Recommended Posts

configure: error: Your compiler does not follow the C++ specification for temporary object destruction order.
*** Fix above errors and then restart with "make -f client.mk build"
make: *** [/cygdrive/c/src/cpp/firefox/mozilla/objbin/Makefile] Error 1

Got this compiling FF. Any ideas? Google doesn't help.

I thought maybe out of date platform SDK.. but I have the latest.

Again. You posted a screenshot and said there are THOUSANDS of memory leaks in Firefox.

Then you uninstalled all extensions and only got 7/8.

Yet you still say there are THOUSANDS of memory leaks in Firefox.

Not true.

There are 7/8.

It's not Firefox, if it was Firefox's memory management, and the extensions were the cause, then the TBE Extension you had would be having problems. Not so. It was one of the other, or multiple other extensions that you had that caused this problem.

This is what we are trying to have you understand, that a program and an extension is NOT the same thing. It's obviously the extensions fault, not Firefox. I love also how you ignored the rest of my post. I still say blame Windows for not having better memory management, and allowing allocated memory to stay allocated after the program is closed.

It's the extensions, you proved that with your tests.

(Note: I'm not saying firefox does not have memory leaks. I havn't said that at all. I'm just saying it does not have thousands, and it is not really major. Hell, millions of people now are using it and not complaining, it can't be that big of a deal.)

This might sway me to put Firefox on the back burner and swith to using Opera and IE together more. VS now, I am currently using all 3 almost equally.

585517811[/snapback]

To me that sounds stupid, if you needed this thread to point it out to you then it wasn't an issue for you in the first place. :rolleyes:

configure: error: Your compiler does not follow the C++ specification for temporary object destruction order.
*** Fix above errors and then restart with "make -f client.mk build"
make: *** [/cygdrive/c/src/cpp/firefox/mozilla/objbin/Makefile] Error 1

Got this compiling FF.  Any ideas?  Google doesn't help.

I thought maybe out of date platform SDK.. but I have the latest.

585518942[/snapback]

Post your mozconfig?

FF Nightly / TBE Extension

Created new tab, clicked home page, closed original tab, closed FF.

2881 leaks found

hmm, how'd I ever guess. Thats exactly why any and every Firefox dev, hacker, or longtime user of firefox completly hates that extension. Exactly why its not on Update.mozilla.org. Exactly why in every thread that asked about tabs, I always tell them to tune the setting through the options menu and I always say not to use TBE. It uses unsupported code and changes so much of firefox's code its not funny.

Please update your original post to add your results on what extension caused the problem and that with a clean profile there were only 23 or however many leaks.

and could a mod edit the title to something more fitting. The "WARNING!" is not necessary and by the outcome of this thread its clear that is that godforesaken extension that is causing his "problem"

things you should never discuss in public:

1. politics

2. religion

3. choice of web browser

man poor vcv, people are so harsh when even when you explained you use firefox and like it but they still think your a IE fanboy, when wil people read post first before hitting the replay button. Good find by the way, long time since i seen something research on here ( the popular Paris Hilton numbers leak doesnt count as research :pinch: )

Again. You posted a screenshot and said there are THOUSANDS of memory leaks in Firefox.

Then you uninstalled all extensions and only got 7/8.

Yet you still say there are THOUSANDS of memory leaks in Firefox.

Not true.

There are 7/8.

THIS IS THE LAST TIME I WILL SAY THIS:

I GOT THE THOUSANDS OF LEAKS WITHOUT ANY EXTENSIONS INSTALLED.

It's not Firefox, if it was Firefox's memory management, and the extensions were the cause, then the TBE Extension you had would be having problems. Not so. It was one of the other, or multiple other extensions that you had that caused this problem.
Yes it is firefox's memory management. You cannot allocate memory manually in JS, the engine is responsible for all memory.
This is what we are trying to have you understand, that a program and an extension is NOT the same thing. It's obviously the extensions fault, not Firefox.
Firefox is responsible for the extensions memory.
I love also how you ignored the rest of my post.
I love how you ignored my multiple posts that pointed out that (a) Firefox uses JS, not java, and does not allow manually memory allocation (b) I got thousands of leaks WITHOUT any extensions.
I still say blame Windows for not having better memory management, and allowing allocated memory to stay allocated after the program is closed.
The same thing is possible in linux. Windows will do a decent job of cleaning it up, but it can't be perfect.
It's the extensions, you proved that with your tests.
Please learn to read.

Now let me reiterate an important point that people have missed over and over and over and over.

I GOT THE THOUSANDS OF LEAKS WITH NO EXTENSIONS INSTALLED.

As well, Firefox IS responsible for an extesnions memory since an extension is nothing more than JS+XUL.

Post your mozconfig?

585519563[/snapback]

I tried using this:

. $topsrcdir/browser/config/mozconfig

GLIB_PREFIX=C:/src/cpp/firefox/vc71
LIBIDL_PREFIX=C:/src/cpp/firefox/vc71

export MOZ_PHOENIX=1
export MOZ_OPTIMIZE_LDFLAGS="-opt:ref,icf,nowin98"

mk_add_options MOZ_PHOENIX=1
mk_add_options MOZ_OPTIMIZE_LDFLAGS="-opt:ref,icf,nowin98"

mk_add_options MOZ_OBJDIR=@TOPSRCDIR@/objbin
ac_add_options --enable-debug
ac_add_options --enable-static
ac_add_options --enable-strip
ac_add_options --enable-single-profile
ac_add_options --disable-shared
ac_add_options --disable-tests
ac_add_options --disable-crypto
ac_add_options --disable-pedantic
ac_add_options --disable-activex
ac_add_options --disable-activex-scripting
ac_add_options --disable-mailnews
ac_add_options --disable-composer
ac_add_options --disable-ldap
ac_add_options --disable-profilesharing
ac_add_options --disable-shared
ac_add_options --disable-accessibility

Figured maybe it was too much, so I tried with this as well:

. $topsrcdir/browser/config/mozconfig

GLIB_PREFIX=C:/src/cpp/firefox/vc71
LIBIDL_PREFIX=C:/src/cpp/firefox/vc71

mk_add_options MOZ_OBJDIR=@TOPSRCDIR@/objbin
ac_add_options --enable-debug
ac_add_options --enable-static
ac_add_options --disable-shared
ac_add_options --disable-tests
ac_add_options --disable-crypto
ac_add_options --disable-pedantic

Same error either way.

---------

they still do

btw a crap load of memory leak fixes were just patched in firefox.

585519900[/snapback]

Glad to hear :) I'll try the nightly when it's out and see if there is any difference.

It's not Firefox, if it was Firefox's memory management, and the extensions were the cause, then the TBE Extension you had would be having problems. Not so. It was one of the other, or multiple other extensions that you had that caused this problem.

This is what we are trying to have you understand, that a program and an extension is NOT the same thing. It's obviously the extensions fault, not Firefox.

585519543[/snapback]

While you COULD argue that the extension is poorly written, it IS still Firefox's responsibility to clean up ALL resources used by its scripting engine (which the extension runs under.)

I love also how you ignored the rest of my post. I still say blame Windows for not having better memory management, and allowing allocated memory to stay allocated after the program is closed.

585519543[/snapback]

This comment is just plain dumb and/or ignorant.

Windows does not "allow allocated memory to stay allocated after the program is closed." The profilers that people use to find memory leaks log each allocation/deallocation made by the program, and then when the program has completed execution, it scans the allocations made for matching deallocations. If one is not found, that is a memory leak. This leaked memory does NOT stay allocated after the program exits, it is flagged as available by Windows and reused. The problem is that the memory stays allocated and is effectively rendered unusable for as long as the program remains running.

It's the extensions, you proved that with your tests.

585519543[/snapback]

He proved that the scripting engine is buggy, it should not mess up no matter how buggy the code it runs is.

It's not as big an issue as he wants to make it though, the "thousands" of leaks are most likely from the exact same piece of code, being executed thousands of times. So thousands of logged leaks could all be due to the one line of buggy code.

As long as people don't run buggy extensions, they'll be fine. And there are other more serious bugs to worry about in Firefox that deserve a higher priority than this. When it comes to software of this magnitude, you have to prioritize.

Im Not a IE Fanboy, (Note Sig) and I use firefox daily, but i just did a small test.

I have installed the most recent nightly built (02-22) and NO extentions, the only theme being Noia 2.0 Extreme, and this is what i see in the task manager after some heavy browsing.

firefoxleak7mg.jpg

This doesnt seem right...

Im Not a IE Fanboy, (Note Sig) and I use firefox daily, but i just did a small test.

I have installed the most recent nightly built (02-22) and NO extentions, the only theme being Noia 2.0 Extreme, and this is what i see in the task manager after some heavy browsing.

firefoxleak7mg.jpg

This doesnt seem right...

585522288[/snapback]

Could it be because you didn't set the memory cache limit?

about:config > browser.cache.memory.capacity 16000

Ever since setting that I don't get the "memory leak" anymore.

Even worse is if you used the old user.js tweak that set the cache to 65500, which just sets you up for a crash.

Is this every extension, or only certain ones?

585522120[/snapback]

Every extension.

585522770[/snapback]

Actually, you're wrong.  Extensions can include XPCOM componenets which may contain compiled dll's (NOT just javascript and XUL).

585524299[/snapback]

If you knew that fact in advance then don't bear-bait users into answering a trick question. It is a form of flame-baiting.

Actually, you're wrong.? Extensions can include XPCOM componenets which may contain compiled dll's (NOT just javascript and XUL).

585524299[/snapback]

Mind finding a specific page that says that?

I know the JS can use XPCOM components, but can it use ones not included with mozilla or firefox?

If so, are there ANY extensions that include their own XPCOM components?

This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.
  • Posts

    • It would be surprising if even 3 year old Apple Watches (and not SE models at that) cannot run watchOS 27. Granted, it doesn't mean it would work as well. We'll see.
    • Apple launches new website for parents and a revamped Screen Time experience by Aditya Tiwari At WWDC 2026, Apple announced new parental control features for iOS 27, iPadOS 27, and macOS 27 to keep kids' device usage in check and keep them safe online. As a parent, you'll get access to a simpler setup experience, Ask to Browse, Time Allowances, and a redesigned Screen Time. You'll be able to pick exactly which apps your child can access on their device, choosing from just a few essential apps, a curated set, or the apps you feel are appropriate. There will be an option to gradually add more apps. Ask to Browse is a new feature that requires kids to request permission before visiting a new website in Safari on iPhone, iPad, or Mac. You can also turn on a setting that requires your kid to ask for approval before connecting with an unknown contact via Messages, FaceTime, or the Phone app. You can manage your child's screen time more effectively with Time Allowances that work across categories, including Games, Entertainment, and Social Media. You can set time limits based on your kid's age and get suggestions informed by expert research. Speaking of expert guidance, Apple added that it's working with the American Academy of Pediatrics (AAP) to adapt its Family Media Plan to create a guide parents can refer to when using Apple products. Moreover, the company has also set up a new dedicated website, where you can find tools, resources, and answers to common questions around parental controls and child safety. Apple also allows parents to set daily schedules to manage screen time, configuring access to different apps at different times of the day and across the week. Overall, the Screen Time section has been revamped and provides a bird's-eye view of your kid's average device usage and most-used apps. "For example, to help protect important family moments, parents can quickly limit access during meals, outdoor play, and other times that deserve full attention. If kids need a little extra time to finish something in an app, parents can also easily extend access," Apple said. Apart from these, Communication Safety has been updated to block gore or violent content when detected in shared images or videos. The feature already blurs nudity in Messages and FaceTime calls, and is enabled by default for users under 18.
    • They already threw the Ultra 1st gen under the bus? It really feels like someone screwed up, and that advanced AI features will only be available on the list above but someone screwed up and said that Watch OS 27 in general will only be available for those models. I think they will pull it back. I did see someone also put a snap shot of their series 9 watch being offered the watch os 27 beta. As the backlash online is already in full swing.
    • Right, Google... because that's who this is going to. Google, the company who's entire business model is geared around ad revenue and your personal data.
    • That is the natural course for regulations. First to protect the consumer. Then to protect the business/industry from upstart competition. There are countless examples across history.
  • Recent Achievements

    • Very Popular
      Captain_Eric earned a badge
      Very Popular
    • One Month Later
      amusc earned a badge
      One Month Later
    • One Month Later
      DJC50PLUS earned a badge
      One Month Later
    • Week One Done
      DJC50PLUS earned a badge
      Week One Done
    • Proficient
      Eric Biran went up a rank
      Proficient
  • Popular Contributors

    1. 1
      +primortal
      513
    2. 2
      PsYcHoKiLLa
      231
    3. 3
      ATLien_0
      87
    4. 4
      +Edouard
      84
    5. 5
      Steven P.
      80
  • Tell a friend

    Love Neowin? Tell a friend!