Closed Bug 192989 Opened 23 years ago Closed 23 years ago

crash on close after reload of page fails

Categories

(SeaMonkey :: General, defect)

x86
Windows 2000
defect
Not set
critical

Tracking

(Not tracked)

RESOLVED WORKSFORME

People

(Reporter: joebrown, Assigned: asa)

Details

(Keywords: crash, stackwanted, Whiteboard: TB17104764E)

User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.3a) Gecko/20021212 Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.3a) Gecko/20021212 Last night, before I reboot the firewall/router box, mozilla was crashing virtually every time I closed it. I suspect it doesn't like opening sockets and not recieving an answer. At first I thought it had to do with mutiple tabs open and reloading more than one page at a time. Pages weren't reloading 90% of the time. The link double click syndrom had been driving me cazy anyway and usual fixes didn't have effect. I sent a bug report because mozilla asked me too (after about the 4th time). Usually after 3rd or 4th try, I might get a page to reload. I finally reboot the firewall/router, and mozilla quit crashing on close for me. I tried to avoid the crash on close by stopping failed requests (click the stop) button but that had no effect. After a page failed to be reloaded, mozilla would crash on close. Since I don't really have a way to test this now that my router has been reboot, I can't provide much more specifics, but hopefully it will help you isolate the cause of others mysterious crashes on close. Reproducible: Always Steps to Reproduce: 1. Hard part -- find a router that eats traffic 2. Reload web pages that fail 3. close mozilla Actual Results: mozilla crashed Expected Results: exited nicely Unfortunatly, I don't have specifics. If the report I tried to send lastnight got filed somewhere, I don't know where, it should have that specific info. I'm pretty sure I filed it as joebrown@rclooke.com, might be joebrown@podiatryfl.com or joe@overdrivepc.com, not likely joebrown@rlooke.com because that's invalid afaik. If it's filed on my harddrive somewhere, tell me where, I'll dig it up and for additional info.
Run Talkback.exe: It displays a list of past crashes.
Keywords: crash
TB17104764E is the Incedent ID according to QFA. I couldnt' get any more info from talckback, HTH.
Keywords: stackwanted
Whiteboard: TB17104764E
Joe, are you still seeing crashed with Mozilla 1.3b?
No, but I haven't had the same router isses crop back up. I've seen it crash for other reasons, I suspect my computer being overckocked or other stuff I'm hacking on, but could be the same issue. The crash in question is reproducable, which was why I logged the bug. Unfortunatly, it's not necissiarly the easiest to reproduce. Dropping packets randomly into /dev/null between the browser and a remote host, seems to be the ticket. I suspect that when the browser initiates a conversation, it doesn't like it when it doesn't recieve a full responce. Don't know how many sockets could be opened, b4 a crash might ensue, once one is opened and the browser is closed before it is completed, it crashes, this is bad (me thinks). As far as I know, the router in question has not hiccupped since the original incident. I'd try and figure out which pieces of packes need to be missing action, but that is a beyond ability. I wish I could do more to fix this bug, because it could be responsible for over 50% of crashes that don't seem reproducable. If the router acts up again, I'll grep the network for debugging info, but that might not happen for a couple more months, if ever.
Last report I was on 1.3a I've recently upgraded to 1.3b I think my router went out into the weeds again but I was unable to reproduce the crash when closing mozilla, so this bug may be squashed.
->WFM based on reporter's comment Feel free to reopen this bug report if you experience this problem with Mozilla 1.3 final or later.
Status: UNCONFIRMED → RESOLVED
Closed: 23 years ago
Resolution: --- → WORKSFORME
Product: Browser → Seamonkey
You need to log in before you can comment on or make changes to this bug.