[mirror_ubuntu-artful-kernel.git] / fs / ext2 / file.c

/*
 *  linux/fs/ext2/file.c
 *
 * Copyright (C) 1992, 1993, 1994, 1995
 * Remy Card (card@masi.ibp.fr)
 * Laboratoire MASI - Institut Blaise Pascal
 * Universite Pierre et Marie Curie (Paris VI)
 *
 *  from
 *
 *  linux/fs/minix/file.c
 *
 *  Copyright (C) 1991, 1992  Linus Torvalds
 *
 *  ext2 fs regular file handling primitives
 *
 *  64-bit file support on 64-bit platforms by Jakub Jelinek
 * 	(jj@sunsite.ms.mff.cuni.cz)
 */

#include <linux/time.h>
#include <linux/pagemap.h>
#include <linux/dax.h>
#include <linux/quotaops.h>
#include <linux/iomap.h>
#include <linux/uio.h>
#include "ext2.h"
#include "xattr.h"
#include "acl.h"

#ifdef CONFIG_FS_DAX
static ssize_t ext2_dax_read_iter(struct kiocb *iocb, struct iov_iter *to)
{
	struct inode *inode = iocb->ki_filp->f_mapping->host;
	ssize_t ret;

	if (!iov_iter_count(to))
		return 0; /* skip atime */

	inode_lock_shared(inode);
	ret = iomap_dax_rw(iocb, to, &ext2_iomap_ops);
	inode_unlock_shared(inode);

	file_accessed(iocb->ki_filp);
	return ret;
}

static ssize_t ext2_dax_write_iter(struct kiocb *iocb, struct iov_iter *from)
{
	struct file *file = iocb->ki_filp;
	struct inode *inode = file->f_mapping->host;
	ssize_t ret;

	inode_lock(inode);
	ret = generic_write_checks(iocb, from);
	if (ret <= 0)
		goto out_unlock;
	ret = file_remove_privs(file);
	if (ret)
		goto out_unlock;
	ret = file_update_time(file);
	if (ret)
		goto out_unlock;

	ret = iomap_dax_rw(iocb, from, &ext2_iomap_ops);
	if (ret > 0 && iocb->ki_pos > i_size_read(inode)) {
		i_size_write(inode, iocb->ki_pos);
		mark_inode_dirty(inode);
	}

out_unlock:
	inode_unlock(inode);
	if (ret > 0)
		ret = generic_write_sync(iocb, ret);
	return ret;
}

/*
 * The lock ordering for ext2 DAX fault paths is:
 *
 * mmap_sem (MM)
 *   sb_start_pagefault (vfs, freeze)
 *     ext2_inode_info->dax_sem
 *       address_space->i_mmap_rwsem or page_lock (mutually exclusive in DAX)
 *         ext2_inode_info->truncate_mutex
 *
 * The default page_lock and i_size verification done by non-DAX fault paths
 * is sufficient because ext2 doesn't support hole punching.
 */
static int ext2_dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
{
	struct inode *inode = file_inode(vma->vm_file);
	struct ext2_inode_info *ei = EXT2_I(inode);
	int ret;

	if (vmf->flags & FAULT_FLAG_WRITE) {
		sb_start_pagefault(inode->i_sb);
		file_update_time(vma->vm_file);
	}
	down_read(&ei->dax_sem);

	ret = iomap_dax_fault(vma, vmf, &ext2_iomap_ops);

	up_read(&ei->dax_sem);
	if (vmf->flags & FAULT_FLAG_WRITE)
		sb_end_pagefault(inode->i_sb);
	return ret;
}

static int ext2_dax_pfn_mkwrite(struct vm_area_struct *vma,
		struct vm_fault *vmf)
{
	struct inode *inode = file_inode(vma->vm_file);
	struct ext2_inode_info *ei = EXT2_I(inode);
	loff_t size;
	int ret;

	sb_start_pagefault(inode->i_sb);
	file_update_time(vma->vm_file);
	down_read(&ei->dax_sem);

	/* check that the faulting page hasn't raced with truncate */
	size = (i_size_read(inode) + PAGE_SIZE - 1) >> PAGE_SHIFT;
	if (vmf->pgoff >= size)
		ret = VM_FAULT_SIGBUS;
	else
		ret = dax_pfn_mkwrite(vma, vmf);

	up_read(&ei->dax_sem);
	sb_end_pagefault(inode->i_sb);
	return ret;
}

static const struct vm_operations_struct ext2_dax_vm_ops = {
	.fault		= ext2_dax_fault,
	/*
	 * .pmd_fault is not supported for DAX because allocation in ext2
	 * cannot be reliably aligned to huge page sizes and so pmd faults
	 * will always fail and fail back to regular faults.
	 */
	.page_mkwrite	= ext2_dax_fault,
	.pfn_mkwrite	= ext2_dax_pfn_mkwrite,
};

static int ext2_file_mmap(struct file *file, struct vm_area_struct *vma)
{
	if (!IS_DAX(file_inode(file)))
		return generic_file_mmap(file, vma);

	file_accessed(file);
	vma->vm_ops = &ext2_dax_vm_ops;
	vma->vm_flags |= VM_MIXEDMAP;
	return 0;
}
#else
#define ext2_file_mmap	generic_file_mmap
#endif

/*
 * Called when filp is released. This happens when all file descriptors
 * for a single struct file are closed. Note that different open() calls
 * for the same file yield different struct file structures.
 */
static int ext2_release_file (struct inode * inode, struct file * filp)
{
	if (filp->f_mode & FMODE_WRITE) {
		mutex_lock(&EXT2_I(inode)->truncate_mutex);
		ext2_discard_reservation(inode);
		mutex_unlock(&EXT2_I(inode)->truncate_mutex);
	}
	return 0;
}

int ext2_fsync(struct file *file, loff_t start, loff_t end, int datasync)
{
	int ret;
	struct super_block *sb = file->f_mapping->host->i_sb;
	struct address_space *mapping = sb->s_bdev->bd_inode->i_mapping;

	ret = generic_file_fsync(file, start, end, datasync);
	if (ret == -EIO || test_and_clear_bit(AS_EIO, &mapping->flags)) {
		/* We don't really know where the IO error happened... */
		ext2_error(sb, __func__,
			   "detected IO error when writing metadata buffers");
		ret = -EIO;
	}
	return ret;
}

static ssize_t ext2_file_read_iter(struct kiocb *iocb, struct iov_iter *to)
{
#ifdef CONFIG_FS_DAX
	if (IS_DAX(iocb->ki_filp->f_mapping->host))
		return ext2_dax_read_iter(iocb, to);
#endif
	return generic_file_read_iter(iocb, to);
}

static ssize_t ext2_file_write_iter(struct kiocb *iocb, struct iov_iter *from)
{
#ifdef CONFIG_FS_DAX
	if (IS_DAX(iocb->ki_filp->f_mapping->host))
		return ext2_dax_write_iter(iocb, from);
#endif
	return generic_file_write_iter(iocb, from);
}

const struct file_operations ext2_file_operations = {
	.llseek		= generic_file_llseek,
	.read_iter	= ext2_file_read_iter,
	.write_iter	= ext2_file_write_iter,
	.unlocked_ioctl = ext2_ioctl,
#ifdef CONFIG_COMPAT
	.compat_ioctl	= ext2_compat_ioctl,
#endif
	.mmap		= ext2_file_mmap,
	.open		= dquot_file_open,
	.release	= ext2_release_file,
	.fsync		= ext2_fsync,
	.get_unmapped_area = thp_get_unmapped_area,
	.splice_read	= generic_file_splice_read,
	.splice_write	= iter_file_splice_write,
};

const struct inode_operations ext2_file_inode_operations = {
#ifdef CONFIG_EXT2_FS_XATTR
	.listxattr	= ext2_listxattr,
#endif
	.setattr	= ext2_setattr,
	.get_acl	= ext2_get_acl,
	.set_acl	= ext2_set_acl,
	.fiemap		= ext2_fiemap,
};
Commit	Line	Data
1da177e4 LT	1	/*
	2	* linux/fs/ext2/file.c
	3	*
	4	* Copyright (C) 1992, 1993, 1994, 1995
	5	* Remy Card (card@masi.ibp.fr)
	6	* Laboratoire MASI - Institut Blaise Pascal
	7	* Universite Pierre et Marie Curie (Paris VI)
	8	*
	9	* from
	10	*
	11	* linux/fs/minix/file.c
	12	*
	13	* Copyright (C) 1991, 1992 Linus Torvalds
	14	*
	15	* ext2 fs regular file handling primitives
	16	*
	17	* 64-bit file support on 64-bit platforms by Jakub Jelinek
	18	* (jj@sunsite.ms.mff.cuni.cz)
	19	*/
	20
	21	#include <linux/time.h>
48bde86d	22	#include <linux/pagemap.h>
c94c2acf	23	#include <linux/dax.h>
871a2931	24	#include <linux/quotaops.h>
25f4e702 CH	25	#include <linux/iomap.h>
25f4e702 CH	26	#include <linux/uio.h>
1da177e4 LT	27	#include "ext2.h"
	28	#include "xattr.h"
	29	#include "acl.h"
	30
6cd176a5	31	#ifdef CONFIG_FS_DAX
25f4e702 CH	32	static ssize_t ext2_dax_read_iter(struct kiocb iocb, struct iov_iter to)
	33	{
	34	struct inode *inode = iocb->ki_filp->f_mapping->host;
	35	ssize_t ret;
	36
	37	if (!iov_iter_count(to))
	38	return 0; /* skip atime */
	39
	40	inode_lock_shared(inode);
	41	ret = iomap_dax_rw(iocb, to, &ext2_iomap_ops);
	42	inode_unlock_shared(inode);
	43
	44	file_accessed(iocb->ki_filp);
	45	return ret;
	46	}
	47
	48	static ssize_t ext2_dax_write_iter(struct kiocb iocb, struct iov_iter from)
	49	{
	50	struct file *file = iocb->ki_filp;
	51	struct inode *inode = file->f_mapping->host;
	52	ssize_t ret;
	53
	54	inode_lock(inode);
	55	ret = generic_write_checks(iocb, from);
	56	if (ret <= 0)
	57	goto out_unlock;
	58	ret = file_remove_privs(file);
	59	if (ret)
	60	goto out_unlock;
	61	ret = file_update_time(file);
	62	if (ret)
	63	goto out_unlock;
	64
	65	ret = iomap_dax_rw(iocb, from, &ext2_iomap_ops);
	66	if (ret > 0 && iocb->ki_pos > i_size_read(inode)) {
	67	i_size_write(inode, iocb->ki_pos);
	68	mark_inode_dirty(inode);
	69	}
	70
	71	out_unlock:
	72	inode_unlock(inode);
	73	if (ret > 0)
	74	ret = generic_write_sync(iocb, ret);
	75	return ret;
	76	}
	77
5726b27b RZ	78	/*
	79	* The lock ordering for ext2 DAX fault paths is:
	80	*
	81	* mmap_sem (MM)
	82	* sb_start_pagefault (vfs, freeze)
	83	* ext2_inode_info->dax_sem
	84	* address_space->i_mmap_rwsem or page_lock (mutually exclusive in DAX)
	85	* ext2_inode_info->truncate_mutex
	86	*
	87	* The default page_lock and i_size verification done by non-DAX fault paths
	88	* is sufficient because ext2 doesn't support hole punching.
	89	*/
f7ca90b1 MW	90	static int ext2_dax_fault(struct vm_area_struct vma, struct vm_fault vmf)
f7ca90b1 MW	91	{
5726b27b RZ	92	struct inode *inode = file_inode(vma->vm_file);
	93	struct ext2_inode_info *ei = EXT2_I(inode);
	94	int ret;
	95
	96	if (vmf->flags & FAULT_FLAG_WRITE) {
	97	sb_start_pagefault(inode->i_sb);
	98	file_update_time(vma->vm_file);
	99	}
	100	down_read(&ei->dax_sem);
	101
25f4e702	102	ret = iomap_dax_fault(vma, vmf, &ext2_iomap_ops);
5726b27b RZ	103
	104	up_read(&ei->dax_sem);
	105	if (vmf->flags & FAULT_FLAG_WRITE)
	106	sb_end_pagefault(inode->i_sb);
	107	return ret;
f7ca90b1 MW	108	}
f7ca90b1 MW	109
5726b27b RZ	110	static int ext2_dax_pfn_mkwrite(struct vm_area_struct *vma,
	111	struct vm_fault *vmf)
	112	{
	113	struct inode *inode = file_inode(vma->vm_file);
	114	struct ext2_inode_info *ei = EXT2_I(inode);
5726b27b	115	loff_t size;
80b4adca	116	int ret;
5726b27b RZ	117
	118	sb_start_pagefault(inode->i_sb);
	119	file_update_time(vma->vm_file);
	120	down_read(&ei->dax_sem);
	121
	122	/* check that the faulting page hasn't raced with truncate */
	123	size = (i_size_read(inode) + PAGE_SIZE - 1) >> PAGE_SHIFT;
	124	if (vmf->pgoff >= size)
	125	ret = VM_FAULT_SIGBUS;
80b4adca RZ	126	else
80b4adca RZ	127	ret = dax_pfn_mkwrite(vma, vmf);
5726b27b RZ	128
	129	up_read(&ei->dax_sem);
	130	sb_end_pagefault(inode->i_sb);
	131	return ret;
f7ca90b1 MW	132	}
	133
	134	static const struct vm_operations_struct ext2_dax_vm_ops = {
	135	.fault = ext2_dax_fault,
03e0990f RZ	136	/*
	137	* .pmd_fault is not supported for DAX because allocation in ext2
	138	* cannot be reliably aligned to huge page sizes and so pmd faults
	139	* will always fail and fail back to regular faults.
	140	*/
1e9d180b	141	.page_mkwrite = ext2_dax_fault,
5726b27b	142	.pfn_mkwrite = ext2_dax_pfn_mkwrite,
f7ca90b1 MW	143	};
	144
	145	static int ext2_file_mmap(struct file file, struct vm_area_struct vma)
	146	{
	147	if (!IS_DAX(file_inode(file)))
	148	return generic_file_mmap(file, vma);
	149
	150	file_accessed(file);
	151	vma->vm_ops = &ext2_dax_vm_ops;
03e0990f	152	vma->vm_flags \|= VM_MIXEDMAP;
f7ca90b1 MW	153	return 0;
	154	}
	155	#else
	156	#define ext2_file_mmap generic_file_mmap
	157	#endif
	158
1da177e4	159	/*
a6739af8 JK	160	* Called when filp is released. This happens when all file descriptors
	161	* for a single struct file are closed. Note that different open() calls
	162	* for the same file yield different struct file structures.
1da177e4 LT	163	*/
	164	static int ext2_release_file (struct inode * inode, struct file * filp)
	165	{
a686cd89 MB	166	if (filp->f_mode & FMODE_WRITE) {
	167	mutex_lock(&EXT2_I(inode)->truncate_mutex);
	168	ext2_discard_reservation(inode);
	169	mutex_unlock(&EXT2_I(inode)->truncate_mutex);
	170	}
1da177e4 LT	171	return 0;
	172	}
	173
02c24a82	174	int ext2_fsync(struct file *file, loff_t start, loff_t end, int datasync)
48bde86d JK	175	{
48bde86d JK	176	int ret;
7ea80859	177	struct super_block *sb = file->f_mapping->host->i_sb;
48bde86d JK	178	struct address_space *mapping = sb->s_bdev->bd_inode->i_mapping;
48bde86d JK	179
02c24a82	180	ret = generic_file_fsync(file, start, end, datasync);
48bde86d JK	181	if (ret == -EIO \|\| test_and_clear_bit(AS_EIO, &mapping->flags)) {
	182	/* We don't really know where the IO error happened... */
	183	ext2_error(sb, __func__,
	184	"detected IO error when writing metadata buffers");
	185	ret = -EIO;
	186	}
	187	return ret;
	188	}
	189
25f4e702 CH	190	static ssize_t ext2_file_read_iter(struct kiocb iocb, struct iov_iter to)
	191	{
	192	#ifdef CONFIG_FS_DAX
	193	if (IS_DAX(iocb->ki_filp->f_mapping->host))
	194	return ext2_dax_read_iter(iocb, to);
	195	#endif
	196	return generic_file_read_iter(iocb, to);
	197	}
	198
	199	static ssize_t ext2_file_write_iter(struct kiocb iocb, struct iov_iter from)
	200	{
	201	#ifdef CONFIG_FS_DAX
	202	if (IS_DAX(iocb->ki_filp->f_mapping->host))
	203	return ext2_dax_write_iter(iocb, from);
	204	#endif
	205	return generic_file_write_iter(iocb, from);
	206	}
	207
4b6f5d20	208	const struct file_operations ext2_file_operations = {
1da177e4	209	.llseek = generic_file_llseek,
25f4e702 CH	210	.read_iter = ext2_file_read_iter,
25f4e702 CH	211	.write_iter = ext2_file_write_iter,
14f9f7b2	212	.unlocked_ioctl = ext2_ioctl,
e322ff07 DH	213	#ifdef CONFIG_COMPAT
	214	.compat_ioctl = ext2_compat_ioctl,
	215	#endif
f7ca90b1	216	.mmap = ext2_file_mmap,
907f4554	217	.open = dquot_file_open,
1da177e4	218	.release = ext2_release_file,
48bde86d	219	.fsync = ext2_fsync,
dbe6ec81	220	.get_unmapped_area = thp_get_unmapped_area,
5274f052	221	.splice_read = generic_file_splice_read,
8d020765	222	.splice_write = iter_file_splice_write,
1da177e4 LT	223	};
1da177e4 LT	224
754661f1	225	const struct inode_operations ext2_file_inode_operations = {
1da177e4	226	#ifdef CONFIG_EXT2_FS_XATTR
1da177e4	227	.listxattr = ext2_listxattr,
1da177e4 LT	228	#endif
1da177e4 LT	229	.setattr = ext2_setattr,
4e34e719	230	.get_acl = ext2_get_acl,
64e178a7	231	.set_acl = ext2_set_acl,
68c9d702	232	.fiemap = ext2_fiemap,
1da177e4	233	};