[linux-block.git] / mm / process_vm_access.c

// SPDX-License-Identifier: GPL-2.0-or-later
/*
 * linux/mm/process_vm_access.c
 *
 * Copyright (C) 2010-2011 Christopher Yeoh <cyeoh@au1.ibm.com>, IBM Corp.
 */

#include <linux/compat.h>
#include <linux/mm.h>
#include <linux/uio.h>
#include <linux/sched.h>
#include <linux/sched/mm.h>
#include <linux/highmem.h>
#include <linux/ptrace.h>
#include <linux/slab.h>
#include <linux/syscalls.h>

/**
 * process_vm_rw_pages - read/write pages from task specified
 * @pages: array of pointers to pages we want to copy
 * @offset: offset in page to start copying from/to
 * @len: number of bytes to copy
 * @iter: where to copy to/from locally
 * @vm_write: 0 means copy from, 1 means copy to
 * Returns 0 on success, error code otherwise
 */
static int process_vm_rw_pages(struct page **pages,
			       unsigned offset,
			       size_t len,
			       struct iov_iter *iter,
			       int vm_write)
{
	/* Do the copy for each page */
	while (len && iov_iter_count(iter)) {
		struct page *page = *pages++;
		size_t copy = PAGE_SIZE - offset;
		size_t copied;

		if (copy > len)
			copy = len;

		if (vm_write)
			copied = copy_page_from_iter(page, offset, copy, iter);
		else
			copied = copy_page_to_iter(page, offset, copy, iter);

		len -= copied;
		if (copied < copy && iov_iter_count(iter))
			return -EFAULT;
		offset = 0;
	}
	return 0;
}

/* Maximum number of pages kmalloc'd to hold struct page's during copy */
#define PVM_MAX_KMALLOC_PAGES 2

/* Maximum number of pages that can be stored at a time */
#define PVM_MAX_USER_PAGES (PVM_MAX_KMALLOC_PAGES * PAGE_SIZE / sizeof(struct page *))

/**
 * process_vm_rw_single_vec - read/write pages from task specified
 * @addr: start memory address of target process
 * @len: size of area to copy to/from
 * @iter: where to copy to/from locally
 * @process_pages: struct pages area that can store at least
 *  nr_pages_to_copy struct page pointers
 * @mm: mm for task
 * @task: task to read/write from
 * @vm_write: 0 means copy from, 1 means copy to
 * Returns 0 on success or on failure error code
 */
static int process_vm_rw_single_vec(unsigned long addr,
				    unsigned long len,
				    struct iov_iter *iter,
				    struct page **process_pages,
				    struct mm_struct *mm,
				    struct task_struct *task,
				    int vm_write)
{
	unsigned long pa = addr & PAGE_MASK;
	unsigned long start_offset = addr - pa;
	unsigned long nr_pages;
	ssize_t rc = 0;
	unsigned int flags = 0;

	/* Work out address and page range required */
	if (len == 0)
		return 0;
	nr_pages = (addr + len - 1) / PAGE_SIZE - addr / PAGE_SIZE + 1;

	if (vm_write)
		flags |= FOLL_WRITE;

	while (!rc && nr_pages && iov_iter_count(iter)) {
		int pinned_pages = min_t(unsigned long, nr_pages, PVM_MAX_USER_PAGES);
		int locked = 1;
		size_t bytes;

		/*
		 * Get the pages we're interested in.  We must
		 * access remotely because task/mm might not
		 * current/current->mm
		 */
		mmap_read_lock(mm);
		pinned_pages = pin_user_pages_remote(mm, pa, pinned_pages,
						     flags, process_pages,
						     &locked);
		if (locked)
			mmap_read_unlock(mm);
		if (pinned_pages <= 0)
			return -EFAULT;

		bytes = pinned_pages * PAGE_SIZE - start_offset;
		if (bytes > len)
			bytes = len;

		rc = process_vm_rw_pages(process_pages,
					 start_offset, bytes, iter,
					 vm_write);
		len -= bytes;
		start_offset = 0;
		nr_pages -= pinned_pages;
		pa += pinned_pages * PAGE_SIZE;

		/* If vm_write is set, the pages need to be made dirty: */
		unpin_user_pages_dirty_lock(process_pages, pinned_pages,
					    vm_write);
	}

	return rc;
}

/* Maximum number of entries for process pages array
   which lives on stack */
#define PVM_MAX_PP_ARRAY_COUNT 16

/**
 * process_vm_rw_core - core of reading/writing pages from task specified
 * @pid: PID of process to read/write from/to
 * @iter: where to copy to/from locally
 * @rvec: iovec array specifying where to copy to/from in the other process
 * @riovcnt: size of rvec array
 * @flags: currently unused
 * @vm_write: 0 if reading from other process, 1 if writing to other process
 *
 * Returns the number of bytes read/written or error code. May
 *  return less bytes than expected if an error occurs during the copying
 *  process.
 */
static ssize_t process_vm_rw_core(pid_t pid, struct iov_iter *iter,
				  const struct iovec *rvec,
				  unsigned long riovcnt,
				  unsigned long flags, int vm_write)
{
	struct task_struct *task;
	struct page *pp_stack[PVM_MAX_PP_ARRAY_COUNT];
	struct page **process_pages = pp_stack;
	struct mm_struct *mm;
	unsigned long i;
	ssize_t rc = 0;
	unsigned long nr_pages = 0;
	unsigned long nr_pages_iov;
	ssize_t iov_len;
	size_t total_len = iov_iter_count(iter);

	/*
	 * Work out how many pages of struct pages we're going to need
	 * when eventually calling get_user_pages
	 */
	for (i = 0; i < riovcnt; i++) {
		iov_len = rvec[i].iov_len;
		if (iov_len > 0) {
			nr_pages_iov = ((unsigned long)rvec[i].iov_base
					+ iov_len - 1)
				/ PAGE_SIZE - (unsigned long)rvec[i].iov_base
				/ PAGE_SIZE + 1;
			nr_pages = max(nr_pages, nr_pages_iov);
		}
	}

	if (nr_pages == 0)
		return 0;

	if (nr_pages > PVM_MAX_PP_ARRAY_COUNT) {
		/* For reliability don't try to kmalloc more than
		   2 pages worth */
		process_pages = kmalloc(min_t(size_t, PVM_MAX_KMALLOC_PAGES * PAGE_SIZE,
					      sizeof(struct page *)*nr_pages),
					GFP_KERNEL);

		if (!process_pages)
			return -ENOMEM;
	}

	/* Get process information */
	task = find_get_task_by_vpid(pid);
	if (!task) {
		rc = -ESRCH;
		goto free_proc_pages;
	}

	mm = mm_access(task, PTRACE_MODE_ATTACH_REALCREDS);
	if (!mm || IS_ERR(mm)) {
		rc = IS_ERR(mm) ? PTR_ERR(mm) : -ESRCH;
		/*
		 * Explicitly map EACCES to EPERM as EPERM is a more
		 * appropriate error code for process_vw_readv/writev
		 */
		if (rc == -EACCES)
			rc = -EPERM;
		goto put_task_struct;
	}

	for (i = 0; i < riovcnt && iov_iter_count(iter) && !rc; i++)
		rc = process_vm_rw_single_vec(
			(unsigned long)rvec[i].iov_base, rvec[i].iov_len,
			iter, process_pages, mm, task, vm_write);

	/* copied = space before - space after */
	total_len -= iov_iter_count(iter);

	/* If we have managed to copy any data at all then
	   we return the number of bytes copied. Otherwise
	   we return the error code */
	if (total_len)
		rc = total_len;

	mmput(mm);

put_task_struct:
	put_task_struct(task);

free_proc_pages:
	if (process_pages != pp_stack)
		kfree(process_pages);
	return rc;
}

/**
 * process_vm_rw - check iovecs before calling core routine
 * @pid: PID of process to read/write from/to
 * @lvec: iovec array specifying where to copy to/from locally
 * @liovcnt: size of lvec array
 * @rvec: iovec array specifying where to copy to/from in the other process
 * @riovcnt: size of rvec array
 * @flags: currently unused
 * @vm_write: 0 if reading from other process, 1 if writing to other process
 *
 * Returns the number of bytes read/written or error code. May
 *  return less bytes than expected if an error occurs during the copying
 *  process.
 */
static ssize_t process_vm_rw(pid_t pid,
			     const struct iovec __user *lvec,
			     unsigned long liovcnt,
			     const struct iovec __user *rvec,
			     unsigned long riovcnt,
			     unsigned long flags, int vm_write)
{
	struct iovec iovstack_l[UIO_FASTIOV];
	struct iovec iovstack_r[UIO_FASTIOV];
	struct iovec *iov_l = iovstack_l;
	struct iovec *iov_r;
	struct iov_iter iter;
	ssize_t rc;
	int dir = vm_write ? ITER_SOURCE : ITER_DEST;

	if (flags != 0)
		return -EINVAL;

	/* Check iovecs */
	rc = import_iovec(dir, lvec, liovcnt, UIO_FASTIOV, &iov_l, &iter);
	if (rc < 0)
		return rc;
	if (!iov_iter_count(&iter))
		goto free_iov_l;
	iov_r = iovec_from_user(rvec, riovcnt, UIO_FASTIOV, iovstack_r,
				in_compat_syscall());
	if (IS_ERR(iov_r)) {
		rc = PTR_ERR(iov_r);
		goto free_iov_l;
	}
	rc = process_vm_rw_core(pid, &iter, iov_r, riovcnt, flags, vm_write);
	if (iov_r != iovstack_r)
		kfree(iov_r);
free_iov_l:
	kfree(iov_l);
	return rc;
}

SYSCALL_DEFINE6(process_vm_readv, pid_t, pid, const struct iovec __user *, lvec,
		unsigned long, liovcnt, const struct iovec __user *, rvec,
		unsigned long, riovcnt,	unsigned long, flags)
{
	return process_vm_rw(pid, lvec, liovcnt, rvec, riovcnt, flags, 0);
}

SYSCALL_DEFINE6(process_vm_writev, pid_t, pid,
		const struct iovec __user *, lvec,
		unsigned long, liovcnt, const struct iovec __user *, rvec,
		unsigned long, riovcnt,	unsigned long, flags)
{
	return process_vm_rw(pid, lvec, liovcnt, rvec, riovcnt, flags, 1);
}
Commit	Line	Data
2874c5fd	1	// SPDX-License-Identifier: GPL-2.0-or-later
fcf63409 CY	2	/*
	3	* linux/mm/process_vm_access.c
	4	*
	5	* Copyright (C) 2010-2011 Christopher Yeoh <cyeoh@au1.ibm.com>, IBM Corp.
fcf63409 CY	6	*/
fcf63409 CY	7
f78f63da	8	#include <linux/compat.h>
fcf63409 CY	9	#include <linux/mm.h>
	10	#include <linux/uio.h>
	11	#include <linux/sched.h>
6e84f315	12	#include <linux/sched/mm.h>
fcf63409 CY	13	#include <linux/highmem.h>
	14	#include <linux/ptrace.h>
	15	#include <linux/slab.h>
	16	#include <linux/syscalls.h>
	17
fcf63409 CY	18	/**
fcf63409 CY	19	* process_vm_rw_pages - read/write pages from task specified
4bafbec7	20	* @pages: array of pointers to pages we want to copy
f144c390	21	* @offset: offset in page to start copying from/to
fcf63409	22	* @len: number of bytes to copy
4bafbec7	23	* @iter: where to copy to/from locally
fcf63409	24	* @vm_write: 0 means copy from, 1 means copy to
fcf63409 CY	25	* Returns 0 on success, error code otherwise
fcf63409 CY	26	*/
70eca12d AV	27	static int process_vm_rw_pages(struct page **pages,
70eca12d AV	28	unsigned offset,
e21345f9	29	size_t len,
9f78bdfa	30	struct iov_iter *iter,
9acc1a0f	31	int vm_write)
fcf63409	32	{
fcf63409	33	/* Do the copy for each page */
9acc1a0f	34	while (len && iov_iter_count(iter)) {
70eca12d	35	struct page page = pages++;
e21345f9	36	size_t copy = PAGE_SIZE - offset;
70eca12d	37	size_t copied;
240f3905	38
e21345f9 AV	39	if (copy > len)
	40	copy = len;
	41
803e4572	42	if (vm_write)
e7c24607	43	copied = copy_page_from_iter(page, offset, copy, iter);
803e4572	44	else
70eca12d	45	copied = copy_page_to_iter(page, offset, copy, iter);
803e4572	46
70eca12d AV	47	len -= copied;
	48	if (copied < copy && iov_iter_count(iter))
	49	return -EFAULT;
	50	offset = 0;
fcf63409	51	}
70eca12d	52	return 0;
fcf63409 CY	53	}
	54
	55	/* Maximum number of pages kmalloc'd to hold struct page's during copy */
932b59e3 YJN	56	#define PVM_MAX_KMALLOC_PAGES 2
	57
	58	/* Maximum number of pages that can be stored at a time */
	59	#define PVM_MAX_USER_PAGES (PVM_MAX_KMALLOC_PAGES * PAGE_SIZE / sizeof(struct page *))
fcf63409 CY	60
	61	/**
	62	* process_vm_rw_single_vec - read/write pages from task specified
	63	* @addr: start memory address of target process
	64	* @len: size of area to copy to/from
4bafbec7	65	* @iter: where to copy to/from locally
fcf63409 CY	66	* @process_pages: struct pages area that can store at least
	67	* nr_pages_to_copy struct page pointers
	68	* @mm: mm for task
	69	* @task: task to read/write from
	70	* @vm_write: 0 means copy from, 1 means copy to
fcf63409 CY	71	* Returns 0 on success or on failure error code
	72	*/
	73	static int process_vm_rw_single_vec(unsigned long addr,
	74	unsigned long len,
9f78bdfa	75	struct iov_iter *iter,
fcf63409 CY	76	struct page **process_pages,
	77	struct mm_struct *mm,
	78	struct task_struct *task,
9acc1a0f	79	int vm_write)
fcf63409 CY	80	{
	81	unsigned long pa = addr & PAGE_MASK;
	82	unsigned long start_offset = addr - pa;
	83	unsigned long nr_pages;
fcf63409	84	ssize_t rc = 0;
8b7457ef	85	unsigned int flags = 0;
fcf63409	86
fcf63409 CY	87	/* Work out address and page range required */
	88	if (len == 0)
	89	return 0;
	90	nr_pages = (addr + len - 1) / PAGE_SIZE - addr / PAGE_SIZE + 1;
	91
d4944b0e LS	92	if (vm_write)
	93	flags \|= FOLL_WRITE;
	94
4bafbec7	95	while (!rc && nr_pages && iov_iter_count(iter)) {
932b59e3	96	int pinned_pages = min_t(unsigned long, nr_pages, PVM_MAX_USER_PAGES);
8b7457ef	97	int locked = 1;
4bafbec7	98	size_t bytes;
fcf63409	99
1e987790 DH	100	/*
1e987790 DH	101	* Get the pages we're interested in. We must
8b7457ef	102	* access remotely because task/mm might not
1e987790 DH	103	* current/current->mm
1e987790 DH	104	*/
d8ed45c5	105	mmap_read_lock(mm);
64019a2e	106	pinned_pages = pin_user_pages_remote(mm, pa, pinned_pages,
803e4572	107	flags, process_pages,
0b295316	108	&locked);
8b7457ef	109	if (locked)
d8ed45c5	110	mmap_read_unlock(mm);
803e4572	111	if (pinned_pages <= 0)
70eca12d AV	112	return -EFAULT;
70eca12d AV	113
803e4572	114	bytes = pinned_pages * PAGE_SIZE - start_offset;
4bafbec7 AV	115	if (bytes > len)
4bafbec7 AV	116	bytes = len;
e21345f9	117
70eca12d	118	rc = process_vm_rw_pages(process_pages,
4bafbec7	119	start_offset, bytes, iter,
9acc1a0f	120	vm_write);
4bafbec7	121	len -= bytes;
fcf63409	122	start_offset = 0;
803e4572 JH	123	nr_pages -= pinned_pages;
	124	pa += pinned_pages * PAGE_SIZE;
	125
	126	/* If vm_write is set, the pages need to be made dirty: */
f1f6a7dd JH	127	unpin_user_pages_dirty_lock(process_pages, pinned_pages,
f1f6a7dd JH	128	vm_write);
fcf63409 CY	129	}
	130
	131	return rc;
	132	}
	133
	134	/* Maximum number of entries for process pages array
	135	which lives on stack */
	136	#define PVM_MAX_PP_ARRAY_COUNT 16
	137
	138	/**
	139	* process_vm_rw_core - core of reading/writing pages from task specified
	140	* @pid: PID of process to read/write from/to
4bafbec7	141	* @iter: where to copy to/from locally
fcf63409 CY	142	* @rvec: iovec array specifying where to copy to/from in the other process
	143	* @riovcnt: size of rvec array
	144	* @flags: currently unused
	145	* @vm_write: 0 if reading from other process, 1 if writing to other process
a5d09bed	146	*
fcf63409 CY	147	* Returns the number of bytes read/written or error code. May
	148	* return less bytes than expected if an error occurs during the copying
	149	* process.
	150	*/
9f78bdfa	151	static ssize_t process_vm_rw_core(pid_t pid, struct iov_iter *iter,
fcf63409 CY	152	const struct iovec *rvec,
	153	unsigned long riovcnt,
	154	unsigned long flags, int vm_write)
	155	{
	156	struct task_struct *task;
	157	struct page *pp_stack[PVM_MAX_PP_ARRAY_COUNT];
	158	struct page **process_pages = pp_stack;
	159	struct mm_struct *mm;
	160	unsigned long i;
	161	ssize_t rc = 0;
fcf63409 CY	162	unsigned long nr_pages = 0;
fcf63409 CY	163	unsigned long nr_pages_iov;
fcf63409	164	ssize_t iov_len;
9acc1a0f	165	size_t total_len = iov_iter_count(iter);
fcf63409 CY	166
	167	/*
	168	* Work out how many pages of struct pages we're going to need
	169	* when eventually calling get_user_pages
	170	*/
	171	for (i = 0; i < riovcnt; i++) {
	172	iov_len = rvec[i].iov_len;
	173	if (iov_len > 0) {
	174	nr_pages_iov = ((unsigned long)rvec[i].iov_base
932b59e3	175	+ iov_len - 1)
fcf63409 CY	176	/ PAGE_SIZE - (unsigned long)rvec[i].iov_base
	177	/ PAGE_SIZE + 1;
	178	nr_pages = max(nr_pages, nr_pages_iov);
	179	}
	180	}
	181
	182	if (nr_pages == 0)
	183	return 0;
	184
	185	if (nr_pages > PVM_MAX_PP_ARRAY_COUNT) {
	186	/* For reliability don't try to kmalloc more than
	187	2 pages worth */
932b59e3 YJN	188	process_pages = kmalloc(min_t(size_t, PVM_MAX_KMALLOC_PAGES * PAGE_SIZE,
932b59e3 YJN	189	sizeof(struct page )nr_pages),
fcf63409 CY	190	GFP_KERNEL);
	191
	192	if (!process_pages)
	193	return -ENOMEM;
	194	}
	195
	196	/* Get process information */
2ee08260	197	task = find_get_task_by_vpid(pid);
fcf63409 CY	198	if (!task) {
	199	rc = -ESRCH;
	200	goto free_proc_pages;
	201	}
	202
caaee623	203	mm = mm_access(task, PTRACE_MODE_ATTACH_REALCREDS);
8cdb878d CY	204	if (!mm \|\| IS_ERR(mm)) {
	205	rc = IS_ERR(mm) ? PTR_ERR(mm) : -ESRCH;
	206	/*
8d09db80	207	* Explicitly map EACCES to EPERM as EPERM is a more
8cdb878d CY	208	* appropriate error code for process_vw_readv/writev
	209	*/
	210	if (rc == -EACCES)
	211	rc = -EPERM;
fcf63409 CY	212	goto put_task_struct;
	213	}
	214
9acc1a0f	215	for (i = 0; i < riovcnt && iov_iter_count(iter) && !rc; i++)
fcf63409 CY	216	rc = process_vm_rw_single_vec(
fcf63409 CY	217	(unsigned long)rvec[i].iov_base, rvec[i].iov_len,
9acc1a0f AV	218	iter, process_pages, mm, task, vm_write);
	219
	220	/* copied = space before - space after */
	221	total_len -= iov_iter_count(iter);
	222
	223	/* If we have managed to copy any data at all then
	224	we return the number of bytes copied. Otherwise
	225	we return the error code */
	226	if (total_len)
	227	rc = total_len;
fcf63409	228
fcf63409 CY	229	mmput(mm);
	230
	231	put_task_struct:
	232	put_task_struct(task);
	233
	234	free_proc_pages:
	235	if (process_pages != pp_stack)
	236	kfree(process_pages);
	237	return rc;
	238	}
	239
	240	/**
	241	* process_vm_rw - check iovecs before calling core routine
	242	* @pid: PID of process to read/write from/to
	243	* @lvec: iovec array specifying where to copy to/from locally
	244	* @liovcnt: size of lvec array
	245	* @rvec: iovec array specifying where to copy to/from in the other process
	246	* @riovcnt: size of rvec array
	247	* @flags: currently unused
	248	* @vm_write: 0 if reading from other process, 1 if writing to other process
a5d09bed	249	*
fcf63409 CY	250	* Returns the number of bytes read/written or error code. May
	251	* return less bytes than expected if an error occurs during the copying
	252	* process.
	253	*/
	254	static ssize_t process_vm_rw(pid_t pid,
	255	const struct iovec __user *lvec,
	256	unsigned long liovcnt,
	257	const struct iovec __user *rvec,
	258	unsigned long riovcnt,
	259	unsigned long flags, int vm_write)
	260	{
	261	struct iovec iovstack_l[UIO_FASTIOV];
	262	struct iovec iovstack_r[UIO_FASTIOV];
	263	struct iovec *iov_l = iovstack_l;
95c9ae14	264	struct iovec *iov_r;
9f78bdfa	265	struct iov_iter iter;
fcf63409	266	ssize_t rc;
de4eda9d	267	int dir = vm_write ? ITER_SOURCE : ITER_DEST;
fcf63409 CY	268
	269	if (flags != 0)
	270	return -EINVAL;
	271
	272	/* Check iovecs */
17d17e72 AV	273	rc = import_iovec(dir, lvec, liovcnt, UIO_FASTIOV, &iov_l, &iter);
	274	if (rc < 0)
	275	return rc;
	276	if (!iov_iter_count(&iter))
bfdc5970	277	goto free_iov_l;
38dc5079 JA	278	iov_r = iovec_from_user(rvec, riovcnt, UIO_FASTIOV, iovstack_r,
38dc5079 JA	279	in_compat_syscall());
bfdc5970 CH	280	if (IS_ERR(iov_r)) {
	281	rc = PTR_ERR(iov_r);
	282	goto free_iov_l;
	283	}
9f78bdfa	284	rc = process_vm_rw_core(pid, &iter, iov_r, riovcnt, flags, vm_write);
fcf63409 CY	285	if (iov_r != iovstack_r)
fcf63409 CY	286	kfree(iov_r);
bfdc5970	287	free_iov_l:
17d17e72	288	kfree(iov_l);
fcf63409 CY	289	return rc;
	290	}
	291
	292	SYSCALL_DEFINE6(process_vm_readv, pid_t, pid, const struct iovec __user *, lvec,
	293	unsigned long, liovcnt, const struct iovec __user *, rvec,
	294	unsigned long, riovcnt, unsigned long, flags)
	295	{
	296	return process_vm_rw(pid, lvec, liovcnt, rvec, riovcnt, flags, 0);
	297	}
	298
	299	SYSCALL_DEFINE6(process_vm_writev, pid_t, pid,
	300	const struct iovec __user *, lvec,
	301	unsigned long, liovcnt, const struct iovec __user *, rvec,
	302	unsigned long, riovcnt, unsigned long, flags)
	303	{
	304	return process_vm_rw(pid, lvec, liovcnt, rvec, riovcnt, flags, 1);
	305	}