WebCrypto: Add ECDSA algorithm (Blink side).

WebCrypto: Add ECDSA algorithm (Blink side).

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

This changelist
 * Defines a new key type for EC keys
 * Implements structured clone for EC keys
 * Adds algorithm parameter parsing for WebCrypto's EcdsaParams, EcKeyGenParams and EcKeyImportParams

This change does not include any LayoutTests, as the tests depend on the Chromium side landing first.

The LayoutTests are part of: https://codereview.chromium.org/707753002/ and will be landed last.

BUG=399094
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=185319

[eroman, 2014-11-06 04:42:33]

eroman@chromium.org changed reviewers:
+ abarth@chromium.org

[abarth-chromium, 2014-11-11 05:52:09]

Would you be willing to find another reviewer for these CLs?  I'm not reviewing
much code in Blink anymore and I'm afraid my knowledge will become outdated...

[eroman, 2014-11-11 18:06:42]

Absolutely! Thanks for all of your past reviews!

[eroman, 2014-11-11 18:34:41]

eroman@chromium.org changed reviewers:
- abarth@chromium.org

[eroman, 2014-11-11 18:43:38]

eroman@chromium.org changed reviewers:
+ jww@chromium.org

[eroman, 2014-11-11 18:43:38]

Hi Joel, would you be able to review this?

Adam Barth has moved off of Blink, and you have experience with WebCrypto and
crypto in general so I thought you would make a good primary reviewer for
blink-side of WebCrypto. (Review load should be small for the Blink side).

Thanks!

[jww, 2014-11-12 08:10:27]

On 2014/11/11 18:43:38, eroman wrote:
> Hi Joel, would you be able to review this?
> 
> Adam Barth has moved off of Blink, and you have experience with WebCrypto and
> crypto in general so I thought you would make a good primary reviewer for
> blink-side of WebCrypto. (Review load should be small for the Blink side).
> 
Sure, sgtm.
> Thanks!

[jww, 2014-11-12 08:10:54]

This lgtm, modulo one question about a comment. I feel like you may have written
similar CLs in the past... :-)

https://codereview.chromium.org/707743002/diff/20001/Source/bindings/core/v8/...
File Source/bindings/core/v8/SerializedScriptValue.cpp (right):

https://codereview.chromium.org/707743002/diff/20001/Source/bindings/core/v8/...
Source/bindings/core/v8/SerializedScriptValue.cpp:290: // Maximum allowed value
is 2^32-1
Okay, I'm feeling really ignorant here, but what does this comment mean in
relation to this enum? It may be obvious, and I just don't know Ecdsa, but it's
not immediately clear to me what this comment means.

[eroman, 2014-11-12 19:11:13]

https://codereview.chromium.org/707743002/diff/20001/Source/bindings/core/v8/...
File Source/bindings/core/v8/SerializedScriptValue.cpp (right):

https://codereview.chromium.org/707743002/diff/20001/Source/bindings/core/v8/...
Source/bindings/core/v8/SerializedScriptValue.cpp:290: // Maximum allowed value
is 2^32-1
On 2014/11/12 08:10:54, jww wrote:
> Okay, I'm feeling really ignorant here, but what does this comment mean in
> relation to this enum? It may be obvious, and I just don't know Ecdsa, but
it's
> not immediately clear to me what this comment means.

Perhaps the comment is not necessary.

My intent is to indicate that no value in the enum should be larger than 2^32-1

My thinking is as follows:

C++ enum (without using features from C++11) does not specify what integral type
is used.

However when serialized these values are being converted to a uint32_t, so it
must safely fit inside of that. Maybe this isn't possible though, since the C++
spec says no enum value should be larger than an int. But what if ints are not
32 bits. Maybe I am just paranoid.

[eroman, 2014-11-12 19:15:11]

eroman@chromium.org changed reviewers:
+ jochen@chromium.org

[eroman, 2014-11-12 19:15:12]

@jochen: Could you please review public/platform/* and
SerializedScriptValue.cpp? Thanks!

[haraken, 2014-11-12 23:28:18]

On 2014/11/12 19:15:12, eroman wrote:
> @jochen: Could you please review public/platform/* and
> SerializedScriptValue.cpp? Thanks!

SerializedScriptValue.cpp LGTM.

[jww, 2014-11-13 06:15:09]

https://codereview.chromium.org/707743002/diff/20001/Source/bindings/core/v8/...
File Source/bindings/core/v8/SerializedScriptValue.cpp (right):

https://codereview.chromium.org/707743002/diff/20001/Source/bindings/core/v8/...
Source/bindings/core/v8/SerializedScriptValue.cpp:290: // Maximum allowed value
is 2^32-1
On 2014/11/12 19:11:13, eroman wrote:
> On 2014/11/12 08:10:54, jww wrote:
> > Okay, I'm feeling really ignorant here, but what does this comment mean in
> > relation to this enum? It may be obvious, and I just don't know Ecdsa, but
> it's
> > not immediately clear to me what this comment means.
> 
> Perhaps the comment is not necessary.
> 
> My intent is to indicate that no value in the enum should be larger than
2^32-1
> 
> My thinking is as follows:
> 
> C++ enum (without using features from C++11) does not specify what integral
type
> is used.
> 
> However when serialized these values are being converted to a uint32_t, so it
> must safely fit inside of that. Maybe this isn't possible though, since the
C++
> spec says no enum value should be larger than an int. But what if ints are not
> 32 bits. Maybe I am just paranoid.

Got it. I actually didn't see that the other enums in this file have that
comment as well. Given that this is a universal issue in Blink, and from what I
can tell, enums elsewhere in Blink don't have similar comments, I'd say drop
this comment (in fact, I'd say drop them all from this file). It just seems a
bit confusing to me, and we need to be aware across all of Blink anyway.

[jochen (gone - plz use gerrit), 2014-11-13 15:16:09]

jochen@chromium.org changed reviewers:
+ haraken@chromium.org

[jochen (gone - plz use gerrit), 2014-11-13 15:16:10]

lgtm

[eroman, 2014-11-13 19:55:00]

The CQ bit was checked by eroman@chromium.org

[commit-bot: I haz the power, 2014-11-13 19:56:16]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/707743002/60001

[commit-bot: I haz the power, 2014-11-13 21:00:26]

Committed patchset #4 (id:60001) as 185319