[linux-2.6-block.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 <linux/seqlock.h>
#include <linux/mutex.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)
			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;
	unsigned count;
	int locked = 0;

	count = read_seqcount_begin(&xip_sparse_seq);

	page = __xip_sparse_page;
	if (!page)
		return;

retry:
	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);

	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 */
		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:
		printk("%s insert %lx@%lx\n", current->comm, (unsigned long)vmf->virtual_address, xip_pfn);
		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 {
		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 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>
538f8ea6 NP	18	#include <linux/seqlock.h>
538f8ea6 NP	19	#include <linux/mutex.h>
ceffc078	20	#include <asm/tlbflush.h>
70688e4d	21	#include <asm/io.h>
ceffc078	22
a76c0b97 CO	23	/*
	24	* We do use our own empty page to avoid interference with other users
	25	* of ZERO_PAGE(), such as /dev/zero
	26	*/
538f8ea6 NP	27	static DEFINE_MUTEX(xip_sparse_mutex);
538f8ea6 NP	28	static seqcount_t xip_sparse_seq = SEQCNT_ZERO;
a76c0b97 CO	29	static struct page *__xip_sparse_page;
a76c0b97 CO	30
538f8ea6	31	/* called under xip_sparse_mutex */
a76c0b97 CO	32	static struct page *xip_sparse_page(void)
	33	{
	34	if (!__xip_sparse_page) {
c51b1a16 AM	35	struct page *page = alloc_page(GFP_HIGHUSER \| __GFP_ZERO);
c51b1a16 AM	36
538f8ea6 NP	37	if (page)
538f8ea6 NP	38	__xip_sparse_page = page;
a76c0b97 CO	39	}
	40	return __xip_sparse_page;
	41	}
	42
ceffc078 CO	43	/*
ceffc078 CO	44	* This is a file read routine for execute in place files, and uses
70688e4d	45	* the mapping->a_ops->get_xip_mem() function for the actual low-level
ceffc078 CO	46	* stuff.
	47	*
	48	* Note the struct file* is not used at all. It may be NULL.
	49	*/
70688e4d	50	static ssize_t
ceffc078 CO	51	do_xip_mapping_read(struct address_space *mapping,
	52	struct file_ra_state *_ra,
	53	struct file *filp,
70688e4d NP	54	char __user *buf,
	55	size_t len,
	56	loff_t *ppos)
ceffc078 CO	57	{
ceffc078 CO	58	struct inode *inode = mapping->host;
2004dc8e JK	59	pgoff_t index, end_index;
2004dc8e JK	60	unsigned long offset;
70688e4d NP	61	loff_t isize, pos;
70688e4d NP	62	size_t copied = 0, error = 0;
ceffc078	63
70688e4d	64	BUG_ON(!mapping->a_ops->get_xip_mem);
ceffc078	65
70688e4d NP	66	pos = *ppos;
	67	index = pos >> PAGE_CACHE_SHIFT;
	68	offset = pos & ~PAGE_CACHE_MASK;
ceffc078 CO	69
	70	isize = i_size_read(inode);
	71	if (!isize)
	72	goto out;
	73
	74	end_index = (isize - 1) >> PAGE_CACHE_SHIFT;
70688e4d NP	75	do {
	76	unsigned long nr, left;
	77	void *xip_mem;
	78	unsigned long xip_pfn;
	79	int zero = 0;
ceffc078 CO	80
	81	/* nr is the maximum number of bytes to copy from this page */
	82	nr = PAGE_CACHE_SIZE;
	83	if (index >= end_index) {
	84	if (index > end_index)
	85	goto out;
	86	nr = ((isize - 1) & ~PAGE_CACHE_MASK) + 1;
	87	if (nr <= offset) {
	88	goto out;
	89	}
	90	}
	91	nr = nr - offset;
70688e4d NP	92	if (nr > len)
70688e4d NP	93	nr = len;
ceffc078	94
70688e4d NP	95	error = mapping->a_ops->get_xip_mem(mapping, index, 0,
	96	&xip_mem, &xip_pfn);
	97	if (unlikely(error)) {
	98	if (error == -ENODATA) {
ceffc078	99	/* sparse */
70688e4d NP	100	zero = 1;
70688e4d NP	101	} else
ceffc078	102	goto out;
afa597ba	103	}
ceffc078 CO	104
	105	/* If users can be writing to this page using arbitrary
	106	* virtual addresses, take care about potential aliasing
	107	* before reading the page on the kernel side.
	108	*/
	109	if (mapping_writably_mapped(mapping))
70688e4d	110	/* address based flush */ ;
ceffc078 CO	111
ceffc078 CO	112	/*
70688e4d	113	* Ok, we have the mem, so now we can copy it to user space...
ceffc078 CO	114	*
	115	* The actor routine returns how many bytes were actually used..
	116	* NOTE! This may not be the same as how much of a user buffer
	117	* we filled up (we may be padding etc), so we can only update
	118	* "pos" here (the actor routine has to update the user buffer
	119	* pointers and the remaining count).
	120	*/
70688e4d NP	121	if (!zero)
	122	left = __copy_to_user(buf+copied, xip_mem+offset, nr);
	123	else
	124	left = __clear_user(buf + copied, nr);
ceffc078	125
70688e4d NP	126	if (left) {
	127	error = -EFAULT;
	128	goto out;
	129	}
ceffc078	130
70688e4d NP	131	copied += (nr - left);
	132	offset += (nr - left);
	133	index += offset >> PAGE_CACHE_SHIFT;
	134	offset &= ~PAGE_CACHE_MASK;
	135	} while (copied < len);
ceffc078 CO	136
ceffc078 CO	137	out:
70688e4d	138	*ppos = pos + copied;
ceffc078 CO	139	if (filp)
ceffc078 CO	140	file_accessed(filp);
70688e4d NP	141
70688e4d NP	142	return (copied ? copied : error);
ceffc078 CO	143	}
ceffc078 CO	144
ceffc078	145	ssize_t
eb6fe0c3	146	xip_file_read(struct file filp, char __user buf, size_t len, loff_t *ppos)
ceffc078	147	{
eb6fe0c3 CO	148	if (!access_ok(VERIFY_WRITE, buf, len))
eb6fe0c3 CO	149	return -EFAULT;
ceffc078	150
70688e4d NP	151	return do_xip_mapping_read(filp->f_mapping, &filp->f_ra, filp,
70688e4d NP	152	buf, len, ppos);
ceffc078	153	}
eb6fe0c3	154	EXPORT_SYMBOL_GPL(xip_file_read);
ceffc078	155
ceffc078 CO	156	/*
	157	* __xip_unmap is invoked from xip_unmap and
	158	* xip_write
	159	*
	160	* This function walks all vmas of the address_space and unmaps the
a76c0b97	161	* __xip_sparse_page when found at pgoff.
ceffc078 CO	162	*/
	163	static void
	164	__xip_unmap (struct address_space * mapping,
	165	unsigned long pgoff)
	166	{
	167	struct vm_area_struct *vma;
	168	struct mm_struct *mm;
	169	struct prio_tree_iter iter;
	170	unsigned long address;
	171	pte_t *pte;
	172	pte_t pteval;
c0718806	173	spinlock_t *ptl;
67b02f11	174	struct page *page;
538f8ea6 NP	175	unsigned count;
	176	int locked = 0;
	177
	178	count = read_seqcount_begin(&xip_sparse_seq);
ceffc078	179
a76c0b97 CO	180	page = __xip_sparse_page;
	181	if (!page)
	182	return;
	183
538f8ea6	184	retry:
ceffc078 CO	185	spin_lock(&mapping->i_mmap_lock);
	186	vma_prio_tree_foreach(vma, &iter, &mapping->i_mmap, pgoff, pgoff) {
	187	mm = vma->vm_mm;
	188	address = vma->vm_start +
	189	((pgoff - vma->vm_pgoff) << PAGE_SHIFT);
	190	BUG_ON(address < vma->vm_start \|\| address >= vma->vm_end);
479db0bf	191	pte = page_check_address(page, mm, address, &ptl, 1);
c0718806	192	if (pte) {
ceffc078	193	/* Nuke the page table entry. */
082ff0a9	194	flush_cache_page(vma, address, pte_pfn(*pte));
cddb8a5c	195	pteval = ptep_clear_flush_notify(vma, address, pte);
7de6b805	196	page_remove_rmap(page, vma);
b5810039	197	dec_mm_counter(mm, file_rss);
ceffc078	198	BUG_ON(pte_dirty(pteval));
c0718806	199	pte_unmap_unlock(pte, ptl);
b5810039	200	page_cache_release(page);
ceffc078 CO	201	}
	202	}
	203	spin_unlock(&mapping->i_mmap_lock);
538f8ea6 NP	204
	205	if (locked) {
	206	mutex_unlock(&xip_sparse_mutex);
	207	} else if (read_seqcount_retry(&xip_sparse_seq, count)) {
	208	mutex_lock(&xip_sparse_mutex);
	209	locked = 1;
	210	goto retry;
	211	}
ceffc078 CO	212	}
	213
	214	/*
54cb8821	215	* xip_fault() is invoked via the vma operations vector for a
ceffc078 CO	216	* mapped memory region to read in file data during a page fault.
ceffc078 CO	217	*
54cb8821	218	* This function is derived from filemap_fault, but used for execute in place
ceffc078	219	*/
70688e4d	220	static int xip_file_fault(struct vm_area_struct vma, struct vm_fault vmf)
ceffc078	221	{
70688e4d	222	struct file *file = vma->vm_file;
ceffc078 CO	223	struct address_space *mapping = file->f_mapping;
ceffc078 CO	224	struct inode *inode = mapping->host;
54cb8821	225	pgoff_t size;
70688e4d NP	226	void *xip_mem;
	227	unsigned long xip_pfn;
	228	struct page *page;
	229	int error;
ceffc078	230
54cb8821	231	/* XXX: are VM_FAULT_ codes OK? */
538f8ea6	232	again:
ceffc078	233	size = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
d0217ac0 NP	234	if (vmf->pgoff >= size)
d0217ac0 NP	235	return VM_FAULT_SIGBUS;
ceffc078	236
70688e4d NP	237	error = mapping->a_ops->get_xip_mem(mapping, vmf->pgoff, 0,
	238	&xip_mem, &xip_pfn);
	239	if (likely(!error))
	240	goto found;
	241	if (error != -ENODATA)
d0217ac0	242	return VM_FAULT_OOM;
ceffc078 CO	243
ceffc078 CO	244	/* sparse block */
70688e4d NP	245	if ((vma->vm_flags & (VM_WRITE \| VM_MAYWRITE)) &&
70688e4d NP	246	(vma->vm_flags & (VM_SHARED \| VM_MAYSHARE)) &&
ceffc078	247	(!(mapping->host->i_sb->s_flags & MS_RDONLY))) {
70688e4d NP	248	int err;
70688e4d NP	249
ceffc078	250	/* maybe shared writable, allocate new block */
70688e4d NP	251	error = mapping->a_ops->get_xip_mem(mapping, vmf->pgoff, 1,
	252	&xip_mem, &xip_pfn);
	253	if (error)
d0217ac0	254	return VM_FAULT_SIGBUS;
70688e4d	255	/* unmap sparse mappings at pgoff from all other vmas */
d0217ac0	256	__xip_unmap(mapping, vmf->pgoff);
70688e4d NP	257
70688e4d NP	258	found:
538f8ea6	259	printk("%s insert %lx@%lx\n", current->comm, (unsigned long)vmf->virtual_address, xip_pfn);
70688e4d NP	260	err = vm_insert_mixed(vma, (unsigned long)vmf->virtual_address,
	261	xip_pfn);
	262	if (err == -ENOMEM)
	263	return VM_FAULT_OOM;
	264	BUG_ON(err);
	265	return VM_FAULT_NOPAGE;
ceffc078	266	} else {
538f8ea6 NP	267	int err, ret = VM_FAULT_OOM;
	268
	269	mutex_lock(&xip_sparse_mutex);
	270	write_seqcount_begin(&xip_sparse_seq);
	271	error = mapping->a_ops->get_xip_mem(mapping, vmf->pgoff, 0,
	272	&xip_mem, &xip_pfn);
	273	if (unlikely(!error)) {
	274	write_seqcount_end(&xip_sparse_seq);
	275	mutex_unlock(&xip_sparse_mutex);
	276	goto again;
	277	}
	278	if (error != -ENODATA)
	279	goto out;
a76c0b97 CO	280	/* not shared and writable, use xip_sparse_page() */
a76c0b97 CO	281	page = xip_sparse_page();
d0217ac0	282	if (!page)
538f8ea6 NP	283	goto out;
	284	err = vm_insert_page(vma, (unsigned long)vmf->virtual_address,
	285	page);
	286	if (err == -ENOMEM)
	287	goto out;
	288
	289	ret = VM_FAULT_NOPAGE;
	290	out:
	291	write_seqcount_end(&xip_sparse_seq);
	292	mutex_unlock(&xip_sparse_mutex);
ceffc078	293
538f8ea6	294	return ret;
70688e4d	295	}
ceffc078 CO	296	}
	297
	298	static struct vm_operations_struct xip_file_vm_ops = {
54cb8821	299	.fault = xip_file_fault,
ceffc078 CO	300	};
	301
	302	int xip_file_mmap(struct file * file, struct vm_area_struct * vma)
	303	{
70688e4d	304	BUG_ON(!file->f_mapping->a_ops->get_xip_mem);
ceffc078 CO	305
	306	file_accessed(file);
	307	vma->vm_ops = &xip_file_vm_ops;
70688e4d	308	vma->vm_flags \|= VM_CAN_NONLINEAR \| VM_MIXEDMAP;
ceffc078 CO	309	return 0;
	310	}
	311	EXPORT_SYMBOL_GPL(xip_file_mmap);
	312
	313	static ssize_t
eb6fe0c3 CO	314	__xip_file_write(struct file filp, const char __user buf,
eb6fe0c3 CO	315	size_t count, loff_t pos, loff_t *ppos)
ceffc078	316	{
eb6fe0c3	317	struct address_space * mapping = filp->f_mapping;
f5e54d6e	318	const struct address_space_operations *a_ops = mapping->a_ops;
ceffc078 CO	319	struct inode *inode = mapping->host;
ceffc078 CO	320	long status = 0;
ceffc078	321	size_t bytes;
ceffc078 CO	322	ssize_t written = 0;
ceffc078 CO	323
70688e4d	324	BUG_ON(!mapping->a_ops->get_xip_mem);
ceffc078	325
ceffc078 CO	326	do {
	327	unsigned long index;
	328	unsigned long offset;
	329	size_t copied;
70688e4d NP	330	void *xip_mem;
70688e4d NP	331	unsigned long xip_pfn;
ceffc078 CO	332
	333	offset = (pos & (PAGE_CACHE_SIZE -1)); /* Within page */
	334	index = pos >> PAGE_CACHE_SHIFT;
	335	bytes = PAGE_CACHE_SIZE - offset;
	336	if (bytes > count)
	337	bytes = count;
	338
70688e4d NP	339	status = a_ops->get_xip_mem(mapping, index, 0,
	340	&xip_mem, &xip_pfn);
	341	if (status == -ENODATA) {
ceffc078	342	/* we allocate a new page unmap it */
70688e4d NP	343	status = a_ops->get_xip_mem(mapping, index, 1,
	344	&xip_mem, &xip_pfn);
	345	if (!status)
eb6fe0c3 CO	346	/* unmap page at pgoff from all other vmas */
eb6fe0c3 CO	347	__xip_unmap(mapping, index);
ceffc078 CO	348	}
ceffc078 CO	349
70688e4d	350	if (status)
ceffc078	351	break;
ceffc078	352
4a9e5ef1	353	copied = bytes -
70688e4d	354	__copy_from_user_nocache(xip_mem + offset, buf, bytes);
4a9e5ef1	355
ceffc078 CO	356	if (likely(copied > 0)) {
	357	status = copied;
	358
	359	if (status >= 0) {
	360	written += status;
	361	count -= status;
	362	pos += status;
	363	buf += status;
ceffc078 CO	364	}
	365	}
	366	if (unlikely(copied != bytes))
	367	if (status >= 0)
	368	status = -EFAULT;
	369	if (status < 0)
	370	break;
	371	} while (count);
	372	*ppos = pos;
	373	/*
	374	* No need to use i_size_read() here, the i_size
1b1dcc1b	375	* cannot change under us because we hold i_mutex.
ceffc078 CO	376	*/
	377	if (pos > inode->i_size) {
	378	i_size_write(inode, pos);
	379	mark_inode_dirty(inode);
	380	}
	381
	382	return written ? written : status;
	383	}
	384
eb6fe0c3 CO	385	ssize_t
	386	xip_file_write(struct file filp, const char __user buf, size_t len,
	387	loff_t *ppos)
ceffc078	388	{
eb6fe0c3 CO	389	struct address_space *mapping = filp->f_mapping;
	390	struct inode *inode = mapping->host;
	391	size_t count;
	392	loff_t pos;
	393	ssize_t ret;
ceffc078	394
1b1dcc1b	395	mutex_lock(&inode->i_mutex);
ceffc078	396
eb6fe0c3 CO	397	if (!access_ok(VERIFY_READ, buf, len)) {
	398	ret=-EFAULT;
	399	goto out_up;
ceffc078 CO	400	}
ceffc078 CO	401
ceffc078	402	pos = *ppos;
eb6fe0c3	403	count = len;
ceffc078 CO	404
	405	vfs_check_frozen(inode->i_sb, SB_FREEZE_WRITE);
	406
eb6fe0c3 CO	407	/* We can write back this queue in page reclaim */
eb6fe0c3 CO	408	current->backing_dev_info = mapping->backing_dev_info;
ceffc078	409
eb6fe0c3 CO	410	ret = generic_write_checks(filp, &pos, &count, S_ISBLK(inode->i_mode));
	411	if (ret)
	412	goto out_backing;
ceffc078	413	if (count == 0)
eb6fe0c3	414	goto out_backing;
ceffc078	415
2f1936b8	416	ret = file_remove_suid(filp);
eb6fe0c3 CO	417	if (ret)
eb6fe0c3 CO	418	goto out_backing;
ceffc078	419
870f4817	420	file_update_time(filp);
ceffc078	421
eb6fe0c3	422	ret = __xip_file_write (filp, buf, count, pos, ppos);
ceffc078	423
eb6fe0c3 CO	424	out_backing:
	425	current->backing_dev_info = NULL;
	426	out_up:
1b1dcc1b	427	mutex_unlock(&inode->i_mutex);
ceffc078 CO	428	return ret;
ceffc078 CO	429	}
eb6fe0c3	430	EXPORT_SYMBOL_GPL(xip_file_write);
ceffc078 CO	431
	432	/*
	433	* truncate a page used for execute in place
70688e4d	434	* functionality is analog to block_truncate_page but does use get_xip_mem
ceffc078 CO	435	* to get the page instead of page cache
	436	*/
	437	int
	438	xip_truncate_page(struct address_space *mapping, loff_t from)
	439	{
	440	pgoff_t index = from >> PAGE_CACHE_SHIFT;
	441	unsigned offset = from & (PAGE_CACHE_SIZE-1);
	442	unsigned blocksize;
	443	unsigned length;
70688e4d NP	444	void *xip_mem;
	445	unsigned long xip_pfn;
	446	int err;
ceffc078	447
70688e4d	448	BUG_ON(!mapping->a_ops->get_xip_mem);
ceffc078 CO	449
	450	blocksize = 1 << mapping->host->i_blkbits;
	451	length = offset & (blocksize - 1);
	452
	453	/* Block boundary? Nothing to do */
	454	if (!length)
	455	return 0;
	456
	457	length = blocksize - length;
	458
70688e4d NP	459	err = mapping->a_ops->get_xip_mem(mapping, index, 0,
	460	&xip_mem, &xip_pfn);
	461	if (unlikely(err)) {
	462	if (err == -ENODATA)
ceffc078 CO	463	/* Hole? No need to truncate */
ceffc078 CO	464	return 0;
eb6fe0c3	465	else
70688e4d	466	return err;
afa597ba	467	}
70688e4d	468	memset(xip_mem + offset, 0, length);
eb6fe0c3	469	return 0;
ceffc078 CO	470	}
ceffc078 CO	471	EXPORT_SYMBOL_GPL(xip_truncate_page);