WebAuthn objects should use [SameObject]

RESOLVED FIXED in Firefox 56

Status

()

P2
enhancement
RESOLVED FIXED
a year ago
a year ago

People

(Reporter: jcj, Assigned: jcj)

Tracking

(Blocks: 1 bug)

55 Branch
mozilla57
Points:
---
Dependency tree / graph

Firefox Tracking Flags

(firefox56 fixed)

Details

(Whiteboard: [webauthn] [webauthn-interop])

Attachments

(1 attachment)

(Assignee)

Description

a year ago
Most of the WebIDL objects for the current editor's draft of Web Authentication are annotated as [SameObject], which won't affect interop based on WD-05 [1]. We should go ahead and annotate ours that way.

[1] https://w3c.github.io/webauthn/#idl-index
Comment hidden (mozreview-request)
(Assignee)

Updated

a year ago
Status: NEW → ASSIGNED
(Assignee)

Comment 3

a year ago
Thanks for the review! Try runs previously looked good, so marking checkin-needed.
Keywords: checkin-needed

Comment 4

a year ago
Pushed by kmachulis@mozilla.com:
https://hg.mozilla.org/integration/autoland/rev/811510fdb51a
WebAuthn objects should use [SameObject] r=qdot
Keywords: checkin-needed

Comment 5

a year ago
bugherder
https://hg.mozilla.org/mozilla-central/rev/811510fdb51a
Status: ASSIGNED → RESOLVED
Last Resolved: a year ago
status-firefox56: --- → fixed
Resolution: --- → FIXED
Does the implementation actually return the same object all the time? For example, PublicKeyCredential.rawId seems to end up in CryptoBuffer::ToUint8Array, which returns a new object every time it's called. Maybe I'm missing something, where do we ensure that the same object is returned?
Flags: needinfo?(jjones)
(Assignee)

Comment 7

a year ago
I was assuming [SameObject] instructed the object to be cached at the JS-layer, but digging into it, I see it does not. I'm opening a follow-on that I'll tackle ASAP to make sure these are cached objects.

Thanks, :peterv.
Flags: needinfo?(jjones)
(Assignee)

Updated

a year ago
Blocks: 1384623
You need to log in before you can comment on or make changes to this bug.