[tor-bugs] #18741 [Applications/Tor Browser]: OCSP and favicon isolation is only partly working in ESR 45
Tor Bug Tracker & Wiki
blackhole at torproject.org
Thu Apr 21 15:46:46 UTC 2016
#18741: OCSP and favicon isolation is only partly working in ESR 45
-------------------------------------------------+-------------------------
Reporter: gk | Owner: tbb-
Type: defect | team
Priority: High | Status:
Component: Applications/Tor Browser | needs_information
Severity: Major | Milestone:
Keywords: ff45-esr, tbb-6.0a5, | Version:
TorBrowserTeam201604R | Resolution:
Parent ID: | Actual Points:
Reviewer: | Points:
| Sponsor:
-------------------------------------------------+-------------------------
Changes (by mcs):
* status: needs_review => needs_information
Comment:
Replying to [comment:8 arthuredelstein]:
> OK, here's the new branch. I tested on Ubuntu and got all favicon and
OCSP requests running through the first party domain:
> https://github.com/arthuredelstein/tor-browser/commits/16326+3
> Note there are three commits here.
> * 483bd1684d437f0e03743b9573990240d77e8acc adds a fix for #16326
> * 4117c6b544e4fba93d192262aeabc0be4f38c4d7 fixes favicon cache and
network isolation
Can you explain why the above patch is needed? Why aren't we passing the
correct aNode in all cases? I am worried that we will poke around in the
ancestor elements looking for a "firstparty" attribute in a lot more cases
now, and I am not sure what the implications are (but I have not run the
code yet).
> * 8317e098f0b880eded1301fe40e3e9fd1b813fc3 adds network isolation
testing to our cache isolation regression test patch
It would have helped me if there was a comment inside the
observeChannels() callback that explained how the check worked, e.g.,
{{{
// All "thirdPartyChild" resources are loaded from example.net, so we
expect
// the first party host to be .com or .org.
}}}
--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/18741#comment:9>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online
More information about the tor-bugs
mailing list