Summary: | Kmail crash (maximized) hang (unmaximized) | ||
---|---|---|---|
Product: | [Unmaintained] kmail | Reporter: | David Southwell <david> |
Component: | encryption | Assignee: | kdepim bugs <kdepim-bugs> |
Status: | RESOLVED WORKSFORME | ||
Severity: | crash | CC: | lemma |
Priority: | NOR | Keywords: | triaged |
Version: | 1.9.7 | ||
Target Milestone: | --- | ||
Platform: | FreeBSD Ports | ||
OS: | FreeBSD | ||
Latest Commit: | Version Fixed In: | ||
Sentry Crash Report: |
Description
David Southwell
2007-08-19 17:08:23 UTC
The following kde ports are installed on this system: [root@dns1 ~]# pkg_info |grep kde kde-3.5.7 The "meta-port" for KDE kde-xdg-env-1.0_3,1 Script which hooks into startkde and helps KDE pick up XDG kde3-apireference-3.4 API reference for KDE 3 kdeaccessibility-3.5.7 Accessibility applications for KDE kdeadmin-3.5.7 KDE applications related to system administration kdeartwork-3.5.7 Additional themes, sounds, wallpapers and window styles for kdebase-3.5.7_1 Basic applications for the KDE system kdebase-kompmgr-3.5.7 Utility needed to enable XComposite support in KDE kdeedu-3.5.7 Collection of entertaining, educational programs for KDE kdegames-3.5.7 Games for the KDE integrated X11 desktop kdegraphics-3.5.7_1 Graphics utilities for the KDE3 integrated X11 desktop kdegraphics-kamera-3.5.7 Digital camera support for KDE kdegraphics-kooka-3.5.7 Raster image scan program for KDE kdegraphics-kuickshow-3.5.7 KDE image viewer kdehier-1.0_11 Utility port which installs a hierarchy of shared KDE direc kdelibs-3.5.7 Base set of libraries needed by KDE programs kdemultimedia-3.5.7_1 Multimedia utilities for the KDE integrated X11 desktop kdemultimedia-mpeglib_artsplug-3.5.7_1 Legacy KDE audio backend kdemultimedia-xine_artsplugin-3.5.7 Xine-based multimedia backend for KDE kdenetwork-3.5.7 Network-related programs and modules for KDE kdenetwork-kopete-0.12.5 KDE multi-protocol instant messenger (IM) kdepim-3.5.7 Personal Information Management tools for KDE kdesdk-3.5.7 KDE Software Development Kit kdetoys-3.5.7 Small applications for KDE kdeutils-3.5.7 Utilities for the KDE integrated X11 desktop kdevelop-3.4.1_1 IDE for a wide variety of programming tasks kdewebdev-3.5.7,2 Comprehensive html/website development environment Output of uname -a 6.1-RELEASE FreeBSD 6.1-RELEASE #0: Sun May 7 04:15:57 UTC 2006 root@bloom.cse.buffalo.edu:/usr/obj/usr/src/sys/SMP amd64 Looks like some things go horribly wrong on FreeBSD. *** This bug has been marked as a duplicate of 147358 *** Humph Some detailed reasoning would be helpful. I do not know on what basis this has been marked as a duplicate. I respectfully do not agree. The bug has some similarities but, on the face of it, it does not appear to be a duplicate. Comments such as "Looks like some things go horribly wrong on FreeBSD." are not helpful nor do they inspire condfidence. Are you saying this bug is something to do with the freebsd environement. If so why is this the only application out of 800+ ports (including 26 kde applications and the kde meta-port) that crashes on this system? Or are you saying this is a bug in kdepim bug in kdepim that only shows up in a Freebsd environment??? What exactly are you saying? Please be a little clearer. Whatever this needs resolving. I do not consider this a duplicate of 147358 because the behaviours are different. 1.147358 was always a hang. 149003 is clearly a crash when the kmail window is maximized but a hang when it is not. 2. In 147358 you had to click on the email to produce the hang. In 149003 you do not have to click on the email. Thanks David The backtrace you posted is the same as in bug 147358 and clearly shows that KMail hangs while trying to verify a signature. This seems to happen only on FreeBSD systems, as the reporter of the other bug is also using FreeBSD. Now the only difference I see is that you claim that KMail crashes when the windows is maximized, which could be another issue. Please post the backtrace of this crash as well. Please do not paste the KMail output here (it is mostly useless), only the backtrace. To add further grist to this mill the following item was posted to 147358 by Matt Blissett which adds to ones lack of confidenece in the way in which this bug report has been handled. ------- Additional Comment #15 From Matt Blissett 2007-08-19 18:48 ------- You mention in #149003 that it's only reported on FreeBSD -- I had the problem under Linux (#147395) It looks as though there is a track here of insufficient attention and action only when it results in marking bugs as resolved when there has in fact been no real bug resolution!! Two additions: * I certainly did not claim that we did OK and FreeBSD did wrong. It could be our code which may be incompatible with the FreeBSD environment; * Identical backtraces often show the same problem. There may be several symptoms of such problems, which may look like separate bugs from the outside. So the reasoning stated by Thomas and these additions made me mark it as dupe. Re Thomas McGuire Matt Blisset reports the same bug occurs on Linux so the notion that this is Freebsd only is clearly out of step with reality. I have posted the backtrace from the hang. Unfortunately the whole system crashes when Kmail is maximized necessitating a power down and reboot. In view of the risks involved in that procedure I am extremely reluctant to do that. I would be happy to work with someone who is trying to solve the problem and run any tests that may help to resolve it. Thanks > It looks as though there is a track here of insufficient attention and action
> only when it results in marking bugs as resolved when there has in fact been
> no real bug resolution!!
Look, we are not here to offend people. We are trying to help both users and developers. We get about 100 new bugs and wishes each day, and only a few people read them and sort things out. I hope you can imagine that mistakes can always happen in a few cases.
> We get about 100 new bugs and wishes each day
I missed an important detail: *many* of them are invalid (configuration issues, distro issues) or are duplicates.
I do not know why you are so determined to suggest this as freebsd only in view of Mark's report that it is also apparent on Linux. Marks comment relied on the notion that this is freebsd only when it is clearly not and your suggestion that it is a duplicate flies in the face of reality. I accept that resolving 147358 ccould light the way to solving 149003 but they are clearly different. 149003 is much more tricky to resolve because we get a total SYSTEM CRASH when kmail is maximized. That never occured with 147358. So they are clearly different. The other thing you will see from careful reading of the two reports is that 147358 was on an earlier version of kdepim whereas 149003 is on the latest version. David version. OK I accept a mistake was made. The difference in versions is also highly significant. My guess is that 147358 was never resolved in the earlier version of kdepim. The code that generated that bug probably remains in the latest version BUT the later version has compounded the problem and made it worse. All the more reason why they need to be looked at separately other wise some poor developer is going to get confused between the bugs on two different versions of kdepim David You may contact kde-freebsd@kde.org or freebsd-questions@freebsd.org about this, they can guide you to the source of the problem better than I can. Hi Bram IMHO you are not paying attention to the facts. I am disappointed the reality this is more than a freebsd issue is being ignored. The same bug is apparent on linux as evidenced by Matt Blissett's report in which he says. ""Matt Blissett 2007-08-19 18:48 ------- You mention in #149003 that it's only reported on FreeBSD -- I had the problem under Linux (#147395)"" This issue has been around for quite some time and IMHO the need for developer attention is being excused at every turn. This is clearly a bug in kde core and cannot be resolved by freebsd. Noone has any evidence of any work being done it by the developers only that the problem is getting worse in the latest version. I request this issue be given a higher priority and more attention than it seems to be getting. Thanks This marking as dup[licates does lead to problems!! Mat Bissetts 147395 bug was reported on Linux debian. It was then marked as resolved on the grounds it is a duplicate of 147358 even though the version of kdepim being run by Matt is different than 147358. There was no record made at that time that the bug has therefore been reported on more than one operating system!! Despite this reality there are persistent attempts to flag these types of bugs as being freebsd specific. Maybe more care is needed before things are marked as duplicate. OK, some points: - Bram is doing a great job looking at incoming bug reports. If the backtraces match, it is almost guarenteed that they are the same bugs. - We don't insist that this is a FreeBSD problem, it was rather a guess. - If bugs are reported against different versions, they are still marked as duplicates if they have not been marked as fixed. After all, they are the same bugs - Same happens if two identical bugs are reported with two different operating systems, they are the same bug. In this case, it is an external gpg process which hangs. >the need for developer attention is being excused at every turn >[..] >Noone has any evidence of any work being done it by the developers Where did we make excuses? The unfortunate fact is that there are very few KDEPIM developers, who don't find the time to look at bug reports. >I request this issue be given a higher priority and more attention See above, it does not get attention because we don't have enough developers for it. Nobody uses bugzilla's priority field anyway. >The difference in versions is also highly significant. I doubt that. It is the same old bug, apart from the crash when maximized, which probably is another issue. >My guess is that 147358 was never resolved in the earlier version of kdepim. Right. Resolved bugs are marked as RESOLVED/FIXED :) I'll leave this bug open because you say that the system crashes when KMail is maximized. I don't know if this is related to the deadlock when viewing signed messages. Does KMail always hang when being maximised or only when an signed message is viewed? Also, it seems strange to me that the whole system hangs, an application shouldn't be able to cause this. OK, Thomas did a part of my initial answer already. But here's an addition explaining comment 13. The reason I pointed you to FreeBSD specific mailing lists is that they can help you track down why your system completely freezes, with FreeBSD specific tools. See, a desktop enviroment itself does not handle the hardware, it's your OS, and it seems something goes wrong in there. Yep but the same thing happens with linux!!!! So the problem must be in the way kmail makes its calls through kde .core to grab the focus of kb and mouse. No other kde program has the same problem -- but kmail has it on both freebsd and linux!! david Just one small point may be helpful.. kmail currently does not, in the version reported in 149003, generally have problems with signed messages. Problems are much rarer than they used to be with previous versions. So my question is has there been some change in the code in recent versions affecting the way signed messages are handled? There seem to be two symptoms -- a reduction in the volume of incidents accompanied by an increase in their severity when they do occur. Perhaps these symptoms are in some way related. I agree crashes should not occur BUT as the crashes affect more than one operating system (freebsd & linux) it is highly likely the problem is generic to the code rather than generic to an individual operating system. Freebsd ports maintainer has determined this is not a ports problem or caused by the freebsd OS but a bug in kdepim. Or to be more precise, we are unable to determine whether it is a FreeBSD-specific problem. We are aware of the bug and will track its progress in bugs.kde.org, but we are not kmail or gnupg or gpgme developers, we are, first and foremost, packagers. I have just reviewed the emails that generate the hang/crash phenomenon and found an interesting correlation. The hung/crash mails have all been subject to filtering as a result of which they have been moved from an incoming mailbox to another. Secondly every mail has come to me twice via a cc mechanism. For example one copy directly to me and one copy from a mailing list. No mails sent to me as an individual have hung/crashed. This makes me wonder if the mail filter system could be implicated. Some attention to this problem would be appreciated. Thanks David Re 21 Thanks Michael for your careful attention to this ( and your assiduous help to me in my efforts to get this particularly nasty bug traced). In the context of your note, it is worth repeating that this bug has also the subject of a bug report on Linux so any suggestion it is freebsd specific flies against the facts. David Does none ever try and track bugs and solve them ???????? Do I need to use another office package?? Problem still around Still no response Hi David, same as for bug 147358: Did the bug get resolved with KMail on KDE4? No feedback for 4 monthes. Closing this report. |