Closed MaxGyver83 closed 1 year ago
The git mirror of libtickit is not up-to-date. Not sure if that'll fix it, but it's a good place to start. If you can't grab the bzr repo directly, I've attached a tarball of rev 810 that you can use.
bzr branch https://bazaar.leonerd.org.uk/c/libvterm/
Please let me know, thanks! -Ross libtickit_r810.tar.gz
Thank you. After building and installing libtickit
from your tarball, I got the same segfault.
So to be sure, I have installed bzr
and cloned the repository and installed again. Same error:
Maybe this helps: I have commented in all occurrences of DEBUG_LOGF
in vts_sb_popline
and logged the "Uvp" flag:
Thanks, here are some questions that might help me:
ldd /usr/local/bin/a4
.$ ldd /usr/local/bin/a4
linux-vdso.so.1 (0x00007ffc37187000)
libtickit.so.3 => /usr/local/lib/libtickit.so.3 (0x00007feb2a123000)
libvterm.so.0 => /usr/local/lib/libvterm.so.0 (0x00007feb2a10b000)
libc.so.6 => /usr/lib/libc.so.6 (0x00007feb29f24000)
libunibilium.so.4 => /usr/lib/libunibilium.so.4 (0x00007feb29f0f000)
libtermkey.so.1 => /usr/lib/libtermkey.so.1 (0x00007feb29f03000)
/lib64/ld-linux-x86-64.so.2 => /usr/lib64/ld-linux-x86-64.so.2 (0x00007feb2a195000)
I use Arch Linux.
I couldn't reproduce this with xterm or in a console. Only with st, both my custom build and a clean build of the latest commit (e5e9598).
And to be more specific about the segfault I get:
Core was generated by `a4/a4'.
Program terminated with signal SIGSEGV, Segmentation fault.
#0 0x00005589fa60753a in vts_sb_popline (cols=148, cells=0x5589fad45ac0, user=0x5589facabaf0) at /home/max/repos/a4/vt.c:167
167 .fg = tframe->cs->fg,
>>> p tframe->cs
$1 = (ColorScheme *) 0x0
>>>
tframe->cs
is a null pointer. Maybe you have an idea what could have caused this.
a4
from bash or fish.)[#1]
) when I run a4
in st
.Screenshot: (top: st, bottom: xterm)
Please apply and test this one line patch. If that resolves the segfault, as I suspect it will, then I'll make a couple more similar updates to make the whole thing more robust. Thanks!
patch <default_cs.txt
Thank you. No crashes so far! tframe->cs
is not NULL anymore (as expected :-) ). But the titlebars are still blank.
Titlebar text is typically controlled in bash by setting PROMPT_COMMAND in your .bashrc file, and then you can make color rules in a4.ini to match the text and change the color scheme of individual terminal windows. Here's the standard example, and you can change the titlebar text in other ways as well for additional control.
PROMPT_COMMAND='echo -ne "\e]0;$USER@$HOSTNAME:${PWD/$HOME/\~}\a"'
Here's a fun example screenshot showing an extreme use of multiple terminal color schemes:
Actually, that patch is not good as it inhibits the initial colorscheme from being applied. This updated one line patch will both set and apply the initial default colors. Please let me know if you're able to validate, thanks!
patch < applydefaultcs.txt
I have tested it and it works as expected! Thank you.
Reclosing with improved fix bce5e14.
Hi Ross,
I just tried a4 (after seeing its announcement on the suckless mailing list). After playing around with it for a minute or two, I get a segmentation fault. I can't reproduce it reliably with a certain sequence of the key strokes, but it happened several times after using these commands (in random order):
(I think
Ctrl-g .
andCtrl-g Enter
were the last before the crash.)Maybe these coredumps help:
Update (with a debug build):
Versions