| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <20151228140440.GA13374@1wt.eu>
Date: Mon, 28 Dec 2015 15:04:40 +0100
From: Willy Tarreau <w@....eu>
To: Alexander Viro <viro@...iv.linux.org.uk>
Cc: linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org,
Linus Torvalds <torvalds@...ux-foundation.org>,
socketpair@...il.com
Subject: [PATCH] pipe: limit the per-user amount of pages allocated in pipes
On no-so-small systems, it is possible for a single process to cause an
OOM condition by filling large pipes with data that are never read. A
typical process filling 4000 pipes with 1 MB of data will use 4 GB of
memory. On small systems it may be tricky to set the pipe max size to
prevent this from happening.
This patch makes it possible to enforce a per-user limit above which
new pipes will be limited to a single page, effectively limiting them
to 4 kB each. This has the effect of protecting the system against
memory abuse without hurting other users, and still allowing pipes to
work correctly though with less data at once.
The limit is controlled by the new sysctl user-max-pipe-pages, and may
be disabled by setting it to zero. The default limit allows the default
number of FDs per process (1024) to create pipes of the default size
(64kB), thus reaching a limit of 64MB before starting to create only
smaller pipes. With a process enforcing the FD limit to 4096, this
results in 64MB + 3072 * 4kB = 72 MB.
Reported-by: socketpair@...il.com
Suggested-by: Linus Torvalds <torvalds@...ux-foundation.org>
Signed-off-by: Willy Tarreau <w@....eu>
---
I don't know if it's better that the limit is not enforced by default
(easy to change), nor do I know if it's better to specify the limit in
bytes instead of pages. Some settings use pages already so I don't
think it's a big deal.
---
Documentation/sysctl/fs.txt | 10 ++++++++++
fs/pipe.c | 35 ++++++++++++++++++++++++++++++++---
include/linux/pipe_fs_i.h | 2 ++
include/linux/sched.h | 1 +
kernel/sysctl.c | 7 +++++++
5 files changed, 52 insertions(+), 3 deletions(-)
diff --git a/Documentation/sysctl/fs.txt b/Documentation/sysctl/fs.txt
index 88152f2..a9751b1 100644
--- a/Documentation/sysctl/fs.txt
+++ b/Documentation/sysctl/fs.txt
@@ -37,6 +37,7 @@ Currently, these files are in /proc/sys/fs:
- suid_dumpable
- super-max
- super-nr
+- user-max-pipe-pages
==============================================================
@@ -234,6 +235,15 @@ allows you to.
==============================================================
+user-max-pipe-pages:
+
+Maximum total number of pages a non-privileged user may allocate for pipes.
+When this limit is reached, new pipes will be limited to a single page in size
+in order to limit total memory usage. The default value allows to allocate up
+to 1024 pipes at their default size. When set to 0, no limit is enforced.
+
+==============================================================
+
aio-nr & aio-max-nr:
aio-nr shows the current system-wide number of asynchronous io
diff --git a/fs/pipe.c b/fs/pipe.c
index 42cf8dd..cc4874b 100644
--- a/fs/pipe.c
+++ b/fs/pipe.c
@@ -38,6 +38,10 @@ unsigned int pipe_max_size = 1048576;
*/
unsigned int pipe_min_size = PAGE_SIZE;
+/* Maximum allocatable pages per user, matches default values */
+unsigned long user_max_pipe_pages = PIPE_DEF_BUFFERS * INR_OPEN_CUR;
+EXPORT_SYMBOL(user_max_pipe_pages);
+
/*
* We use a start+len construction, which provides full use of the
* allocated memory.
@@ -583,20 +587,41 @@ pipe_fasync(int fd, struct file *filp, int on)
return retval;
}
+static void account_pipe_buffers(struct pipe_inode_info *pipe,
+ unsigned long old, unsigned long new)
+{
+ atomic_long_add(new - old, &pipe->user->pipe_bufs);
+}
+
+static bool too_many_pipe_buffers(struct user_struct *user)
+{
+ return user_max_pipe_pages &&
+ atomic_long_read(&user->pipe_bufs) >= user_max_pipe_pages;
+}
+
struct pipe_inode_info *alloc_pipe_info(void)
{
struct pipe_inode_info *pipe;
pipe = kzalloc(sizeof(struct pipe_inode_info), GFP_KERNEL);
if (pipe) {
- pipe->bufs = kzalloc(sizeof(struct pipe_buffer) * PIPE_DEF_BUFFERS, GFP_KERNEL);
+ unsigned long pipe_bufs = PIPE_DEF_BUFFERS;
+ struct user_struct *user = get_current_user();
+
+ if (too_many_pipe_buffers(user))
+ pipe_bufs = 1;
+
+ pipe->bufs = kzalloc(sizeof(struct pipe_buffer) * pipe_bufs, GFP_KERNEL);
if (pipe->bufs) {
init_waitqueue_head(&pipe->wait);
pipe->r_counter = pipe->w_counter = 1;
- pipe->buffers = PIPE_DEF_BUFFERS;
+ pipe->buffers = pipe_bufs;
+ pipe->user = user;
+ account_pipe_buffers(pipe, 0, pipe_bufs);
mutex_init(&pipe->mutex);
return pipe;
}
+ free_uid(user);
kfree(pipe);
}
@@ -607,6 +632,8 @@ void free_pipe_info(struct pipe_inode_info *pipe)
{
int i;
+ account_pipe_buffers(pipe, pipe->buffers, 0);
+ free_uid(pipe->user);
for (i = 0; i < pipe->buffers; i++) {
struct pipe_buffer *buf = pipe->bufs + i;
if (buf->ops)
@@ -998,6 +1025,7 @@ static long pipe_set_size(struct pipe_inode_info *pipe, unsigned long nr_pages)
memcpy(bufs + head, pipe->bufs, tail * sizeof(struct pipe_buffer));
}
+ account_pipe_buffers(pipe, pipe->buffers, nr_pages);
pipe->curbuf = 0;
kfree(pipe->bufs);
pipe->bufs = bufs;
@@ -1066,7 +1094,8 @@ long pipe_fcntl(struct file *file, unsigned int cmd, unsigned long arg)
if (!nr_pages)
goto out;
- if (!capable(CAP_SYS_RESOURCE) && size > pipe_max_size) {
+ if (!capable(CAP_SYS_RESOURCE) &&
+ (size > pipe_max_size || too_many_pipe_buffers(pipe->user))) {
ret = -EPERM;
goto out;
}
diff --git a/include/linux/pipe_fs_i.h b/include/linux/pipe_fs_i.h
index eb8b8ac..a4cee7c 100644
--- a/include/linux/pipe_fs_i.h
+++ b/include/linux/pipe_fs_i.h
@@ -57,6 +57,7 @@ struct pipe_inode_info {
struct fasync_struct *fasync_readers;
struct fasync_struct *fasync_writers;
struct pipe_buffer *bufs;
+ struct user_struct *user;
};
/*
@@ -123,6 +124,7 @@ void pipe_unlock(struct pipe_inode_info *);
void pipe_double_lock(struct pipe_inode_info *, struct pipe_inode_info *);
extern unsigned int pipe_max_size, pipe_min_size;
+extern unsigned long user_max_pipe_pages;
int pipe_proc_fn(struct ctl_table *, int, void __user *, size_t *, loff_t *);
diff --git a/include/linux/sched.h b/include/linux/sched.h
index fbf25f1..93643762 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -831,6 +831,7 @@ struct user_struct {
#endif
unsigned long locked_shm; /* How many pages of mlocked shm ? */
unsigned long unix_inflight; /* How many files in flight in unix sockets */
+ atomic_long_t pipe_bufs; /* how many pages are allocated in pipe buffers */
#ifdef CONFIG_KEYS
struct key *uid_keyring; /* UID specific keyring */
diff --git a/kernel/sysctl.c b/kernel/sysctl.c
index dc6858d..a288edf 100644
--- a/kernel/sysctl.c
+++ b/kernel/sysctl.c
@@ -1735,6 +1735,13 @@ static struct ctl_table fs_table[] = {
.proc_handler = &pipe_proc_fn,
.extra1 = &pipe_min_size,
},
+ {
+ .procname = "user-max-pipe-pages",
+ .data = &user_max_pipe_pages,
+ .maxlen = sizeof(user_max_pipe_pages),
+ .mode = 0644,
+ .proc_handler = proc_doulongvec_minmax,
+ },
{ }
};
--
1.7.12.1
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists