← Back to team overview

kicad-developers team mailing list archive

Crash invoking pcbnew Board settings



I've been using the nightly build of kicad via flatpak (because of the WxWidgets issues on Linux) for a while, and have encountered a fairly consistent crash when invoking board settings from the File menu of pcbnew.

The build is from this evening, using up to date flatpak and libs, was done on ubuntu 21.04. However I have also experienced a very similar crash with the same project using a recent (~week old) MacOS build.

The crash results in this appearing in the terminal window:

terminate called after throwing an instance of 'std::bad_alloc'
  what():  std::bad_alloc
[aside: could Kicad catch such exceptions rather than having the window just disappear?]

Invoking gdb produces a backtrace of 54 frames of which the top 27 are somewhat interesting:

(gdb) bt
#0  0x00007ffff5a6d4bb in raise () at /usr/lib/x86_64-linux-gnu/libc.so.6
#1  0x00007ffff5a56867 in abort () at /usr/lib/x86_64-linux-gnu/libc.so.6
#2  0x00007ffff5e308f9 in __gnu_cxx::__verbose_terminate_handler() [clone .cold] () at /usr/lib/x86_64-linux-gnu/libstdc++.so.6 #3  0x00007ffff5e3c62a in __cxxabiv1::__terminate(void (*)()) () at /usr/lib/x86_64-linux-gnu/libstdc++.so.6
#4  0x00007ffff5e3c695 in  () at /usr/lib/x86_64-linux-gnu/libstdc++.so.6
#5  0x00007ffff5e3c971 in  () at /usr/lib/x86_64-linux-gnu/libstdc++.so.6
#6  0x00007ffff5e3050e in operator new(unsigned long, std::nothrow_t const&) [clone .cold] () at /usr/lib/x86_64-linux-gnu/libstdc++.so.6 #7  0x0000555555858695 in std::__cxx11::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> >::_M_assign(std::__cxx11::basic_string<wchar_t, std::char_traits<wchar_t>, std::allocator<wchar_t> > const&) () #8  0x00007ffff7be21e2 in wxGrid::SetCellValue(int, int, wxString const&) () at /app/lib/libwx_gtk3u_core-3.1.so.5
#9  0x00007fffcaefe280 in  () at /app/bin/_pcbnew.kiface
#10 0x00007fffcaf034f3 in  () at /app/bin/_pcbnew.kiface
#11 0x00007ffff7b21728 in wxWindowBase::TransferDataToWindow() () at /app/lib/libwx_gtk3u_core-3.1.so.5 #12 0x00007ffff7b21728 in wxWindowBase::TransferDataToWindow() () at /app/lib/libwx_gtk3u_core-3.1.so.5
#13 0x00007fffcaf40227 in  () at /app/bin/_pcbnew.kiface
#14 0x00007ffff7b21b32 in wxWindowBase::OnInitDialog(wxInitDialogEvent&) () at /app/lib/libwx_gtk3u_core-3.1.so.5 #15 0x00007ffff746229e in wxEvtHandler::ProcessEventIfMatchesId(wxEventTableEntryBase const&, wxEvtHandler*, wxEvent&) ()
    at /app/lib/libwx_baseu-3.1.so.5
#16 0x00007ffff7462443 in wxEventHashTable::HandleEvent(wxEvent&, wxEvtHandler*) () at /app/lib/libwx_baseu-3.1.so.5 #17 0x00007ffff7462b56 in wxEvtHandler::TryHereOnly(wxEvent&) () at /app/lib/libwx_baseu-3.1.so.5 #18 0x00007ffff7462be0 in wxEvtHandler::ProcessEventLocally(wxEvent&) () at /app/lib/libwx_baseu-3.1.so.5 #19 0x00007ffff7462cf2 in wxEvtHandler::ProcessEvent(wxEvent&) () at /app/lib/libwx_baseu-3.1.so.5 #20 0x00007ffff7b1a2ce in wxWindowBase::InitDialog() () at /app/lib/libwx_gtk3u_core-3.1.so.5 #21 0x00007ffff7972114 in wxDialog::Show(bool) () at /app/lib/libwx_gtk3u_core-3.1.so.5
#22 0x00007fffcafc7e7a in  () at /app/bin/_pcbnew.kiface
#23 0x00007fffcafc68a6 in  () at /app/bin/_pcbnew.kiface
#24 0x00007fffca7a74ad in  () at /app/bin/_pcbnew.kiface
#25 0x00007fffca823117 in  () at /app/bin/_pcbnew.kiface
#26 0x00007fffcb0fd3a0 in  () at /app/bin/_pcbnew.kiface
#27 0x00007fffcb19e211 in make_fcontext () at /app/bin/_pcbnew.kiface
#28 0x0000007800000031 in  ()

Suggestions online are that the crash is to do with using a the address of a local variable out of scope. Having looked at the code, and using SetCellValue as the anchor point, I believe the problem lies in the code in:

dialog_global_edit_tracks_and_vias.cpp: void DIALOG_GLOBAL_EDIT_TRACKS_AND_VIAS::buildNetclassesGrid()

and specifically in the netclass loop:

    for( const auto& netclass : netclasses )
        m_netclassGrid->SetCellValue( row, GRID_NAME, netclass.first );

I'm not quite following where 'first' is defined in 'auto& netclass', however, so cannot pinpoint the actual problem.

Hope this is enough to be useful.


Tel: 01223 414180
Blog: http://www.ivimey.org/blog
LinkedIn: http://uk.linkedin.com/in/ruthivimeycook/

Follow ups