1930501 - Implement bfcache clearing for `Clear-Site-Data: cache`-header

Status: RESOLVED FIXED

(Toolkit :: Data Sanitization, enhancement, P2)

Comment 1

[Manuel Bucher [:manuel]]

Attachment: Bug 1930501 - Implement bfcache clearing for `Clear-Site-Data: "cache"`-header r=#anti-tracking,smaug

Comment 2

[Phabricator Automation]

Comment on attachment 9474811 [details]
Bug 1930501 - Implement bfcache clearing for Clear-Site-Data: "cache"-header r=#anti-tracking,smaug

Revision D243176 was moved to bug 1967474. Setting attachment 9474811 [details] to obsolete.

Comment 3

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/54690808a80d
Implement bfcache clearing for `Clear-Site-Data: "cache"`-header r=smaug,timhuang

Comment 4

[Manuel Bucher [:manuel]]

With this change we'll be clearing the bfcache with the clear-site-data: cache directive. Might be good to add the cache to the list in https://developer.mozilla.org/en-US/docs/Web/HTTP/Reference/Headers/Clear-Site-Data#cache

Also this patch somewhat introduces partitioning for bfcache due to exposing it to web content somehow. It could potentially be added to https://developer.mozilla.org/en-US/docs/Web/Privacy/Guides/State_Partitioning#network_partitioning

Comment 5

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Created web-platform-tests PR https://github.com/web-platform-tests/wpt/pull/52831 for changes under testing/web-platform/tests

Comment 6

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/fb6df3918f14
Revert "Bug 1930501 - Implement bfcache clearing for `Clear-Site-Data: "cache"`-header r=smaug,timhuang" for causing bc failures on browser_cookiePermission.js.

Comment 7

[Iulian Moraru]

Revert for causing bc failures on browser_cookiePermission.js. and for causing multiple failures.

• Push with failures

• Failure log example 1

• Failure log example 2

• Failure log example 3

• Backout link

Comment 8

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Upstream PR merged by moz-wptsync-bot

Comment 9

[Pulsebot]

Pushed by [e-mail]:
https://github.com/mozilla-firefox/firefox/commit/0203d8d57228
https://hg.mozilla.org/integration/autoland/rev/9cfd289176fd
Implement bfcache clearing for `Clear-Site-Data: "cache"`-header r=smaug,timhuang

Comment 10

[Pulsebot]

Pushed by [e-mail]:
https://github.com/mozilla-firefox/firefox/commit/9e3fbae366fa
https://hg.mozilla.org/integration/autoland/rev/625988823050
Revert "Bug 1930501 - Implement bfcache clearing for `Clear-Site-Data: "cache"`-header r=smaug,timhuang" for causing BfCached related wpt failures.

Comment 11

[Cosmin Sabou [:CosminS]]

Backed out for causing BfCached related wpt failures.

Push with failures

Failure log

Backout link

Comment 12

[Manuel Bucher [:manuel]]

Bfcache cleaning is only implemented when ship (session history in parent) is enabled. I need to disable / accept failure other CI configurations (nofis / noship).

Comment 13

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Created web-platform-tests PR https://github.com/web-platform-tests/wpt/pull/53126 for changes under testing/web-platform/tests

Comment 14

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Upstream PR merged by moz-wptsync-bot

Comment 15

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Created web-platform-tests PR https://github.com/web-platform-tests/wpt/pull/53127 for changes under testing/web-platform/tests

Comment 16

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Upstream PR merged by moz-wptsync-bot

Comment 17

[Pulsebot]

Pushed by [e-mail]:
https://github.com/mozilla-firefox/firefox/commit/ccd10b84c405
https://hg.mozilla.org/integration/autoland/rev/fa43957d8b0b
Implement bfcache clearing for `Clear-Site-Data: "cache"`-header r=smaug,timhuang

Comment 18

[amarc]

https://hg.mozilla.org/mozilla-central/rev/fa43957d8b0b

Comment 19

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Created web-platform-tests PR https://github.com/web-platform-tests/wpt/pull/53153 for changes under testing/web-platform/tests

Comment 20

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Upstream PR merged by moz-wptsync-bot

Comment 21

[Manuel Bucher [:manuel]]

Release Note Request (optional, but appreciated)
[Why is this notable]: This is noteworthy for two reasons:

• Less potential privacy leaks on websites that use this header. E.g. If a bank sends this header on logout, the account balance won't be visible when using the back button after logout.
• This exposes the bfcache programmatically to websites and therefore introduced a partitioning concept for it

Potentially too technical and more approriate for Firefox release notes for developers.
[Affects Firefox for Android]: Yes
[Suggested wording]: Firefox will now also clear the back-forward-cache when receiving the Clear-Site-Data: "cache" response header.
[Links (documentation, blog post, etc)]:

• https://developer.mozilla.org/en-US/docs/Web/HTTP/Reference/Headers/Clear-Site-Data#cache
• https://w3c.github.io/webappsec-clear-site-data/#grammardef-cache

Comment 22

[Pascal Chevrel:pascalc]

Note added to the Developer section of our Nightly 141 release notes. I am keeping the relnote flag open until we ship it to release, thanks.

Comment 23

[Hamish Willee]

FF141 docs work for this can be tracked in https://github.com/mdn/content/issues/40020

1. Currently this is just a release note update and adding mention of the bfcache to the "cache" directive - see: https://github.com/mdn/content/pull/40033

2. I didn't update compatibility data because as far as I can tell the interface has no expectation that a particular cache will be cleared, and we don't track any of the other kinds that might be (see compatibility data https://developer.mozilla.org/en-US/docs/Web/HTTP/Reference/Headers/Clear-Site-Data).

3. YOu say:

   This exposes the bfcache programmatically to websites and therefore introduced a partitioning concept for it

   What difference does this make to a website (the partitioning concept)? I mean "what would/could a website do differently by knowing about this".
   I'm trying to work out whether it is useful for website developers to mention the partitioning thing in the developer release note.

Comment 24

[Manuel Bucher [:manuel]]

(In reply to Hamish Willee from comment #23)

What difference does this make to a website (the partitioning concept)? I mean "what would/could a website do differently by knowing about this".
I'm trying to work out whether it is useful for website developers to mention the partitioning thing in the developer release note.

I mean non really, in the end. But the use-cases for this API are slim anyway. If you are embedded by a third party with an iframe, your bfcache (and other caches) are unaffected in first party (=unpartitioned) context.

So from a website perspective: if you have an enpoint like https://example.com/clear-all that always returns the header, your cache (including bfcache) would stay unaffected if this was embedded as an iframe from https://thirdparty.com for whatever reason.

The partitioning is mostly there to isolate sites from each other to prevent cross-site tracking users. The bfcache would be really hard to use (potentially impossible) for cross-site tracking. However, by also partitioning the bfcache, it is guaranteed to not introduce another cross-site tracking mechanism.

IMHO it the partitioning part isn't developer release note worthy, the bfcache part might.

Comment 25

[Pascal Chevrel:pascalc]

Added to our 141 final release notes.