Currently the spec calls for 2 error pages: one when the TRR provider is unavailable and one when the provider tried and couldn't resolve the hostname. The former would offer a button that temporarily completely disables DoH, and the latter would offer a button to whitelist that domain.

This is currently not possible to implement since we don't have any way to distinguish between the two scenarios. Pending necko enhancements to support this, here are a few ideas that Bryan and I are considering for a single error page:

1. Offer both buttons on the same error page.
2. Offer a button to reload the page with DoH disabled.
3. Option 2, but also, offer a button to temporarily completely disable DoH after this happens N times.

Bugbug thinks this bug should belong to this component, but please revert this change in case of error.

The product::component has been changed since the backlog priority was decided, so we're resetting it.
For more information, please visit auto_nag documentation.

No, this is not a necko bug, this belongs in Firefox. Maybe I'll request a DoH component for tracking frontend bugs and then we can move it over from General.

Depends on D164345

Depends on D164346

Depends on D164347

Previosuly we'd only pass the TRRService::ProviderKey() into the content
process. But not we need the full domain for the error page in the content
process, so we now pass the full domain. The ChildDNSService now holds on
to the full domain, but calls into TRRService to update the key for
telemetry and returns that when necessary.

Depends on D164348

Depends on D164347

This allows us to use a consistent size for the dnsFlags field.
across different files (previously some would use uint16_t and some uint32_t).
It also improves type safety - making sure we don't pass in the wrong value
to DNSFlags.

Depends on D164856

We encountered data races in unit tests.

Depends on D164346

Backed out for causing mochitest failures in browser/base/content/test/about/browser_aboutCertError_telemetry.js

Backout link: https://hg.mozilla.org/integration/autoland/rev/a5e4362f37d6b2acbe72a62518855e80370444b1

Push with failures

Failure log

INFO - Buffered messages finished
[task 2022-12-22T22:14:34.613Z] 22:14:34     INFO - TEST-UNEXPECTED-FAIL | browser/base/content/test/about/browser_aboutCertError_telemetry.js | Uncaught exception in test - at chrome://global/content/elements/browser-custom-element.js:708 - TypeError: can't access property "userTyped", this.urlbarChangeTracker is undefined
[task 2022-12-22T22:14:34.613Z] 22:14:34     INFO - Stack trace:

Fix incoming, small change.

Backed out 9 changesets (Bug 1596845) for causing xpcshell failures on test_trr_enterprise_policy.js.
Backout link
Push with failures
Failure Log

(In reply to Marian-Vasile Laza from comment #18)

Backed out 9 changesets (Bug 1596845) for causing xpcshell failures on test_trr_enterprise_policy.js.
Backout link
Push with failures
Failure Log

Not sure why this test still failed with socket process. However, we can disable this test for socket process for now and fix it later.

https://hg.mozilla.org/mozilla-central/rev/726e877bdd46
https://hg.mozilla.org/mozilla-central/rev/13d383bde6cf
https://hg.mozilla.org/mozilla-central/rev/0c74e206883c
https://hg.mozilla.org/mozilla-central/rev/30f730eba0c4
https://hg.mozilla.org/mozilla-central/rev/c4400bef7b19
https://hg.mozilla.org/mozilla-central/rev/ac391762ae01
https://hg.mozilla.org/mozilla-central/rev/236b2bde7397
https://hg.mozilla.org/mozilla-central/rev/c1d4e09eca69
https://hg.mozilla.org/mozilla-central/rev/c9c9844f1f2f