[Webkit-unassigned] [Bug 159638] Improve test coverage for imported/w3c/WebCryptoAPI/

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=159638

--- Comment #10 from Jiewen Tan <jiewen_tan at apple.com> ---
Failing tests of encrypt_decrypt/test_aes_cbc.html are:
FAIL AES-CBC 128-bit key with mismatched key and algorithm assert_equals: Mismatch should cause InvalidAccessError instead of The operation is not supported. expected "InvalidAccessError" but got "NotSupportedError"
FAIL AES-CBC 192-bit key with mismatched key and algorithm assert_equals: Mismatch should cause InvalidAccessError instead of The operation is not supported. expected "InvalidAccessError" but got "NotSupportedError"
FAIL AES-CBC 256-bit key with mismatched key and algorithm assert_equals: Mismatch should cause InvalidAccessError instead of The operation is not supported. expected "InvalidAccessError" but got "NotSupportedError"

FAIL AES-CBC 128-bit key, zeroPadChar assert_unreached: should have thrown exception for test AES-CBC 128-bit key, zeroPadChar Reached unreachable code
FAIL AES-CBC 128-bit key, bigPadChar assert_unreached: should have thrown exception for test AES-CBC 128-bit key, bigPadChar Reached unreachable code
FAIL AES-CBC 128-bit key, inconsistentPadChars assert_unreached: should have thrown exception for test AES-CBC 128-bit key, inconsistentPadChars Reached unreachable code
FAIL AES-CBC 192-bit key, zeroPadChar assert_unreached: should have thrown exception for test AES-CBC 192-bit key, zeroPadChar Reached unreachable code
FAIL AES-CBC 192-bit key, bigPadChar assert_unreached: should have thrown exception for test AES-CBC 192-bit key, bigPadChar Reached unreachable code
FAIL AES-CBC 192-bit key, inconsistentPadChars assert_unreached: should have thrown exception for test AES-CBC 192-bit key, inconsistentPadChars Reached unreachable code
FAIL AES-CBC 256-bit key, zeroPadChar assert_unreached: should have thrown exception for test AES-CBC 256-bit key, zeroPadChar Reached unreachable code
FAIL AES-CBC 256-bit key, bigPadChar assert_unreached: should have thrown exception for test AES-CBC 256-bit key, bigPadChar Reached unreachable code
FAIL AES-CBC 256-bit key, inconsistentPadChars assert_unreached: should have thrown exception for test AES-CBC 256-bit key, inconsistentPadChars Reached unreachable code

For the first group, the test case use "AES-GCM" as the mismatched algorithm name, which we haven't support yet. Therefore, we return NotSupportedError instead of InvalidAccessError.

For the second group, our underlying SecLibrary silently bypass the failures. Actually I don't think there is a failure in our SecLibrary as adding a wrong padding to the plain text will just result in another new plain text. Since we hand over all the padding jobs to the underlying SecLibrary, we should not follow the spec to remove bad padding. Otherwise, it will produce corrupted decrypted text. Mark as NTF.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20161209/13e7439b/attachment-0001.html>