[deliverable/linux.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 "ext2.h"
#include "xattr.h"
#include "acl.h"

#ifdef CONFIG_FS_DAX
/*
 * 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 = dax_fault(vma, vmf, ext2_get_block);

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

static int ext2_dax_pmd_fault(struct vm_area_struct *vma, unsigned long addr,
						pmd_t *pmd, unsigned int flags)
{
	struct inode *inode = file_inode(vma->vm_file);
	struct ext2_inode_info *ei = EXT2_I(inode);
	int ret;

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

	ret = dax_pmd_fault(vma, addr, pmd, flags, ext2_get_block);

	up_read(&ei->dax_sem);
	if (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	= ext2_dax_pmd_fault,
	.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 | VM_HUGEPAGE;
	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;
}

/*
 * We have mostly NULL's here: the current defaults are ok for
 * the ext2 filesystem.
 */
const struct file_operations ext2_file_operations = {
	.llseek		= generic_file_llseek,
	.read_iter	= generic_file_read_iter,
	.write_iter	= generic_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,
	.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
	.setxattr	= generic_setxattr,
	.getxattr	= generic_getxattr,
	.listxattr	= ext2_listxattr,
	.removexattr	= generic_removexattr,
#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>
1da177e4 LT	25	#include "ext2.h"
	26	#include "xattr.h"
	27	#include "acl.h"
	28
6cd176a5	29	#ifdef CONFIG_FS_DAX
5726b27b RZ	30	/*
	31	* The lock ordering for ext2 DAX fault paths is:
	32	*
	33	* mmap_sem (MM)
	34	* sb_start_pagefault (vfs, freeze)
	35	* ext2_inode_info->dax_sem
	36	* address_space->i_mmap_rwsem or page_lock (mutually exclusive in DAX)
	37	* ext2_inode_info->truncate_mutex
	38	*
	39	* The default page_lock and i_size verification done by non-DAX fault paths
	40	* is sufficient because ext2 doesn't support hole punching.
	41	*/
f7ca90b1 MW	42	static int ext2_dax_fault(struct vm_area_struct vma, struct vm_fault vmf)
f7ca90b1 MW	43	{
5726b27b RZ	44	struct inode *inode = file_inode(vma->vm_file);
	45	struct ext2_inode_info *ei = EXT2_I(inode);
	46	int ret;
	47
	48	if (vmf->flags & FAULT_FLAG_WRITE) {
	49	sb_start_pagefault(inode->i_sb);
	50	file_update_time(vma->vm_file);
	51	}
	52	down_read(&ei->dax_sem);
	53
6b524995	54	ret = dax_fault(vma, vmf, ext2_get_block);
5726b27b RZ	55
	56	up_read(&ei->dax_sem);
	57	if (vmf->flags & FAULT_FLAG_WRITE)
	58	sb_end_pagefault(inode->i_sb);
	59	return ret;
f7ca90b1 MW	60	}
f7ca90b1 MW	61
e7b1ea2a MW	62	static int ext2_dax_pmd_fault(struct vm_area_struct *vma, unsigned long addr,
	63	pmd_t *pmd, unsigned int flags)
	64	{
5726b27b RZ	65	struct inode *inode = file_inode(vma->vm_file);
	66	struct ext2_inode_info *ei = EXT2_I(inode);
	67	int ret;
	68
	69	if (flags & FAULT_FLAG_WRITE) {
	70	sb_start_pagefault(inode->i_sb);
	71	file_update_time(vma->vm_file);
	72	}
	73	down_read(&ei->dax_sem);
	74
6b524995	75	ret = dax_pmd_fault(vma, addr, pmd, flags, ext2_get_block);
5726b27b RZ	76
	77	up_read(&ei->dax_sem);
	78	if (flags & FAULT_FLAG_WRITE)
	79	sb_end_pagefault(inode->i_sb);
	80	return ret;
e7b1ea2a MW	81	}
e7b1ea2a MW	82
5726b27b RZ	83	static int ext2_dax_pfn_mkwrite(struct vm_area_struct *vma,
	84	struct vm_fault *vmf)
	85	{
	86	struct inode *inode = file_inode(vma->vm_file);
	87	struct ext2_inode_info *ei = EXT2_I(inode);
5726b27b	88	loff_t size;
80b4adca	89	int ret;
5726b27b RZ	90
	91	sb_start_pagefault(inode->i_sb);
	92	file_update_time(vma->vm_file);
	93	down_read(&ei->dax_sem);
	94
	95	/* check that the faulting page hasn't raced with truncate */
	96	size = (i_size_read(inode) + PAGE_SIZE - 1) >> PAGE_SHIFT;
	97	if (vmf->pgoff >= size)
	98	ret = VM_FAULT_SIGBUS;
80b4adca RZ	99	else
80b4adca RZ	100	ret = dax_pfn_mkwrite(vma, vmf);
5726b27b RZ	101
	102	up_read(&ei->dax_sem);
	103	sb_end_pagefault(inode->i_sb);
	104	return ret;
f7ca90b1 MW	105	}
	106
	107	static const struct vm_operations_struct ext2_dax_vm_ops = {
	108	.fault = ext2_dax_fault,
e7b1ea2a	109	.pmd_fault = ext2_dax_pmd_fault,
1e9d180b	110	.page_mkwrite = ext2_dax_fault,
5726b27b	111	.pfn_mkwrite = ext2_dax_pfn_mkwrite,
f7ca90b1 MW	112	};
	113
	114	static int ext2_file_mmap(struct file file, struct vm_area_struct vma)
	115	{
	116	if (!IS_DAX(file_inode(file)))
	117	return generic_file_mmap(file, vma);
	118
	119	file_accessed(file);
	120	vma->vm_ops = &ext2_dax_vm_ops;
e7b1ea2a	121	vma->vm_flags \|= VM_MIXEDMAP \| VM_HUGEPAGE;
f7ca90b1 MW	122	return 0;
	123	}
	124	#else
	125	#define ext2_file_mmap generic_file_mmap
	126	#endif
	127
1da177e4	128	/*
a6739af8 JK	129	* Called when filp is released. This happens when all file descriptors
	130	* for a single struct file are closed. Note that different open() calls
	131	* for the same file yield different struct file structures.
1da177e4 LT	132	*/
	133	static int ext2_release_file (struct inode * inode, struct file * filp)
	134	{
a686cd89 MB	135	if (filp->f_mode & FMODE_WRITE) {
	136	mutex_lock(&EXT2_I(inode)->truncate_mutex);
	137	ext2_discard_reservation(inode);
	138	mutex_unlock(&EXT2_I(inode)->truncate_mutex);
	139	}
1da177e4 LT	140	return 0;
	141	}
	142
02c24a82	143	int ext2_fsync(struct file *file, loff_t start, loff_t end, int datasync)
48bde86d JK	144	{
48bde86d JK	145	int ret;
7ea80859	146	struct super_block *sb = file->f_mapping->host->i_sb;
48bde86d JK	147	struct address_space *mapping = sb->s_bdev->bd_inode->i_mapping;
48bde86d JK	148
02c24a82	149	ret = generic_file_fsync(file, start, end, datasync);
48bde86d JK	150	if (ret == -EIO \|\| test_and_clear_bit(AS_EIO, &mapping->flags)) {
	151	/* We don't really know where the IO error happened... */
	152	ext2_error(sb, __func__,
	153	"detected IO error when writing metadata buffers");
	154	ret = -EIO;
	155	}
	156	return ret;
	157	}
	158
1da177e4 LT	159	/*
	160	* We have mostly NULL's here: the current defaults are ok for
	161	* the ext2 filesystem.
	162	*/
4b6f5d20	163	const struct file_operations ext2_file_operations = {
1da177e4	164	.llseek = generic_file_llseek,
aad4f8bb	165	.read_iter = generic_file_read_iter,
8174202b	166	.write_iter = generic_file_write_iter,
14f9f7b2	167	.unlocked_ioctl = ext2_ioctl,
e322ff07 DH	168	#ifdef CONFIG_COMPAT
	169	.compat_ioctl = ext2_compat_ioctl,
	170	#endif
f7ca90b1	171	.mmap = ext2_file_mmap,
907f4554	172	.open = dquot_file_open,
1da177e4	173	.release = ext2_release_file,
48bde86d	174	.fsync = ext2_fsync,
5274f052	175	.splice_read = generic_file_splice_read,
8d020765	176	.splice_write = iter_file_splice_write,
1da177e4 LT	177	};
1da177e4 LT	178
754661f1	179	const struct inode_operations ext2_file_inode_operations = {
1da177e4 LT	180	#ifdef CONFIG_EXT2_FS_XATTR
	181	.setxattr = generic_setxattr,
	182	.getxattr = generic_getxattr,
	183	.listxattr = ext2_listxattr,
	184	.removexattr = generic_removexattr,
	185	#endif
	186	.setattr = ext2_setattr,
4e34e719	187	.get_acl = ext2_get_acl,
64e178a7	188	.set_acl = ext2_set_acl,
68c9d702	189	.fiemap = ext2_fiemap,
1da177e4	190	};