Opened 5 years ago

Closed 5 years ago

#13186 closed defect (fixed)

performance.timing leak in ESR31

Reported by: arthuredelstein Owned by: gk
Priority: Medium Milestone:
Component: Applications/Tor Browser Version:
Severity: Keywords: ff31-esr, tbb-fingerprinting, TorBrowserTeam201409, tbb-testcase, MikePerry201409R, boklm201410R
Cc: boklm Actual Points:
Parent ID: Points:
Reviewer: Sponsor:

Description (last modified by arthuredelstein)

In TBB/ESR31, although we set dom.enable_preformance to false, I still see nonzero performance.timing properties: For example:

> performance.timing
< PerformanceTiming { navigationStart: 0, unloadEventStart: 0, 
unloadEventEnd: 0, redirectStart: 1411030797789, redirectEnd: 1411030797789, 
fetchStart: 1411030797789, domainLookupStart: 1411030797789, 
domainLookupEnd: 1411030797789, connectStart: 1411030797789, 
connectEnd: 1411030797789 }

I suppose this is a regression in Firefox.

Child Tickets

Change History (12)

comment:1 Changed 5 years ago by arthuredelstein

Another example:

PerformanceTiming {
 navigationStart: 1411031157263,
 unloadEventStart: 1411031159920,
 unloadEventEnd: 1411031159920,
 redirectStart: 1411031157264,
 redirectEnd: 1411031158179,
 fetchStart: 1411031158185,
 domainLookupStart: 1411031158185,
 domainLookupEnd: 1411031158185,
 connectStart: 1411031158185,
 connectEnd: 1411031158185
}

comment:2 Changed 5 years ago by arthuredelstein

Description: modified (diff)

comment:3 Changed 5 years ago by gk

Keywords: TorBrowserTeam201409 added
Owner: changed from tbb-team to gk
Status: newassigned

The regression is in https://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=5010b38abf18&tochange=c962bde5ac0b. I can look at it and try to come up with something... How did you get the values in comment:1?

Last edited 5 years ago by gk (previous) (diff)

comment:4 in reply to:  3 Changed 5 years ago by arthuredelstein

Replying to gk:

The regression is in https://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=5010b38abf18&tochange=c962bde5ac0b. I can look at it and try to come up with something...

Thanks!

How did you get the values in comment:1?

Now I don't seem to be able to get nonzero values for navigationStart, unloadEventStart, and unloadEventEnd unless I set dom.enable_performance to true. It's possible I mistakenly had that flag set to true in comment:1.

comment:5 Changed 5 years ago by gk

The values in the description are explained by https://hg.mozilla.org/mozilla-central/rev/a5679c8bc3fe. The changes (return 0 -> return mZeroTime) in nsPerformance.cpp are causing this. What bothers me is the example in comment:1 as it shows the behavior for dom.enable_performance = false looking at the last 5 values...

Last edited 5 years ago by gk (previous) (diff)

comment:6 in reply to:  5 Changed 5 years ago by arthuredelstein

Replying to gk:

The values in the description are explained by https://hg.mozilla.org/mozilla-central/rev/a5679c8bc3fe. The changes (return 0 -> return mZeroTime) in nsPerformance.cpp are causing this. What bothers me is the example in comment:1 as it shows the behavior for dom.enable_performance = false looking at the last 5 values...

Yeah, it's weird. Unfortunately I don't know how I did it! :P Anyhow, seems like it's probably best to set everything to 0?

comment:8 Changed 5 years ago by gk

Cc: boklm added
Keywords: tbb-testcase MikePerry201409R added
Status: assignedneeds_review

The test is in bug_13186 in my tor-browser-bundle-testsuite repo.

comment:9 Changed 5 years ago by mikeperry

Ok, I backported the Firefox patch side of this to the shiny new tor-browser-31.1.1esr-4.x-1 in tor-browser.git.

comment:10 Changed 5 years ago by gk

Keywords: boklm201410R added

comment:11 Changed 5 years ago by boklm

The branch bug_13186 on tor-browser-bundle-testsuite has been merged.

comment:12 Changed 5 years ago by gk

Resolution: fixed
Status: needs_reviewclosed

Closing then, thanks!

Note: See TracTickets for help on using tickets.