Am Mittwoch, 3. Mai 2017, 18:16:49 CEST schrieb Jefferson Ferreira:
> Hi,
> 
> I found the real problem. The problem occurs when I add the following line
> in darktablerc:
> 
> lighttable/ui/single_module=TRUE
> 
> If I remove this line, my lua code works perfectly, and the darktable does
> not break when I click on the export module.

Should be fixed in master now.

> []'s
> 
> On Wed, May 3, 2017 at 3:46 AM, Tobias Ellinghaus <m...@houz.org> wrote:
> > Am Dienstag, 2. Mai 2017, 16:37:20 CEST schrieb Jefferson Ferreira:
> > > Hi,
> > > 
> > > This link has the content of gdb.txt with the output of gdb. My luarc
> > > has
> > > just the code that I mentioned at the first email.
> > > 
> > > https://www.dropbox.com/s/pjydo45jrndu8o5/gdb.txt?dl=0
> > 
> > Thanks, that looks familiar. Initially I had that, too, but somehow never
> > got
> > it again. It seems to be due to how lua handles its objects internally. I
> > will
> > ask Jeremy for advice.
> > 
> > For those curious, the relevant part is
> > 
> > Thread 1 (Thread 0x7ffff7ee0ac0 (LWP 15239)):
> > #0  get_position_description (cur_view=0x2295300, gui_data=0x26c73d0) at /
> > home/jeff/Tools/Darktable/darktable/src/lua/lualib.c:94
> > 
> >         position_description = <optimized out>
> >         iter = 0x1
> > 
> > #1  0x00007ffff7ab7314 in container_wrapper (self=<optimized out>) at
> > /home/
> > jeff/Tools/Darktable/darktable/src/lua/lualib.c:104
> > 
> >         cur_view = <optimized out>
> >         gui_data = 0x26c73d0
> >         position_description = <optimized out>
> > 
> > #2  0x00007ffff7a9d36b in _lib_plugin_header_button_press (w=<optimized
> > out>,
> > e=<optimized out>, user_data=0x26c4c00) at /home/jeff/Tools/Darktable/
> > darktable/src/libs/lib.c:929
> > 
> >         m = 0x2cb0b70
> >         it = 0x2067840
> >         v = 0x2295300
> >         all_other_closed = 1
> >         container = 4
> >         w = <optimized out>
> >         user_data = 0x26c4c00
> >         e = <optimized out>
> >         module = 0x26c4c00
> > 
> > where iter is suddenly "1" instead of a valid pointer. To me it looks like
> > somewhere the struct holding the lib data is copied without taking into
> > account that there are pointers inside.
> > 
> > > Thanks in advance.
> > 
> > Tobias
> > 
> > [...]

Attachment: signature.asc
Description: This is a digitally signed message part.

Reply via email to