diff options
| author | Mathias Krause <minipli@googlemail.com> | 2014-12-30 22:50:54 +0100 |
|---|---|---|
| committer | Herbert Xu <herbert@gondor.apana.org.au> | 2015-01-05 21:35:02 +1100 |
| commit | 0b1e95b2fa0934c3a08db483979c70d3b287f50e (patch) | |
| tree | 3494012be5e6e4911c9069e1dcaccdf2203ab2f9 /arch/x86/crypto | |
| parent | 7e77bdebff5cb1e9876c561f69710b9ab8fa1f7e (diff) | |
| download | talos-obmc-linux-0b1e95b2fa0934c3a08db483979c70d3b287f50e.tar.gz talos-obmc-linux-0b1e95b2fa0934c3a08db483979c70d3b287f50e.zip | |
crypto: aesni - fix "by8" variant for 128 bit keys
The "by8" counter mode optimization is broken for 128 bit keys with
input data longer than 128 bytes. It uses the wrong key material for
en- and decryption.
The key registers xkey0, xkey4, xkey8 and xkey12 need to be preserved
in case we're handling more than 128 bytes of input data -- they won't
get reloaded after the initial load. They must therefore be (a) loaded
on the first iteration and (b) be preserved for the latter ones. The
implementation for 128 bit keys does not comply with (a) nor (b).
Fix this by bringing the implementation back to its original source
and correctly load the key registers and preserve their values by
*not* re-using the registers for other purposes.
Kudos to James for reporting the issue and providing a test case
showing the discrepancies.
Reported-by: James Yonan <james@openvpn.net>
Cc: Chandramouli Narayanan <mouli@linux.intel.com>
Cc: <stable@vger.kernel.org> # v3.18
Signed-off-by: Mathias Krause <minipli@googlemail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Diffstat (limited to 'arch/x86/crypto')
| -rw-r--r-- | arch/x86/crypto/aes_ctrby8_avx-x86_64.S | 46 |
1 files changed, 35 insertions, 11 deletions
diff --git a/arch/x86/crypto/aes_ctrby8_avx-x86_64.S b/arch/x86/crypto/aes_ctrby8_avx-x86_64.S index 2df2a0298f5a..a916c4a61165 100644 --- a/arch/x86/crypto/aes_ctrby8_avx-x86_64.S +++ b/arch/x86/crypto/aes_ctrby8_avx-x86_64.S @@ -208,7 +208,7 @@ ddq_add_8: .if (klen == KEY_128) .if (load_keys) - vmovdqa 3*16(p_keys), xkeyA + vmovdqa 3*16(p_keys), xkey4 .endif .else vmovdqa 3*16(p_keys), xkeyA @@ -224,7 +224,7 @@ ddq_add_8: add $(16*by), p_in .if (klen == KEY_128) - vmovdqa 4*16(p_keys), xkey4 + vmovdqa 4*16(p_keys), xkeyB .else .if (load_keys) vmovdqa 4*16(p_keys), xkey4 @@ -234,7 +234,12 @@ ddq_add_8: .set i, 0 .rept by club XDATA, i - vaesenc xkeyA, var_xdata, var_xdata /* key 3 */ + /* key 3 */ + .if (klen == KEY_128) + vaesenc xkey4, var_xdata, var_xdata + .else + vaesenc xkeyA, var_xdata, var_xdata + .endif .set i, (i +1) .endr @@ -243,13 +248,18 @@ ddq_add_8: .set i, 0 .rept by club XDATA, i - vaesenc xkey4, var_xdata, var_xdata /* key 4 */ + /* key 4 */ + .if (klen == KEY_128) + vaesenc xkeyB, var_xdata, var_xdata + .else + vaesenc xkey4, var_xdata, var_xdata + .endif .set i, (i +1) .endr .if (klen == KEY_128) .if (load_keys) - vmovdqa 6*16(p_keys), xkeyB + vmovdqa 6*16(p_keys), xkey8 .endif .else vmovdqa 6*16(p_keys), xkeyB @@ -267,12 +277,17 @@ ddq_add_8: .set i, 0 .rept by club XDATA, i - vaesenc xkeyB, var_xdata, var_xdata /* key 6 */ + /* key 6 */ + .if (klen == KEY_128) + vaesenc xkey8, var_xdata, var_xdata + .else + vaesenc xkeyB, var_xdata, var_xdata + .endif .set i, (i +1) .endr .if (klen == KEY_128) - vmovdqa 8*16(p_keys), xkey8 + vmovdqa 8*16(p_keys), xkeyB .else .if (load_keys) vmovdqa 8*16(p_keys), xkey8 @@ -288,7 +303,7 @@ ddq_add_8: .if (klen == KEY_128) .if (load_keys) - vmovdqa 9*16(p_keys), xkeyA + vmovdqa 9*16(p_keys), xkey12 .endif .else vmovdqa 9*16(p_keys), xkeyA @@ -297,7 +312,12 @@ ddq_add_8: .set i, 0 .rept by club XDATA, i - vaesenc xkey8, var_xdata, var_xdata /* key 8 */ + /* key 8 */ + .if (klen == KEY_128) + vaesenc xkeyB, var_xdata, var_xdata + .else + vaesenc xkey8, var_xdata, var_xdata + .endif .set i, (i +1) .endr @@ -306,7 +326,12 @@ ddq_add_8: .set i, 0 .rept by club XDATA, i - vaesenc xkeyA, var_xdata, var_xdata /* key 9 */ + /* key 9 */ + .if (klen == KEY_128) + vaesenc xkey12, var_xdata, var_xdata + .else + vaesenc xkeyA, var_xdata, var_xdata + .endif .set i, (i +1) .endr @@ -412,7 +437,6 @@ ddq_add_8: /* main body of aes ctr load */ .macro do_aes_ctrmain key_len - cmp $16, num_bytes jb .Ldo_return2\key_len |
