Opening Fast Preview Window crashes Hugin

[Jonathan Sambrook]

Whenever I open the FPW, Hugin segfaults. I've attached the full backtrace of the segfault triggered by manually opening the FPW, but here are the top 19 lines (of 62):

#0 0x00f7eae1 in wxWindow::DoSetSize(int, int, int, int, int) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#1 0x01072584 in wxSizerItem::SetDimension(wxPoint const&, wxSize const&) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#2 0x01072f6b in wxGridSizer::SetItemBounds(wxSizerItem*, int, int, int, int) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#3 0x010740ff in wxFlexGridSizer::RecalcSizes() () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#4 0x010722d4 in wxSizer::Layout() () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#5 0x01072409 in wxSizer::SetDimension(int, int, int, int) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#6 0x010725c1 in wxSizerItem::SetDimension(wxPoint const&, wxSize const&) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#7 0x01073512 in wxBoxSizer::RecalcSizes() () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#8 0x010722d4 in wxSizer::Layout() () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#9 0x01072409 in wxSizer::SetDimension(int, int, int, int) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#10 0x01084f55 in wxWindowBase::Layout() () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#11 0x0107ee45 in wxTopLevelWindowBase::DoLayout() () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#12 0x0107fd2d in wxTopLevelWindowBase::OnSize(wxSizeEvent&) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#13 0x00d3dc2f in wxAppConsole::HandleEvent(wxEvtHandler*, void (wxEvtHandler::*)(wxEvent&), wxEvent&) const () from /usr/lib/libwx_baseu-2.8.so.0
#14 0x00ddc089 in wxEvtHandler::ProcessEventIfMatches(wxEventTableEntryBase const&, wxEvtHandler*, wxEvent&) () from /usr/lib/libwx_baseu-2.8.so.0
#15 0x00ddd134 in wxEventHashTable::HandleEvent(wxEvent&, wxEvtHandler*) () from /usr/lib/libwx_baseu-2.8.so.0
#16 0x00ddd233 in wxEvtHandler::ProcessEvent(wxEvent&) () from /usr/lib/libwx_baseu-2.8.so.0
#17 0x00fcaf65 in wxFrame::GtkOnSize() () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#18 0x00f73e03 in wxTopLevelWindowGTK::Show(bool) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#19 0x080bb5f1 in MainFrame::OnToggleGLPreviewFrame(wxCommandEvent&) ()

Running Ubuntu Maverick on an Intel DG45ID mobo with the G45 chipset and builtin GMA X4500HD GPU

About→System says:

 Operating System: Linux 2.6.35-25-generic i686
 Architecture: 32 bit
 Free memory: -1738088 kiB

Hugin
Version: 2010.4.0.854952d82c8f
Path to ressources: /usr/share/hugin/xrc/
Path to data: /usr/share/hugin/data/

[Yuv (yuv)]

Not sure if it is the same case,
2011.0.0.5338677e592c segfaults on me when I open the fast preview after loading a relatively big project.

The output I see in gdb is

$ gdb hugin
GNU gdb (GDB) 7.1-ubuntu
Copyright (C) 2010 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law. Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-linux-gnu".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /usr/local/bin/hugin...done.
(gdb) run
Starting program: /usr/local/bin/hugin
[Thread debugging using libthread_db enabled]
MainFrame::RestoreLayoutOnNextResize()
Number of images 98
[New Thread 0x7fffe3515700 (LWP 9877)]
[Thread 0x7fffe3515700 (LWP 9877) exited]
[New Thread 0x7fffe3515700 (LWP 9878)]
[Thread 0x7fffe3515700 (LWP 9878) exited]

Program received signal SIGSEGV, Segmentation fault.
PreviewIdentifyTool::ShowImageNumber (this=0x0, image=<value optimized out>)
    at /home/yuv/src/hugin/hugin.hg/src/hugin1/hugin/PreviewIdentifyTool.cpp:476
476 DEBUG_ASSERT(image < helper->GetPanoramaPtr()->getNrOfImages());

[Jonathan Sambrook (jonathan-hmmn)]

Latest nightly build (2010.5.0+2011.0.0+hg5191-0ubuntu1~maverick ) now crashes on startup on Ubuntu Maverick and so not even the old style UI is available.

$ gdb /usr/bin/hugin
GNU gdb (GDB) 7.2-ubuntu
Copyright (C) 2010 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law. Type "show copying"
and "show warranty" for details.
This GDB was configured as "i686-linux-gnu".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /usr/bin/hugin...(no debugging symbols found)...done.
(gdb) r
Starting program: /usr/bin/hugin
[Thread debugging using libthread_db enabled]
Xlib: extension "NV-GLX" missing on display ":0.0".

Program received signal SIGSEGV, Segmentation fault.
0x00f80ae1 in wxWindow::DoSetSize(int, int, int, int, int) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
(gdb) bt
#0 0x00f80ae1 in wxWindow::DoSetSize(int, int, int, int, int) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#1 0x010745a4 in wxSizerItem::SetDimension(wxPoint const&, wxSize const&) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#2 0x01075532 in wxBoxSizer::RecalcSizes() () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#3 0x010742f4 in wxSizer::Layout() () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#4 0x01074429 in wxSizer::SetDimension(int, int, int, int) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#5 0x01086f75 in wxWindowBase::Layout() () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#6 0x010b4e5d in wxPanel::OnSize(wxSizeEvent&) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#7 0x00d3fc2f in wxAppConsole::HandleEvent(wxEvtHandler*, void (wxEvtHandler::*)(wxEvent&), wxEvent&) const () from /usr/lib/libwx_baseu-2.8.so.0
#8 0x00dde089 in wxEvtHandler::ProcessEventIfMatches(wxEventTableEntryBase const&, wxEvtHandler*, wxEvent&) () from /usr/lib/libwx_baseu-2.8.so.0
#9 0x00ddf134 in wxEventHashTable::HandleEvent(wxEvent&, wxEvtHandler*) () from /usr/lib/libwx_baseu-2.8.so.0
#10 0x00ddf233 in wxEvtHandler::ProcessEvent(wxEvent&) () from /usr/lib/libwx_baseu-2.8.so.0
#11 0x00f80ca6 in wxWindow::DoSetSize(int, int, int, int, int) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#12 0x010745a4 in wxSizerItem::SetDimension(wxPoint const&, wxSize const&) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#13 0x01075532 in wxBoxSizer::RecalcSizes() () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#14 0x010742f4 in wxSizer::Layout() () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#15 0x01074429 in wxSizer::SetDimension(int, int, int, int) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#16 0x010745e1 in wxSizerItem::SetDimension(wxPoint const&, wxSize const&) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#17 0x010754ab in wxBoxSizer::RecalcSizes() () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#18 0x010742f4 in wxSizer::Layout() () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#19 0x01074429 in wxSizer::SetDimension(int, int, int, int) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#20 0x010745e1 in wxSizerItem::SetDimension(wxPoint const&, wxSize const&) () from /usr/lib/libwx_gtk2u_core-2.8.so.0
#21 0x01075532 i...

[Tristan Burtenshaw (tburtenshaw)]

Also crashes for me on Windows 2000.

It doesn't seem to cause any exception, but just runs as a "non-responding" application. "End task" quits it easily without giving any information.

My current workaround is to avoid Fast Preview (I have turned this off from coming up automatically), but this is not satisfactory, so I'll download the old version.

[Vladimir Nadvornik (nadvornik-suse)]

I have the same problem.
I have reproduced it with valgrind.

It seems that TextureManager::SetPhotometricCorrect deletes some data that are used elsewhere. Full stack trace is in the log.

[Vladimir Nadvornik (nadvornik-suse)]

It looks like that boost::signals::trackable does not work at all. Even the boost example fails for me.

http://svn.boost.org/svn/boost/trunk/libs/signals/example/button_click.cpp

Tested on openSUSE 11.3 and 11.4 with boost 1.42, 1.44, 1.46.
Can anybody test other systems?

nadvornik@sphinx2:~> wget http://svn.boost.org/svn/boost/trunk/libs/signals/example/button_click.cpp
nadvornik@sphinx2:~> gcc button_click.cpp -o button_click -lboost_signals
nadvornik@sphinx2:~> ./button_click
OK!
OK!

# "OK!" should be printed just once

[Felix Hagemann (flixh)]

> Tested on openSUSE 11.3 and 11.4 with boost 1.42, 1.44, 1.46.
> Can anybody test other systems?

I'm not too sure what I'm testing here, but:
$ ./button_click
OK!
$

Tested on Debian testing, boost 1.46

Fast preview has not been crashing for me for a long time now.
Hope this helpss,
Felix

[Stefan Peter (s-peter-deactivatedaccount)]

Hi Vladimir

On 25.08.2011 21:18, Vladimir Nadvornik wrote:
> It looks like that boost::signals::trackable does not work at all. Even
> the boost example fails for me.
...
> # "OK!" should be printed just once
>

This is under Ubuntu Natty X86_64:

stefan@nbstepet:~/src$ gcc button_click.cpp -o button_click -lboost_signals
stefan@nbstepet:~/src$ ./button_click
OK!
stefan@nbstepet:~/src$

I got the same results under Maverick, both for 32 and 64 bits.

Regards

Stefan Peter
--
In theory there is no difference between theory and practice. In
practice there is.

[Vladimir Nadvornik (nadvornik-suse)]

Thanks for testing.
Filled https://bugzilla.novell.com/show_bug.cgi?id=714373

[KFJ (kfj)]

I posted this to hugin-ptx, maybe what I found is helpful:

Hi group!

I've recently been annoyed by the openGL preview crashing when I had previously disabled the panosphere view and then restarted hugin. Hugin would crash a few times, but sometimes after some blind trying this and that would come round again. So I launched hugin under gdb from a console window and got a hint at where the problem occurs:

Program received signal SIGSEGV, Segmentation fault.
PreviewIdentifyTool::UpdateWithNewImageSet (this=0x0, new_image_set=...)
    at /home/kfj/src/hugin/hugin_clean.hg/src/hugin1/hugin/PreviewIdentifyTool.cpp:567
567 std::vector<unsigned int> difference(new_image_set.size());

this = 0x0 definitely looks bad, but I also found that of new_image_set.size() was zero. After a bit of trial and error I ended up simply putting the whole body of code inside the method in question into a conditional:

void PreviewIdentifyTool::UpdateWithNewImageSet(std::set<unsigned int> new_image_set)
{
    if ( new_image_set.size() )
     {
      ...
     }
}

now the error was gone. I might just have been lucky, but since then I could not get the bug to occur. Since I haven't really got a clue what the code does, I'm not sure if my quick shot will be a solution, but maybe it's a start :)

Kay

[tmodes (tmodes)]

Hi Kay,

I've taken your findings as starting point. I implemented a slightly different solution in changeset ed931cca4d44. Have you tested it yet? Does it still crashes, or does it works?

[Mark Winter (thewinterstudio)]

I'm running windows 7 and having this problem also. I've uninstalled and reinstalled this several times. After three tries at this I did get it to work for about a week. I would open the fast view window first then load the images. This worked for a while then it went back to crashing upon opening the fast view.

[Giulio Canevari (giulio-canevari)]

Using 12.04 and hugin 2011.4.0+dfsg-1, here is a video of what happens (due to this hugin is of course unusable):

http://ubuntuone.com/5szJQu1i6YTRk6BpA3S6pc

Linux computer-System-Product-Name 3.2.0-27-generic-pae #43-Ubuntu SMP Fri Jul 6 15:06:05 UTC 2012 i686 i686 i386 GNU/Linux

00:00.0 Host bridge: Silicon Integrated Systems [SiS] 661FX/M661FX/M661MX Host (rev 11)
00:01.0 PCI bridge: Silicon Integrated Systems [SiS] AGP Port (virtual PCI-to-PCI bridge)
00:02.0 ISA bridge: Silicon Integrated Systems [SiS] SiS964 [MuTIOL Media IO] LPC Controller (rev 36)
00:02.5 IDE interface: Silicon Integrated Systems [SiS] 5513 IDE Controller (rev 01)
00:02.7 Multimedia audio controller: Silicon Integrated Systems [SiS] SiS7012 AC'97 Sound Controller (rev a0)
00:03.0 USB controller: Silicon Integrated Systems [SiS] USB 1.1 Controller (rev 0f)
00:03.1 USB controller: Silicon Integrated Systems [SiS] USB 1.1 Controller (rev 0f)
00:03.2 USB controller: Silicon Integrated Systems [SiS] USB 1.1 Controller (rev 0f)
00:03.3 USB controller: Silicon Integrated Systems [SiS] USB 2.0 Controller
00:05.0 IDE interface: Silicon Integrated Systems [SiS] SATA (rev 01)
00:09.0 USB controller: ALi Corporation USB 1.1 Controller (rev 03)
00:09.1 USB controller: ALi Corporation USB 1.1 Controller (rev 03)
00:09.2 USB controller: ALi Corporation USB 1.1 Controller (rev 03)
00:09.3 USB controller: ALi Corporation USB 2.0 Controller (rev 01)
00:0b.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139/8139C/8139C+ (rev 10)
01:00.0 VGA compatible controller: Advanced Micro Devices [AMD] nee ATI RV280 [Radeon 9200 SE] (rev 01)
01:00.1 Display controller: Advanced Micro Devices [AMD] nee ATI RV280 [Radeon 9200 SE] (Secondary) (rev 01)

[tmodes (tmodes)]

Does this still happens with current versions? (2014.0 rc1 was already released)

[Andreas Metzler (k-launchpad-downhill-at-eu-org)]

Hello,

I just got a similar report on Debian 2014.0 rc1 http://bugs.debian.org/741466

cu Andreas
-------------------------------
Philipp Marek writes:

With "Interface - Simple", clicking on _any_ tab (except the first visible,
"Assistant") crashes Hugin with a segfault; please see the backtrace attached.

The "Expert" Interface works better, but I cannot do any previews.

After comparing with #711997 I can offer further this information:

* glxgears just works:
  # glxgears -info
  Running synchronized to the vertical refresh. The framerate should be
  approximately the same as the monitor refresh rate.
  GL_RENDERER = Mesa DRI Intel(R) Ironlake Mobile
  GL_VERSION = 2.1 Mesa 9.2.2
  GL_VENDOR = Intel Open Source Technology Center

* No Nvidia traces in /var/log/Xorg.0.log; this is a notebook with an Intel
graphics card, and has never seen the Nvidia proprietary driver.

  # grep module -i /var/log/Xorg.0.log looks clean:
  [ 59.155] (II) LoadModule: "glx"
  [ 59.164] (II) Loading /usr/lib/xorg/modules/extensions/libglx.so
  [ 59.215] (II) Module glx: vendor="X.Org Foundation"
  [ 59.215] compiled for 1.15.0, module version = 1.0.0
  [ 59.215] (II) LoadModule: "dbe"
  [ 59.215] (II) Module "dbe" already built-in
  [ 59.215] (II) LoadModule: "extmod"
  [ 59.215] (II) Module "extmod" already built-in
  [ 59.215] (II) LoadModule: "intel"
  [ 59.215] (II) Loading /usr/lib/xorg/modules/drivers/intel_drv.so
  [ 59.274] (II) Module intel: vendor="X.Org Foundation"
  [ 59.274] compiled for 1.15.0, module version = 2.21.15
  [ 59.274] Module class: X.Org Video Driver
  [ 59.448] (II) Loading sub module "fb"
  [ 59.448] (II) LoadModule: "fb"
  [ 59.448] (II) Loading /usr/lib/xorg/modules/libfb.so
  [ 59.459] (II) Module fb: vendor="X.Org Foundation"
  [ 59.459] compiled for 1.15.0, module version = 1.0.0
  [ 59.459] (II) Loading sub module "dri2"
  [ 59.459] (II) LoadModule: "dri2"
  [ 59.459] (II) Module "dri2" already built-in

-- System Information:
Debian Release: jessie/sid
  APT prefers testing
  APT policy: (990, 'testing'), (500, 'unstable'), (500, 'stable'), (1,
'experimental')
Architecture: amd64 (x86_64)

Kernel: Linux 3.13-1-amd64 (SMP w/4 CPU cores)
Locale: LANG=de_AT.UTF-8, LC_CTYPE=de_AT.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages hugin depends on:
ii enblend 4.1.2+dfsg-2
ii enfuse 4.1.2+dfsg-2
ii hugin-tools 2014.0.0~rc1+dfsg-2
ii libboost-system1.55.0 1.55.0-1
ii libboost-thread1.55.0 1.55.0-1
ii libc6 2.18-4
ii libexiv2-12 0.23-1
ii libgcc1 1:4.8.2-16
ii libgl1-mesa-glx [libgl1] 9.2.2-1
ii libglew1.10 1.10.0-3
ii libglu1-mesa [libglu1] 9.0.0-2
ii libimage-exiftool-perl 9.53-1
ii libpano13-3 2.9.19~beta1+dfsg-2
ii libstdc++6 4.8.2-16
ii libtiff5 4.0.3-8
ii libwxbase2.8-0 2.8.12.1+dfsg-2
ii libwxgtk2.8-0 2.8.12.1+dfsg-2
ii make 3.82-1

hugin recommends no packages.

hugin sug...

[tmodes (tmodes)]

I have no idea what goes here wrong:

Program received signal SIGSEGV, Segmentation fault.
0x000000000058b1f3 in PreviewIdentifyTool::setConstantOn(bool) ()

PreviewIdentifyTool::setConstantOn is a very simple member function:

void PreviewIdentifyTool::setConstantOn(bool constant_on_in) {
    constantOn = constant_on_in;
    if (constant_on_in) {
        stopUpdating = false;
        ContinueUpdating();
    } else {
        stopUpdating = true;
        StopUpdating();
    }
}

So there are only 2 bool assignments and then it calls another function (which is not mentioned in the log).

Are there any possibilities to encircle the crash further?

[Erik V (e7ik-deactivatedaccount)]

I am new to Hugin. I spend a whole trying things. Several times the program just stopped all of a sudden. And the last time, I started browsing this forum and I can confirm I have the same problem.
Mac OS X 10.9.4

I am just a user, so I have no clue what to do....

[Frantisek Boranek (fboranek)]

The problem stopped as soon as I'am disabled Overview (Menu->View->Overview).
It crashes each time when I load more then one photos.

Process: Hugin [2234]
Path: /Users/USER/*/Hugin.app/Contents/MacOS/Hugin
Identifier: net.sourceforge.hugin.Hugin
Version: 2014.0.0 (483c40467492)
Code Type: X86-64 (Native)
Parent Process: ??? [1]
Responsible: Hugin [2234]
User ID: 501

Date/Time: 2015-05-31 22:14:19.184 +0200
OS Version: Mac OS X 10.10.3 (14D136)
Report Version: 11
Anonymous UUID: 355FF23F-066D-60D7-4364-7D59C3DC7FA6

Sleep/Wake UUID: 3FF2FFDB-12D5-408E-8A8B-83EF2248DB1B

Time Awake Since Boot: 12000 seconds
Time Since Wake: 900 seconds

Crashed Thread: 0 Dispatch queue: com.apple.main-thread

Exception Type: EXC_BAD_ACCESS (SIGSEGV)
Exception Codes: KERN_INVALID_ADDRESS at 0x000000010f700000

VM Regions Near 0x10f700000:
    MALLOC_LARGE 000000010f400000-000000010f700000 [ 3072K] rw-/rwx SM=PRV
-->
    MALLOC_TINY 000000010fb00000-000000010fc00000 [ 1024K] rw-/rwx SM=PRV

Thread 0 Crashed:: Dispatch queue: com.apple.main-thread
0 libGLU.dylib 0x00007fff938a5512 halveImage_ubyte(int, unsigned int, unsigned int, unsigned char*, unsigned char*, int, int, int) + 146
1 libGLU.dylib 0x00007fff9389c2f7 gluBuild2DMipmapsCTXCore(_CGLContextObject*, unsigned int, int, int, int, int, int, unsigned int, unsigned int, int, int, int, void const*) + 3673
2 libGLU.dylib 0x00007fff938a8c2e gluBuild2DMipmapsCTX + 300
3 libGLU.dylib 0x00007fff938a8c71 gluBuild2DMipmaps + 52
4 net.sourceforge.hugin.Hugin 0x00000001001b13f3 TextureManager::TextureInfo::DefineLevels(int, int, bool, HuginBase::PanoramaOptions const&, HuginBase::SrcPanoImage const&) + 3603
5 libhugin_base-2014.0.dylib 0x0000000101e064d8 boost::function3<void, boost::shared_ptr<HuginBase::ImageCache::Entry>, std::string, bool>::operator()(boost::shared_ptr<HuginBase::ImageCache::Entry>, std::string, bool) const + 104
6 libhugin_base-2014.0.dylib 0x0000000101e0636b boost::signals2::detail::signal3_impl<void, boost::shared_ptr<HuginBase::ImageCache::Entry>, std::string, bool, boost::signals2::optional_last_value<void>, int, std::less<int>, boost::function<void (boost::shared_ptr<HuginBase::ImageCache::Entry>, std::string, bool)>, boost::function<void (boost::signals2::connection const&, boost::shared_ptr<HuginBase::ImageCache::Entry>, std::string, bool)>, boost::signals2::mutex>::slot_invoker::m_invoke(boost::shared_ptr<boost::signals2::detail::connection_body<std::pair<boost::signals2::detail::slot_meta_group, boost::optional<int> >, boost::signals2::slot3<void, boost::shared_ptr<HuginBase::ImageCache::Entry>, std::string, bool, boost::function<void (boost::shared_ptr<HuginBase::ImageCache::Entry>, std::string, bool)> >, boost::signals2::mutex> > const&, boost::signals2::detail::void_type const*) const + 91
7 libhugin_base-2014.0.dylib 0x0000000101e062a8 boost::signals2::detail::slot_call_iterator_t<boost::s...

[tmodes (tmodes)]

Several uninitialized variables fixed in default repository. This should fix the crash.