(&bh, address, "pfn unaligned");
|
2016-01-16 01:55:59 +01:00
|
|
|
dax_unmap_atomic(bdev, &dax);
|
2015-09-08 23:58:57 +02:00
|
|
|
goto fallback;
|
2016-01-16 01:55:59 +01:00
|
|
|
}
|
2015-09-08 23:58:57 +02:00
|
|
|
|
2016-01-16 01:57:01 +01:00
|
|
|
if (!pfn_t_devmap(dax.pfn)) {
|
2016-01-16 01:55:59 +01:00
|
|
|
dax_unmap_atomic(bdev, &dax);
|
2016-01-16 01:56:58 +01:00
|
|
|
dax_pmd_dbg(&bh, address, "pfn not in memmap");
|
2015-11-13 03:33:54 +01:00
|
|
|
goto fallback;
|
2016-01-16 01:55:59 +01:00
|
|
|
}
|
2015-11-13 03:33:54 +01:00
|
|
|
|
2015-10-16 00:28:32 +02:00
|
|
|
if (buffer_unwritten(&bh) || buffer_new(&bh)) {
|
2016-01-16 01:55:59 +01:00
|
|
|
clear_pmem(dax.addr, PMD_SIZE);
|
2015-10-16 00:28:32 +02:00
|
|
|
wmb_pmem();
|
|
|
|
count_vm_event(PGMAJFAULT);
|
|
|
|
mem_cgroup_count_vm_event(vma->vm_mm, PGMAJFAULT);
|
|
|
|
result |= VM_FAULT_MAJOR;
|
|
|
|
}
|
2016-01-16 01:55:59 +01:00
|
|
|
dax_unmap_atomic(bdev, &dax);
|
2015-10-16 00:28:32 +02:00
|
|
|
|
2016-01-16 01:56:58 +01:00
|
|
|
dev_dbg(part_to_dev(bdev->bd_part),
|
|
|
|
"%s: %s addr: %lx pfn: %lx sect: %llx\n",
|
|
|
|
__func__, current->comm, address,
|
|
|
|
pfn_t_to_pfn(dax.pfn),
|
|
|
|
(unsigned long long) dax.sector);
|
2016-01-16 01:56:14 +01:00
|
|
|
result |= vmf_insert_pfn_pmd(vma, address, pmd,
|
2016-01-16 01:56:43 +01:00
|
|
|
dax.pfn, write);
|
2015-09-08 23:58:57 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
out:
|
2015-10-16 00:28:32 +02:00
|
|
|
i_mmap_unlock_read(mapping);
|
|
|
|
|
2015-09-08 23:58:57 +02:00
|
|
|
if (buffer_unwritten(&bh))
|
|
|
|
complete_unwritten(&bh, !(result & VM_FAULT_ERROR));
|
|
|
|
|
|
|
|
return result;
|
|
|
|
|
|
|
|
fallback:
|
|
|
|
count_vm_event(THP_FAULT_FALLBACK);
|
|
|
|
result = VM_FAULT_FALLBACK;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(__dax_pmd_fault);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* dax_pmd_fault - handle a PMD fault on a DAX file
|
|
|
|
* @vma: The virtual memory area where the fault occurred
|
|
|
|
* @vmf: The description of the fault
|
|
|
|
* @get_block: The filesystem method used to translate file offsets to blocks
|
|
|
|
*
|
|
|
|
* When a page fault occurs, filesystems may call this helper in their
|
|
|
|
* pmd_fault handler for DAX files.
|
|
|
|
*/
|
|
|
|
int dax_pmd_fault(struct vm_area_struct *vma, unsigned long address,
|
|
|
|
pmd_t *pmd, unsigned int flags, get_block_t get_block,
|
|
|
|
dax_iodone_t complete_unwritten)
|
|
|
|
{
|
|
|
|
int result;
|
|
|
|
struct super_block *sb = file_inode(vma->vm_file)->i_sb;
|
|
|
|
|
|
|
|
if (flags & FAULT_FLAG_WRITE) {
|
|
|
|
sb_start_pagefault(sb);
|
|
|
|
file_update_time(vma->vm_file);
|
|
|
|
}
|
|
|
|
result = __dax_pmd_fault(vma, address, pmd, flags, get_block,
|
|
|
|
complete_unwritten);
|
|
|
|
if (flags & FAULT_FLAG_WRITE)
|
|
|
|
sb_end_pagefault(sb);
|
|
|
|
|
|
|
|
return result;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dax_pmd_fault);
|
2015-09-08 23:59:09 +02:00
|
|
|
#endif /* CONFIG_TRANSPARENT_HUGEPAGE */
|
2015-09-08 23:58:57 +02:00
|
|
|
|
2015-04-16 01:15:14 +02:00
|
|
|
/**
|
|
|
|
* dax_pfn_mkwrite - handle first write to DAX page
|
|
|
|
* @vma: The virtual memory area where the fault occurred
|
|
|
|
* @vmf: The description of the fault
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
int dax_pfn_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
|
|
|
|
{
|
|
|
|
struct super_block *sb = file_inode(vma->vm_file)->i_sb;
|
|
|
|
|
|
|
|
sb_start_pagefault(sb);
|
|
|
|
file_update_time(vma->vm_file);
|
|
|
|
sb_end_pagefault(sb);
|
|
|
|
return VM_FAULT_NOPAGE;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dax_pfn_mkwrite);
|
|
|
|
|
2015-02-17 00:59:06 +01:00
|
|
|
/**
|
2015-02-17 00:59:35 +01:00
|
|
|
* dax_zero_page_range - zero a range within a page of a DAX file
|
2015-02-17 00:59:06 +01:00
|
|
|
* @inode: The file being truncated
|
|
|
|
* @from: The file offset that is being truncated to
|
2015-02-17 00:59:35 +01:00
|
|
|
* @length: The number of bytes to zero
|
2015-02-17 00:59:06 +01:00
|
|
|
* @get_block: The filesystem method used to translate file offsets to blocks
|
|
|
|
*
|
2015-02-17 00:59:35 +01:00
|
|
|
* This function can be called by a filesystem when it is zeroing part of a
|
|
|
|
* page in a DAX file. This is intended for hole-punch operations. If
|
|
|
|
* you are truncating a file, the helper function dax_truncate_page() may be
|
|
|
|
* more convenient.
|
2015-02-17 00:59:06 +01:00
|
|
|
*
|
|
|
|
* We work in terms of PAGE_CACHE_SIZE here for commonality with
|
|
|
|
* block_truncate_page(), but we could go down to PAGE_SIZE if the filesystem
|
|
|
|
* took care of disposing of the unnecessary blocks. Even if the filesystem
|
|
|
|
* block size is smaller than PAGE_SIZE, we have to zero the rest of the page
|
2015-02-17 00:59:35 +01:00
|
|
|
* since the file might be mmapped.
|
2015-02-17 00:59:06 +01:00
|
|
|
*/
|
2015-02-17 00:59:35 +01:00
|
|
|
int dax_zero_page_range(struct inode *inode, loff_t from, unsigned length,
|
|
|
|
get_block_t get_block)
|
2015-02-17 00:59:06 +01:00
|
|
|
{
|
|
|
|
struct buffer_head bh;
|
|
|
|
pgoff_t index = from >> PAGE_CACHE_SHIFT;
|
|
|
|
unsigned offset = from & (PAGE_CACHE_SIZE-1);
|
|
|
|
int err;
|
|
|
|
|
|
|
|
/* Block boundary? Nothing to do */
|
|
|
|
if (!length)
|
|
|
|
return 0;
|
2015-02-17 00:59:35 +01:00
|
|
|
BUG_ON((offset + length) > PAGE_CACHE_SIZE);
|
2015-02-17 00:59:06 +01:00
|
|
|
|
|
|
|
memset(&bh, 0, sizeof(bh));
|
|
|
|
bh.b_size = PAGE_CACHE_SIZE;
|
|
|
|
err = get_block(inode, index, &bh, 0);
|
|
|
|
if (err < 0)
|
|
|
|
return err;
|
|
|
|
if (buffer_written(&bh)) {
|
2016-01-16 01:55:59 +01:00
|
|
|
struct block_device *bdev = bh.b_bdev;
|
|
|
|
struct blk_dax_ctl dax = {
|
|
|
|
.sector = to_sector(&bh, inode),
|
|
|
|
.size = PAGE_CACHE_SIZE,
|
|
|
|
};
|
|
|
|
|
|
|
|
if (dax_map_atomic(bdev, &dax) < 0)
|
|
|
|
return PTR_ERR(dax.addr);
|
|
|
|
clear_pmem(dax.addr + offset, length);
|
2015-08-18 21:55:40 +02:00
|
|
|
wmb_pmem();
|
2016-01-16 01:55:59 +01:00
|
|
|
dax_unmap_atomic(bdev, &dax);
|
2015-02-17 00:59:06 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
2015-02-17 00:59:35 +01:00
|
|
|
EXPORT_SYMBOL_GPL(dax_zero_page_range);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* dax_truncate_page - handle a partial page being truncated in a DAX file
|
|
|
|
* @inode: The file being truncated
|
|
|
|
* @from: The file offset that is being truncated to
|
|
|
|
* @get_block: The filesystem method used to translate file offsets to blocks
|
|
|
|
*
|
|
|
|
* Similar to block_truncate_page(), this function can be called by a
|
|
|
|
* filesystem when it is truncating a DAX file to handle the partial page.
|
|
|
|
*
|
|
|
|
* We work in terms of PAGE_CACHE_SIZE here for commonality with
|
|
|
|
* block_truncate_page(), but we could go down to PAGE_SIZE if the filesystem
|
|
|
|
* took care of disposing of the unnecessary blocks. Even if the filesystem
|
|
|
|
* block size is smaller than PAGE_SIZE, we have to zero the rest of the page
|
|
|
|
* since the file might be mmapped.
|
|
|
|
*/
|
|
|
|
int dax_truncate_page(struct inode *inode, loff_t from, get_block_t get_block)
|
|
|
|
{
|
|
|
|
unsigned length = PAGE_CACHE_ALIGN(from) - from;
|
|
|
|
return dax_zero_page_range(inode, from, length, get_block);
|
|
|
|
}
|
2015-02-17 00:59:06 +01:00
|
|
|
EXPORT_SYMBOL_GPL(dax_truncate_page);
|