[mirror_ubuntu-jammy-kernel.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/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 (PAGE_SIZE * 2)

/**
 * 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 long max_pages_per_loop = PVM_MAX_KMALLOC_PAGES
		/ sizeof(struct pages *);
	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(nr_pages, max_pages_per_loop);
		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,
						     NULL, &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)
				/ 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,
					      sizeof(struct pages *)*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 = iovstack_r;
	struct iov_iter iter;
	ssize_t rc;
	int dir = vm_write ? WRITE : READ;

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