[linux.git] / mm / filemap_xip.c

/*
 *	linux/mm/filemap_xip.c
 *
 * Copyright (C) 2005 IBM Corporation
 * Author: Carsten Otte <[email protected]>
 *
 * derived from linux/mm/filemap.c - Copyright (C) Linus Torvalds
 *
 */

#include <linux/fs.h>
#include <linux/pagemap.h>
#include <linux/export.h>
#include <linux/uio.h>
#include <linux/rmap.h>
#include <linux/mmu_notifier.h>
#include <linux/sched.h>
#include <linux/seqlock.h>
#include <linux/mutex.h>
#include <linux/gfp.h>
#include <asm/tlbflush.h>
#include <asm/io.h>

/*
 * We do use our own empty page to avoid interference with other users
 * of ZERO_PAGE(), such as /dev/zero
 */
static DEFINE_MUTEX(xip_sparse_mutex);
static seqcount_t xip_sparse_seq = SEQCNT_ZERO;
static struct page *__xip_sparse_page;

/* called under xip_sparse_mutex */
static struct page *xip_sparse_page(void)
{
	if (!__xip_sparse_page) {
		struct page *page = alloc_page(GFP_HIGHUSER | __GFP_ZERO);

		if (page)
			__xip_sparse_page = page;
	}
	return __xip_sparse_page;
}

/*
 * This is a file read routine for execute in place files, and uses
 * the mapping->a_ops->get_xip_mem() function for the actual low-level
 * stuff.
 *
 * Note the struct file* is not used at all.  It may be NULL.
 */
static ssize_t
do_xip_mapping_read(struct address_space *mapping,
		    struct file_ra_state *_ra,
		    struct file *filp,
		    char __user *buf,
		    size_t len,
		    loff_t *ppos)
{
	struct inode *inode = mapping->host;
	pgoff_t index, end_index;
	unsigned long offset;
	loff_t isize, pos;
	size_t copied = 0, error = 0;

	BUG_ON(!mapping->a_ops->get_xip_mem);

	pos = *ppos;
	index = pos >> PAGE_CACHE_SHIFT;
	offset = pos & ~PAGE_CACHE_MASK;

	isize = i_size_read(inode);
	if (!isize)
		goto out;

	end_index = (isize - 1) >> PAGE_CACHE_SHIFT;
	do {
		unsigned long nr, left;
		void *xip_mem;
		unsigned long xip_pfn;
		int zero = 0;

		/* nr is the maximum number of bytes to copy from this page */
		nr = PAGE_CACHE_SIZE;
		if (index >= end_index) {
			if (index > end_index)
				goto out;
			nr = ((isize - 1) & ~PAGE_CACHE_MASK) + 1;
			if (nr <= offset) {
				goto out;
			}
		}
		nr = nr - offset;
		if (nr > len - copied)
			nr = len - copied;

		error = mapping->a_ops->get_xip_mem(mapping, index, 0,
							&xip_mem, &xip_pfn);
		if (unlikely(error)) {
			if (error == -ENODATA) {
				/* sparse */
				zero = 1;
			} else
				goto out;
		}

		/* If users can be writing to this page using arbitrary
		 * virtual addresses, take care about potential aliasing
		 * before reading the page on the kernel side.
		 */
		if (mapping_writably_mapped(mapping))
			/* address based flush */ ;

		/*
		 * Ok, we have the mem, so now we can copy it to user space...
		 *
		 * The actor routine returns how many bytes were actually used..
		 * NOTE! This may not be the same as how much of a user buffer
		 * we filled up (we may be padding etc), so we can only update
		 * "pos" here (the actor routine has to update the user buffer
		 * pointers and the remaining count).
		 */
		if (!zero)
			left = __copy_to_user(buf+copied, xip_mem+offset, nr);
		else
			left = __clear_user(buf + copied, nr);

		if (left) {
			error = -EFAULT;
			goto out;
		}

		copied += (nr - left);
		offset += (nr - left);
		index += offset >> PAGE_CACHE_SHIFT;
		offset &= ~PAGE_CACHE_MASK;
	} while (copied < len);

out:
	*ppos = pos + copied;
	if (filp)
		file_accessed(filp);

	return (copied ? copied : error);
}

ssize_t
xip_file_read(struct file *filp, char __user *buf, size_t len, loff_t *ppos)
{
	if (!access_ok(VERIFY_WRITE, buf, len))
		return -EFAULT;

	return do_xip_mapping_read(filp->f_mapping, &filp->f_ra, filp,
			    buf, len, ppos);
}
EXPORT_SYMBOL_GPL(xip_file_read);

/*
 * __xip_unmap is invoked from xip_unmap and
 * xip_write
 *
 * This function walks all vmas of the address_space and unmaps the
 * __xip_sparse_page when found at pgoff.
 */
static void
__xip_unmap (struct address_space * mapping,
		     unsigned long pgoff)
{
	struct vm_area_struct *vma;
	struct mm_struct *mm;
	struct prio_tree_iter iter;
	unsigned long address;
	pte_t *pte;
	pte_t pteval;
	spinlock_t *ptl;
	struct page *page;
	unsigned count;
	int locked = 0;

	count = read_seqcount_begin(&xip_sparse_seq);

	page = __xip_sparse_page;
	if (!page)
		return;

retry:
	mutex_lock(&mapping->i_mmap_mutex);
	vma_prio_tree_foreach(vma, &iter, &mapping->i_mmap, pgoff, pgoff) {
		mm = vma->vm_mm;
		address = vma->vm_start +
			((pgoff - vma->vm_pgoff) << PAGE_SHIFT);
		BUG_ON(address < vma->vm_start || address >= vma->vm_end);
		pte = page_check_address(page, mm, address, &ptl, 1);
		if (pte) {
			/* Nuke the page table entry. */
			flush_cache_page(vma, address, pte_pfn(*pte));
			pteval = ptep_clear_flush_notify(vma, address, pte);
			page_remove_rmap(page);
			dec_mm_counter(mm, MM_FILEPAGES);
			BUG_ON(pte_dirty(pteval));
			pte_unmap_unlock(pte, ptl);
			page_cache_release(page);
		}
	}
	mutex_unlock(&mapping->i_mmap_mutex);

	if (locked) {
		mutex_unlock(&xip_sparse_mutex);
	} else if (read_seqcount_retry(&xip_sparse_seq, count)) {
		mutex_lock(&xip_sparse_mutex);
		locked = 1;
		goto retry;
	}
}

/*
 * xip_fault() is invoked via the vma operations vector for a
 * mapped memory region to read in file data during a page fault.
 *
 * This function is derived from filemap_fault, but used for execute in place
 */
static int xip_file_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
{
	struct file *file = vma->vm_file;
	struct address_space *mapping = file->f_mapping;
	struct inode *inode = mapping->host;
	pgoff_t size;
	void *xip_mem;
	unsigned long xip_pfn;
	struct page *page;
	int error;

	/* XXX: are VM_FAULT_ codes OK? */
again:
	size = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
	if (vmf->pgoff >= size)
		return VM_FAULT_SIGBUS;

	error = mapping->a_ops->get_xip_mem(mapping, vmf->pgoff, 0,
						&xip_mem, &xip_pfn);
	if (likely(!error))
		goto found;
	if (error != -ENODATA)
		return VM_FAULT_OOM;

	/* sparse block */
	if ((vma->vm_flags & (VM_WRITE | VM_MAYWRITE)) &&
	    (vma->vm_flags & (VM_SHARED | VM_MAYSHARE)) &&
	    (!(mapping->host->i_sb->s_flags & MS_RDONLY))) {
		int err;

		/* maybe shared writable, allocate new block */
		mutex_lock(&xip_sparse_mutex);
		error = mapping->a_ops->get_xip_mem(mapping, vmf->pgoff, 1,
							&xip_mem, &xip_pfn);
		mutex_unlock(&xip_sparse_mutex);
		if (error)
			return VM_FAULT_SIGBUS;
		/* unmap sparse mappings at pgoff from all other vmas */
		__xip_unmap(mapping, vmf->pgoff);

found:
		err = vm_insert_mixed(vma, (unsigned long)vmf->virtual_address,
							xip_pfn);
		if (err == -ENOMEM)
			return VM_FAULT_OOM;
		/*
		 * err == -EBUSY is fine, we've raced against another thread
		 * that faulted-in the same page
		 */
		if (err != -EBUSY)
			BUG_ON(err);
		return VM_FAULT_NOPAGE;
	} else {
		int err, ret = VM_FAULT_OOM;

		mutex_lock(&xip_sparse_mutex);
		write_seqcount_begin(&xip_sparse_seq);
		error = mapping->a_ops->get_xip_mem(mapping, vmf->pgoff, 0,
							&xip_mem, &xip_pfn);
		if (unlikely(!error)) {
			write_seqcount_end(&xip_sparse_seq);
			mutex_unlock(&xip_sparse_mutex);
			goto again;
		}
		if (error != -ENODATA)
			goto out;
		/* not shared and writable, use xip_sparse_page() */
		page = xip_sparse_page();
		if (!page)
			goto out;
		err = vm_insert_page(vma, (unsigned long)vmf->virtual_address,
							page);
		if (err == -ENOMEM)
			goto out;

		ret = VM_FAULT_NOPAGE;
out:
		write_seqcount_end(&xip_sparse_seq);
		mutex_unlock(&xip_sparse_mutex);

		return ret;
	}
}

static const struct vm_operations_struct xip_file_vm_ops = {
	.fault	= xip_file_fault,
};

int xip_file_mmap(struct file * file, struct vm_area_struct * vma)
{
	BUG_ON(!file->f_mapping->a_ops->get_xip_mem);

	file_accessed(file);
	vma->vm_ops = &xip_file_vm_ops;
	vma->vm_flags |= VM_CAN_NONLINEAR | VM_MIXEDMAP;
	return 0;
}
EXPORT_SYMBOL_GPL(xip_file_mmap);

static ssize_t
__xip_file_write(struct file *filp, const char __user *buf,
		  size_t count, loff_t pos, loff_t *ppos)
{
	struct address_space * mapping = filp->f_mapping;
	const struct address_space_operations *a_ops = mapping->a_ops;
	struct inode 	*inode = mapping->host;
	long		status = 0;
	size_t		bytes;
	ssize_t		written = 0;

	BUG_ON(!mapping->a_ops->get_xip_mem);

	do {
		unsigned long index;
		unsigned long offset;
		size_t copied;
		void *xip_mem;
		unsigned long xip_pfn;

		offset = (pos & (PAGE_CACHE_SIZE -1)); /* Within page */
		index = pos >> PAGE_CACHE_SHIFT;
		bytes = PAGE_CACHE_SIZE - offset;
		if (bytes > count)
			bytes = count;

		status = a_ops->get_xip_mem(mapping, index, 0,
						&xip_mem, &xip_pfn);
		if (status == -ENODATA) {
			/* we allocate a new page unmap it */
			mutex_lock(&xip_sparse_mutex);
			status = a_ops->get_xip_mem(mapping, index, 1,
							&xip_mem, &xip_pfn);
			mutex_unlock(&xip_sparse_mutex);
			if (!status)
				/* unmap page at pgoff from all other vmas */
				__xip_unmap(mapping, index);
		}

		if (status)
			break;

		copied = bytes -
			__copy_from_user_nocache(xip_mem + offset, buf, bytes);

		if (likely(copied > 0)) {
			status = copied;

			if (status >= 0) {
				written += status;
				count -= status;
				pos += status;
				buf += status;
			}
		}
		if (unlikely(copied != bytes))
			if (status >= 0)
				status = -EFAULT;
		if (status < 0)
			break;
	} while (count);
	*ppos = pos;
	/*
	 * No need to use i_size_read() here, the i_size
	 * cannot change under us because we hold i_mutex.
	 */
	if (pos > inode->i_size) {
		i_size_write(inode, pos);
		mark_inode_dirty(inode);
	}

	return written ? written : status;
}

ssize_t
xip_file_write(struct file *filp, const char __user *buf, size_t len,
	       loff_t *ppos)
{
	struct address_space *mapping = filp->f_mapping;
	struct inode *inode = mapping->host;
	size_t count;
	loff_t pos;
	ssize_t ret;

	mutex_lock(&inode->i_mutex);

	if (!access_ok(VERIFY_READ, buf, len)) {
		ret=-EFAULT;
		goto out_up;
	}

	pos = *ppos;
	count = len;

	vfs_check_frozen(inode->i_sb, SB_FREEZE_WRITE);

	/* We can write back this queue in page reclaim */
	current->backing_dev_info = mapping->backing_dev_info;

	ret = generic_write_checks(filp, &pos, &count, S_ISBLK(inode->i_mode));
	if (ret)
		goto out_backing;
	if (count == 0)
		goto out_backing;

	ret = file_remove_suid(filp);
	if (ret)
		goto out_backing;

	file_update_time(filp);

	ret = __xip_file_write (filp, buf, count, pos, ppos);

 out_backing:
	current->backing_dev_info = NULL;
 out_up:
	mutex_unlock(&inode->i_mutex);
	return ret;
}
EXPORT_SYMBOL_GPL(xip_file_write);

/*
 * truncate a page used for execute in place
 * functionality is analog to block_truncate_page but does use get_xip_mem
 * to get the page instead of page cache
 */
int
xip_truncate_page(struct address_space *mapping, loff_t from)
{
	pgoff_t index = from >> PAGE_CACHE_SHIFT;
	unsigned offset = from & (PAGE_CACHE_SIZE-1);
	unsigned blocksize;
	unsigned length;
	void *xip_mem;
	unsigned long xip_pfn;
	int err;

	BUG_ON(!mapping->a_ops->get_xip_mem);

	blocksize = 1 << mapping->host->i_blkbits;
	length = offset & (blocksize - 1);

	/* Block boundary? Nothing to do */
	if (!length)
		return 0;

	length = blocksize - length;

	err = mapping->a_ops->get_xip_mem(mapping, index, 0,
						&xip_mem, &xip_pfn);
	if (unlikely(err)) {
		if (err == -ENODATA)
			/* Hole? No need to truncate */
			return 0;
		else
			return err;
	}
	memset(xip_mem + offset, 0, length);
	return 0;
}
EXPORT_SYMBOL_GPL(xip_truncate_page);
Commit	Line	Data
ceffc078 CO	1	/*
	2	* linux/mm/filemap_xip.c
	3	*
	4	* Copyright (C) 2005 IBM Corporation
	5	* Author: Carsten Otte <[email protected]>
	6	*
	7	* derived from linux/mm/filemap.c - Copyright (C) Linus Torvalds
	8	*
	9	*/
	10
	11	#include <linux/fs.h>
	12	#include <linux/pagemap.h>
b95f1b31	13	#include <linux/export.h>
ceffc078 CO	14	#include <linux/uio.h>
ceffc078 CO	15	#include <linux/rmap.h>
cddb8a5c	16	#include <linux/mmu_notifier.h>
e8edc6e0	17	#include <linux/sched.h>
538f8ea6 NP	18	#include <linux/seqlock.h>
538f8ea6 NP	19	#include <linux/mutex.h>
5a0e3ad6	20	#include <linux/gfp.h>
ceffc078	21	#include <asm/tlbflush.h>
70688e4d	22	#include <asm/io.h>
ceffc078	23
a76c0b97 CO	24	/*
	25	* We do use our own empty page to avoid interference with other users
	26	* of ZERO_PAGE(), such as /dev/zero
	27	*/
538f8ea6 NP	28	static DEFINE_MUTEX(xip_sparse_mutex);
538f8ea6 NP	29	static seqcount_t xip_sparse_seq = SEQCNT_ZERO;
a76c0b97 CO	30	static struct page *__xip_sparse_page;
a76c0b97 CO	31
538f8ea6	32	/* called under xip_sparse_mutex */
a76c0b97 CO	33	static struct page *xip_sparse_page(void)
	34	{
	35	if (!__xip_sparse_page) {
c51b1a16 AM	36	struct page *page = alloc_page(GFP_HIGHUSER \| __GFP_ZERO);
c51b1a16 AM	37
538f8ea6 NP	38	if (page)
538f8ea6 NP	39	__xip_sparse_page = page;
a76c0b97 CO	40	}
	41	return __xip_sparse_page;
	42	}
	43
ceffc078 CO	44	/*
ceffc078 CO	45	* This is a file read routine for execute in place files, and uses
70688e4d	46	* the mapping->a_ops->get_xip_mem() function for the actual low-level
ceffc078 CO	47	* stuff.
	48	*
	49	* Note the struct file* is not used at all. It may be NULL.
	50	*/
70688e4d	51	static ssize_t
ceffc078 CO	52	do_xip_mapping_read(struct address_space *mapping,
	53	struct file_ra_state *_ra,
	54	struct file *filp,
70688e4d NP	55	char __user *buf,
	56	size_t len,
	57	loff_t *ppos)
ceffc078 CO	58	{
ceffc078 CO	59	struct inode *inode = mapping->host;
2004dc8e JK	60	pgoff_t index, end_index;
2004dc8e JK	61	unsigned long offset;
70688e4d NP	62	loff_t isize, pos;
70688e4d NP	63	size_t copied = 0, error = 0;
ceffc078	64
70688e4d	65	BUG_ON(!mapping->a_ops->get_xip_mem);
ceffc078	66
70688e4d NP	67	pos = *ppos;
	68	index = pos >> PAGE_CACHE_SHIFT;
	69	offset = pos & ~PAGE_CACHE_MASK;
ceffc078 CO	70
	71	isize = i_size_read(inode);
	72	if (!isize)
	73	goto out;
	74
	75	end_index = (isize - 1) >> PAGE_CACHE_SHIFT;
70688e4d NP	76	do {
	77	unsigned long nr, left;
	78	void *xip_mem;
	79	unsigned long xip_pfn;
	80	int zero = 0;
ceffc078 CO	81
	82	/* nr is the maximum number of bytes to copy from this page */
	83	nr = PAGE_CACHE_SIZE;
	84	if (index >= end_index) {
	85	if (index > end_index)
	86	goto out;
	87	nr = ((isize - 1) & ~PAGE_CACHE_MASK) + 1;
	88	if (nr <= offset) {
	89	goto out;
	90	}
	91	}
	92	nr = nr - offset;
58984ce2 MS	93	if (nr > len - copied)
58984ce2 MS	94	nr = len - copied;
ceffc078	95
70688e4d NP	96	error = mapping->a_ops->get_xip_mem(mapping, index, 0,
	97	&xip_mem, &xip_pfn);
	98	if (unlikely(error)) {
	99	if (error == -ENODATA) {
ceffc078	100	/* sparse */
70688e4d NP	101	zero = 1;
70688e4d NP	102	} else
ceffc078	103	goto out;
afa597ba	104	}
ceffc078 CO	105
	106	/* If users can be writing to this page using arbitrary
	107	* virtual addresses, take care about potential aliasing
	108	* before reading the page on the kernel side.
	109	*/
	110	if (mapping_writably_mapped(mapping))
70688e4d	111	/* address based flush */ ;
ceffc078 CO	112
ceffc078 CO	113	/*
70688e4d	114	* Ok, we have the mem, so now we can copy it to user space...
ceffc078 CO	115	*
	116	* The actor routine returns how many bytes were actually used..
	117	* NOTE! This may not be the same as how much of a user buffer
	118	* we filled up (we may be padding etc), so we can only update
	119	* "pos" here (the actor routine has to update the user buffer
	120	* pointers and the remaining count).
	121	*/
70688e4d NP	122	if (!zero)
	123	left = __copy_to_user(buf+copied, xip_mem+offset, nr);
	124	else
	125	left = __clear_user(buf + copied, nr);
ceffc078	126
70688e4d NP	127	if (left) {
	128	error = -EFAULT;
	129	goto out;
	130	}
ceffc078	131
70688e4d NP	132	copied += (nr - left);
	133	offset += (nr - left);
	134	index += offset >> PAGE_CACHE_SHIFT;
	135	offset &= ~PAGE_CACHE_MASK;
	136	} while (copied < len);
ceffc078 CO	137
ceffc078 CO	138	out:
70688e4d	139	*ppos = pos + copied;
ceffc078 CO	140	if (filp)
ceffc078 CO	141	file_accessed(filp);
70688e4d NP	142
70688e4d NP	143	return (copied ? copied : error);
ceffc078 CO	144	}
ceffc078 CO	145
ceffc078	146	ssize_t
eb6fe0c3	147	xip_file_read(struct file filp, char __user buf, size_t len, loff_t *ppos)
ceffc078	148	{
eb6fe0c3 CO	149	if (!access_ok(VERIFY_WRITE, buf, len))
eb6fe0c3 CO	150	return -EFAULT;
ceffc078	151
70688e4d NP	152	return do_xip_mapping_read(filp->f_mapping, &filp->f_ra, filp,
70688e4d NP	153	buf, len, ppos);
ceffc078	154	}
eb6fe0c3	155	EXPORT_SYMBOL_GPL(xip_file_read);
ceffc078	156
ceffc078 CO	157	/*
	158	* __xip_unmap is invoked from xip_unmap and
	159	* xip_write
	160	*
	161	* This function walks all vmas of the address_space and unmaps the
a76c0b97	162	* __xip_sparse_page when found at pgoff.
ceffc078 CO	163	*/
	164	static void
	165	__xip_unmap (struct address_space * mapping,
	166	unsigned long pgoff)
	167	{
	168	struct vm_area_struct *vma;
	169	struct mm_struct *mm;
	170	struct prio_tree_iter iter;
	171	unsigned long address;
	172	pte_t *pte;
	173	pte_t pteval;
c0718806	174	spinlock_t *ptl;
67b02f11	175	struct page *page;
538f8ea6 NP	176	unsigned count;
	177	int locked = 0;
	178
	179	count = read_seqcount_begin(&xip_sparse_seq);
ceffc078	180
a76c0b97 CO	181	page = __xip_sparse_page;
	182	if (!page)
	183	return;
	184
538f8ea6	185	retry:
3d48ae45	186	mutex_lock(&mapping->i_mmap_mutex);
ceffc078 CO	187	vma_prio_tree_foreach(vma, &iter, &mapping->i_mmap, pgoff, pgoff) {
	188	mm = vma->vm_mm;
	189	address = vma->vm_start +
	190	((pgoff - vma->vm_pgoff) << PAGE_SHIFT);
	191	BUG_ON(address < vma->vm_start \|\| address >= vma->vm_end);
479db0bf	192	pte = page_check_address(page, mm, address, &ptl, 1);
c0718806	193	if (pte) {
ceffc078	194	/* Nuke the page table entry. */
082ff0a9	195	flush_cache_page(vma, address, pte_pfn(*pte));
cddb8a5c	196	pteval = ptep_clear_flush_notify(vma, address, pte);
edc315fd	197	page_remove_rmap(page);
d559db08	198	dec_mm_counter(mm, MM_FILEPAGES);
ceffc078	199	BUG_ON(pte_dirty(pteval));
c0718806	200	pte_unmap_unlock(pte, ptl);
b5810039	201	page_cache_release(page);
ceffc078 CO	202	}
ceffc078 CO	203	}
3d48ae45	204	mutex_unlock(&mapping->i_mmap_mutex);
538f8ea6 NP	205
	206	if (locked) {
	207	mutex_unlock(&xip_sparse_mutex);
	208	} else if (read_seqcount_retry(&xip_sparse_seq, count)) {
	209	mutex_lock(&xip_sparse_mutex);
	210	locked = 1;
	211	goto retry;
	212	}
ceffc078 CO	213	}
	214
	215	/*
54cb8821	216	* xip_fault() is invoked via the vma operations vector for a
ceffc078 CO	217	* mapped memory region to read in file data during a page fault.
ceffc078 CO	218	*
54cb8821	219	* This function is derived from filemap_fault, but used for execute in place
ceffc078	220	*/
70688e4d	221	static int xip_file_fault(struct vm_area_struct vma, struct vm_fault vmf)
ceffc078	222	{
70688e4d	223	struct file *file = vma->vm_file;
ceffc078 CO	224	struct address_space *mapping = file->f_mapping;
ceffc078 CO	225	struct inode *inode = mapping->host;
54cb8821	226	pgoff_t size;
70688e4d NP	227	void *xip_mem;
	228	unsigned long xip_pfn;
	229	struct page *page;
	230	int error;
ceffc078	231
54cb8821	232	/* XXX: are VM_FAULT_ codes OK? */
538f8ea6	233	again:
ceffc078	234	size = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
d0217ac0 NP	235	if (vmf->pgoff >= size)
d0217ac0 NP	236	return VM_FAULT_SIGBUS;
ceffc078	237
70688e4d NP	238	error = mapping->a_ops->get_xip_mem(mapping, vmf->pgoff, 0,
	239	&xip_mem, &xip_pfn);
	240	if (likely(!error))
	241	goto found;
	242	if (error != -ENODATA)
d0217ac0	243	return VM_FAULT_OOM;
ceffc078 CO	244
ceffc078 CO	245	/* sparse block */
70688e4d NP	246	if ((vma->vm_flags & (VM_WRITE \| VM_MAYWRITE)) &&
70688e4d NP	247	(vma->vm_flags & (VM_SHARED \| VM_MAYSHARE)) &&
ceffc078	248	(!(mapping->host->i_sb->s_flags & MS_RDONLY))) {
70688e4d NP	249	int err;
70688e4d NP	250
ceffc078	251	/* maybe shared writable, allocate new block */
14bac5ac	252	mutex_lock(&xip_sparse_mutex);
70688e4d NP	253	error = mapping->a_ops->get_xip_mem(mapping, vmf->pgoff, 1,
70688e4d NP	254	&xip_mem, &xip_pfn);
14bac5ac	255	mutex_unlock(&xip_sparse_mutex);
70688e4d	256	if (error)
d0217ac0	257	return VM_FAULT_SIGBUS;
70688e4d	258	/* unmap sparse mappings at pgoff from all other vmas */
d0217ac0	259	__xip_unmap(mapping, vmf->pgoff);
70688e4d NP	260
	261	found:
	262	err = vm_insert_mixed(vma, (unsigned long)vmf->virtual_address,
	263	xip_pfn);
	264	if (err == -ENOMEM)
	265	return VM_FAULT_OOM;
99f02ef1 CO	266	/*
	267	* err == -EBUSY is fine, we've raced against another thread
	268	* that faulted-in the same page
	269	*/
	270	if (err != -EBUSY)
	271	BUG_ON(err);
70688e4d	272	return VM_FAULT_NOPAGE;
ceffc078	273	} else {
538f8ea6 NP	274	int err, ret = VM_FAULT_OOM;
	275
	276	mutex_lock(&xip_sparse_mutex);
	277	write_seqcount_begin(&xip_sparse_seq);
	278	error = mapping->a_ops->get_xip_mem(mapping, vmf->pgoff, 0,
	279	&xip_mem, &xip_pfn);
	280	if (unlikely(!error)) {
	281	write_seqcount_end(&xip_sparse_seq);
	282	mutex_unlock(&xip_sparse_mutex);
	283	goto again;
	284	}
	285	if (error != -ENODATA)
	286	goto out;
a76c0b97 CO	287	/* not shared and writable, use xip_sparse_page() */
a76c0b97 CO	288	page = xip_sparse_page();
d0217ac0	289	if (!page)
538f8ea6 NP	290	goto out;
	291	err = vm_insert_page(vma, (unsigned long)vmf->virtual_address,
	292	page);
	293	if (err == -ENOMEM)
	294	goto out;
	295
	296	ret = VM_FAULT_NOPAGE;
	297	out:
	298	write_seqcount_end(&xip_sparse_seq);
	299	mutex_unlock(&xip_sparse_mutex);
ceffc078	300
538f8ea6	301	return ret;
70688e4d	302	}
ceffc078 CO	303	}
ceffc078 CO	304
f0f37e2f	305	static const struct vm_operations_struct xip_file_vm_ops = {
54cb8821	306	.fault = xip_file_fault,
ceffc078 CO	307	};
	308
	309	int xip_file_mmap(struct file * file, struct vm_area_struct * vma)
	310	{
70688e4d	311	BUG_ON(!file->f_mapping->a_ops->get_xip_mem);
ceffc078 CO	312
	313	file_accessed(file);
	314	vma->vm_ops = &xip_file_vm_ops;
70688e4d	315	vma->vm_flags \|= VM_CAN_NONLINEAR \| VM_MIXEDMAP;
ceffc078 CO	316	return 0;
	317	}
	318	EXPORT_SYMBOL_GPL(xip_file_mmap);
	319
	320	static ssize_t
eb6fe0c3 CO	321	__xip_file_write(struct file filp, const char __user buf,
eb6fe0c3 CO	322	size_t count, loff_t pos, loff_t *ppos)
ceffc078	323	{
eb6fe0c3	324	struct address_space * mapping = filp->f_mapping;
f5e54d6e	325	const struct address_space_operations *a_ops = mapping->a_ops;
ceffc078 CO	326	struct inode *inode = mapping->host;
ceffc078 CO	327	long status = 0;
ceffc078	328	size_t bytes;
ceffc078 CO	329	ssize_t written = 0;
ceffc078 CO	330
70688e4d	331	BUG_ON(!mapping->a_ops->get_xip_mem);
ceffc078	332
ceffc078 CO	333	do {
	334	unsigned long index;
	335	unsigned long offset;
	336	size_t copied;
70688e4d NP	337	void *xip_mem;
70688e4d NP	338	unsigned long xip_pfn;
ceffc078 CO	339
	340	offset = (pos & (PAGE_CACHE_SIZE -1)); /* Within page */
	341	index = pos >> PAGE_CACHE_SHIFT;
	342	bytes = PAGE_CACHE_SIZE - offset;
	343	if (bytes > count)
	344	bytes = count;
	345
70688e4d NP	346	status = a_ops->get_xip_mem(mapping, index, 0,
	347	&xip_mem, &xip_pfn);
	348	if (status == -ENODATA) {
ceffc078	349	/* we allocate a new page unmap it */
14bac5ac	350	mutex_lock(&xip_sparse_mutex);
70688e4d NP	351	status = a_ops->get_xip_mem(mapping, index, 1,
70688e4d NP	352	&xip_mem, &xip_pfn);
14bac5ac	353	mutex_unlock(&xip_sparse_mutex);
70688e4d	354	if (!status)
eb6fe0c3 CO	355	/* unmap page at pgoff from all other vmas */
eb6fe0c3 CO	356	__xip_unmap(mapping, index);
ceffc078 CO	357	}
ceffc078 CO	358
70688e4d	359	if (status)
ceffc078	360	break;
ceffc078	361
4a9e5ef1	362	copied = bytes -
70688e4d	363	__copy_from_user_nocache(xip_mem + offset, buf, bytes);
4a9e5ef1	364
ceffc078 CO	365	if (likely(copied > 0)) {
	366	status = copied;
	367
	368	if (status >= 0) {
	369	written += status;
	370	count -= status;
	371	pos += status;
	372	buf += status;
ceffc078 CO	373	}
	374	}
	375	if (unlikely(copied != bytes))
	376	if (status >= 0)
	377	status = -EFAULT;
	378	if (status < 0)
	379	break;
	380	} while (count);
	381	*ppos = pos;
	382	/*
	383	* No need to use i_size_read() here, the i_size
1b1dcc1b	384	* cannot change under us because we hold i_mutex.
ceffc078 CO	385	*/
	386	if (pos > inode->i_size) {
	387	i_size_write(inode, pos);
	388	mark_inode_dirty(inode);
	389	}
	390
	391	return written ? written : status;
	392	}
	393
eb6fe0c3 CO	394	ssize_t
	395	xip_file_write(struct file filp, const char __user buf, size_t len,
	396	loff_t *ppos)
ceffc078	397	{
eb6fe0c3 CO	398	struct address_space *mapping = filp->f_mapping;
	399	struct inode *inode = mapping->host;
	400	size_t count;
	401	loff_t pos;
	402	ssize_t ret;
ceffc078	403
1b1dcc1b	404	mutex_lock(&inode->i_mutex);
ceffc078	405
eb6fe0c3 CO	406	if (!access_ok(VERIFY_READ, buf, len)) {
	407	ret=-EFAULT;
	408	goto out_up;
ceffc078 CO	409	}
ceffc078 CO	410
ceffc078	411	pos = *ppos;
eb6fe0c3	412	count = len;
ceffc078 CO	413
	414	vfs_check_frozen(inode->i_sb, SB_FREEZE_WRITE);
	415
eb6fe0c3 CO	416	/* We can write back this queue in page reclaim */
eb6fe0c3 CO	417	current->backing_dev_info = mapping->backing_dev_info;
ceffc078	418
eb6fe0c3 CO	419	ret = generic_write_checks(filp, &pos, &count, S_ISBLK(inode->i_mode));
	420	if (ret)
	421	goto out_backing;
ceffc078	422	if (count == 0)
eb6fe0c3	423	goto out_backing;
ceffc078	424
2f1936b8	425	ret = file_remove_suid(filp);
eb6fe0c3 CO	426	if (ret)
eb6fe0c3 CO	427	goto out_backing;
ceffc078	428
870f4817	429	file_update_time(filp);
ceffc078	430
eb6fe0c3	431	ret = __xip_file_write (filp, buf, count, pos, ppos);
ceffc078	432
eb6fe0c3 CO	433	out_backing:
	434	current->backing_dev_info = NULL;
	435	out_up:
1b1dcc1b	436	mutex_unlock(&inode->i_mutex);
ceffc078 CO	437	return ret;
ceffc078 CO	438	}
eb6fe0c3	439	EXPORT_SYMBOL_GPL(xip_file_write);
ceffc078 CO	440
	441	/*
	442	* truncate a page used for execute in place
70688e4d	443	* functionality is analog to block_truncate_page but does use get_xip_mem
ceffc078 CO	444	* to get the page instead of page cache
	445	*/
	446	int
	447	xip_truncate_page(struct address_space *mapping, loff_t from)
	448	{
	449	pgoff_t index = from >> PAGE_CACHE_SHIFT;
	450	unsigned offset = from & (PAGE_CACHE_SIZE-1);
	451	unsigned blocksize;
	452	unsigned length;
70688e4d NP	453	void *xip_mem;
	454	unsigned long xip_pfn;
	455	int err;
ceffc078	456
70688e4d	457	BUG_ON(!mapping->a_ops->get_xip_mem);
ceffc078 CO	458
	459	blocksize = 1 << mapping->host->i_blkbits;
	460	length = offset & (blocksize - 1);
	461
	462	/* Block boundary? Nothing to do */
	463	if (!length)
	464	return 0;
	465
	466	length = blocksize - length;
	467
70688e4d NP	468	err = mapping->a_ops->get_xip_mem(mapping, index, 0,
	469	&xip_mem, &xip_pfn);
	470	if (unlikely(err)) {
	471	if (err == -ENODATA)
ceffc078 CO	472	/* Hole? No need to truncate */
ceffc078 CO	473	return 0;
eb6fe0c3	474	else
70688e4d	475	return err;
afa597ba	476	}
70688e4d	477	memset(xip_mem + offset, 0, length);
eb6fe0c3	478	return 0;
ceffc078 CO	479	}
ceffc078 CO	480	EXPORT_SYMBOL_GPL(xip_truncate_page);