/* ar-skbuff.c: socket buffer destruction handling
 *
 * Copyright (C) 2007 Red Hat, Inc. All Rights Reserved.
 * Written by David Howells (dhowells@redhat.com)
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License
 * as published by the Free Software Foundation; either version
 * 2 of the License, or (at your option) any later version.
 */

#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

#include <linux/module.h>
#include <linux/net.h>
#include <linux/skbuff.h>
#include <net/sock.h>
#include <net/af_rxrpc.h>
#include "ar-internal.h"

#define select_skb_count(op) (op >= rxrpc_skb_tx_cleaned ? &rxrpc_n_tx_skbs : &rxrpc_n_rx_skbs)

/*
 * Note the allocation or reception of a socket buffer.
 */
void rxrpc_new_skb(struct sk_buff *skb, enum rxrpc_skb_trace op)
{
	const void *here = __builtin_return_address(0);
	int n = atomic_inc_return(select_skb_count(op));
	trace_rxrpc_skb(skb, op, atomic_read(&skb->users), n, here);
}

/*
 * Note the re-emergence of a socket buffer from a queue or buffer.
 */
void rxrpc_see_skb(struct sk_buff *skb, enum rxrpc_skb_trace op)
{
	const void *here = __builtin_return_address(0);
	if (skb) {
		int n = atomic_read(select_skb_count(op));
		trace_rxrpc_skb(skb, op, atomic_read(&skb->users), n, here);
	}
}

/*
 * Note the addition of a ref on a socket buffer.
 */
void rxrpc_get_skb(struct sk_buff *skb, enum rxrpc_skb_trace op)
{
	const void *here = __builtin_return_address(0);
	int n = atomic_inc_return(select_skb_count(op));
	trace_rxrpc_skb(skb, op, atomic_read(&skb->users), n, here);
	skb_get(skb);
}

/*
 * Note the destruction of a socket buffer.
 */
void rxrpc_free_skb(struct sk_buff *skb, enum rxrpc_skb_trace op)
{
	const void *here = __builtin_return_address(0);
	if (skb) {
		int n;
		CHECK_SLAB_OKAY(&skb->users);
		n = atomic_dec_return(select_skb_count(op));
		trace_rxrpc_skb(skb, op, atomic_read(&skb->users), n, here);
		kfree_skb(skb);
	}
}

/*
 * Note the injected loss of a socket buffer.
 */
void rxrpc_lose_skb(struct sk_buff *skb, enum rxrpc_skb_trace op)
{
	const void *here = __builtin_return_address(0);
	if (skb) {
		int n;
		CHECK_SLAB_OKAY(&skb->users);
		n = atomic_dec_return(select_skb_count(op));
		trace_rxrpc_skb(skb, op, atomic_read(&skb->users), n, here);
		kfree_skb(skb);
	}
}

/*
 * Clear a queue of socket buffers.
 */
void rxrpc_purge_queue(struct sk_buff_head *list)
{
	const void *here = __builtin_return_address(0);
	struct sk_buff *skb;
	while ((skb = skb_dequeue((list))) != NULL) {
		int n = atomic_dec_return(select_skb_count(rxrpc_skb_rx_purged));
		trace_rxrpc_skb(skb, rxrpc_skb_rx_purged,
				atomic_read(&skb->users), n, here);
		kfree_skb(skb);
	}
}
d=966d2b04e070bc040319aaebfec09e0144dc3341'>root</a>/<a href='/cgit.cgi/linux/net-next.git/commit/net?id=966d2b04e070bc040319aaebfec09e0144dc3341'>net</a>/<a href='/cgit.cgi/linux/net-next.git/commit/net/ipv6?id=966d2b04e070bc040319aaebfec09e0144dc3341'>ipv6</a>/<a href='/cgit.cgi/linux/net-next.git/commit/net/ipv6/ipcomp6.c?id=966d2b04e070bc040319aaebfec09e0144dc3341'>ipcomp6.c</a></div><div class='content'><div class='cgit-panel'><b>diff options</b><form method='get'><input type='hidden' name='id' value='966d2b04e070bc040319aaebfec09e0144dc3341'/><table><tr><td colspan='2'/></tr><tr><td class='label'>context:</td><td class='ctrl'><select name='context' onchange='this.form.submit();'><option value='1'>1</option><option value='2'>2</option><option value='3' selected='selected'>3</option><option value='4'>4</option><option value='5'>5</option><option value='6'>6</option><option value='7'>7</option><option value='8'>8</option><option value='9'>9</option><option value='10'>10</option><option value='15'>15</option><option value='20'>20</option><option value='25'>25</option><option value='30'>30</option><option value='35'>35</option><option value='40'>40</option></select></td></tr><tr><td class='label'>space:</td><td class='ctrl'><select name='ignorews' onchange='this.form.submit();'><option value='0' selected='selected'>include</option><option value='1'>ignore</option></select></td></tr><tr><td class='label'>mode:</td><td class='ctrl'><select name='dt' onchange='this.form.submit();'><option value='0' selected='selected'>unified</option><option value='1'>ssdiff</option><option value='2'>stat only</option></select></td></tr><tr><td/><td class='ctrl'><noscript><input type='submit' value='reload'/></noscript></td></tr></table></form></div><table summary='commit info' class='commit-info'>
<tr><th>author</th><td>Douglas Miller &lt;dougmill@linux.vnet.ibm.com&gt;</td><td class='right'>2017-01-28 06:42:20 -0600</td></tr>
<tr><th>committer</th><td>Tejun Heo &lt;tj@kernel.org&gt;</td><td class='right'>2017-01-28 07:49:42 -0500</td></tr>
<tr><th>commit</th><td colspan='2' class='oid'><a href='/cgit.cgi/linux/net-next.git/commit/net/ipv6/ipcomp6.c?id=966d2b04e070bc040319aaebfec09e0144dc3341'>966d2b04e070bc040319aaebfec09e0144dc3341</a> (<a href='/cgit.cgi/linux/net-next.git/patch/net/ipv6/ipcomp6.c?id=966d2b04e070bc040319aaebfec09e0144dc3341'>patch</a>)</td></tr>
<tr><th>tree</th><td colspan='2' class='oid'><a href='/cgit.cgi/linux/net-next.git/tree/?id=966d2b04e070bc040319aaebfec09e0144dc3341'>4b96156e3d1dd4dfd6039b7c219c9dc4616da52d</a> /<a href='/cgit.cgi/linux/net-next.git/tree/net/ipv6/ipcomp6.c?id=966d2b04e070bc040319aaebfec09e0144dc3341'>net/ipv6/ipcomp6.c</a></td></tr>
<tr><th>parent</th><td colspan='2' class='oid'><a href='/cgit.cgi/linux/net-next.git/commit/net/ipv6/ipcomp6.c?id=1b1bc42c1692e9b62756323c675a44cb1a1f9dbd'>1b1bc42c1692e9b62756323c675a44cb1a1f9dbd</a> (<a href='/cgit.cgi/linux/net-next.git/diff/net/ipv6/ipcomp6.c?id=966d2b04e070bc040319aaebfec09e0144dc3341&amp;id2=1b1bc42c1692e9b62756323c675a44cb1a1f9dbd'>diff</a>)</td></tr></table>
<div class='commit-subject'>percpu-refcount: fix reference leak during percpu-atomic transition</div><div class='commit-msg'>percpu_ref_tryget() and percpu_ref_tryget_live() should return
"true" IFF they acquire a reference. But the return value from
atomic_long_inc_not_zero() is a long and may have high bits set,
e.g. PERCPU_COUNT_BIAS, and the return value of the tryget routines
is bool so the reference may actually be acquired but the routines
return "false" which results in a reference leak since the caller
assumes it does not need to do a corresponding percpu_ref_put().

This was seen when performing CPU hotplug during I/O, as hangs in
blk_mq_freeze_queue_wait where percpu_ref_kill (blk_mq_freeze_queue_start)
raced with percpu_ref_tryget (blk_mq_timeout_work).
Sample stack trace:

__switch_to+0x2c0/0x450
__schedule+0x2f8/0x970
schedule+0x48/0xc0
blk_mq_freeze_queue_wait+0x94/0x120
blk_mq_queue_reinit_work+0xb8/0x180
blk_mq_queue_reinit_prepare+0x84/0xa0
cpuhp_invoke_callback+0x17c/0x600
cpuhp_up_callbacks+0x58/0x150
_cpu_up+0xf0/0x1c0
do_cpu_up+0x120/0x150
cpu_subsys_online+0x64/0xe0
device_online+0xb4/0x120
online_store+0xb4/0xc0
dev_attr_store+0x68/0xa0
sysfs_kf_write+0x80/0xb0
kernfs_fop_write+0x17c/0x250
__vfs_write+0x6c/0x1e0
vfs_write+0xd0/0x270
SyS_write+0x6c/0x110
system_call+0x38/0xe0

Examination of the queue showed a single reference (no PERCPU_COUNT_BIAS,
and __PERCPU_REF_DEAD, __PERCPU_REF_ATOMIC set) and no requests.
However, conditions at the time of the race are count of PERCPU_COUNT_BIAS + 0
and __PERCPU_REF_DEAD and __PERCPU_REF_ATOMIC set.

The fix is to make the tryget routines use an actual boolean internally instead
of the atomic long result truncated to a int.

Fixes: e625305b3907 percpu-refcount: make percpu_ref based on longs instead of ints
Link: https://bugzilla.kernel.org/show_bug.cgi?id=190751
Signed-off-by: Douglas Miller &lt;dougmill@linux.vnet.ibm.com&gt;
Reviewed-by: Jens Axboe &lt;axboe@fb.com&gt;
Signed-off-by: Tejun Heo &lt;tj@kernel.org&gt;
Fixes: e625305b3907 ("percpu-refcount: make percpu_ref based on longs instead of ints")
Cc: stable@vger.kernel.org # v3.18+
</div><div class='diffstat-header'><a href='/cgit.cgi/linux/net-next.git/diff/?id=966d2b04e070bc040319aaebfec09e0144dc3341'>Diffstat</a> (limited to 'net/ipv6/ipcomp6.c')</div><table summary='diffstat' class='diffstat'>