NoScript versions 2.6.8.7 and newer on Firefox 3.6 result in Firefox's memory usage quickly rising to two times normal levels after a bit of browsing and having 15-30 tabs open. Every so often, maybe every 0.5 to 2 minutes, Firefox will freeze for a short moment. Freezes seem to take longer and appear more often when Firefox uses more memory. Firefox's CPU usage jumps to 100% during freezes. Even though I can't quantify it, newer versions of NS seem to be getting worse with every release.
I first noticed the problem with NS 2.6.8.7 and I managed to hold out for a few days before downgrading. With NS 2.6.8.12, the freezes annoyed me too much after maybe 10 minutes.
Downgrading to NS 2.6.8.6 always brings the memory usage back to normal levels and the freezes are gone.
Performance regression: NS 2.6.8.7+ hogs memory on Fx 3.6
Performance regression: NS 2.6.8.7+ hogs memory on Fx 3.6
Mozilla/5.0 (Windows; U; Windows NT 5.1; en-GB; rv:1.9.2.12) Gecko/20101026 Firefox/3.6.12
Re: Performance regression: NS 2.6.8.7+ hogs memory on Fx 3.
Does it help if you upgrade Fx to 3.6.28 (if you can)?
https://ftp.mozilla.org/pub/mozilla.org ... n32/en-GB/
https://ftp.mozilla.org/pub/mozilla.org ... n32/en-GB/
*Always* check the changelogs BEFORE updating that important software!
Mozilla/5.0 (X11; Linux i686; rv:29.0) Gecko/20100101 Firefox/29.0 SeaMonkey/2.26a1
Re: Performance regression: NS 2.6.8.7+ hogs memory on Fx 3.
Not possible at the moment, sorry. 

Mozilla/5.0 (Windows; U; Windows NT 5.1; en-GB; rv:1.9.2.12) Gecko/20101026 Firefox/3.6.12
- Giorgio Maone
- Site Admin
- Posts: 9528
- Joined: Wed Mar 18, 2009 11:22 pm
- Location: Palermo - Italy
- Contact:
Re: Performance regression: NS 2.6.8.7+ hogs memory on Fx 3.
I tried to reproduce with no success.
Maybe an extensions conflict? Could you try Standard Diagnostic?
Maybe an extensions conflict? Could you try Standard Diagnostic?
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:26.0) Gecko/20100101 Firefox/26.0
Re: Performance regression: NS 2.6.8.7+ hogs memory on Fx 3.
Sorry for the delay. I finally found the time to somewhat reliably test this.
I performed the following tests like this:
1. Always had 15 tabs open (non-trivial content).
2. Opened a few additional tabs (non-trivial content).
3. Closed additional tabs.
4. Checked memory usage.
5. Goto 2.
I tested NS 2.6.8.6, 2.6.8.7 and 2.6.8.12. I made sure not to change any other settings besides changing NS versions and enabling/disabling additional extensions.
Scenario A: NS 2.6.8.7 and 2.6.8.12 with all other extensions enabled
- Freshly started Fx with 15 tabs takes up about 100 MiB.
- Opening and closing tabs increases memory usage; Fx becomes more and more jerky as memory usage increases.
- At about 300 MiB, Fx freezes for several seconds and memory usage goes down to about 200 MiB. Jerkiness is somewhat reduced. This happened after opening and closing roughly 50-100 tabs in total.
- Opening and closing tabs increases memory usage and jerkiness again.
Scenario B: NS 2.6.8.7 and 2.6.8.12 with all other extensions disabled
- Same as scenario A, but all memory usage figures about 10 MiB lower.
Scenario C: NS 2.6.8.6 with all other extensions enabled
- Freshly started Fx with 15 tabs takes up about 100 MiB.
- Opening tabs increases memory usage to somewhere above 150 MiB, depending on content.
- Shortly after closing the additional tabs, memory usage always goes down to about 150 MiB. Memory usage was still back to 150 MiB after opening and closing about 300 tabs in total, at which point I stopped.
The results of the different scenarios are reproducible every time (for me). I haven't tested NS 2.6.8.6 with all extensions disabled because it already works fine with everything enabled.
Edit: I just found the development versions and tried 2.6.8.7rc3 and rc4.
- 2.6.8.7rc3 behaves the same as in scenario C (aka. fine).
- 2.6.8.7rc4 behaves the same as in scenarios A and B (aka. affected).
I performed the following tests like this:
1. Always had 15 tabs open (non-trivial content).
2. Opened a few additional tabs (non-trivial content).
3. Closed additional tabs.
4. Checked memory usage.
5. Goto 2.
I tested NS 2.6.8.6, 2.6.8.7 and 2.6.8.12. I made sure not to change any other settings besides changing NS versions and enabling/disabling additional extensions.
Scenario A: NS 2.6.8.7 and 2.6.8.12 with all other extensions enabled
- Freshly started Fx with 15 tabs takes up about 100 MiB.
- Opening and closing tabs increases memory usage; Fx becomes more and more jerky as memory usage increases.
- At about 300 MiB, Fx freezes for several seconds and memory usage goes down to about 200 MiB. Jerkiness is somewhat reduced. This happened after opening and closing roughly 50-100 tabs in total.
- Opening and closing tabs increases memory usage and jerkiness again.
Scenario B: NS 2.6.8.7 and 2.6.8.12 with all other extensions disabled
- Same as scenario A, but all memory usage figures about 10 MiB lower.
Scenario C: NS 2.6.8.6 with all other extensions enabled
- Freshly started Fx with 15 tabs takes up about 100 MiB.
- Opening tabs increases memory usage to somewhere above 150 MiB, depending on content.
- Shortly after closing the additional tabs, memory usage always goes down to about 150 MiB. Memory usage was still back to 150 MiB after opening and closing about 300 tabs in total, at which point I stopped.
The results of the different scenarios are reproducible every time (for me). I haven't tested NS 2.6.8.6 with all extensions disabled because it already works fine with everything enabled.
Edit: I just found the development versions and tried 2.6.8.7rc3 and rc4.
- 2.6.8.7rc3 behaves the same as in scenario C (aka. fine).
- 2.6.8.7rc4 behaves the same as in scenarios A and B (aka. affected).
Last edited by ometa on Thu Jan 23, 2014 10:20 pm, edited 1 time in total.
Mozilla/5.0 (Windows; U; Windows NT 5.1; en-GB; rv:1.9.2.12) Gecko/20101026 Firefox/3.6.12
Re: Performance regression: NS 2.6.8.7+ hogs memory on Fx 3.
Well, that release was performance-related - although ironically it was actually fixing a performance regression.
======
Thrawn
------------
Religion is not the opium of the masses. Daily life is the opium of the masses.
True religion, which dares to acknowledge death and challenge the way we live, is an attempt to wake up.
Thrawn
------------
Religion is not the opium of the masses. Daily life is the opium of the masses.
True religion, which dares to acknowledge death and challenge the way we live, is an attempt to wake up.
Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:26.0) Gecko/20100101 Firefox/26.0
Re: Performance regression: NS 2.6.8.7+ hogs memory on Fx 3.
So... it doesn't look like NS 2.6.8.14 addressed the issue. Do you need any other information from me to debug this?
Edit 2015-02-24:
Nvm. This is obviously not going to get fixed. I'm a bit disappointed to see that NS is still shown as being compatible with FF 3.0.9+ on AMO, though.
Edit 2015-02-24:
Nvm. This is obviously not going to get fixed. I'm a bit disappointed to see that NS is still shown as being compatible with FF 3.0.9+ on AMO, though.
Last edited by ometa on Tue Feb 24, 2015 11:13 pm, edited 1 time in total.
Mozilla/5.0 (Windows; U; Windows NT 5.1; en-GB; rv:1.9.2.12) Gecko/20101026 Firefox/3.6.12
- Giorgio Maone
- Site Admin
- Posts: 9528
- Joined: Wed Mar 18, 2009 11:22 pm
- Location: Palermo - Italy
- Contact:
Re: Performance regression: NS 2.6.8.7+ hogs memory on Fx 3.
No thank you, I'm investigating it but had higher priority bugs to tackle until now.ometa wrote:So... it doesn't look like NS 2.6.8.14 addressed the issue. Do you need any other information from me to debug this?
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:27.0) Gecko/20100101 Firefox/27.0