Am Mittwoch, 15. September 2004 10:23 schrieb Frederik Himpe:
> I'm using KDE 3.2.3 with KDEPim 3.3 (cvs of 13 september 2004) on
> Mandrakelinux 10.1 Cooker (all included RPM packages).

KDEPIM as in CVS HEAD requires KDELIBS 3.3 however the KDE_3_3_BRANCH of 
KDEPIM should work fine on KDELIBS 3.2.x.

Please close the report if there is a version problem on your side.

Yes, the Mandrakelinux Cooker KDEPim 3.3.0 RPM is build from the KDE_3_3_BRANCH.

Actually, the message view becomes empty after Kontact has been minimzed and restored, probably while it is checking for mails in the IMAP folder while kontact is in minimzed state.

Could it be that this is fixed by this commit? http://lists.kde.org/?l=kde-cvs&m=109683551407295&w=2

iirc that commit fixed a problem created just a few days earlier by Till.

That's a valid bug, I've seen it as well. Working with Waldo on resolving it.

*** Bug 93440 has been marked as a duplicate of this bug. ***

I have the impression that each time this problem happens (very often here), these messages are put on the console:

kmail: connections to server imap.domain.com now 0
kio_imap: IMAP4::dispatch - command=67
kio_imap: IMAP4::get -  newimap://login@imap.domain.com:143/INBOX/;UID=822;SECTION=BODY.PEEK[1]
kio_imap: URL: box= INBOX, section= BODY.PEEK[1], type= , uid= 822, validity=
kio_imap: IMAP4::parseURL: box INBOX
kio_imap: IMAP4::parseURL - return 5
kio_imap: IMAP4Protocol::assureBox - reusing box
kio_imap: imapParser::parseBody - discarding 822
kio_imap: IMAP4::get -  finished
kmail: ImapJob::slotGetMessageResult - retrieved part 1
kmail: KMMessage::updateBodyPart 1
kmail: ISubject::notify 0
kio_imap: IMAP4::dispatch - command=67
kio_imap: IMAP4::get -  newimap://login@imap.domain.com:143/INBOX/;UID=822;SECTION=BODY.PEEK[2.MIME]
kio_imap: URL: box= INBOX, section= BODY.PEEK[2.MIME], type= , uid= 822, validity=
kio_imap: IMAP4::parseURL: box INBOX
kio_imap: IMAP4::parseURL - return 5
kio_imap: IMAP4Protocol::assureBox - reusing box
kio_imap: mimeHeader::parseHeader - starting parsing
kio_imap: mimeHeader::parseHeader - finished parsing
kio_imap: IMAP4::get -  finished
kmail: ImapJob::slotGetMessageResult - retrieved part 2.MIME
kmail: KMMessage::updateBodyPart 2
kmail: ISubject::notify 0
kio_imap: IMAP4::dispatch - command=80
kio_imap: IMAP4::dispatch - command=77
kio_imap: IMAP4Protocol::special
kio_imap: IMAP4::dispatch - command=77
kio_imap: IMAP4Protocol::special
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
kio_imap: IMAP4::dispatch - command=80
kio_imap: IMAP4::dispatch - command=77
kio_imap: IMAP4Protocol::special
kio_imap: IMAP4::dispatch - command=77
kio_imap: IMAP4Protocol::special
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 32 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 44 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 48 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 52 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 56 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range
QGArray::at: Absolute index 60 out of range

On Friday 14 January 2005 16:30, Frederik Himpe wrote:

> ------- I have the impression that each time this problem happens (very
> often here), these messages are put on the console:

That's correct. I have a pretty good idea what's happening by now, but still 
no idea why. I've not given up on it. :)

Till

I've committed a fix in this area last night, can someone maybe try with head if that helped?

Could it be possible to have a patch for 3_3_BRANCH or proko2 branch?

The fix was just backported to both branches.

The problem occured to me again with Mandrakelinux' Cooker kdepim-3.3.2-26mdk, which includes this patch.

*** Bug 99097 has been marked as a duplicate of this bug. ***

Hello,

with KDE 3.4.0 (SuSE 9.2 RPMs) this just happened again. When the rows disappeared, I started gdb on the kontact process and got this:

0xffffe410 in ?? ()
(gdb) cont
Continuing.

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 1105432128 (LWP 17240)]
0x427792c5 in KMMsgBase::transferInProgress ()
   from /opt/kde3/lib/libkmailprivate.so
(gdb) bt 10
#0  0x427792c5 in KMMsgBase::transferInProgress ()
   from /opt/kde3/lib/libkmailprivate.so
#1  0x427f5071 in KMHeaders::selectMessage ()
   from /opt/kde3/lib/libkmailprivate.so
#2  0x429312fe in KMHeaders::qt_invoke () from /opt/kde3/lib/libkmailprivate.so
#3  0x4155031e in QObject::activate_signal ()
   from /usr/lib/qt3/lib/libqt-mt.so.3
#4  0x418b62a5 in QListView::doubleClicked ()
   from /usr/lib/qt3/lib/libqt-mt.so.3
#5  0x40ee3d8e in KListView::contentsMouseDoubleClickEvent ()
   from /opt/kde3/lib/libkdeui.so.4
#6  0x4166d1f8 in QScrollView::viewportMouseDoubleClickEvent ()
   from /usr/lib/qt3/lib/libqt-mt.so.3
#7  0x4166e6e6 in QScrollView::eventFilter ()
   from /usr/lib/qt3/lib/libqt-mt.so.3
#8  0x4163d608 in QListView::eventFilter () from /usr/lib/qt3/lib/libqt-mt.so.3
#9  0x42931c66 in KMHeaders::eventFilter ()
   from /opt/kde3/lib/libkmailprivate.so

Created attachment 10217 [details]
valgrind log using "--tool=memcheck --num-callers=50 kmail --nofork"

Hi, I ran kmail in --nofork mode for about three hours while new mail kept
coming in faster than I could answer it. I didn't catch this bug, but I did
catch some other things like uninitialized variables and so on.

Maybe it helps. Maybe the crash is just a special case of one of the errors I
did catch.

At work (SuSE 9.1, KDE 3.4.0 RPMs from SuSE) I am having this bug *all the time*, ie. once every ten minutes. Here are some things to corner the bug:

- It always happens when the IMAP message list is updated, either when I click on a new folder, or when I delete mails and the tree (thread view) must be updated. 

- It has never happened in any folder other than the INBOX.

- Here's a recent backtrace:

Using host libthread_db library "/lib/libthread_db.so.1".
[Thread debugging using libthread_db enabled]
[New Thread 16384 (LWP 28377)]
[KCrash handler]
#5  0x428cd2a6 in KMail::HeaderItem::text ()
   from /opt/kde3/lib/libkmailprivate.so
#6  0x4161274f in QListViewToolTip::maybeTip ()
   from /usr/lib/qt3/lib/libqt-mt.so.3
#7  0x4165bc5f in QTipManager::showTip () from /usr/lib/qt3/lib/libqt-mt.so.3
#8  0x4165cdaa in QTipManager::qt_invoke ()
   from /usr/lib/qt3/lib/libqt-mt.so.3
#9  0x4151d2ee in QObject::activate_signal ()
   from /usr/lib/qt3/lib/libqt-mt.so.3
#10 0x4151da96 in QObject::activate_signal ()
   from /usr/lib/qt3/lib/libqt-mt.so.3
#11 0x418664f9 in QTimer::timeout () from /usr/lib/qt3/lib/libqt-mt.so.3
#12 0x415420df in QTimer::event () from /usr/lib/qt3/lib/libqt-mt.so.3
#13 0x414bb71f in QApplication::internalNotify ()
   from /usr/lib/qt3/lib/libqt-mt.so.3
#14 0x414bd2c9 in QApplication::notify () from /usr/lib/qt3/lib/libqt-mt.so.3
#15 0x4114f87e in KApplication::notify () from /opt/kde3/lib/libkdecore.so.4
#16 0x414affdc in QEventLoop::activateTimers ()
   from /usr/lib/qt3/lib/libqt-mt.so.3
#17 0x41469dd3 in QEventLoop::processEvents ()
   from /usr/lib/qt3/lib/libqt-mt.so.3
#18 0x414d33c1 in QEventLoop::enterLoop () from /usr/lib/qt3/lib/libqt-mt.so.3
#19 0x414d3206 in QEventLoop::exec () from /usr/lib/qt3/lib/libqt-mt.so.3
#20 0x414bd19f in QApplication::exec () from /usr/lib/qt3/lib/libqt-mt.so.3
#21 0x08063b05 in main ()


Jens

... and (forgot to mention) I also get these, like above:

kontact: WARNING: QuotedPrintableDecoder: illegally formed hex char! Outputting verbatim.
kontact: WARNING: QuotedPrintableDecoder: illegally formed hex char! Outputting verbatim.
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range
QGArray::at: Absolute index 36 out of range

(... also for 40, 44, ..., 72)

QGArray::at: Absolute index 76 out of range
QGArray::at: Absolute index 76 out of range
QGArray::at: Absolute index 76 out of range
QGArray::at: Absolute index 76 out of range
*** KMail got signal 11 (Crashing)
KCrash: Application 'kontact' crashing...
X Error: BadWindow (invalid Window parameter) 3
  Major opcode:  20
  Minor opcode:  0
  Resource id:  0x2600009
kio (KIOConnection): ERROR: Header read failed, errno=104
kio (KIOConnection): ERROR: Header has invalid size (-1)

*** Bug 102931 has been marked as a duplicate of this bug. ***

Can we do something to help debugging this? I'm still being bugged almost each time I check my IMAP account (i.e. each ten minutes) with KDE 3.4.2.

I have the impression that this is solved in KDE 3.5.

OK, one day after I thought it was fixed, I encounter again the same problem with KDE 3.5RC. But still, it seems much more rare...

I changed to another IMAP mailbox, and a newly arrived mail was automatically selected. The mail appears in the preview pane, but at the same time, all the lines in the message view become empty.

There is open folder->open() missing somewhere but as I'm not sure where 
exactly - here is a hotfix that should at least fix the bug.


Created an attachment (id=15526)
kmfolderimap.diff

*** Bug 125311 has been marked as a duplicate of this bug. ***

SVN commit 529696 by burghard:

Do not close a selected folder, this is a hotfix for the messagelist corruption.
The source of the problem persists though - probably a missing folder->open() somewhere.
BUGS:89549


 M  +5 -0      kmfolderimap.cpp  


--- branches/KDE/3.5/kdepim/kmail/kmfolderimap.cpp #529695:529696
@@ -110,6 +110,11 @@
   if (mOpenCount <= 0 ) return;
   if (mOpenCount > 0) mOpenCount--;
   if (mOpenCount > 0 && !aForced) return;
+  if (isSelected() && !aForced) {
+      kdWarning(5006) << "Trying to close the selected folder " << label() << 
+          " - ignoring!" << endl;
+      return;
+  }
   // FIXME is this still needed?
   if (mAccount)
     mAccount->ignoreJobsForFolder( folder() );

*** Bug 120590 has been marked as a duplicate of this bug. ***

*** Bug 117926 has been marked as a duplicate of this bug. ***

It looks like the problem persists in yesterdays (Mai, 9th 
2006) svn version at least in a maildir tree. And it seems 
pretty related to filtering/copying something into the 
current mailbox. There is no warning message (as 
expected ...).





On Friday 14 April 2006 09:47, Carsten Burghardt wrote:
[bugs.kde.org quoted mail]

The following patch lets the message list view survive 
filtering into the current folder; since there was anyway a 
comment re. opening/closing of folders that's probably 
fixed that way and the comment could be removed?

Michael

Index: kmfilteraction.cpp
===================================================================
--- kmfilteraction.cpp  (revision 539146)
+++ kmfilteraction.cpp  (working copy)
@@ -1373,12 +1373,21 @@

 KMFilterAction::ReturnCode 
KMFilterActionCopy::process(KMMessage* msg) const
 {
+  bool opened = false;
+
   // TODO opening and closing the folder is a trade off.
   // Perhaps Copy is a seldomly used action for now,
   // but I gonna look at improvements ASAP.
-  if ( !mFolder && mFolder->open() != 0 )
-    return ErrorButGoOn;
-
+  if ( !mFolder ) {
+    if (!mFolder->isOpened()) {
+      int rc;
+      opened = true;
+      rc = mFolder->open();
+      kdDebug(5006) << "KMFolderMaildir::addMsg-open: " << 
rc << " of folder: " << label() << endl;
+      if (rc) return ErrorButGoOn;
+    }
+  }
+
   // copy the message 1:1
   KMMessage* msgCopy = new KMMessage;
   msgCopy->fromDwString(msg->asDwString());
@@ -1387,8 +1396,9 @@
   int rc = mFolder->addMsg(msgCopy, &index);
   if (rc == 0 && index != -1)
     mFolder->unGetMsg( index );
-  mFolder->close();

+  if (opened) mFolder->close();
+
   return GoOn;
 }



On Friday 14 April 2006 09:47, Carsten Burghardt wrote:
[bugs.kde.org quoted mail]

Hi,

I still see this problem occasionally (as in, every couple days) in KMail 1.9.1, as compiled by SuSE for SuSE 10.0, in KDE 3.5.2. Most of the time it happens when switching to my INBOX, it never happened on any other mailboxes yet. The inbox is the only box that gets updated server-side, I don't use serverside filters. Switching to another folder when I see the empty rows before selecting them, and then switching back to the INBOX, avoids the crash.

Also, I could not pinpoint this problem when running KMail under gdb or valgrind, so it might be a race. 

Maybe that helps.

Jens :-)

Created attachment 16029 [details]
kmail/kmfilteraction.cpp diff

Do not (re-)open and close an already opened folder during message copy while
filtering

Hm, open/close on a folder should be just refcounting, so the problem is not that the folder is openend multiple times, it's more that it is closed, when it should be kept open, in case it was not open before, I would guess. Also the  if ( !mFolder && mFolder->open() != 0 ) should be a ||. 

I get similar behavior to this in a very reproducable way on KDE 3.5.2, Kmail 1.9.1, though I'm relatively certain mine has nothing to do with filtering.

While checking mail and ergo updating the message list, if I move the scrollbar at  all, or if I click any message before the task is 100% complete (status display lower right goes idle), I end up with the blank lines or, even better, a "hung" message list where that widget will not refresh. Occasionally this happens without apparently touching anyting at all, and either way it happens probably 10 times per day.

Selecting a different folder, letting that refresh without touching anything, then re-selecting my Inbox folder fixes the gui issue without restarting kmail.   

I can reliably reproduce this on 2 different Gentoo systems using:
Qt: 3.3.6
KDE: 3.5.2
KMail: 1.9.1

and a Debian system using:
Qt: 3.3.5
KDE: 3.5.1
KMail: 1.9.1