2007-05-05 20:45:53 +02:00
|
|
|
/*
|
|
|
|
* Copyright 2002-2004, Instant802 Networks, Inc.
|
|
|
|
* Copyright 2005, Devicescape Software, Inc.
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or modify
|
|
|
|
* it under the terms of the GNU General Public License version 2 as
|
|
|
|
* published by the Free Software Foundation.
|
|
|
|
*/
|
|
|
|
#include <linux/kernel.h>
|
2008-05-02 22:47:45 +02:00
|
|
|
#include <linux/bitops.h>
|
2007-05-05 20:45:53 +02:00
|
|
|
#include <linux/types.h>
|
|
|
|
#include <linux/netdevice.h>
|
2011-07-15 17:47:34 +02:00
|
|
|
#include <linux/export.h>
|
2008-05-02 22:47:45 +02:00
|
|
|
#include <asm/unaligned.h>
|
2007-05-05 20:45:53 +02:00
|
|
|
|
|
|
|
#include <net/mac80211.h>
|
2009-04-23 18:52:52 +02:00
|
|
|
#include "driver-ops.h"
|
2008-04-08 21:14:40 +02:00
|
|
|
#include "key.h"
|
2007-05-05 20:45:53 +02:00
|
|
|
#include "tkip.h"
|
|
|
|
#include "wep.h"
|
|
|
|
|
|
|
|
#define PHASE1_LOOP_COUNT 8
|
|
|
|
|
2008-05-02 22:47:45 +02:00
|
|
|
/*
|
|
|
|
* 2-byte by 2-byte subset of the full AES S-box table; second part of this
|
|
|
|
* table is identical to first part but byte-swapped
|
|
|
|
*/
|
2007-05-05 20:45:53 +02:00
|
|
|
static const u16 tkip_sbox[256] =
|
|
|
|
{
|
|
|
|
0xC6A5, 0xF884, 0xEE99, 0xF68D, 0xFF0D, 0xD6BD, 0xDEB1, 0x9154,
|
|
|
|
0x6050, 0x0203, 0xCEA9, 0x567D, 0xE719, 0xB562, 0x4DE6, 0xEC9A,
|
|
|
|
0x8F45, 0x1F9D, 0x8940, 0xFA87, 0xEF15, 0xB2EB, 0x8EC9, 0xFB0B,
|
|
|
|
0x41EC, 0xB367, 0x5FFD, 0x45EA, 0x23BF, 0x53F7, 0xE496, 0x9B5B,
|
|
|
|
0x75C2, 0xE11C, 0x3DAE, 0x4C6A, 0x6C5A, 0x7E41, 0xF502, 0x834F,
|
|
|
|
0x685C, 0x51F4, 0xD134, 0xF908, 0xE293, 0xAB73, 0x6253, 0x2A3F,
|
|
|
|
0x080C, 0x9552, 0x4665, 0x9D5E, 0x3028, 0x37A1, 0x0A0F, 0x2FB5,
|
|
|
|
0x0E09, 0x2436, 0x1B9B, 0xDF3D, 0xCD26, 0x4E69, 0x7FCD, 0xEA9F,
|
|
|
|
0x121B, 0x1D9E, 0x5874, 0x342E, 0x362D, 0xDCB2, 0xB4EE, 0x5BFB,
|
|
|
|
0xA4F6, 0x764D, 0xB761, 0x7DCE, 0x527B, 0xDD3E, 0x5E71, 0x1397,
|
|
|
|
0xA6F5, 0xB968, 0x0000, 0xC12C, 0x4060, 0xE31F, 0x79C8, 0xB6ED,
|
|
|
|
0xD4BE, 0x8D46, 0x67D9, 0x724B, 0x94DE, 0x98D4, 0xB0E8, 0x854A,
|
|
|
|
0xBB6B, 0xC52A, 0x4FE5, 0xED16, 0x86C5, 0x9AD7, 0x6655, 0x1194,
|
|
|
|
0x8ACF, 0xE910, 0x0406, 0xFE81, 0xA0F0, 0x7844, 0x25BA, 0x4BE3,
|
|
|
|
0xA2F3, 0x5DFE, 0x80C0, 0x058A, 0x3FAD, 0x21BC, 0x7048, 0xF104,
|
|
|
|
0x63DF, 0x77C1, 0xAF75, 0x4263, 0x2030, 0xE51A, 0xFD0E, 0xBF6D,
|
|
|
|
0x814C, 0x1814, 0x2635, 0xC32F, 0xBEE1, 0x35A2, 0x88CC, 0x2E39,
|
|
|
|
0x9357, 0x55F2, 0xFC82, 0x7A47, 0xC8AC, 0xBAE7, 0x322B, 0xE695,
|
|
|
|
0xC0A0, 0x1998, 0x9ED1, 0xA37F, 0x4466, 0x547E, 0x3BAB, 0x0B83,
|
|
|
|
0x8CCA, 0xC729, 0x6BD3, 0x283C, 0xA779, 0xBCE2, 0x161D, 0xAD76,
|
|
|
|
0xDB3B, 0x6456, 0x744E, 0x141E, 0x92DB, 0x0C0A, 0x486C, 0xB8E4,
|
|
|
|
0x9F5D, 0xBD6E, 0x43EF, 0xC4A6, 0x39A8, 0x31A4, 0xD337, 0xF28B,
|
|
|
|
0xD532, 0x8B43, 0x6E59, 0xDAB7, 0x018C, 0xB164, 0x9CD2, 0x49E0,
|
|
|
|
0xD8B4, 0xACFA, 0xF307, 0xCF25, 0xCAAF, 0xF48E, 0x47E9, 0x1018,
|
|
|
|
0x6FD5, 0xF088, 0x4A6F, 0x5C72, 0x3824, 0x57F1, 0x73C7, 0x9751,
|
|
|
|
0xCB23, 0xA17C, 0xE89C, 0x3E21, 0x96DD, 0x61DC, 0x0D86, 0x0F85,
|
|
|
|
0xE090, 0x7C42, 0x71C4, 0xCCAA, 0x90D8, 0x0605, 0xF701, 0x1C12,
|
|
|
|
0xC2A3, 0x6A5F, 0xAEF9, 0x69D0, 0x1791, 0x9958, 0x3A27, 0x27B9,
|
|
|
|
0xD938, 0xEB13, 0x2BB3, 0x2233, 0xD2BB, 0xA970, 0x0789, 0x33A7,
|
|
|
|
0x2DB6, 0x3C22, 0x1592, 0xC920, 0x8749, 0xAAFF, 0x5078, 0xA57A,
|
|
|
|
0x038F, 0x59F8, 0x0980, 0x1A17, 0x65DA, 0xD731, 0x84C6, 0xD0B8,
|
|
|
|
0x82C3, 0x29B0, 0x5A77, 0x1E11, 0x7BCB, 0xA8FC, 0x6DD6, 0x2C3A,
|
|
|
|
};
|
|
|
|
|
2008-05-02 22:47:45 +02:00
|
|
|
static u16 tkipS(u16 val)
|
2007-05-05 20:45:53 +02:00
|
|
|
{
|
2008-05-02 22:47:45 +02:00
|
|
|
return tkip_sbox[val & 0xff] ^ swab16(tkip_sbox[val >> 8]);
|
2007-05-05 20:45:53 +02:00
|
|
|
}
|
|
|
|
|
2008-06-11 23:22:00 +02:00
|
|
|
static u8 *write_tkip_iv(u8 *pos, u16 iv16)
|
|
|
|
{
|
|
|
|
*pos++ = iv16 >> 8;
|
|
|
|
*pos++ = ((iv16 >> 8) | 0x20) & 0x7f;
|
|
|
|
*pos++ = iv16 & 0xFF;
|
|
|
|
return pos;
|
|
|
|
}
|
|
|
|
|
2008-05-02 22:47:45 +02:00
|
|
|
/*
|
|
|
|
* P1K := Phase1(TA, TK, TSC)
|
2007-05-05 20:45:53 +02:00
|
|
|
* TA = transmitter address (48 bits)
|
|
|
|
* TK = dot11DefaultKeyValue or dot11KeyMappingValue (128 bits)
|
|
|
|
* TSC = TKIP sequence counter (48 bits, only 32 msb bits used)
|
|
|
|
* P1K: 80 bits
|
|
|
|
*/
|
2008-06-11 23:22:00 +02:00
|
|
|
static void tkip_mixing_phase1(const u8 *tk, struct tkip_ctx *ctx,
|
|
|
|
const u8 *ta, u32 tsc_IV32)
|
2007-05-05 20:45:53 +02:00
|
|
|
{
|
|
|
|
int i, j;
|
2008-05-15 01:26:19 +02:00
|
|
|
u16 *p1k = ctx->p1k;
|
2007-05-05 20:45:53 +02:00
|
|
|
|
2008-05-02 22:47:45 +02:00
|
|
|
p1k[0] = tsc_IV32 & 0xFFFF;
|
|
|
|
p1k[1] = tsc_IV32 >> 16;
|
|
|
|
p1k[2] = get_unaligned_le16(ta + 0);
|
|
|
|
p1k[3] = get_unaligned_le16(ta + 2);
|
|
|
|
p1k[4] = get_unaligned_le16(ta + 4);
|
2007-05-05 20:45:53 +02:00
|
|
|
|
|
|
|
for (i = 0; i < PHASE1_LOOP_COUNT; i++) {
|
|
|
|
j = 2 * (i & 1);
|
2008-05-02 22:47:45 +02:00
|
|
|
p1k[0] += tkipS(p1k[4] ^ get_unaligned_le16(tk + 0 + j));
|
|
|
|
p1k[1] += tkipS(p1k[0] ^ get_unaligned_le16(tk + 4 + j));
|
|
|
|
p1k[2] += tkipS(p1k[1] ^ get_unaligned_le16(tk + 8 + j));
|
|
|
|
p1k[3] += tkipS(p1k[2] ^ get_unaligned_le16(tk + 12 + j));
|
|
|
|
p1k[4] += tkipS(p1k[3] ^ get_unaligned_le16(tk + 0 + j)) + i;
|
2007-05-05 20:45:53 +02:00
|
|
|
}
|
2009-12-09 23:25:05 +01:00
|
|
|
ctx->state = TKIP_STATE_PHASE1_DONE;
|
mac80211: fix TKIP races, make API easier to use
Our current TKIP code races against itself on TX
since we can process multiple packets at the same
time on different ACs, but they all share the TX
context for TKIP. This can lead to bad IVs etc.
Also, the crypto offload helper code just obtains
the P1K/P2K from the cache, and can update it as
well, but there's no guarantee that packets are
really processed in order.
To fix these issues, first introduce a spinlock
that will protect the IV16/IV32 values in the TX
context. This first step makes sure that we don't
assign the same IV multiple times or get confused
in other ways.
Secondly, change the way the P1K cache works. I
add a field "p1k_iv32" that stores the value of
the IV32 when the P1K was last recomputed, and
if different from the last time, then a new P1K
is recomputed. This can cause the P1K computation
to flip back and forth if packets are processed
out of order. All this also happens under the new
spinlock.
Finally, because there are argument differences,
split up the ieee80211_get_tkip_key() API into
ieee80211_get_tkip_p1k() and ieee80211_get_tkip_p2k()
and give them the correct arguments.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
2011-07-07 22:28:01 +02:00
|
|
|
ctx->p1k_iv32 = tsc_IV32;
|
2007-05-05 20:45:53 +02:00
|
|
|
}
|
|
|
|
|
2008-06-11 23:22:00 +02:00
|
|
|
static void tkip_mixing_phase2(const u8 *tk, struct tkip_ctx *ctx,
|
2008-05-15 01:26:20 +02:00
|
|
|
u16 tsc_IV16, u8 *rc4key)
|
2007-05-05 20:45:53 +02:00
|
|
|
{
|
|
|
|
u16 ppk[6];
|
2008-05-15 01:26:20 +02:00
|
|
|
const u16 *p1k = ctx->p1k;
|
2007-05-05 20:45:53 +02:00
|
|
|
int i;
|
|
|
|
|
|
|
|
ppk[0] = p1k[0];
|
|
|
|
ppk[1] = p1k[1];
|
|
|
|
ppk[2] = p1k[2];
|
|
|
|
ppk[3] = p1k[3];
|
|
|
|
ppk[4] = p1k[4];
|
|
|
|
ppk[5] = p1k[4] + tsc_IV16;
|
|
|
|
|
2008-05-02 22:47:45 +02:00
|
|
|
ppk[0] += tkipS(ppk[5] ^ get_unaligned_le16(tk + 0));
|
|
|
|
ppk[1] += tkipS(ppk[0] ^ get_unaligned_le16(tk + 2));
|
|
|
|
ppk[2] += tkipS(ppk[1] ^ get_unaligned_le16(tk + 4));
|
|
|
|
ppk[3] += tkipS(ppk[2] ^ get_unaligned_le16(tk + 6));
|
|
|
|
ppk[4] += tkipS(ppk[3] ^ get_unaligned_le16(tk + 8));
|
|
|
|
ppk[5] += tkipS(ppk[4] ^ get_unaligned_le16(tk + 10));
|
|
|
|
ppk[0] += ror16(ppk[5] ^ get_unaligned_le16(tk + 12), 1);
|
|
|
|
ppk[1] += ror16(ppk[0] ^ get_unaligned_le16(tk + 14), 1);
|
|
|
|
ppk[2] += ror16(ppk[1], 1);
|
|
|
|
ppk[3] += ror16(ppk[2], 1);
|
|
|
|
ppk[4] += ror16(ppk[3], 1);
|
|
|
|
ppk[5] += ror16(ppk[4], 1);
|
|
|
|
|
2008-06-11 23:22:00 +02:00
|
|
|
rc4key = write_tkip_iv(rc4key, tsc_IV16);
|
|
|
|
*rc4key++ = ((ppk[5] ^ get_unaligned_le16(tk)) >> 1) & 0xFF;
|
2008-05-02 22:47:45 +02:00
|
|
|
|
|
|
|
for (i = 0; i < 6; i++)
|
|
|
|
put_unaligned_le16(ppk[i], rc4key + 2 * i);
|
2007-05-05 20:45:53 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Add TKIP IV and Ext. IV at @pos. @iv0, @iv1, and @iv2 are the first octets
|
|
|
|
* of the IV. Returns pointer to the octet following IVs (i.e., beginning of
|
|
|
|
* the packet payload). */
|
mac80211: fix TKIP races, make API easier to use
Our current TKIP code races against itself on TX
since we can process multiple packets at the same
time on different ACs, but they all share the TX
context for TKIP. This can lead to bad IVs etc.
Also, the crypto offload helper code just obtains
the P1K/P2K from the cache, and can update it as
well, but there's no guarantee that packets are
really processed in order.
To fix these issues, first introduce a spinlock
that will protect the IV16/IV32 values in the TX
context. This first step makes sure that we don't
assign the same IV multiple times or get confused
in other ways.
Secondly, change the way the P1K cache works. I
add a field "p1k_iv32" that stores the value of
the IV32 when the P1K was last recomputed, and
if different from the last time, then a new P1K
is recomputed. This can cause the P1K computation
to flip back and forth if packets are processed
out of order. All this also happens under the new
spinlock.
Finally, because there are argument differences,
split up the ieee80211_get_tkip_key() API into
ieee80211_get_tkip_p1k() and ieee80211_get_tkip_p2k()
and give them the correct arguments.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
2011-07-07 22:28:01 +02:00
|
|
|
u8 *ieee80211_tkip_add_iv(u8 *pos, struct ieee80211_key *key)
|
2007-05-05 20:45:53 +02:00
|
|
|
{
|
mac80211: fix TKIP races, make API easier to use
Our current TKIP code races against itself on TX
since we can process multiple packets at the same
time on different ACs, but they all share the TX
context for TKIP. This can lead to bad IVs etc.
Also, the crypto offload helper code just obtains
the P1K/P2K from the cache, and can update it as
well, but there's no guarantee that packets are
really processed in order.
To fix these issues, first introduce a spinlock
that will protect the IV16/IV32 values in the TX
context. This first step makes sure that we don't
assign the same IV multiple times or get confused
in other ways.
Secondly, change the way the P1K cache works. I
add a field "p1k_iv32" that stores the value of
the IV32 when the P1K was last recomputed, and
if different from the last time, then a new P1K
is recomputed. This can cause the P1K computation
to flip back and forth if packets are processed
out of order. All this also happens under the new
spinlock.
Finally, because there are argument differences,
split up the ieee80211_get_tkip_key() API into
ieee80211_get_tkip_p1k() and ieee80211_get_tkip_p2k()
and give them the correct arguments.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
2011-07-07 22:28:01 +02:00
|
|
|
lockdep_assert_held(&key->u.tkip.txlock);
|
|
|
|
|
|
|
|
pos = write_tkip_iv(pos, key->u.tkip.tx.iv16);
|
2007-08-28 23:01:54 +02:00
|
|
|
*pos++ = (key->conf.keyidx << 6) | (1 << 5) /* Ext IV */;
|
2008-05-15 01:26:19 +02:00
|
|
|
put_unaligned_le32(key->u.tkip.tx.iv32, pos);
|
2008-05-02 22:47:45 +02:00
|
|
|
return pos + 4;
|
2007-05-05 20:45:53 +02:00
|
|
|
}
|
|
|
|
|
mac80211: fix TKIP races, make API easier to use
Our current TKIP code races against itself on TX
since we can process multiple packets at the same
time on different ACs, but they all share the TX
context for TKIP. This can lead to bad IVs etc.
Also, the crypto offload helper code just obtains
the P1K/P2K from the cache, and can update it as
well, but there's no guarantee that packets are
really processed in order.
To fix these issues, first introduce a spinlock
that will protect the IV16/IV32 values in the TX
context. This first step makes sure that we don't
assign the same IV multiple times or get confused
in other ways.
Secondly, change the way the P1K cache works. I
add a field "p1k_iv32" that stores the value of
the IV32 when the P1K was last recomputed, and
if different from the last time, then a new P1K
is recomputed. This can cause the P1K computation
to flip back and forth if packets are processed
out of order. All this also happens under the new
spinlock.
Finally, because there are argument differences,
split up the ieee80211_get_tkip_key() API into
ieee80211_get_tkip_p1k() and ieee80211_get_tkip_p2k()
and give them the correct arguments.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
2011-07-07 22:28:01 +02:00
|
|
|
static void ieee80211_compute_tkip_p1k(struct ieee80211_key *key, u32 iv32)
|
|
|
|
{
|
|
|
|
struct ieee80211_sub_if_data *sdata = key->sdata;
|
|
|
|
struct tkip_ctx *ctx = &key->u.tkip.tx;
|
|
|
|
const u8 *tk = &key->conf.key[NL80211_TKIP_DATA_OFFSET_ENCR_KEY];
|
|
|
|
|
|
|
|
lockdep_assert_held(&key->u.tkip.txlock);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Update the P1K when the IV32 is different from the value it
|
|
|
|
* had when we last computed it (or when not initialised yet).
|
|
|
|
* This might flip-flop back and forth if packets are processed
|
|
|
|
* out-of-order due to the different ACs, but then we have to
|
|
|
|
* just compute the P1K more often.
|
|
|
|
*/
|
|
|
|
if (ctx->p1k_iv32 != iv32 || ctx->state == TKIP_STATE_NOT_INIT)
|
|
|
|
tkip_mixing_phase1(tk, ctx, sdata->vif.addr, iv32);
|
|
|
|
}
|
|
|
|
|
2011-07-07 18:58:01 +02:00
|
|
|
void ieee80211_get_tkip_p1k_iv(struct ieee80211_key_conf *keyconf,
|
|
|
|
u32 iv32, u16 *p1k)
|
2008-03-20 14:06:41 +01:00
|
|
|
{
|
|
|
|
struct ieee80211_key *key = (struct ieee80211_key *)
|
|
|
|
container_of(keyconf, struct ieee80211_key, conf);
|
mac80211: fix TKIP races, make API easier to use
Our current TKIP code races against itself on TX
since we can process multiple packets at the same
time on different ACs, but they all share the TX
context for TKIP. This can lead to bad IVs etc.
Also, the crypto offload helper code just obtains
the P1K/P2K from the cache, and can update it as
well, but there's no guarantee that packets are
really processed in order.
To fix these issues, first introduce a spinlock
that will protect the IV16/IV32 values in the TX
context. This first step makes sure that we don't
assign the same IV multiple times or get confused
in other ways.
Secondly, change the way the P1K cache works. I
add a field "p1k_iv32" that stores the value of
the IV32 when the P1K was last recomputed, and
if different from the last time, then a new P1K
is recomputed. This can cause the P1K computation
to flip back and forth if packets are processed
out of order. All this also happens under the new
spinlock.
Finally, because there are argument differences,
split up the ieee80211_get_tkip_key() API into
ieee80211_get_tkip_p1k() and ieee80211_get_tkip_p2k()
and give them the correct arguments.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
2011-07-07 22:28:01 +02:00
|
|
|
struct tkip_ctx *ctx = &key->u.tkip.tx;
|
|
|
|
|
2013-02-13 18:03:00 +01:00
|
|
|
spin_lock_bh(&key->u.tkip.txlock);
|
mac80211: fix TKIP races, make API easier to use
Our current TKIP code races against itself on TX
since we can process multiple packets at the same
time on different ACs, but they all share the TX
context for TKIP. This can lead to bad IVs etc.
Also, the crypto offload helper code just obtains
the P1K/P2K from the cache, and can update it as
well, but there's no guarantee that packets are
really processed in order.
To fix these issues, first introduce a spinlock
that will protect the IV16/IV32 values in the TX
context. This first step makes sure that we don't
assign the same IV multiple times or get confused
in other ways.
Secondly, change the way the P1K cache works. I
add a field "p1k_iv32" that stores the value of
the IV32 when the P1K was last recomputed, and
if different from the last time, then a new P1K
is recomputed. This can cause the P1K computation
to flip back and forth if packets are processed
out of order. All this also happens under the new
spinlock.
Finally, because there are argument differences,
split up the ieee80211_get_tkip_key() API into
ieee80211_get_tkip_p1k() and ieee80211_get_tkip_p2k()
and give them the correct arguments.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
2011-07-07 22:28:01 +02:00
|
|
|
ieee80211_compute_tkip_p1k(key, iv32);
|
|
|
|
memcpy(p1k, ctx->p1k, sizeof(ctx->p1k));
|
2013-02-13 18:03:00 +01:00
|
|
|
spin_unlock_bh(&key->u.tkip.txlock);
|
mac80211: fix TKIP races, make API easier to use
Our current TKIP code races against itself on TX
since we can process multiple packets at the same
time on different ACs, but they all share the TX
context for TKIP. This can lead to bad IVs etc.
Also, the crypto offload helper code just obtains
the P1K/P2K from the cache, and can update it as
well, but there's no guarantee that packets are
really processed in order.
To fix these issues, first introduce a spinlock
that will protect the IV16/IV32 values in the TX
context. This first step makes sure that we don't
assign the same IV multiple times or get confused
in other ways.
Secondly, change the way the P1K cache works. I
add a field "p1k_iv32" that stores the value of
the IV32 when the P1K was last recomputed, and
if different from the last time, then a new P1K
is recomputed. This can cause the P1K computation
to flip back and forth if packets are processed
out of order. All this also happens under the new
spinlock.
Finally, because there are argument differences,
split up the ieee80211_get_tkip_key() API into
ieee80211_get_tkip_p1k() and ieee80211_get_tkip_p2k()
and give them the correct arguments.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
2011-07-07 22:28:01 +02:00
|
|
|
}
|
2011-07-07 18:58:01 +02:00
|
|
|
EXPORT_SYMBOL(ieee80211_get_tkip_p1k_iv);
|
2008-03-20 14:06:41 +01:00
|
|
|
|
2011-07-13 19:50:34 +02:00
|
|
|
void ieee80211_get_tkip_rx_p1k(struct ieee80211_key_conf *keyconf,
|
|
|
|
const u8 *ta, u32 iv32, u16 *p1k)
|
|
|
|
{
|
|
|
|
const u8 *tk = &keyconf->key[NL80211_TKIP_DATA_OFFSET_ENCR_KEY];
|
|
|
|
struct tkip_ctx ctx;
|
|
|
|
|
|
|
|
tkip_mixing_phase1(tk, &ctx, ta, iv32);
|
|
|
|
memcpy(p1k, ctx.p1k, sizeof(ctx.p1k));
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(ieee80211_get_tkip_rx_p1k);
|
|
|
|
|
mac80211: fix TKIP races, make API easier to use
Our current TKIP code races against itself on TX
since we can process multiple packets at the same
time on different ACs, but they all share the TX
context for TKIP. This can lead to bad IVs etc.
Also, the crypto offload helper code just obtains
the P1K/P2K from the cache, and can update it as
well, but there's no guarantee that packets are
really processed in order.
To fix these issues, first introduce a spinlock
that will protect the IV16/IV32 values in the TX
context. This first step makes sure that we don't
assign the same IV multiple times or get confused
in other ways.
Secondly, change the way the P1K cache works. I
add a field "p1k_iv32" that stores the value of
the IV32 when the P1K was last recomputed, and
if different from the last time, then a new P1K
is recomputed. This can cause the P1K computation
to flip back and forth if packets are processed
out of order. All this also happens under the new
spinlock.
Finally, because there are argument differences,
split up the ieee80211_get_tkip_key() API into
ieee80211_get_tkip_p1k() and ieee80211_get_tkip_p2k()
and give them the correct arguments.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
2011-07-07 22:28:01 +02:00
|
|
|
void ieee80211_get_tkip_p2k(struct ieee80211_key_conf *keyconf,
|
|
|
|
struct sk_buff *skb, u8 *p2k)
|
|
|
|
{
|
|
|
|
struct ieee80211_key *key = (struct ieee80211_key *)
|
|
|
|
container_of(keyconf, struct ieee80211_key, conf);
|
|
|
|
const u8 *tk = &key->conf.key[NL80211_TKIP_DATA_OFFSET_ENCR_KEY];
|
|
|
|
struct tkip_ctx *ctx = &key->u.tkip.tx;
|
|
|
|
struct ieee80211_hdr *hdr = (struct ieee80211_hdr *)skb->data;
|
|
|
|
const u8 *data = (u8 *)hdr + ieee80211_hdrlen(hdr->frame_control);
|
|
|
|
u32 iv32 = get_unaligned_le32(&data[4]);
|
|
|
|
u16 iv16 = data[2] | (data[0] << 8);
|
|
|
|
|
2013-05-06 15:54:58 +02:00
|
|
|
spin_lock(&key->u.tkip.txlock);
|
mac80211: fix TKIP races, make API easier to use
Our current TKIP code races against itself on TX
since we can process multiple packets at the same
time on different ACs, but they all share the TX
context for TKIP. This can lead to bad IVs etc.
Also, the crypto offload helper code just obtains
the P1K/P2K from the cache, and can update it as
well, but there's no guarantee that packets are
really processed in order.
To fix these issues, first introduce a spinlock
that will protect the IV16/IV32 values in the TX
context. This first step makes sure that we don't
assign the same IV multiple times or get confused
in other ways.
Secondly, change the way the P1K cache works. I
add a field "p1k_iv32" that stores the value of
the IV32 when the P1K was last recomputed, and
if different from the last time, then a new P1K
is recomputed. This can cause the P1K computation
to flip back and forth if packets are processed
out of order. All this also happens under the new
spinlock.
Finally, because there are argument differences,
split up the ieee80211_get_tkip_key() API into
ieee80211_get_tkip_p1k() and ieee80211_get_tkip_p2k()
and give them the correct arguments.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
2011-07-07 22:28:01 +02:00
|
|
|
ieee80211_compute_tkip_p1k(key, iv32);
|
|
|
|
tkip_mixing_phase2(tk, ctx, iv16, p2k);
|
2013-05-06 15:54:58 +02:00
|
|
|
spin_unlock(&key->u.tkip.txlock);
|
2008-03-20 14:06:41 +01:00
|
|
|
}
|
mac80211: fix TKIP races, make API easier to use
Our current TKIP code races against itself on TX
since we can process multiple packets at the same
time on different ACs, but they all share the TX
context for TKIP. This can lead to bad IVs etc.
Also, the crypto offload helper code just obtains
the P1K/P2K from the cache, and can update it as
well, but there's no guarantee that packets are
really processed in order.
To fix these issues, first introduce a spinlock
that will protect the IV16/IV32 values in the TX
context. This first step makes sure that we don't
assign the same IV multiple times or get confused
in other ways.
Secondly, change the way the P1K cache works. I
add a field "p1k_iv32" that stores the value of
the IV32 when the P1K was last recomputed, and
if different from the last time, then a new P1K
is recomputed. This can cause the P1K computation
to flip back and forth if packets are processed
out of order. All this also happens under the new
spinlock.
Finally, because there are argument differences,
split up the ieee80211_get_tkip_key() API into
ieee80211_get_tkip_p1k() and ieee80211_get_tkip_p2k()
and give them the correct arguments.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
2011-07-07 22:28:01 +02:00
|
|
|
EXPORT_SYMBOL(ieee80211_get_tkip_p2k);
|
2008-03-20 14:06:41 +01:00
|
|
|
|
2010-01-17 01:47:58 +01:00
|
|
|
/*
|
|
|
|
* Encrypt packet payload with TKIP using @key. @pos is a pointer to the
|
2007-05-05 20:45:53 +02:00
|
|
|
* beginning of the buffer containing payload. This payload must include
|
2010-01-17 01:47:58 +01:00
|
|
|
* the IV/Ext.IV and space for (taildroom) four octets for ICV.
|
|
|
|
* @payload_len is the length of payload (_not_ including IV/ICV length).
|
|
|
|
* @ta is the transmitter addresses.
|
|
|
|
*/
|
2011-03-25 21:39:19 +01:00
|
|
|
int ieee80211_tkip_encrypt_data(struct crypto_cipher *tfm,
|
2010-07-07 21:07:49 +02:00
|
|
|
struct ieee80211_key *key,
|
mac80211: fix TKIP races, make API easier to use
Our current TKIP code races against itself on TX
since we can process multiple packets at the same
time on different ACs, but they all share the TX
context for TKIP. This can lead to bad IVs etc.
Also, the crypto offload helper code just obtains
the P1K/P2K from the cache, and can update it as
well, but there's no guarantee that packets are
really processed in order.
To fix these issues, first introduce a spinlock
that will protect the IV16/IV32 values in the TX
context. This first step makes sure that we don't
assign the same IV multiple times or get confused
in other ways.
Secondly, change the way the P1K cache works. I
add a field "p1k_iv32" that stores the value of
the IV32 when the P1K was last recomputed, and
if different from the last time, then a new P1K
is recomputed. This can cause the P1K computation
to flip back and forth if packets are processed
out of order. All this also happens under the new
spinlock.
Finally, because there are argument differences,
split up the ieee80211_get_tkip_key() API into
ieee80211_get_tkip_p1k() and ieee80211_get_tkip_p2k()
and give them the correct arguments.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
2011-07-07 22:28:01 +02:00
|
|
|
struct sk_buff *skb,
|
|
|
|
u8 *payload, size_t payload_len)
|
2007-05-05 20:45:53 +02:00
|
|
|
{
|
|
|
|
u8 rc4key[16];
|
2008-06-11 23:22:00 +02:00
|
|
|
|
mac80211: fix TKIP races, make API easier to use
Our current TKIP code races against itself on TX
since we can process multiple packets at the same
time on different ACs, but they all share the TX
context for TKIP. This can lead to bad IVs etc.
Also, the crypto offload helper code just obtains
the P1K/P2K from the cache, and can update it as
well, but there's no guarantee that packets are
really processed in order.
To fix these issues, first introduce a spinlock
that will protect the IV16/IV32 values in the TX
context. This first step makes sure that we don't
assign the same IV multiple times or get confused
in other ways.
Secondly, change the way the P1K cache works. I
add a field "p1k_iv32" that stores the value of
the IV32 when the P1K was last recomputed, and
if different from the last time, then a new P1K
is recomputed. This can cause the P1K computation
to flip back and forth if packets are processed
out of order. All this also happens under the new
spinlock.
Finally, because there are argument differences,
split up the ieee80211_get_tkip_key() API into
ieee80211_get_tkip_p1k() and ieee80211_get_tkip_p2k()
and give them the correct arguments.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
2011-07-07 22:28:01 +02:00
|
|
|
ieee80211_get_tkip_p2k(&key->conf, skb, rc4key);
|
2007-05-05 20:45:53 +02:00
|
|
|
|
mac80211: fix TKIP races, make API easier to use
Our current TKIP code races against itself on TX
since we can process multiple packets at the same
time on different ACs, but they all share the TX
context for TKIP. This can lead to bad IVs etc.
Also, the crypto offload helper code just obtains
the P1K/P2K from the cache, and can update it as
well, but there's no guarantee that packets are
really processed in order.
To fix these issues, first introduce a spinlock
that will protect the IV16/IV32 values in the TX
context. This first step makes sure that we don't
assign the same IV multiple times or get confused
in other ways.
Secondly, change the way the P1K cache works. I
add a field "p1k_iv32" that stores the value of
the IV32 when the P1K was last recomputed, and
if different from the last time, then a new P1K
is recomputed. This can cause the P1K computation
to flip back and forth if packets are processed
out of order. All this also happens under the new
spinlock.
Finally, because there are argument differences,
split up the ieee80211_get_tkip_key() API into
ieee80211_get_tkip_p1k() and ieee80211_get_tkip_p2k()
and give them the correct arguments.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
2011-07-07 22:28:01 +02:00
|
|
|
return ieee80211_wep_encrypt_data(tfm, rc4key, 16,
|
|
|
|
payload, payload_len);
|
2007-05-05 20:45:53 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Decrypt packet payload with TKIP using @key. @pos is a pointer to the
|
|
|
|
* beginning of the buffer containing IEEE 802.11 header payload, i.e.,
|
|
|
|
* including IV, Ext. IV, real data, Michael MIC, ICV. @payload_len is the
|
|
|
|
* length of payload, including IV, Ext. IV, MIC, ICV. */
|
2011-03-25 21:39:19 +01:00
|
|
|
int ieee80211_tkip_decrypt_data(struct crypto_cipher *tfm,
|
2007-05-05 20:45:53 +02:00
|
|
|
struct ieee80211_key *key,
|
|
|
|
u8 *payload, size_t payload_len, u8 *ta,
|
2008-03-20 14:06:42 +01:00
|
|
|
u8 *ra, int only_iv, int queue,
|
2007-09-26 15:19:45 +02:00
|
|
|
u32 *out_iv32, u16 *out_iv16)
|
2007-05-05 20:45:53 +02:00
|
|
|
{
|
|
|
|
u32 iv32;
|
|
|
|
u32 iv16;
|
|
|
|
u8 rc4key[16], keyid, *pos = payload;
|
|
|
|
int res;
|
2008-06-21 16:02:46 +02:00
|
|
|
const u8 *tk = &key->conf.key[NL80211_TKIP_DATA_OFFSET_ENCR_KEY];
|
2007-05-05 20:45:53 +02:00
|
|
|
|
|
|
|
if (payload_len < 12)
|
|
|
|
return -1;
|
|
|
|
|
|
|
|
iv16 = (pos[0] << 8) | pos[2];
|
|
|
|
keyid = pos[3];
|
2008-05-02 22:47:45 +02:00
|
|
|
iv32 = get_unaligned_le32(pos + 4);
|
2007-05-05 20:45:53 +02:00
|
|
|
pos += 8;
|
|
|
|
|
|
|
|
if (!(keyid & (1 << 5)))
|
|
|
|
return TKIP_DECRYPT_NO_EXT_IV;
|
|
|
|
|
2007-08-28 23:01:54 +02:00
|
|
|
if ((keyid >> 6) != key->conf.keyidx)
|
2007-05-05 20:45:53 +02:00
|
|
|
return TKIP_DECRYPT_INVALID_KEYIDX;
|
|
|
|
|
2009-12-09 23:25:05 +01:00
|
|
|
if (key->u.tkip.rx[queue].state != TKIP_STATE_NOT_INIT &&
|
2008-05-15 01:26:19 +02:00
|
|
|
(iv32 < key->u.tkip.rx[queue].iv32 ||
|
|
|
|
(iv32 == key->u.tkip.rx[queue].iv32 &&
|
2012-06-22 11:26:47 +02:00
|
|
|
iv16 <= key->u.tkip.rx[queue].iv16)))
|
2007-05-05 20:45:53 +02:00
|
|
|
return TKIP_DECRYPT_REPLAY;
|
|
|
|
|
|
|
|
if (only_iv) {
|
|
|
|
res = TKIP_DECRYPT_OK;
|
2009-12-09 23:25:05 +01:00
|
|
|
key->u.tkip.rx[queue].state = TKIP_STATE_PHASE1_HW_UPLOADED;
|
2007-05-05 20:45:53 +02:00
|
|
|
goto done;
|
|
|
|
}
|
|
|
|
|
2009-12-09 23:25:05 +01:00
|
|
|
if (key->u.tkip.rx[queue].state == TKIP_STATE_NOT_INIT ||
|
2008-05-15 01:26:19 +02:00
|
|
|
key->u.tkip.rx[queue].iv32 != iv32) {
|
2007-05-05 20:45:53 +02:00
|
|
|
/* IV16 wrapped around - perform TKIP phase 1 */
|
2008-06-11 23:22:00 +02:00
|
|
|
tkip_mixing_phase1(tk, &key->u.tkip.rx[queue], ta, iv32);
|
2009-12-09 23:25:05 +01:00
|
|
|
}
|
|
|
|
if (key->local->ops->update_tkip_key &&
|
|
|
|
key->flags & KEY_FLAG_UPLOADED_TO_HARDWARE &&
|
|
|
|
key->u.tkip.rx[queue].state != TKIP_STATE_PHASE1_HW_UPLOADED) {
|
2010-01-21 11:40:47 +01:00
|
|
|
struct ieee80211_sub_if_data *sdata = key->sdata;
|
2009-12-09 23:25:05 +01:00
|
|
|
|
2010-01-21 11:40:47 +01:00
|
|
|
if (sdata->vif.type == NL80211_IFTYPE_AP_VLAN)
|
|
|
|
sdata = container_of(key->sdata->bss,
|
|
|
|
struct ieee80211_sub_if_data, u.ap);
|
|
|
|
drv_update_tkip_key(key->local, sdata, &key->conf, key->sta,
|
2009-12-09 23:25:05 +01:00
|
|
|
iv32, key->u.tkip.rx[queue].p1k);
|
|
|
|
key->u.tkip.rx[queue].state = TKIP_STATE_PHASE1_HW_UPLOADED;
|
2007-05-05 20:45:53 +02:00
|
|
|
}
|
|
|
|
|
2008-06-11 23:22:00 +02:00
|
|
|
tkip_mixing_phase2(tk, &key->u.tkip.rx[queue], iv16, rc4key);
|
2007-05-05 20:45:53 +02:00
|
|
|
|
|
|
|
res = ieee80211_wep_decrypt_data(tfm, rc4key, 16, pos, payload_len - 12);
|
|
|
|
done:
|
|
|
|
if (res == TKIP_DECRYPT_OK) {
|
2007-09-26 15:19:45 +02:00
|
|
|
/*
|
|
|
|
* Record previously received IV, will be copied into the
|
|
|
|
* key information after MIC verification. It is possible
|
|
|
|
* that we don't catch replays of fragments but that's ok
|
|
|
|
* because the Michael MIC verication will then fail.
|
|
|
|
*/
|
|
|
|
*out_iv32 = iv32;
|
|
|
|
*out_iv16 = iv16;
|
2007-05-05 20:45:53 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
return res;
|
|
|
|
}
|