I'm sorry, I jumped to conclusions. It still crashes. It doesn't crash
if just run it in TUI mode without any program for debugging (just
`gdb -tui <enter>`)
But if I run it with even this simple program:
int main(void)
{
return 0;
}
it crashes with this backtrace:
#0 0x0000000000ebe6ed in handle_fatal_signal(int)::{lambda(char
const*)#1}::operator()(char const*) const [clone .constprop.0] ()
[Current thread is 1 (process 28340)]
(gdb) bt
#0 0x0000000000ebe6ed in handle_fatal_signal(int)::{lambda(char
const*)#1}::operator()(char const*) const [clone .constprop.0] ()
#1 0x0000000000ebe788 in handle_fatal_signal(int) ()
#2 0x0000000000ebe8fd in handle_sigsegv(int) ()
#3 <signal handler called>
#4 0x000078ad2334bde3 in getmaxx () from /usr/lib/libcurses.so.9
#5 0x0000000000d8e0bd in tui_source_window_base::validate_scroll_offsets() ()
#6 0x0000000000d8e56e in
tui_source_window_base::update_source_window_as_is(gdbarch*,
symtab_and_line const&) ()
#7 0x0000000000d8ed14 in tui_source_window_base::rerender() ()
#8 0x0000000000d96dbf in tui_layout_split::apply(int, int, int, int, bool) ()
#9 0x0000000000d95620 in tui_apply_current_layout(bool) ()
#10 0x0000000000d95933 in tui_set_layout(tui_layout_split*) ()
#11 0x0000000000d8a131 in tui_enable() ()
#12 0x0000000000d8a3d8 in tui_rl_switch_mode(int, int) ()
#13 0x000000000122e598 in _rl_dispatch_subseq ()
#14 0x000000000122efdb in _rl_dispatch_callback ()
#15 0x000000000120e234 in rl_callback_read_char ()
#16 0x0000000000ebfaf7 in gdb_rl_callback_read_char_wrapper_noexcept() ()
#17 0x0000000000ebfd4f in gdb_rl_callback_read_char_wrapper(void*) ()
#18 0x0000000000d86d8a in stdin_event_handler(int, void*) ()
#19 0x0000000001197d06 in gdb_wait_for_event(int) [clone .part.0] ()
#20 0x0000000001198430 in gdb_do_one_event(int) ()
#21 0x0000000000e4081e in captured_command_loop() ()
#22 0x0000000000e431b8 in gdb_main(captured_main_args*) ()
#23 0x000000000127d198 in main ()
It seems there is something wrong with getmaxx() from /usr/lib/libcurses.so.9
On Mon, Dec 16, 2024 at 4:49 PM Vitaly Shevtsov <shev.vt1...@gmail.com> wrote:
>
> No more crashes in tui mode. Thanks, Martin.
>
> On Wed, Dec 11, 2024 at 6:06 PM Martin Husemann <mar...@duskware.de> wrote:
> >
> > Sorry, I was reporting the -current state from a machine with a local
> > patch. Actually -current has 15.1 (and it is not working very well).
> >
> > Martin
