[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/module.h>
#include <linux/uio.h>
#include <linux/rmap.h>
#include <linux/sched.h>
#include <asm/tlbflush.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_page() function for the actual low-level
 * stuff.
 *
 * Note the struct file* is not used at all.  It may be NULL.
 */
static void
do_xip_mapping_read(struct address_space *mapping,
		    struct file_ra_state *_ra,
		    struct file *filp,
		    loff_t *ppos,
		    read_descriptor_t *desc,
		    read_actor_t actor)
{
	struct inode *inode = mapping->host;
	pgoff_t index, end_index;
	unsigned long offset;
	loff_t isize;

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

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

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

	end_index = (isize - 1) >> PAGE_CACHE_SHIFT;
	for (;;) {
		struct page *page;
		unsigned long nr, ret;

		/* 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;

		page = mapping->a_ops->get_xip_page(mapping,
			index*(PAGE_SIZE/512), 0);
		if (!page)
			goto no_xip_page;
		if (unlikely(IS_ERR(page))) {
			if (PTR_ERR(page) == -ENODATA) {
				/* sparse */
				page = ZERO_PAGE(0);
			} else {
				desc->error = PTR_ERR(page);
				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))
			flush_dcache_page(page);

		/*
		 * Ok, we have the page, 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).
		 */
		ret = actor(desc, page, offset, nr);
		offset += ret;
		index += offset >> PAGE_CACHE_SHIFT;
		offset &= ~PAGE_CACHE_MASK;

		if (ret == nr && desc->count)
			continue;
		goto out;

no_xip_page:
		/* Did not get the page. Report it */
		desc->error = -EIO;
		goto out;
	}

out:
	*ppos = ((loff_t) index << PAGE_CACHE_SHIFT) + offset;
	if (filp)
		file_accessed(filp);
}

ssize_t
xip_file_read(struct file *filp, char __user *buf, size_t len, loff_t *ppos)
{
	read_descriptor_t desc;

	if (!access_ok(VERIFY_WRITE, buf, len))
		return -EFAULT;

	desc.written = 0;
	desc.arg.buf = buf;
	desc.count = len;
	desc.error = 0;

	do_xip_mapping_read(filp->f_mapping, &filp->f_ra, filp,
			    ppos, &desc, file_read_actor);

	if (desc.written)
		return desc.written;
	else
		return desc.error;
}
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);
		if (pte) {
			/* Nuke the page table entry. */
			flush_cache_page(vma, address, pte_pfn(*pte));
			pteval = ptep_clear_flush(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 *area, struct vm_fault *vmf)
{
	struct file *file = area->vm_file;
	struct address_space *mapping = file->f_mapping;
	struct inode *inode = mapping->host;
	struct page *page;
	pgoff_t size;

	/* 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;

	page = mapping->a_ops->get_xip_page(mapping,
					vmf->pgoff*(PAGE_SIZE/512), 0);
	if (!IS_ERR(page))
		goto out;
	if (PTR_ERR(page) != -ENODATA)
		return VM_FAULT_OOM;

	/* sparse block */
	if ((area->vm_flags & (VM_WRITE | VM_MAYWRITE)) &&
	    (area->vm_flags & (VM_SHARED| VM_MAYSHARE)) &&
	    (!(mapping->host->i_sb->s_flags & MS_RDONLY))) {
		/* maybe shared writable, allocate new block */
		page = mapping->a_ops->get_xip_page(mapping,
					vmf->pgoff*(PAGE_SIZE/512), 1);
		if (IS_ERR(page))
			return VM_FAULT_SIGBUS;
		/* unmap page at pgoff from all other vmas */
		__xip_unmap(mapping, vmf->pgoff);
	} else {
		/* not shared and writable, use xip_sparse_page() */
		page = xip_sparse_page();
		if (!page)
			return VM_FAULT_OOM;
	}

out:
	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_page);

	file_accessed(file);
	vma->vm_ops = &xip_file_vm_ops;
	vma->vm_flags |= VM_CAN_NONLINEAR;
	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;
	struct page	*page;
	size_t		bytes;
	ssize_t		written = 0;

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

	do {
		unsigned long index;
		unsigned long offset;
		size_t copied;
		char *kaddr;

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

		page = a_ops->get_xip_page(mapping,
					   index*(PAGE_SIZE/512), 0);
		if (IS_ERR(page) && (PTR_ERR(page) == -ENODATA)) {
			/* we allocate a new page unmap it */
			page = a_ops->get_xip_page(mapping,
						   index*(PAGE_SIZE/512), 1);
			if (!IS_ERR(page))
				/* unmap page at pgoff from all other vmas */
				__xip_unmap(mapping, index);
		}

		if (IS_ERR(page)) {
			status = PTR_ERR(page);
			break;
		}

		fault_in_pages_readable(buf, bytes);
		kaddr = kmap_atomic(page, KM_USER0);
		copied = bytes -
			__copy_from_user_inatomic_nocache(kaddr + offset, buf, bytes);
		kunmap_atomic(kaddr, KM_USER0);
		flush_dcache_page(page);

		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 = remove_suid(filp->f_path.dentry);
	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_page
 * 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;
	struct page *page;

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

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

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

	length = blocksize - length;

	page = mapping->a_ops->get_xip_page(mapping,
					    index*(PAGE_SIZE/512), 0);
	if (!page)
		return -ENOMEM;
	if (unlikely(IS_ERR(page))) {
		if (PTR_ERR(page) == -ENODATA)
			/* Hole? No need to truncate */
			return 0;
		else
			return PTR_ERR(page);
	}
	zero_user(page, offset, 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>
	13	#include <linux/module.h>
	14	#include <linux/uio.h>
	15	#include <linux/rmap.h>
e8edc6e0	16	#include <linux/sched.h>
ceffc078	17	#include <asm/tlbflush.h>
ceffc078	18
a76c0b97 CO	19	/*
	20	* We do use our own empty page to avoid interference with other users
	21	* of ZERO_PAGE(), such as /dev/zero
	22	*/
	23	static struct page *__xip_sparse_page;
	24
	25	static struct page *xip_sparse_page(void)
	26	{
	27	if (!__xip_sparse_page) {
c51b1a16 AM	28	struct page *page = alloc_page(GFP_HIGHUSER \| __GFP_ZERO);
	29
	30	if (page) {
a76c0b97 CO	31	static DEFINE_SPINLOCK(xip_alloc_lock);
	32	spin_lock(&xip_alloc_lock);
	33	if (!__xip_sparse_page)
c51b1a16	34	__xip_sparse_page = page;
a76c0b97	35	else
c51b1a16	36	__free_page(page);
a76c0b97 CO	37	spin_unlock(&xip_alloc_lock);
	38	}
	39	}
	40	return __xip_sparse_page;
	41	}
	42
ceffc078 CO	43	/*
	44	* This is a file read routine for execute in place files, and uses
	45	* the mapping->a_ops->get_xip_page() function for the actual low-level
	46	* stuff.
	47	*
	48	* Note the struct file* is not used at all. It may be NULL.
	49	*/
	50	static void
	51	do_xip_mapping_read(struct address_space *mapping,
	52	struct file_ra_state *_ra,
	53	struct file *filp,
	54	loff_t *ppos,
	55	read_descriptor_t *desc,
	56	read_actor_t actor)
	57	{
	58	struct inode *inode = mapping->host;
2004dc8e JK	59	pgoff_t index, end_index;
2004dc8e JK	60	unsigned long offset;
ceffc078 CO	61	loff_t isize;
	62
	63	BUG_ON(!mapping->a_ops->get_xip_page);
	64
	65	index = *ppos >> PAGE_CACHE_SHIFT;
	66	offset = *ppos & ~PAGE_CACHE_MASK;
	67
	68	isize = i_size_read(inode);
	69	if (!isize)
	70	goto out;
	71
	72	end_index = (isize - 1) >> PAGE_CACHE_SHIFT;
	73	for (;;) {
	74	struct page *page;
	75	unsigned long nr, ret;
	76
	77	/* nr is the maximum number of bytes to copy from this page */
	78	nr = PAGE_CACHE_SIZE;
	79	if (index >= end_index) {
	80	if (index > end_index)
	81	goto out;
	82	nr = ((isize - 1) & ~PAGE_CACHE_MASK) + 1;
	83	if (nr <= offset) {
	84	goto out;
	85	}
	86	}
	87	nr = nr - offset;
	88
	89	page = mapping->a_ops->get_xip_page(mapping,
	90	index*(PAGE_SIZE/512), 0);
	91	if (!page)
	92	goto no_xip_page;
	93	if (unlikely(IS_ERR(page))) {
	94	if (PTR_ERR(page) == -ENODATA) {
	95	/* sparse */
afa597ba	96	page = ZERO_PAGE(0);
ceffc078 CO	97	} else {
	98	desc->error = PTR_ERR(page);
	99	goto out;
	100	}
afa597ba	101	}
ceffc078 CO	102
	103	/* If users can be writing to this page using arbitrary
	104	* virtual addresses, take care about potential aliasing
	105	* before reading the page on the kernel side.
	106	*/
	107	if (mapping_writably_mapped(mapping))
	108	flush_dcache_page(page);
	109
	110	/*
afa597ba	111	* Ok, we have the page, so now we can copy it to user space...
ceffc078 CO	112	*
	113	* The actor routine returns how many bytes were actually used..
	114	* NOTE! This may not be the same as how much of a user buffer
	115	* we filled up (we may be padding etc), so we can only update
	116	* "pos" here (the actor routine has to update the user buffer
	117	* pointers and the remaining count).
	118	*/
	119	ret = actor(desc, page, offset, nr);
	120	offset += ret;
	121	index += offset >> PAGE_CACHE_SHIFT;
	122	offset &= ~PAGE_CACHE_MASK;
	123
	124	if (ret == nr && desc->count)
	125	continue;
	126	goto out;
	127
	128	no_xip_page:
	129	/* Did not get the page. Report it */
	130	desc->error = -EIO;
	131	goto out;
	132	}
	133
	134	out:
	135	*ppos = ((loff_t) index << PAGE_CACHE_SHIFT) + offset;
	136	if (filp)
	137	file_accessed(filp);
	138	}
	139
ceffc078	140	ssize_t
eb6fe0c3	141	xip_file_read(struct file filp, char __user buf, size_t len, loff_t *ppos)
ceffc078	142	{
eb6fe0c3	143	read_descriptor_t desc;
ceffc078	144
eb6fe0c3 CO	145	if (!access_ok(VERIFY_WRITE, buf, len))
eb6fe0c3 CO	146	return -EFAULT;
ceffc078	147
eb6fe0c3 CO	148	desc.written = 0;
	149	desc.arg.buf = buf;
	150	desc.count = len;
	151	desc.error = 0;
ceffc078	152
eb6fe0c3 CO	153	do_xip_mapping_read(filp->f_mapping, &filp->f_ra, filp,
	154	ppos, &desc, file_read_actor);
	155
	156	if (desc.written)
	157	return desc.written;
	158	else
	159	return desc.error;
ceffc078	160	}
eb6fe0c3	161	EXPORT_SYMBOL_GPL(xip_file_read);
ceffc078	162
ceffc078 CO	163	/*
	164	* __xip_unmap is invoked from xip_unmap and
	165	* xip_write
	166	*
	167	* This function walks all vmas of the address_space and unmaps the
a76c0b97	168	* __xip_sparse_page when found at pgoff.
ceffc078 CO	169	*/
	170	static void
	171	__xip_unmap (struct address_space * mapping,
	172	unsigned long pgoff)
	173	{
	174	struct vm_area_struct *vma;
	175	struct mm_struct *mm;
	176	struct prio_tree_iter iter;
	177	unsigned long address;
	178	pte_t *pte;
	179	pte_t pteval;
c0718806	180	spinlock_t *ptl;
67b02f11	181	struct page *page;
ceffc078	182
a76c0b97 CO	183	page = __xip_sparse_page;
	184	if (!page)
	185	return;
	186
ceffc078 CO	187	spin_lock(&mapping->i_mmap_lock);
	188	vma_prio_tree_foreach(vma, &iter, &mapping->i_mmap, pgoff, pgoff) {
	189	mm = vma->vm_mm;
	190	address = vma->vm_start +
	191	((pgoff - vma->vm_pgoff) << PAGE_SHIFT);
	192	BUG_ON(address < vma->vm_start \|\| address >= vma->vm_end);
c0718806 HD	193	pte = page_check_address(page, mm, address, &ptl);
c0718806 HD	194	if (pte) {
ceffc078	195	/* Nuke the page table entry. */
082ff0a9	196	flush_cache_page(vma, address, pte_pfn(*pte));
ceffc078	197	pteval = ptep_clear_flush(vma, address, pte);
7de6b805	198	page_remove_rmap(page, vma);
b5810039	199	dec_mm_counter(mm, file_rss);
ceffc078	200	BUG_ON(pte_dirty(pteval));
c0718806	201	pte_unmap_unlock(pte, ptl);
b5810039	202	page_cache_release(page);
ceffc078 CO	203	}
	204	}
	205	spin_unlock(&mapping->i_mmap_lock);
	206	}
	207
	208	/*
54cb8821	209	* xip_fault() is invoked via the vma operations vector for a
ceffc078 CO	210	* mapped memory region to read in file data during a page fault.
ceffc078 CO	211	*
54cb8821	212	* This function is derived from filemap_fault, but used for execute in place
ceffc078	213	*/
d0217ac0	214	static int xip_file_fault(struct vm_area_struct area, struct vm_fault vmf)
ceffc078 CO	215	{
	216	struct file *file = area->vm_file;
	217	struct address_space *mapping = file->f_mapping;
	218	struct inode *inode = mapping->host;
	219	struct page *page;
54cb8821	220	pgoff_t size;
ceffc078	221
54cb8821	222	/* XXX: are VM_FAULT_ codes OK? */
ceffc078 CO	223
ceffc078 CO	224	size = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
d0217ac0 NP	225	if (vmf->pgoff >= size)
d0217ac0 NP	226	return VM_FAULT_SIGBUS;
ceffc078	227
54cb8821	228	page = mapping->a_ops->get_xip_page(mapping,
d0217ac0	229	vmf->pgoff*(PAGE_SIZE/512), 0);
a76c0b97	230	if (!IS_ERR(page))
b5810039	231	goto out;
d0217ac0 NP	232	if (PTR_ERR(page) != -ENODATA)
d0217ac0 NP	233	return VM_FAULT_OOM;
ceffc078 CO	234
	235	/* sparse block */
	236	if ((area->vm_flags & (VM_WRITE \| VM_MAYWRITE)) &&
	237	(area->vm_flags & (VM_SHARED\| VM_MAYSHARE)) &&
	238	(!(mapping->host->i_sb->s_flags & MS_RDONLY))) {
	239	/* maybe shared writable, allocate new block */
54cb8821	240	page = mapping->a_ops->get_xip_page(mapping,
d0217ac0 NP	241	vmf->pgoff*(PAGE_SIZE/512), 1);
	242	if (IS_ERR(page))
	243	return VM_FAULT_SIGBUS;
ceffc078	244	/* unmap page at pgoff from all other vmas */
d0217ac0	245	__xip_unmap(mapping, vmf->pgoff);
ceffc078	246	} else {
a76c0b97 CO	247	/* not shared and writable, use xip_sparse_page() */
a76c0b97 CO	248	page = xip_sparse_page();
d0217ac0 NP	249	if (!page)
d0217ac0 NP	250	return VM_FAULT_OOM;
ceffc078 CO	251	}
ceffc078 CO	252
b5810039 NP	253	out:
b5810039 NP	254	page_cache_get(page);
d0217ac0	255	vmf->page = page;
83c54070	256	return 0;
ceffc078 CO	257	}
	258
	259	static struct vm_operations_struct xip_file_vm_ops = {
54cb8821	260	.fault = xip_file_fault,
ceffc078 CO	261	};
	262
	263	int xip_file_mmap(struct file * file, struct vm_area_struct * vma)
	264	{
	265	BUG_ON(!file->f_mapping->a_ops->get_xip_page);
	266
	267	file_accessed(file);
	268	vma->vm_ops = &xip_file_vm_ops;
54cb8821	269	vma->vm_flags \|= VM_CAN_NONLINEAR;
ceffc078 CO	270	return 0;
	271	}
	272	EXPORT_SYMBOL_GPL(xip_file_mmap);
	273
	274	static ssize_t
eb6fe0c3 CO	275	__xip_file_write(struct file filp, const char __user buf,
eb6fe0c3 CO	276	size_t count, loff_t pos, loff_t *ppos)
ceffc078	277	{
eb6fe0c3	278	struct address_space * mapping = filp->f_mapping;
f5e54d6e	279	const struct address_space_operations *a_ops = mapping->a_ops;
ceffc078 CO	280	struct inode *inode = mapping->host;
	281	long status = 0;
	282	struct page *page;
	283	size_t bytes;
ceffc078 CO	284	ssize_t written = 0;
	285
	286	BUG_ON(!mapping->a_ops->get_xip_page);
	287
ceffc078 CO	288	do {
	289	unsigned long index;
	290	unsigned long offset;
	291	size_t copied;
4a9e5ef1	292	char *kaddr;
ceffc078 CO	293
	294	offset = (pos & (PAGE_CACHE_SIZE -1)); /* Within page */
	295	index = pos >> PAGE_CACHE_SHIFT;
	296	bytes = PAGE_CACHE_SIZE - offset;
	297	if (bytes > count)
	298	bytes = count;
	299
ceffc078	300	page = a_ops->get_xip_page(mapping,
eb6fe0c3	301	index*(PAGE_SIZE/512), 0);
ceffc078 CO	302	if (IS_ERR(page) && (PTR_ERR(page) == -ENODATA)) {
	303	/* we allocate a new page unmap it */
	304	page = a_ops->get_xip_page(mapping,
eb6fe0c3	305	index*(PAGE_SIZE/512), 1);
ceffc078	306	if (!IS_ERR(page))
eb6fe0c3 CO	307	/* unmap page at pgoff from all other vmas */
eb6fe0c3 CO	308	__xip_unmap(mapping, index);
ceffc078 CO	309	}
	310
	311	if (IS_ERR(page)) {
	312	status = PTR_ERR(page);
	313	break;
	314	}
	315
4a9e5ef1 NP	316	fault_in_pages_readable(buf, bytes);
	317	kaddr = kmap_atomic(page, KM_USER0);
	318	copied = bytes -
369b8f5a	319	__copy_from_user_inatomic_nocache(kaddr + offset, buf, bytes);
4a9e5ef1	320	kunmap_atomic(kaddr, KM_USER0);
ceffc078	321	flush_dcache_page(page);
4a9e5ef1	322
ceffc078 CO	323	if (likely(copied > 0)) {
	324	status = copied;
	325
	326	if (status >= 0) {
	327	written += status;
	328	count -= status;
	329	pos += status;
	330	buf += status;
ceffc078 CO	331	}
	332	}
	333	if (unlikely(copied != bytes))
	334	if (status >= 0)
	335	status = -EFAULT;
	336	if (status < 0)
	337	break;
	338	} while (count);
	339	*ppos = pos;
	340	/*
	341	* No need to use i_size_read() here, the i_size
1b1dcc1b	342	* cannot change under us because we hold i_mutex.
ceffc078 CO	343	*/
	344	if (pos > inode->i_size) {
	345	i_size_write(inode, pos);
	346	mark_inode_dirty(inode);
	347	}
	348
	349	return written ? written : status;
	350	}
	351
eb6fe0c3 CO	352	ssize_t
	353	xip_file_write(struct file filp, const char __user buf, size_t len,
	354	loff_t *ppos)
ceffc078	355	{
eb6fe0c3 CO	356	struct address_space *mapping = filp->f_mapping;
	357	struct inode *inode = mapping->host;
	358	size_t count;
	359	loff_t pos;
	360	ssize_t ret;
ceffc078	361
1b1dcc1b	362	mutex_lock(&inode->i_mutex);
ceffc078	363
eb6fe0c3 CO	364	if (!access_ok(VERIFY_READ, buf, len)) {
	365	ret=-EFAULT;
	366	goto out_up;
ceffc078 CO	367	}
ceffc078 CO	368
ceffc078	369	pos = *ppos;
eb6fe0c3	370	count = len;
ceffc078 CO	371
	372	vfs_check_frozen(inode->i_sb, SB_FREEZE_WRITE);
	373
eb6fe0c3 CO	374	/* We can write back this queue in page reclaim */
eb6fe0c3 CO	375	current->backing_dev_info = mapping->backing_dev_info;
ceffc078	376
eb6fe0c3 CO	377	ret = generic_write_checks(filp, &pos, &count, S_ISBLK(inode->i_mode));
	378	if (ret)
	379	goto out_backing;
ceffc078	380	if (count == 0)
eb6fe0c3	381	goto out_backing;
ceffc078	382
d3ac7f89	383	ret = remove_suid(filp->f_path.dentry);
eb6fe0c3 CO	384	if (ret)
eb6fe0c3 CO	385	goto out_backing;
ceffc078	386
870f4817	387	file_update_time(filp);
ceffc078	388
eb6fe0c3	389	ret = __xip_file_write (filp, buf, count, pos, ppos);
ceffc078	390
eb6fe0c3 CO	391	out_backing:
	392	current->backing_dev_info = NULL;
	393	out_up:
1b1dcc1b	394	mutex_unlock(&inode->i_mutex);
ceffc078 CO	395	return ret;
ceffc078 CO	396	}
eb6fe0c3	397	EXPORT_SYMBOL_GPL(xip_file_write);
ceffc078 CO	398
	399	/*
	400	* truncate a page used for execute in place
	401	* functionality is analog to block_truncate_page but does use get_xip_page
	402	* to get the page instead of page cache
	403	*/
	404	int
	405	xip_truncate_page(struct address_space *mapping, loff_t from)
	406	{
	407	pgoff_t index = from >> PAGE_CACHE_SHIFT;
	408	unsigned offset = from & (PAGE_CACHE_SIZE-1);
	409	unsigned blocksize;
	410	unsigned length;
	411	struct page *page;
ceffc078 CO	412
	413	BUG_ON(!mapping->a_ops->get_xip_page);
	414
	415	blocksize = 1 << mapping->host->i_blkbits;
	416	length = offset & (blocksize - 1);
	417
	418	/* Block boundary? Nothing to do */
	419	if (!length)
	420	return 0;
	421
	422	length = blocksize - length;
	423
	424	page = mapping->a_ops->get_xip_page(mapping,
	425	index*(PAGE_SIZE/512), 0);
ceffc078	426	if (!page)
eb6fe0c3	427	return -ENOMEM;
ceffc078	428	if (unlikely(IS_ERR(page))) {
eb6fe0c3	429	if (PTR_ERR(page) == -ENODATA)
ceffc078 CO	430	/* Hole? No need to truncate */
ceffc078 CO	431	return 0;
eb6fe0c3 CO	432	else
eb6fe0c3 CO	433	return PTR_ERR(page);
afa597ba	434	}
eebd2aa3	435	zero_user(page, offset, length);
eb6fe0c3	436	return 0;
ceffc078 CO	437	}
ceffc078 CO	438	EXPORT_SYMBOL_GPL(xip_truncate_page);