Implement PBKDF2 in WebCrypto

• Component: DOM
• Status: RESOLVED FIXED

The Password Based Key Derivation Function 2.0 (PBKDF2) is crucial for sites that wish to use a password to derive a key versus generating one for a user:


https://dvcs.w3.org/hg/webcrypto-api/raw-file/tip/spec/Overview.html#pbkdf2

[Yingle]

WebKit is lagging on web API support, this is one example. Can someone work on this?

+1

[Radar WebKit Bug Importer]

<rdar://problem/27311703>

[Jiewen Tan]

Created attachment 303906 [details]
Patch

[Jiewen Tan]

90% of the patch is w3c tests rebaseline.

[WebKit Commit Bot]

Attachment 303906 [details] did not pass style-queue:


ERROR: Source/WebCore/crypto/CryptoKey.h:48:  enum members should use InterCaps with an initial capital letter or initial 'k' for C-style enums.  [readability/enum_casing] [4]
Total errors found: 1 in 49 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Build Bot]

Comment on attachment 303906 [details]
Patch

Attachment 303906 [details] did not pass mac-ews (mac):
Output: http://webkit-queues.webkit.org/results/3276047

New failing tests:
crypto/subtle/derive-key-malformed-parameters.html

[Build Bot]

Created attachment 303910 [details]
Archive of layout-test-results from ews103 for mac-elcapitan

The attached test failures were seen while running run-webkit-tests on the mac-ews.
Bot: ews103  Port: mac-elcapitan  Platform: Mac OS X 10.11.6

[Build Bot]

Comment on attachment 303906 [details]
Patch

Attachment 303906 [details] did not pass mac-wk2-ews (mac-wk2):
Output: http://webkit-queues.webkit.org/results/3276078

New failing tests:
crypto/subtle/derive-key-malformed-parameters.html

[Build Bot]

Created attachment 303912 [details]
Archive of layout-test-results from ews105 for mac-elcapitan-wk2

The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews.
Bot: ews105  Port: mac-elcapitan-wk2  Platform: Mac OS X 10.11.6

[Build Bot]

Comment on attachment 303906 [details]
Patch

Attachment 303906 [details] did not pass mac-debug-ews (mac):
Output: http://webkit-queues.webkit.org/results/3276051

New failing tests:
crypto/subtle/derive-key-malformed-parameters.html

[Build Bot]

Created attachment 303913 [details]
Archive of layout-test-results from ews113 for mac-elcapitan

The attached test failures were seen while running run-webkit-tests on the mac-debug-ews.
Bot: ews113  Port: mac-elcapitan  Platform: Mac OS X 10.11.6

[Build Bot]

Comment on attachment 303906 [details]
Patch

Attachment 303906 [details] did not pass ios-sim-ews (ios-simulator-wk2):
Output: http://webkit-queues.webkit.org/results/3276084

New failing tests:
crypto/subtle/derive-key-malformed-parameters.html

[Build Bot]

Created attachment 303915 [details]
Archive of layout-test-results from ews126 for ios-simulator-wk2

The attached test failures were seen while running run-webkit-tests on the ios-sim-ews.
Bot: ews126  Port: ios-simulator-wk2  Platform: Mac OS X 10.11.6

[Brent Fulgham]

Test failure on Mac (64-bit):

crypto/subtle/derive-key-malformed-parameters.html

[Brent Fulgham]

Some kind of whitespace issue:

--- /Volumes/Data/EWS/WebKit/WebKitBuild/Release/layout-test-results/crypto/subtle/derive-key-malformed-parameters-expected.txt
+++ /Volumes/Data/EWS/WebKit/WebKitBuild/Release/layout-test-results/crypto/subtle/derive-key-malformed-parameters-actual.txt
@@ -17,7 +17,7 @@
 PASS crypto.subtle.deriveKey({ name:"ECDH", public:publicKey }, privateKey, {name: "aes-cbc", length: 1}, extractable, ["encrypt"]) rejected promise  with OperationError (DOM Exception 34): Cannot get key length from derivedKeyType.
 PASS crypto.subtle.deriveKey({ name:"ECDH", public:publicKey }, privateKey, {name: "hmac", hash: "hmac"}, extractable, ["sign"]) rejected promise  with NotSupportedError (DOM Exception 9): The operation is not supported..
 PASS crypto.subtle.deriveKey({ name:"ECDH", public:publicKey }, privateKey, {name: "hmac", hash: "sha-1", length: 0}, extractable, ["sign"]) rejected promise  with TypeError: Cannot get key length from derivedKeyType.
-PASS crypto.subtle.deriveKey({ name:"ECDH", public:publicKey }, privateKey, {name: "aes-cbc", length: 128}, extractable, []) rejected promise  with SyntaxError (DOM Exception 12): A required parameter was missing or out-of-range.
+PASS crypto.subtle.deriveKey({ name:"ECDH", public:publicKey }, privateKey, {name: "aes-cbc", length: 128}, extractable, [ ]) rejected promise  with SyntaxError (DOM Exception 12): A required parameter was missing or out-of-range.
 PASS successfullyParsed is true
 
 TEST COMPLETE

[Brent Fulgham]

Comment on attachment 303906 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=303906&action=review

Looks great! You have a whitespace error in your tests (which I think I noted below). r=me, but please fix the test.

> LayoutTests/crypto/subtle/derive-key-malformed-parameters.html:71
> +    return shouldReject('crypto.subtle.deriveKey({ name:"ECDH", public:publicKey }, privateKey, {name: "aes-cbc", length: 128}, extractable, [ ])');

This is either causing your test failure, or you need to add this space somewhere else to match the test output.

> LayoutTests/imported/w3c/web-platform-tests/WebCryptoAPI/derive_bits_keys/pbkdf2.worker-expected.txt:3
> +PASS Derived key of type name: AES-CBC length: 128  using short password, short salt, SHA-384, with 1 iterations 

Yay!

> LayoutTests/imported/w3c/web-platform-tests/WebCryptoAPI/derive_bits_keys/pbkdf2.worker-expected.txt:35
> +PASS Derived key of type name: AES-GCM length: 256  using short password, short salt, SHA-384, with 1 iterations 

Hooray! So many new passes!

> LayoutTests/imported/w3c/web-platform-tests/WebCryptoAPI/derive_bits_keys/pbkdf2.worker-expected.txt:8689
> +PASS Derived key of type name: HMAC hash: SHA-512 length: 256  using empty password, empty salt, SHA-256, with 0 iterations 

That's a lot of new passes!

> LayoutTests/imported/w3c/web-platform-tests/WebCryptoAPI/derive_bits_keys/test_pbkdf2-expected.txt:8691
> +PASS Derived key of type name: HMAC hash: SHA-512 length: 256  using empty password, empty salt, SHA-256, with 0 iterations 

I'm tired after scanning all of that!

[Alexey Proskuryakov]

One general comment.

> +        * crypto/mac/CryptoAlgorithmPBKDF2Mac.cpp: Added.

Mac is not a good directory or name for these, as they are in use on iOS too. I think that CoreCrypto would be best for most, and maybe Darwin for others?

[Brent Fulgham]

(In reply to comment #18)
> One general comment.
> 
> > +        * crypto/mac/CryptoAlgorithmPBKDF2Mac.cpp: Added.
> 
> Mac is not a good directory or name for these, as they are in use on iOS
> too. I think that CoreCrypto would be best for most, and maybe Darwin for
> others?

That's a good point. Jiewen, can you file a radar to do that reorganization? We should tackle that as clean-up in a few weeks.

[Jiewen Tan]

(In reply to comment #18)
> One general comment.
> 
> > +        * crypto/mac/CryptoAlgorithmPBKDF2Mac.cpp: Added.
> 
> Mac is not a good directory or name for these, as they are in use on iOS
> too. I think that CoreCrypto would be best for most, and maybe Darwin for
> others?

Filed: Bug 169432.

[Jiewen Tan]

Created attachment 303965 [details]
Patch for landing

[WebKit Commit Bot]

Attachment 303965 [details] did not pass style-queue:


ERROR: Source/WebCore/crypto/CryptoKey.h:48:  enum members should use InterCaps with an initial capital letter or initial 'k' for C-style enums.  [readability/enum_casing] [4]
Total errors found: 1 in 50 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Jiewen Tan]

Committed r213671: <http://trac.webkit.org/changeset/213671>

[Javier Fernandez]

This patch caused several tests to fail in GTK. 
see bug #169468 for details.