[Libreoffice-bugs] [Bug 126828] Variables having assigned objects from closed components pretend to still be valid. Attempts to inspect them in the IDE cause crashes.

bugzilla-daemon at bugs.documentfoundation.org bugzilla-daemon at bugs.documentfoundation.org
Mon Apr 13 13:03:49 UTC 2020


https://bugs.documentfoundation.org/show_bug.cgi?id=126828

Julien Nabet <serval2412 at yahoo.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |caolanm at redhat.com

--- Comment #4 from Julien Nabet <serval2412 at yahoo.fr> ---
Caolán: I made different tests and I noticed that I got a crash even before
calling calling close function in macro.
Above all, I noticed it was gtk3 specific (I don't reproduce this with gen
rendering).

Indeed, sometimes in WatchWindow::UpdateWatches, inside
m_xTreeListBox->all_foreach([this, pCurMethod, bBasicStopped](weld::TreeIter&
rEntry), pItem could be null.

To reproduce this, I put "doc" for watch variable, then debug until doc is
initialized.
- Expand doc
- Expand Reference Device
- Expand Info
- Unexpand Info
- Expand FontDescriptor
=> crash

Another remark: with gtk3 rendering, when editing macro to use debug, the
current window isn't macro one but stays the main Writer window.
With gen rendering, the current window becomes macro one (far more practical).

If you want me to create a new bugtracker, don't hesitate to tell. I thought it
could be related to this one.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/libreoffice-bugs/attachments/20200413/36cd091b/attachment.htm>


More information about the Libreoffice-bugs mailing list