User-Agent: Mozilla/5.0 (Macintosh; U; PPC Mac OS X; en-US; rv:1.2a) Gecko/20020906 Build Identifier: Mozilla/5.0 (Macintosh; U; PPC Mac OS X; en-US; rv:1.2a) Gecko/20020906 This morning, FizzillaCFM/2002090608 failed to start up. It reached the splash screen and showed, "Starting up...", but failed to complete startup. I let the process continue for about five minutes, with no success. CPU usage contined between 10% and 60% by Mozilla during this time. Experimenting with removing certain files from my profile (cache, etc.), I was able to get it to start up after removing the XUL FastLoad File. I'm using the Pinstripe theme, a version updated with the changes for bug 161226. Reproducible: Couldn't Reproduce Steps to Reproduce: 1. Start Mozilla Actual Results: Mozilla stalls on startup. Expected Results: Mozilla should have completed startup. I'll make the XUL FastLoad File in question available via my web server. URL to come.
The XUL FastLoad File is available from [http://greg.tcp.com/mozilla/167523/XUL FastLoad File.sit.bin].
-> ben, cc: brendan for fastload issues
Assignee: hyatt → ben
Status: UNCONFIRMED → NEW
Ever confirmed: true
Unless someone catches Mozilla in a debugger, this may not be debugable from the FastLoad file. Greg, do you have a debugger, or a way to trace system calls, or otherwise to see what Mozilla is doing while stalled with the bad FastLoad file in place? /be
I got also this some times. But It is not a real Hang (for me). After 10 or 15 minutes Mozilla worked Correctly (Mac OS 9.1 and many builds)
Created attachment 98499 [details] Sampler data taken while Mozilla is trying to start up Brendan, here is about five seconds worth of Sampler data taken while Mozilla's trying to start up.
Jean-Pierre is right. I let Mozilla continue, and it completed starting up sometime during my 15 minutes away from the machine. I know Mozilla clears the cache after a crash, and that can take some time, but this was happening even after I manually deleted the cache folder.
Ouch, we're suffering some kind of Mac-specific tiny-gram read storm, perhas due to FastLoad file fragmentation(?). Cc'ing necko and Mac gurus. /be
OK, Mac OS 9 and MacOSX. It was not a dream I am not sure, but i think also that opening the script manager and other circumstances. I repeat i am not sure I'll signal it here next time The rule is if Mozilla cause an great disk activity per moment and seems to loop, wait!!!
I'm having what seems to be the same problem. (I arrived here by searching for `XUL FastLoad File'...) For me Mozilla starts up normally, but I get a 100% reproducible hang (spinning wheel and skyrocketing CPU usage, have to force quit) by the following steps: 1. Load some page -- say http://www.mozilla.org 2. Select some text -- say `Mozilla News' 3. Hit `Find' (command-F) I took a 1 minute call graph while Mozilla spinned and will attach it next. Like greg's call graph, it involves nsFastLoadFileReader, and the problem ceases if I delete the XUL FastLoad File before restarting Mozilla. Similar/related(?): bug 168516. (Let me know if you want to see my corrupt FastLoad file, or if my case is different enough that I should file it separately.)
Summary: Mozilla stalled at splash screen on startup → Mozilla stalled at splash screen on startup (XUL FastLoad File corruption)
> If you let Mozilla spin, does it eventually start up (after many minutes)? > Please say yea or nay in the bug. > > /be Well in my case Mozilla starts up quite normally -- only the steps of comment 9 make it hang. (The call graph then is much the same as greg's, so that's why I filed this here for now.) I've just let it spin for over an hour while I was out, and it never stopped. Memory usage went from: PID COMMAND %CPU TIME #TH #PRTS #MREGS RPRVT RSHRD RSIZE VSIZE 1173 Mozilla 0.8% 0:12.85 6 88 341 15.1M 24.1M 30.3M 83.6M to: PID COMMAND %CPU TIME #TH #PRTS #MREGS RPRVT RSHRD RSIZE VSIZE 1173 Mozilla 85.0% 61:34.70 6 87 402 268M+ 7.48M 270M+ 344M
Looks like my FastLoad file became corrupt again. This time it's the same symptom as Greg: Mozilla doesn't start up. Again, deleting the FastLoad file solves the problem. Of possible interest: the cause may well be the "crash while editing prefs" that I experienced just before this (re)start. See bug 150893, comment 13 for the stack trace of that crash; its timestamp coincides with the last modification date of the corrupt FastLoad file. Greg, I think you also had a similar crash?
FZiegler, I've experienced that crash, but I don't know if it precipitated this problem. As I recall, i encountered this problem first thing on starting up Mozilla and my computer that morning so I don't know what I did previously.
*** Bug 176049 has been marked as a duplicate of this bug. ***
*** Bug 133223 has been marked as a duplicate of this bug. ***
OS: MacOS X → All
Hardware: Macintosh → All
Summary: Mozilla stalled at splash screen on startup (XUL FastLoad File corruption) → Mozilla stalled at splash screen on startup (XUL FastLoad File corruption) (XUL.mfasl)
*** Bug 181280 has been marked as a duplicate of this bug. ***
This is not only a Mac issue. I just had the same bug on Linux
We had the same problem on Debian Linux. Only as we deleted the XUL.mfasl file and restored it from CVS (where we save our configs in regular intervals), the Mozilla started to work again. The first things which stopped working was a download. Jörg
> Only as we deleted the XUL.mfasl file and restored it from CVS (where we save > our configs in regular intervals), ... Just a quick note: a fastload file is only used if it is used with the same filepaths for the jar files, and they have exactly the same modified time. Otherwise, the fastload file will be deleted. You can save the XUL.mfasl with the rest of your profile if you want, but I would recommend that you consider it to be a "volatile" part of your profile, and to not bother saving it.
Experiencing this very issue with Mozilla/1.2.1 on Windows 2000 sp3 and 100% CPU Utilization and memory growing without bounds until I kill the process, along with freezes when sending email such as bug 168516 with the same CPU Utilization and memory leakage.
Duplicate of bug 169777?
Yeah, you're right. This is a dupe. In fact, the other one was filed later, but there has been more activity over there. Voters, please move your votes to the other one if you care :) *** This bug has been marked as a duplicate of 169777 ***
Status: NEW → RESOLVED
Last Resolved: 16 years ago
Resolution: --- → DUPLICATE
Status: RESOLVED → VERIFIED
Component: XP Toolkit/Widgets: XUL → XUL
QA Contact: jrgmorrison → xptoolkit.widgets
You need to log in before you can comment on or make changes to this bug.