AAO25.com
Assist => Support => Assist Support => Topic started by: ultr on Tuesday, March 22, 2016, 22:21:31 PM
-
Since yesterday Assist crashes with a segfault in Linux.
Application starts normally, but crashes after a successful login.
It worked perfectly several days ago (I played on 19.03.2016).
I haven't updated any related or system library since the last time it was working.
dmesg:
[14358.835538] 25Assist[12561]: segfault at f4452298 ip 00000000f6ff55de
sp 00000000f4452290 error 7 in libgdk_pixbuf-2.0.so.0.3100.1[f6fed000+26000]
gdb:
[New Thread 0xf4cc0b40 (LWP 7285)]
[New Thread 0xf4cafb40 (LWP 7286)]
[New Thread 0xf4c9eb40 (LWP 7287)]
[Thread 0xf4cafb40 (LWP 7286) exited]
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0xf4c9eb40 (LWP 7287)]
0xf782f5de in load_from_stream (loader=loader@entry=0x876a550,
stream=stream@entry=0xf4907fc8, cancellable=0x0, error=0xf48338cc) at
gdk-pixbuf-io.c:1433
Debian 8 Jessie (stable), x86_64
-
Try to re install "ia32-libs / ia32-libs-gtk" and/or re-download 25 Assist client .
Libs:
http://www.debian-administration.org/article/531/Using_proprietary_i386_apps_on_an_amd64_system (http://www.debian-administration.org/article/531/Using_proprietary_i386_apps_on_an_amd64_system)
25 Assist Client:
http://sourceforge.net/projects/aa25assist/files/Binaries/25Assist-Linux.zip/download (http://sourceforge.net/projects/aa25assist/files/Binaries/25Assist-Linux.zip/download)
-
Debian supports multiarch now, and I have all required i386 libraries installed. As I said, it worked perfectly just some days ago.
Of course I tried reinstalling the Assist Client as well.
Is there source code for the Assist Client available, so I can debug this?
-
libgdk_pixbuf has many reports online for bugs regarding support for jpeg2000 images with transparency.
Maybe (just maybe) such image is now sent to the client after logon (though Wireshark does not report any) or maybe some command triggers its load from the local Assist's resources.
That's all I could find so far. And without the source code I cannot debug this.
-
I've got a feeling it might have something to do with "show poll" critical error as that's the most recent "change" to Assist - the polls section was never really meant for polls with so many choices.
The poll went live Tuesday evening and it damn well suites the timeline as well.
Not entirely sure, though.
-
I've got a feeling it might have something to do with "show poll" critical error as that's the most recent "change" to Assist - the polls section was never really meant for polls with so many choices.
The poll went live Tuesday evening and it damn well suites the timeline as well.
Not entirely sure, though.
if it's about this, I make the poll to expire tomorrow
-
I get also errors because of the poll but it doesnt crash my assist. Well Im nowadays using Windows.
-
I get also errors because of the poll but it doesnt crash my assist. Well Im nowadays using Windows.
It doesn't crash the mac version either, but the error is shown every time Assist is launched. I'll close the poll today to make the error disappear.
-
I guess it was that poll, because the issue is no more now.
Thanks.
-
I guess it was that poll, because the issue is no more now.
Thanks.
See, ronski? Don't do such long polls anymore! :)
-
See, ronski? Don't do such long polls anymore! :)
:sign_shithappens:
It would be interesting to know which part of it failed, since multivote poll could be useful in future as well. Was it the amount of options (20, Assist showed 10) or the multivote itself?
-
First and the most important think that failed is RealBasic.
-
See, ronski? Don't do such long polls anymore! :)
Since I didn't write the code (and if I'm honest, I've never looked at it), I really don't know...