1945596 - Firefox should be still sending headers even if there are missing ones

Status: RESOLVED FIXED

(Core :: Networking: HTTP, defect, P2)

Description

[Sean Kim]

We should avoid returning from the delegate even if there are missing headers

As discussed with Microsoft earlier, the correct approach for handling cookie headers is for the browser to include in its requests and allow the server to deal with them. Previously, we just returned from the delegate if we have missing headers. We should simply let the server deal with it instead of handling it ourselves.

Comment 1

[Sean Kim]

Attachment: Bug 1945596 - Firefox should be still send headers even if there are missing ones r=mkaply

Comment 3

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/d5df64ac441f
Firefox should be still send headers even if there are missing ones r=mkaply,necko-reviewers,jesup

Comment 4

[Sean Kim]

Attachment: Bug 1945596 - Firefox should be still sending headers even if there are missing ones

Original Revision: https://phabricator.services.mozilla.com/D236568

Comment 5

[Phabricator Automation]

beta Uplift Approval Request

• User impact if declined: When Microsoft's broker sends a cookie without device header, the SSO would not complete (on macOS).
• Code covered by automated testing: no
• Fix verified in Nightly: yes
• Needs manual QE test: no
• Steps to reproduce for manual QE testing: N/A
• Risk associated with taking this patch: Minimal
• Explanation of risk level: We are simply appending the headers instead of quickly returning (when there is less than two headers attached. Tested via a local signed Nightly build.
• String changes made/needed: N/A
• Is Android affected?: no

Comment 6

[Cristina Horotan [:chorotan]]

https://hg.mozilla.org/mozilla-central/rev/d5df64ac441f

Comment 7

[Pulsebot]

https://hg.mozilla.org/releases/mozilla-beta/rev/1dc6ea91499c

Comment 8

[Mike Kaply [:mkaply]]

We'll need to uplift to ESR as well for parity.

Comment 9

[Donal Meehan [:dmeehan]]

(In reply to Mike Kaply [:mkaply] from comment #8)

We'll need to uplift to ESR as well for parity.

Adding an NI for Sean re: Comment 8.
Note, the depends on the Bug 1944119 is only in Fx136+

Comment 10

[Sean Kim]

(In reply to Donal Meehan [:dmeehan] from comment #9)

(In reply to Mike Kaply [:mkaply] from comment #8)

We'll need to uplift to ESR as well for parity.

Adding an NI for Sean re: Comment 8.
Note, the depends on the Bug 1944119 is only in Fx136+

Considering that Bug 1944119 is a low risk refactor, we can probably uplift both.

Comment 11

[Sean Kim]

Attachment: Bug 1945596 - Firefox should be still sending headers even if there are missing ones

Original Revision: https://phabricator.services.mozilla.com/D236568

Comment 12

[Phabricator Automation]

esr128 Uplift Approval Request

• User impact if declined: When Microsoft's broker sends a cookie without device header, the SSO would not complete (on macOS).
• Code covered by automated testing: no
• Fix verified in Nightly: yes
• Needs manual QE test: no
• Steps to reproduce for manual QE testing: N/A
• Risk associated with taking this patch: Low
• Explanation of risk level: We are simply appending the headers instead of quickly returning (when there is less than two headers attached. Tested via a local signed Nightly build.
• String changes made/needed: N/A
• Is Android affected?: no

Comment 13

[Pulsebot]

https://hg.mozilla.org/releases/mozilla-esr128/rev/023ef18b4571