[Bug 20340] CryptImportKey fails chromium's base_unittests.exe in HMACTest.*

[wine-bugs at winehq.org]

http://bugs.winehq.org/show_bug.cgi?id=20340





--- Comment #7 from Juan Lang <juan_lang at yahoo.com>  2009-10-21 17:16:01 ---
Here's a little more from a +crypt log:
trace:crypt:CryptImportKey (0x1528c8, 0xca8270, 28, 0x0, 00000100, 0xca67d8)
trace:crypt:RSAENH_CPImportKey (hProv=00000001, pbData=0xca8270, dwDataLen=28,
hPubKey=00000000, dwFlags=00000100, phKey=0x152abc)
trace:crypt:import_key blob type: 8
trace:crypt:new_key alg = "RC2", dwKeyLen = 128
trace:crypt:CryptCreateHash (0x1528c8, 0x8009, 0x152ab8, 00000000, 0xca67d4)

x8009 is CALG_HMAC in CryptCreateHash.

The blob type is 8, which is PLAINTEXTKEYBLOB.  This code in rsaenh.c's
import_plaintext_key is incorrect when CRYPT_IPSEC_HMAC_KEY is passed:
    memcpy(pCryptKey->abKeyValue, pbKeyStream, *pKeyLen);

The reason is abKeyValue is declared as:
    BYTE        abKeyValue[RSAENH_MAX_KEY_SIZE];

and RSAENH_MAX_KEY_SIZE is:
#define RSAENH_MAX_KEY_SIZE        48

This is fine for the current key length of 128 bits, but not for the largest
the app passes (640 = 80 bytes).  Testing on Windows showed that there is no
key length limit when CRYPT_IPSEC_HMAC_KEY is passed, so I think
CRYPT_IPSEC_HMAC_KEY implies the data are passed to an HMAC stream, rather than
imported directly.

-- 
Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email
Do not reply to this email, post in Bugzilla using the
above URL to reply.
------- You are receiving this mail because: -------
You are watching all bug changes.