[RESOLVED] NoScript 2.0 and 2.0.1rc1 performance problems

[therube]

FWIW:

Workaround (Works for me):

In NoScript (2.0.1), I disabled ABE, as suggested in a posting on NoScript (thanks for the link) and the slowdowns stopped.
I did notice some periodic CPU spikes, but they appeared to be marginal and SeaMonkey wouldn't hold up anymore.

Where to disable ABE:
NoScript Options > Advanced tab > ABE subtab > [ ] "Enable ABE (Application Boundaries Enforcer)" — Remove checkbox.

I did not disable WAN IP protection in NoScript, but disabling ABE removed the slowdowns.
I do hope for a better and working solution with ABE re-enabled, but without that bad performance penalty.

http://forums.mozillazine.org/viewtopic ... 5#p9746885

[GµårÐïåñ]

@therube, thanks for that but unfortunately disabling ABE defeats a huge portion of the security offered by NoScript and I would much rather find a solution that does not involve just getting rid of the function. I mean if every time a security software caused such a problem, we disabled it, it defeats the point, no? Which brings me to my next point, if this in fact IS an ABE issue (which I don't know and needs Giorgio input) simply based on the solution posted, then is Giorgio aware of it? Is there a fix in the works? Are we even sure why and what's causing it to interfere like that? Just wondering.

[therube]

My reason for the "FWIW:".
The OP gave no real information as to what problem they had or how it was affecting him.
His "fix" worked for him, & as it could relate to this thread, so here it is.

And as I posted, "An unconfirmed, unresolved thread" (this one). So if ABE or something else is causing problems ... ?

[GµårÐïåñ]

My reason for the "FWIW:".
The OP gave no real information as to what problem they had or how it was affecting him.
His "fix" worked for him, & as it could relate to this thread, so here it is.

And as I posted, "An unconfirmed, unresolved thread" (this one). So if ABE or something else is causing problems ... ?

You misunderstood me, I wasn't criticizing and I understand full well what you did and why you did it and I am with you on that. I was simply stating that before we suggest to people to disable the security features maybe they could give us more information to figure out what's wrong and fix it, that's all After all if a person just simply sees that and decides to do that each time they hit a problem, then it defeats their security solution and they might be sore with us for a whole different set of reasons, that's all buddy, no harm no foul.

[aperson]

Hi... I can report that I'm having this problem as well. I disabled all add-ons and restarted Firefox... cpu is busy averaging 50-70%, but no hang-ups and usable. With only noscript re-enabled, the cpu pegs cyclically about every 12-18 seconds during which time the PC appears to hang or be stuck for a few seconds. Very annoying. I'm running Firefox 3.6.8, noscript 2.0.2.3, Windows XP SP3 on a Pentium 4 2.8 ghz 1GB RAM PC (yeh, I know, it's pretty old). I've relied on noscript for several years now as a security measure but will have to disable it until this bug is fixed as it's hampering my work. BTW, I think this problem just started sometime over the summer, so it shouldn't be too hard to track down what changed...??? Also, I'm having the same problem on my PC at home. Thanks... Art.

[therube]

As a test, create a new, clean Profile, install only NoScript.
Does the problem repeat?

[aperson]

As a test, create a new, clean Profile, install only NoScript.
Does the problem repeat?

Amazing... I made a fresh profile, installed only the noscript add-on, loaded up maybe 150 pages/tabs, and it's running smoothly... no hangups. Then I tried it again, another fresh profile with just noscript and opened all the pages I had previously opened in the old profile that was having problems, it's also running smoothly. Since then I've added several other add-ons I normally use and it's still running smoothly. So... what's with the profile that causes all this grief? I guess this would be a Firefox issue rather than an add-on issue... is there any easy way to transfer all open URL's from all open Firefox windows/tabs from an old profile to a new one? That was a real pain...

Thanks for the suggestion! Art

[dhouwn]

is there any easy way to transfer all open URL's from all open Firefox windows/tabs from an old profile to a new one?

Going File → Exit, saving all tabs, navigating to the old profile folder (in "%appdata%\Mozilla\Firefox\Profiles" under Windows) and copying "sessionstore.js" over to the other profile's folder should work (not tested though).

[aperson]

is there any easy way to transfer all open URL's from all open Firefox windows/tabs from an old profile to a new one?

Going File → Exit, saving all tabs, navigating to the old profile folder (in "%appdata%\Mozilla\Firefox\Profiles" under Windows) and copying "sessionstore.js" over to the other profile's folder should work (not tested though).

Tried this in a new profile... Firefox just deletes the file and doesn't use it.

Art

[therube]

Tried this in a new profile... Firefox just deletes the file and doesn't use it.

Actually it looks like it renames it (backs it up), leaves the original, & then doesn't use it.

Copying sessionstore.json would have been my idea too.

Appears though that if you place the following in prefs.js, it will restore.

Code: Select all

user_pref("browser.sessionstore.resume_session_once", true);


So ...

Exit
Copy sessionstore.json over
Edit (& or create) prefs.js
Insert user_pref("browser.sessionstore.resume_session_once", true); into same & save
Start FF

Don't know if this is a kosher way of doing it, but seems to work.

[aperson]

Tried this in a new profile... Firefox just deletes the file and doesn't use it.

Actually it looks like it renames it (backs it up), leaves the original, & then doesn't use it.

Copying sessionstore.json would have been my idea too.

Appears though that if you place the following in prefs.js, it will restore.

Code: Select all

user_pref("browser.sessionstore.resume_session_once", true);


So ...

Exit
Copy sessionstore.json over
Edit (& or create) prefs.js
Insert user_pref("browser.sessionstore.resume_session_once", true); into same & save
Start FF

Don't know if this is a kosher way of doing it, but seems to work.

Okay... I tried that and it works. That's good, because my browser degraded again... seemed to be hanging up on WOT with runaway js notices and 100% cpu usage plateaus... I'm going to start over again with just the browser and noscript and use it awhile and see how it goes. If no problems develop, I'll slowly start adding back in the other extensions. Thanks again for the help... Art.

[miroslav.policki]

Ok, I have a test case.
1. I used Firefox 3.6.14 (Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.2.14) Gecko/20110218 Firefox/3.6.14).
2. I created a new profile.
3. I disabled all extensions and plugins.
4. I installed NS 2.0.9.8.
5. I opened 50 tabs of cnn.com by executing the following in the address bar of the first tab (which had the Mozilla Firefox start page):

Code: Select all

javascript:for(i = 0; i < 50; i++, window.open('http://www.cnn.com/'));

6. When all the tabs loaded and FF processor usage dropped to 0 (as monitored in Process Explorer), I tried surfing in one of the tabs - clicking links, scrolling etc. (Almost) immediately FF processor usage went to max and FF was unusable. After several seconds it would start responding, but then it would become unresponsive again.
7. With NS disabled, there are no such problems. This is also the case with NS 1.10.
8. The behavior of 2.0.9.9rc4 is the same as 2.0.9.8.

The problematic behavior is not the same as in my original post, but there is definitely a significant performance impact, and it makes FF unusable even in the relatively modest use case of 50 instances of cnn.com. Can anyone confirm?

[Giorgio Maone]

No, I cannot reproduce. As soon as they're loaded, all the 50 tabs are perfectly responsive for me.

[al_9x]

Can anyone confirm?

Confirmed. Fx is effectively unusable due to regular prolonged cpu spikes (with NS), when browsing around in one of the 50 tabs. I've been noticing these spikes myself with many tabs, they were locking up scrolling briefly, but it was not that severe since I had considerably fewer tabs and I didn't pursue it.

Furthermore, it appears to take considerably longer to load the 50 tabs (from bookmarklet execution to cpu subsiding) with NS, than without. I'll time it and post the results.

1 hyperthreaded core (2 virtual cpus)

[therube]

First time through, cnn.com blocked, & so very possibly everything blocked.
Loaded speedily, no problems with tabs or opening links.

Second time, I Allowed Globally.

Took LOTS of time before finally finished loading.
Then Clicked one link, jumped to another tab, clicked on another. The second link seemed as if it is going to take a long time before it responds & so here I am typing this <in SeaMonkey, testing in FF>. Still in SeaMonkey, watching Process Hacker, I'm seeing (FF) CPU usage periodically jumping to 25% CPU (1 core), which I have to assume is related to Flash.

Clicking on some links & FF is struggling. I'm about to give up on it in a moment ...

FF definitely not responding appropriately. Going to kill it, or shut it down if I can catch it at the right moment.
It's in a loop. CPU, relaxes, CPU, relaxes. You have momentary responsiveness during relaxed state, but when CPU kicks in, not much you can do. (Can't even scroll in Error Console to view XSS information.)

I have seen these types of loops before (in SeaMonkey) & I know that once they start, there is no way to stop them. Closing much down, may extent slightly your browsing session, but soon you will say enough is enough & restart.

(Even though this is happening, nothing to say that it is NoScript related. Once I can get disabled, I'll restart without NoScript ...)

Now restart, NoScript (& it turns out FlashGot & Adblock Plus & BarTab too) disabled, I'm getting a much different, better experience. (I shut down & reopened from Session Restore.) Gone is the looping. I click a link, CPU usage goes up (as expected) as the page loads, then returns back down. Jump to another tab, repeat with the same results.

So looks like NoScript could be causing ...