guile-user
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: GNU G-Golf 0.8.0-rc-3 available for testing


From: David Pirotte
Subject: Re: GNU G-Golf 0.8.0-rc-3 available for testing
Date: Sat, 4 May 2024 02:59:29 -0300

Hi Florian,

> > The true error location was given by the gdb backtrace, the
> > gtk-widget-snapshot-child call, presumably with a bad pointer or so
> > ...  

> Hmm my assumption was that all callbacks are broken, not vfuncs
> specifically ...

Very unlikely, but the important thing is to get a proper full
backtrace, as then there is no need to make assumptions ...

> The GTK_A11Y warning I can easily fix by exporting GTK_A11Y=none, but
> it was interesting that the crash happens two function calls away from
> where the warning is printed.

I can't make any useful deduction from the above, a warning is expected
here, till you (the guix team) fix it, let's set this env var, and
concentrate on the cause of the segfault ...


> On a x86_64 GNOME laptop on the distro Guix System on the g-golf devel
> branch with (main '("guile" "--debug")):

> ;;;

this gdb backtrace 'preamble' is, to me, useless as well

> On my aarch64 rock64 the complete backtrace is useless:

> (gdb) bt
> #0  0x0000ffff8066e278 in ?? ()
> #1  0x000000001b976640 in ?? ()

I won't be able to help you help Guix track this problem unless we get a
useful backtrace, so much as because i probably will need it to ask for
some help from upstream devs ...

Could you ask for some help from guix to produce it?

> Your next e-mail:

> I get only
> 
> scheme@(guile-user)> (main '("guile" "--debug"))
> ;; g-application-run 
> ;;          application: #<<gtk-application> 7fce23e673b0> 
> ;;                 argc: 0 
> ;;                 argv: () 
> ;; gtk-window-set-child 
> ;;               window: #<<gtk-application-window> 7fce23e730c0> 
> ;;                child: #<<drawing-widget> 7fce23e7bd10> 
> ;; gtk-window-present 
> ;;               window: #<<gtk-application-window> 7fce23e730c0> 

Yes, i did expect this and wanted to show you ... you don't see
any g-golf trace here because it fails in the gtk engine, as i
understood so far, before it can't find the derived class snapshot
vfunc.

David

ps:     you did not answer my quiz about your guix host distro - i
would like you to - at least temporarily - manually install g-golf
there and see if everything works fine ... Could you do that? thanks

Attachment: pgpPjUzglv6aJ.pgp
Description: OpenPGP digital signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]