[net-next-2.6.git] / mm / filemap_xip.c

/*
 *	linux/mm/filemap_xip.c
 *
 * Copyright (C) 2005 IBM Corporation
 * Author: Carsten Otte <cotte@de.ibm.com>
 *
 * derived from linux/mm/filemap.c - Copyright (C) Linus Torvalds
 *
 */

#include <linux/fs.h>
#include <linux/pagemap.h>
#include <linux/module.h>
#include <linux/uio.h>
#include <linux/rmap.h>
#include <linux/mmu_notifier.h>
#include <linux/sched.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 struct page *__xip_sparse_page;

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

		if (page) {
			static DEFINE_SPINLOCK(xip_alloc_lock);
			spin_lock(&xip_alloc_lock);
			if (!__xip_sparse_page)
				__xip_sparse_page = page;
			else
				__free_page(page);
			spin_unlock(&xip_alloc_lock);
		}
	}
	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)
			nr = len;

		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;

	page = __xip_sparse_page;
	if (!page)
		return;

	spin_lock(&mapping->i_mmap_lock);
	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, vma);
			dec_mm_counter(mm, file_rss);
			BUG_ON(pte_dirty(pteval));
			pte_unmap_unlock(pte, ptl);
			page_cache_release(page);
		}
	}
	spin_unlock(&mapping->i_mmap_lock);
}

/*
 * 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? */

	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 */
		error = mapping->a_ops->get_xip_mem(mapping, vmf->pgoff, 1,
							&xip_mem, &xip_pfn);
		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;
		BUG_ON(err);
		return VM_FAULT_NOPAGE;
	} else {
		/* not shared and writable, use xip_sparse_page() */
		page = xip_sparse_page();
		if (!page)
			return VM_FAULT_OOM;

		page_cache_get(page);
		vmf->page = page;
		return 0;
	}
}

static 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 */
			status = a_ops->get_xip_mem(mapping, index, 1,
							&xip_mem, &xip_pfn);
			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 <cotte@de.ibm.com>
	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>
	13	#include <linux/module.h>
	14	#include <linux/uio.h>
	15	#include <linux/rmap.h>
cddb8a5c	16	#include <linux/mmu_notifier.h>
e8edc6e0	17	#include <linux/sched.h>
ceffc078	18	#include <asm/tlbflush.h>
70688e4d	19	#include <asm/io.h>
ceffc078	20
a76c0b97 CO	21	/*
	22	* We do use our own empty page to avoid interference with other users
	23	* of ZERO_PAGE(), such as /dev/zero
	24	*/
	25	static struct page *__xip_sparse_page;
	26
	27	static struct page *xip_sparse_page(void)
	28	{
	29	if (!__xip_sparse_page) {
c51b1a16 AM	30	struct page *page = alloc_page(GFP_HIGHUSER \| __GFP_ZERO);
	31
	32	if (page) {
a76c0b97 CO	33	static DEFINE_SPINLOCK(xip_alloc_lock);
	34	spin_lock(&xip_alloc_lock);
	35	if (!__xip_sparse_page)
c51b1a16	36	__xip_sparse_page = page;
a76c0b97	37	else
c51b1a16	38	__free_page(page);
a76c0b97 CO	39	spin_unlock(&xip_alloc_lock);
	40	}
	41	}
	42	return __xip_sparse_page;
	43	}
	44
ceffc078 CO	45	/*
ceffc078 CO	46	* This is a file read routine for execute in place files, and uses
70688e4d	47	* the mapping->a_ops->get_xip_mem() function for the actual low-level
ceffc078 CO	48	* stuff.
	49	*
	50	* Note the struct file* is not used at all. It may be NULL.
	51	*/
70688e4d	52	static ssize_t
ceffc078 CO	53	do_xip_mapping_read(struct address_space *mapping,
	54	struct file_ra_state *_ra,
	55	struct file *filp,
70688e4d NP	56	char __user *buf,
	57	size_t len,
	58	loff_t *ppos)
ceffc078 CO	59	{
ceffc078 CO	60	struct inode *inode = mapping->host;
2004dc8e JK	61	pgoff_t index, end_index;
2004dc8e JK	62	unsigned long offset;
70688e4d NP	63	loff_t isize, pos;
70688e4d NP	64	size_t copied = 0, error = 0;
ceffc078	65
70688e4d	66	BUG_ON(!mapping->a_ops->get_xip_mem);
ceffc078	67
70688e4d NP	68	pos = *ppos;
	69	index = pos >> PAGE_CACHE_SHIFT;
	70	offset = pos & ~PAGE_CACHE_MASK;
ceffc078 CO	71
	72	isize = i_size_read(inode);
	73	if (!isize)
	74	goto out;
	75
	76	end_index = (isize - 1) >> PAGE_CACHE_SHIFT;
70688e4d NP	77	do {
	78	unsigned long nr, left;
	79	void *xip_mem;
	80	unsigned long xip_pfn;
	81	int zero = 0;
ceffc078 CO	82
	83	/* nr is the maximum number of bytes to copy from this page */
	84	nr = PAGE_CACHE_SIZE;
	85	if (index >= end_index) {
	86	if (index > end_index)
	87	goto out;
	88	nr = ((isize - 1) & ~PAGE_CACHE_MASK) + 1;
	89	if (nr <= offset) {
	90	goto out;
	91	}
	92	}
	93	nr = nr - offset;
70688e4d NP	94	if (nr > len)
70688e4d NP	95	nr = len;
ceffc078	96
70688e4d NP	97	error = mapping->a_ops->get_xip_mem(mapping, index, 0,
	98	&xip_mem, &xip_pfn);
	99	if (unlikely(error)) {
	100	if (error == -ENODATA) {
ceffc078	101	/* sparse */
70688e4d NP	102	zero = 1;
70688e4d NP	103	} else
ceffc078	104	goto out;
afa597ba	105	}
ceffc078 CO	106
	107	/* If users can be writing to this page using arbitrary
	108	* virtual addresses, take care about potential aliasing
	109	* before reading the page on the kernel side.
	110	*/
	111	if (mapping_writably_mapped(mapping))
70688e4d	112	/* address based flush */ ;
ceffc078 CO	113
ceffc078 CO	114	/*
70688e4d	115	* Ok, we have the mem, so now we can copy it to user space...
ceffc078 CO	116	*
	117	* The actor routine returns how many bytes were actually used..
	118	* NOTE! This may not be the same as how much of a user buffer
	119	* we filled up (we may be padding etc), so we can only update
	120	* "pos" here (the actor routine has to update the user buffer
	121	* pointers and the remaining count).
	122	*/
70688e4d NP	123	if (!zero)
	124	left = __copy_to_user(buf+copied, xip_mem+offset, nr);
	125	else
	126	left = __clear_user(buf + copied, nr);
ceffc078	127
70688e4d NP	128	if (left) {
	129	error = -EFAULT;
	130	goto out;
	131	}
ceffc078	132
70688e4d NP	133	copied += (nr - left);
	134	offset += (nr - left);
	135	index += offset >> PAGE_CACHE_SHIFT;
	136	offset &= ~PAGE_CACHE_MASK;
	137	} while (copied < len);
ceffc078 CO	138
ceffc078 CO	139	out:
70688e4d	140	*ppos = pos + copied;
ceffc078 CO	141	if (filp)
ceffc078 CO	142	file_accessed(filp);
70688e4d NP	143
70688e4d NP	144	return (copied ? copied : error);
ceffc078 CO	145	}
ceffc078 CO	146
ceffc078	147	ssize_t
eb6fe0c3	148	xip_file_read(struct file filp, char __user buf, size_t len, loff_t *ppos)
ceffc078	149	{
eb6fe0c3 CO	150	if (!access_ok(VERIFY_WRITE, buf, len))
eb6fe0c3 CO	151	return -EFAULT;
ceffc078	152
70688e4d NP	153	return do_xip_mapping_read(filp->f_mapping, &filp->f_ra, filp,
70688e4d NP	154	buf, len, ppos);
ceffc078	155	}
eb6fe0c3	156	EXPORT_SYMBOL_GPL(xip_file_read);
ceffc078	157
ceffc078 CO	158	/*
	159	* __xip_unmap is invoked from xip_unmap and
	160	* xip_write
	161	*
	162	* This function walks all vmas of the address_space and unmaps the
a76c0b97	163	* __xip_sparse_page when found at pgoff.
ceffc078 CO	164	*/
	165	static void
	166	__xip_unmap (struct address_space * mapping,
	167	unsigned long pgoff)
	168	{
	169	struct vm_area_struct *vma;
	170	struct mm_struct *mm;
	171	struct prio_tree_iter iter;
	172	unsigned long address;
	173	pte_t *pte;
	174	pte_t pteval;
c0718806	175	spinlock_t *ptl;
67b02f11	176	struct page *page;
ceffc078	177
a76c0b97 CO	178	page = __xip_sparse_page;
	179	if (!page)
	180	return;
	181
ceffc078 CO	182	spin_lock(&mapping->i_mmap_lock);
	183	vma_prio_tree_foreach(vma, &iter, &mapping->i_mmap, pgoff, pgoff) {
	184	mm = vma->vm_mm;
	185	address = vma->vm_start +
	186	((pgoff - vma->vm_pgoff) << PAGE_SHIFT);
	187	BUG_ON(address < vma->vm_start \|\| address >= vma->vm_end);
479db0bf	188	pte = page_check_address(page, mm, address, &ptl, 1);
c0718806	189	if (pte) {
ceffc078	190	/* Nuke the page table entry. */
082ff0a9	191	flush_cache_page(vma, address, pte_pfn(*pte));
cddb8a5c	192	pteval = ptep_clear_flush_notify(vma, address, pte);
7de6b805	193	page_remove_rmap(page, vma);
b5810039	194	dec_mm_counter(mm, file_rss);
ceffc078	195	BUG_ON(pte_dirty(pteval));
c0718806	196	pte_unmap_unlock(pte, ptl);
b5810039	197	page_cache_release(page);
ceffc078 CO	198	}
	199	}
	200	spin_unlock(&mapping->i_mmap_lock);
	201	}
	202
	203	/*
54cb8821	204	* xip_fault() is invoked via the vma operations vector for a
ceffc078 CO	205	* mapped memory region to read in file data during a page fault.
ceffc078 CO	206	*
54cb8821	207	* This function is derived from filemap_fault, but used for execute in place
ceffc078	208	*/
70688e4d	209	static int xip_file_fault(struct vm_area_struct vma, struct vm_fault vmf)
ceffc078	210	{
70688e4d	211	struct file *file = vma->vm_file;
ceffc078 CO	212	struct address_space *mapping = file->f_mapping;
ceffc078 CO	213	struct inode *inode = mapping->host;
54cb8821	214	pgoff_t size;
70688e4d NP	215	void *xip_mem;
	216	unsigned long xip_pfn;
	217	struct page *page;
	218	int error;
ceffc078	219
54cb8821	220	/* XXX: are VM_FAULT_ codes OK? */
ceffc078 CO	221
ceffc078 CO	222	size = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
d0217ac0 NP	223	if (vmf->pgoff >= size)
d0217ac0 NP	224	return VM_FAULT_SIGBUS;
ceffc078	225
70688e4d NP	226	error = mapping->a_ops->get_xip_mem(mapping, vmf->pgoff, 0,
	227	&xip_mem, &xip_pfn);
	228	if (likely(!error))
	229	goto found;
	230	if (error != -ENODATA)
d0217ac0	231	return VM_FAULT_OOM;
ceffc078 CO	232
ceffc078 CO	233	/* sparse block */
70688e4d NP	234	if ((vma->vm_flags & (VM_WRITE \| VM_MAYWRITE)) &&
70688e4d NP	235	(vma->vm_flags & (VM_SHARED \| VM_MAYSHARE)) &&
ceffc078	236	(!(mapping->host->i_sb->s_flags & MS_RDONLY))) {
70688e4d NP	237	int err;
70688e4d NP	238
ceffc078	239	/* maybe shared writable, allocate new block */
70688e4d NP	240	error = mapping->a_ops->get_xip_mem(mapping, vmf->pgoff, 1,
	241	&xip_mem, &xip_pfn);
	242	if (error)
d0217ac0	243	return VM_FAULT_SIGBUS;
70688e4d	244	/* unmap sparse mappings at pgoff from all other vmas */
d0217ac0	245	__xip_unmap(mapping, vmf->pgoff);
70688e4d NP	246
	247	found:
	248	err = vm_insert_mixed(vma, (unsigned long)vmf->virtual_address,
	249	xip_pfn);
	250	if (err == -ENOMEM)
	251	return VM_FAULT_OOM;
	252	BUG_ON(err);
	253	return VM_FAULT_NOPAGE;
ceffc078	254	} else {
a76c0b97 CO	255	/* not shared and writable, use xip_sparse_page() */
a76c0b97 CO	256	page = xip_sparse_page();
d0217ac0 NP	257	if (!page)
d0217ac0 NP	258	return VM_FAULT_OOM;
ceffc078	259
70688e4d NP	260	page_cache_get(page);
	261	vmf->page = page;
	262	return 0;
	263	}
ceffc078 CO	264	}
	265
	266	static struct vm_operations_struct xip_file_vm_ops = {
54cb8821	267	.fault = xip_file_fault,
ceffc078 CO	268	};
	269
	270	int xip_file_mmap(struct file * file, struct vm_area_struct * vma)
	271	{
70688e4d	272	BUG_ON(!file->f_mapping->a_ops->get_xip_mem);
ceffc078 CO	273
	274	file_accessed(file);
	275	vma->vm_ops = &xip_file_vm_ops;
70688e4d	276	vma->vm_flags \|= VM_CAN_NONLINEAR \| VM_MIXEDMAP;
ceffc078 CO	277	return 0;
	278	}
	279	EXPORT_SYMBOL_GPL(xip_file_mmap);
	280
	281	static ssize_t
eb6fe0c3 CO	282	__xip_file_write(struct file filp, const char __user buf,
eb6fe0c3 CO	283	size_t count, loff_t pos, loff_t *ppos)
ceffc078	284	{
eb6fe0c3	285	struct address_space * mapping = filp->f_mapping;
f5e54d6e	286	const struct address_space_operations *a_ops = mapping->a_ops;
ceffc078 CO	287	struct inode *inode = mapping->host;
ceffc078 CO	288	long status = 0;
ceffc078	289	size_t bytes;
ceffc078 CO	290	ssize_t written = 0;
ceffc078 CO	291
70688e4d	292	BUG_ON(!mapping->a_ops->get_xip_mem);
ceffc078	293
ceffc078 CO	294	do {
	295	unsigned long index;
	296	unsigned long offset;
	297	size_t copied;
70688e4d NP	298	void *xip_mem;
70688e4d NP	299	unsigned long xip_pfn;
ceffc078 CO	300
	301	offset = (pos & (PAGE_CACHE_SIZE -1)); /* Within page */
	302	index = pos >> PAGE_CACHE_SHIFT;
	303	bytes = PAGE_CACHE_SIZE - offset;
	304	if (bytes > count)
	305	bytes = count;
	306
70688e4d NP	307	status = a_ops->get_xip_mem(mapping, index, 0,
	308	&xip_mem, &xip_pfn);
	309	if (status == -ENODATA) {
ceffc078	310	/* we allocate a new page unmap it */
70688e4d NP	311	status = a_ops->get_xip_mem(mapping, index, 1,
	312	&xip_mem, &xip_pfn);
	313	if (!status)
eb6fe0c3 CO	314	/* unmap page at pgoff from all other vmas */
eb6fe0c3 CO	315	__xip_unmap(mapping, index);
ceffc078 CO	316	}
ceffc078 CO	317
70688e4d	318	if (status)
ceffc078	319	break;
ceffc078	320
4a9e5ef1	321	copied = bytes -
70688e4d	322	__copy_from_user_nocache(xip_mem + offset, buf, bytes);
4a9e5ef1	323
ceffc078 CO	324	if (likely(copied > 0)) {
	325	status = copied;
	326
	327	if (status >= 0) {
	328	written += status;
	329	count -= status;
	330	pos += status;
	331	buf += status;
ceffc078 CO	332	}
	333	}
	334	if (unlikely(copied != bytes))
	335	if (status >= 0)
	336	status = -EFAULT;
	337	if (status < 0)
	338	break;
	339	} while (count);
	340	*ppos = pos;
	341	/*
	342	* No need to use i_size_read() here, the i_size
1b1dcc1b	343	* cannot change under us because we hold i_mutex.
ceffc078 CO	344	*/
	345	if (pos > inode->i_size) {
	346	i_size_write(inode, pos);
	347	mark_inode_dirty(inode);
	348	}
	349
	350	return written ? written : status;
	351	}
	352
eb6fe0c3 CO	353	ssize_t
	354	xip_file_write(struct file filp, const char __user buf, size_t len,
	355	loff_t *ppos)
ceffc078	356	{
eb6fe0c3 CO	357	struct address_space *mapping = filp->f_mapping;
	358	struct inode *inode = mapping->host;
	359	size_t count;
	360	loff_t pos;
	361	ssize_t ret;
ceffc078	362
1b1dcc1b	363	mutex_lock(&inode->i_mutex);
ceffc078	364
eb6fe0c3 CO	365	if (!access_ok(VERIFY_READ, buf, len)) {
	366	ret=-EFAULT;
	367	goto out_up;
ceffc078 CO	368	}
ceffc078 CO	369
ceffc078	370	pos = *ppos;
eb6fe0c3	371	count = len;
ceffc078 CO	372
	373	vfs_check_frozen(inode->i_sb, SB_FREEZE_WRITE);
	374
eb6fe0c3 CO	375	/* We can write back this queue in page reclaim */
eb6fe0c3 CO	376	current->backing_dev_info = mapping->backing_dev_info;
ceffc078	377
eb6fe0c3 CO	378	ret = generic_write_checks(filp, &pos, &count, S_ISBLK(inode->i_mode));
	379	if (ret)
	380	goto out_backing;
ceffc078	381	if (count == 0)
eb6fe0c3	382	goto out_backing;
ceffc078	383
2f1936b8	384	ret = file_remove_suid(filp);
eb6fe0c3 CO	385	if (ret)
eb6fe0c3 CO	386	goto out_backing;
ceffc078	387
870f4817	388	file_update_time(filp);
ceffc078	389
eb6fe0c3	390	ret = __xip_file_write (filp, buf, count, pos, ppos);
ceffc078	391
eb6fe0c3 CO	392	out_backing:
	393	current->backing_dev_info = NULL;
	394	out_up:
1b1dcc1b	395	mutex_unlock(&inode->i_mutex);
ceffc078 CO	396	return ret;
ceffc078 CO	397	}
eb6fe0c3	398	EXPORT_SYMBOL_GPL(xip_file_write);
ceffc078 CO	399
	400	/*
	401	* truncate a page used for execute in place
70688e4d	402	* functionality is analog to block_truncate_page but does use get_xip_mem
ceffc078 CO	403	* to get the page instead of page cache
	404	*/
	405	int
	406	xip_truncate_page(struct address_space *mapping, loff_t from)
	407	{
	408	pgoff_t index = from >> PAGE_CACHE_SHIFT;
	409	unsigned offset = from & (PAGE_CACHE_SIZE-1);
	410	unsigned blocksize;
	411	unsigned length;
70688e4d NP	412	void *xip_mem;
	413	unsigned long xip_pfn;
	414	int err;
ceffc078	415
70688e4d	416	BUG_ON(!mapping->a_ops->get_xip_mem);
ceffc078 CO	417
	418	blocksize = 1 << mapping->host->i_blkbits;
	419	length = offset & (blocksize - 1);
	420
	421	/* Block boundary? Nothing to do */
	422	if (!length)
	423	return 0;
	424
	425	length = blocksize - length;
	426
70688e4d NP	427	err = mapping->a_ops->get_xip_mem(mapping, index, 0,
	428	&xip_mem, &xip_pfn);
	429	if (unlikely(err)) {
	430	if (err == -ENODATA)
ceffc078 CO	431	/* Hole? No need to truncate */
ceffc078 CO	432	return 0;
eb6fe0c3	433	else
70688e4d	434	return err;
afa597ba	435	}
70688e4d	436	memset(xip_mem + offset, 0, length);
eb6fe0c3	437	return 0;
ceffc078 CO	438	}
ceffc078 CO	439	EXPORT_SYMBOL_GPL(xip_truncate_page);