Closed Bug 283823 Opened 20 years ago Closed 17 years ago

(my !?) MailNews "hangs" when it should be closing: CPU goes to 18%

Categories

(MailNews Core :: Networking: IMAP, defect)

x86
Windows 98
defect
Not set
critical

Tracking

(Not tracked)

RESOLVED WORKSFORME

People

(Reporter: sgautherie, Unassigned)

References

Details

(Keywords: hang, regression, Whiteboard: [Possible "steps": see comment 8])

Attachments

(1 file)

[Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.8b2) Gecko/20050223] (nightly) (W98SE) Everything is fine, until I want to "Exit" Mozilla. "Hangs": have to kill the process, using WinTop. (I found old (stalled) bug 95130, but they should be different !?) [Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.8a5) Gecko/20041122] (release) (W98SE) Closes fine.
Component: General → MailNews: Main Mail Window
Odd: this happens only if I never displayed a message: just started the 3-pane window ! Steps: 0. (MAS being closed) 1. Start MailNews, without displaying a message 2. Close/Exit the window !
See also bug 283523.
(In reply to comment #1) > Odd: > this happens only if I never displayed a message: just started the 3-pane window ! This is not always enough to not get the bug... > Steps: > 0. (MAS being closed) > 1. Start MailNews, without displaying a message > 2. Close/Exit the window ! But this reproduces it all the time. (In reply to comment #2) > See also bug 283523. That bug seems closely related: marking it as "blocked" since that other one states that Navigator hangs too, which is not the case on my computer.
Blocks: 283523
Summary: MailNews "hangs" were it should be closing: CPU goes to 100% → MailNews "hangs" when it should be closing: CPU goes to 100%
Not a mailbug - its has to do with closing windows as such. Behaves just like recent bug 280427 -> bug 280456, but occures in newer builds Recently also reported in bug 283523 and bug 284204
[Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.8b2) Gecko/20050701] (nightly) (W98SE) It still happens: I'm not sure how/why. It can be EXITing MAS, or simply closing MailNews with another window (Browser/JSConsole) still opened. I think I looked at it at some point (with WinTop): there seemed to be "4-5" Windows threads running, and one was having like 15% of my CPU. I can't be very specific ... but the hang IS there from times to times :-(
[Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.8b2) Gecko/20050705 SeaMonkey/1.0a] (nightly) (W98SE) Once again: 8 threads = 7 inactive, 1 taking 17% CPU.
(In reply to comment #6) > Once again: 8 threads = 7 inactive, 1 taking 17% CPU. [Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.8b3) Gecko/20050714 SeaMonkey/1.0a] (nightly) (W98SE) Same.
[Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.8b4) Gecko/20050720 SeaMonkey/1.0a] (nightly) (W98SE) My case (8 threads, 17% CPU) could very much be "network" (caused by IMAP and/or NNTP ?!) related. Today, case 1) I set my Network card to Full Duplex, which my current cable modem does not seem to handle; I clicked on a newsgroup to retrieve the headers; this began but did not progress very much; I exited the application: got the bug. case 2) With the restored & supported Half Duplex setting, I got the bug on one of my exit after reading newsgroups ... I had the idea to check the open network connections: still opened, there was my IMAP ones, and 2-3 NNTP ones. Case 2 could be a consequence, but case 1 makes me guess it could be the cause. Further confirmation from me/others could help ... but that a(nother) way to investigate.
Whiteboard: [Possible "steps": see comment 8]
[Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.8b4) Gecko/20050720 SeaMonkey/1.0a] (nightly) (W98SE) Happened again: this time, only my 2 IMAP connections were opened. (I'm pretty sure I had not used NNTP in this session)
[Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.9a1) Gecko/20050816 SeaMonkey/1.0a] (nightly) (W98SE) (still there)
[Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.9a1) Gecko/20050907 SeaMonkey/1.1a] (nightly) (W98SE) I get this bug repeatedly today; and "worse": I'm closing MailNews, and have to use C-A-D to kill the process, as launching WinTop for example is delayed until Mozilla is killed :-/ Can a log like this one help ? NB: From a few tests I tried, it could be that I never get the bug if I go offline before closing MailNews. (PS: going back online after going offline and before closing doesn't trigger the bug either.)
Attachment #195276 - Attachment is patch: false
[Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.8b4) Gecko/20050910 SeaMonkey/1.0a] (nightly) (W98SE) In SMv1.0a candidate too; but that's still my home computer only so far...
[Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.8) Gecko/20051027 SeaMonkey/1.0b] (release) (W98SE) (still there)
Summary: MailNews "hangs" when it should be closing: CPU goes to 100% → (my !?) MailNews "hangs" when it should be closing: CPU goes to 18%
(In reply to comment #13) > (still there) [Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.9a1) Gecko/20051030 SeaMonkey/1.1a] (nightly) (W98SE) (And on Trunk too.)
[Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.9a1) Gecko/20051124 SeaMonkey/1.5a] (nightly) (W98SE) I did not got this bug for a while, then it happened again today. A wild guess would be that it happens when I close MailNews while Navigator is loading a page.?.
(In reply to comment #12) > [Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.8b4) Gecko/20050910 > SeaMonkey/1.0a] > (nightly) (W98SE) (For the record, that build was a release :->) (In reply to comment #13) > [Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.8) Gecko/20051027 SeaMonkey/1.0b] > (release) (W98SE) (For the record, that build was a nightly :->)
last seen ... 2005-11-27?
(In reply to comment #17) > last seen ... 2005-11-27? No, I simply stopped reporting. I'm pretty sure it still happened until some days/weeks ago with SMv1.5a/Trunk. [Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.8.0.2) Gecko/20060407 SeaMonkey/1.0.1] (nightly) (W98SE) And it happened only minutes ago with SMv1.0.1 branch. I think it happened a few times with SMv1.0 on my previous job Win2K too. (But I still don't have any precise idea on why/when this bug is triggered :-()
(In reply to comment #18) > I think it happened a few times with SMv1.0 on my previous job Win2K too. (That was WinXP, not Win2K.)
bug 283523 is now wfm, so removing this as blocker. has imap log been examined yet?
No longer blocks: 283523
(In reply to comment #20) > bug 283523 is now wfm, so removing this as blocker. Then, adding it back, but in the 'depends on' list, to keep the link. > has imap log been examined yet? It seems that it has not. (At this time, I'm too sure (anymore) about anything like the offline/online trick either :-/)
Depends on: 283523
I'm still seeing this bug in recent SeaMonkey v1.1a nighlies. After reading bug 246909, let me write that: I use 'Clean up ("Expunge") Inbox on Exit'; I don't use 'Empty Trash on Exit'. Then, let's add a dependency to bug 312775 for now.
Depends on: imapdeadlock
[Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.8.1a3) Gecko/20060617 SeaMonkey/1.1a] (nightly) (W98SE) Bug 312775 checkin did not solve the current bug.
Component: MailNews: Main Mail Window → Networking: IMAP
No longer depends on: imapdeadlock
Product: Mozilla Application Suite → Core
Hoping bug 342784 fix solved this bug too...
Depends on: 342784
[Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.8.1a3) Gecko/20060706 SeaMonkey/1.1a] (nightly) (W98SE) I got it once: closed Browser and MailNews, ChatZilla was left but impossible to restore it.
Serge, do you see this in the newer 1.1.xs?
tested on trunk?
Serge, it would be great if we could compare notes on your present situation with this. It may be mail related as you think - TB occasionally fails to close for me. Or, it could be some thing more central to gecko shutdown. also, you have regression noted in keywords but don't have a regression range. Did this problem not happen for you prior to 20050223 nightly build? bug 342784 fixed 2006-07-05 which did not help you so removing dependency. stm bug 283523 should be removed also - there is no relationship afaict
No longer depends on: 342784
(Sorry for not answering this for a long time.) I'm not using Win98SE anymore... I'm not experiencing hanged process with my Win2K (and whatever build(s) I use nowadays). I think I sometimes get a little delay during exit ... assuming some cleanup is taking place (maybe related to IMAP ?). Since the attached log file was never analyzed, I'm R.WorksForMe this bug. I'll reopen a new/cleaner one if need be.
Status: NEW → RESOLVED
Closed: 17 years ago
Keywords: hang
Resolution: --- → WORKSFORME
Product: Core → MailNews Core
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: