[SOLVED] UT99 not start

seth · 2023-01-28 22:31:09

Do you have a backtrace for the patched version? (You'll need lib32-mesa)

freeartist · 2023-01-29 00:58:25

Don't know. Why for patched ut? Debug needed for UNpatched.

upd
how make backtrace for UNpatched ut?

Last edited by freeartist (2023-01-29 02:13:18)

seth · 2023-01-29 07:51:57

You can post both, but the unpatched version will likely be the same meaningless crash in /home/freeartist/ut/System/Core.so
Since the patched version works w/ nvidia, the backtrace might reveal a problem in your GL setup.

freeartist · 2023-01-29 19:06:27

and how do this backtrace, tell me please?

>problem in your GL setup

not my, but arch

Last edited by freeartist (2023-01-29 19:12:20)

seth · 2023-01-29 19:12:01

Same as before: https://bbs.archlinux.org/viewtopic.php … 3#p2081273

not my, but arch

You configured the system, right?

freeartist · 2023-01-29 19:14:25

>You configured the system, right?

I do fresh install and configurated according archwiki and your suggestions

>Same as before

so again dumps needed? its will be same as before?

Last edited by freeartist (2023-01-29 19:16:19)

seth · 2023-01-29 19:32:25

so again dumps needed? its will be same as before?

Yes. It'll be the same process to get those dumps, whether the content is the same we'll see.

freeartist · 2023-01-29 23:37:10

UNpatched ut

gdb

https://pastebin.com/raw/yqKvvK6c

https://pastebin.com/raw/79hm0Xsi

coredump

https://pastebin.com/raw/DcEtRfcn

https://pastebin.com/raw/T5BrxFKs

Last edited by freeartist (2023-01-29 23:54:36)

seth · 2023-01-30 08:06:05

The coredump is 5 days old and looks like from your mixed GL setup.
You initially tried to "bt" in gdb, figured that you first need to run the process, had it segfault somewhere in libX11 when flushing the event queue (which is meaningless, though are we back to a dual keyboard layout?) and then forgot to "bt" *after* the crash.

=> Get an updated coredump from your presently failing nouveau setup.

freeartist · 2023-01-30 12:31:50

ut always freeze system at gdb then crashed i must kill ut process from another tty

may be its all because i do not full xfce4 setup? just
pacman -S xfce4
not all of that https://www.debugpoint.com/xfce-arch-li … tall-4-16/

or may be again keyboard issue, because ut sarts, i see flying demo intro 3d map, but then press anykey ut crash then main menu must appear?

new dumps not created

Last edited by freeartist (2023-01-30 14:54:16)

seth · 2023-01-30 14:35:28

may be its all because i do not full xfce4 setup?

No.

ut always freeze system at gdb then crashed i must kill ut process from another tty

This is not how gdb works.

gdb ./ut 2>&1 | tee ut.gdb
run
# wait for crash
bt
detach
quit

The output should be logged in ut.gdb

or may be again keyboard issue

setxkbmap -print -query

We already know that it doesn't like dual-layout setups.

freeartist · 2023-01-30 20:52:18

>This is not how gdb works.

Its all freeze at crash moment. I need kill ut-bin from another tty.

ut.gdb
https://pastebin.com/raw/qv56f4Wc

[freeartist-arch@home ~]$ setxkbmap -print -query
xkb_keymap {
	xkb_keycodes  { include "evdev+aliases(qwerty)"	};
	xkb_types     { include "complete"	};
	xkb_compat    { include "complete"	};
	xkb_symbols   { include "pc+us+inet(evdev)"	};
	xkb_geometry  { include "pc(pc105)"	};
};
rules:      evdev
model:      pc105
layout:     us
[freeartist-arch@home ~]$

Last edited by freeartist (2023-01-30 20:54:19)

seth · 2023-01-30 21:05:39

Its all freeze at crash moment. I need kill ut-bin from another tty.

Do not run ut99 in gdb, you should really get a coredump for that.

I need kill ut-bin

Wait, "ut-bin"? Is "./ut" a script that runs ut-bin?

Otherwise

i see flying demo intro 3d map, but then press anykey ut crash then main menu must appear?

From a different TTY

echo 0 | sudo tee /proc/sys/kernel/yama/ptrace_scope # this allows you to attach gdb to other proceses
gdb --attach $(pidof ut-bin) 2>&1 | tee ut.gdb # or instead of $(pidof ut-bin), pick the proper PID somehow
% continue # move back to the game, make it crash
% bt
% detach
% quit

freeartist · 2023-01-31 00:15:59

gdb attach but
ut freezes when i move back to tty7

ut.gdb
https://pastebin.com/raw/gX54rPpE

Last edited by freeartist (2023-01-31 00:23:18)

seth · 2023-01-31 09:06:05

There's still no backtrace in that log (which is also not from an atatchement but a subprocess invocation)

ut freezes when i move back to tty7

Did you "continue" in gdb?
If you return to the gdb TTY, is there now a segfault recorded and can you enter "bt" in gdb?

freeartist · 2023-01-31 21:21:47

SDL render
https://pastebin.com/raw/PkvqHmX4

OpenGL render
https://pastebin.com/raw/Y1Mrf90K

Coredump gdb
https://pastebin.com/raw/CqW3vBWU

Last edited by freeartist (2023-01-31 21:33:36)

seth · 2023-01-31 21:39:49

#13 0xf48b69db in ?? () from /usr/lib32/nvidia/libnvidia-glcore.so.340.108

I already kinda suspected that you've not properly switched to nouveau.
Move away /etc/ld.so.conf.d/00-nvidia.conf and /etc/ld.so.conf.d/00-lib32-nvidia.conf, reboot, try again.

freeartist · 2023-01-31 21:59:56

at this way i have 2 files with same content, move both?

/etc/ld.so.conf.d/00-nvidia.conf   
/usr/lib/nvidia/

/etc/ld.so.conf.d/00-lib32-nvidia.conf     
/usr/lib32/nvidia/

Last edited by freeartist (2023-01-31 22:00:11)

seth · 2023-01-31 22:01:47

Yes, they'll cause libGL being resolved as the nvidia library which are incompatible w/ nouveau.
Therefore those files must be moved out of that path (eg. to /etc/ld.so.conf.bak), then reboot to clear the stage and try again.

freeartist · 2023-01-31 22:09:48

same

>I already kinda suspected that you've not properly switched to nouveau.
And how correct? I use yay for nvidia driver, its all done by yay.

Last edited by freeartist (2023-01-31 22:10:30)

seth · 2023-01-31 22:22:00

same

That's not possible - you cannot use the nvidia libraries anymore.
Please post the current xorg log, https://wiki.archlinux.org/title/Xorg#General and the output of

ldd /usr/bin/glxinfo; ldd /usr/bin/glxinfo32

And how correct? I use yay for nvidia driver, its all done by yay.

You would have to completely uninstall all nvidia packages
But if you blacklist nvidia, unblacklist nouveau, make sure the xorg config doesn't reference nvidia and remove /usr/lib*/nvidia/ from the ld path (what should™ be achieved by removing those ld.so.conf.d files) you can keep the packages installed.

freeartist · 2023-01-31 22:30:51

X log
https://pastebin.com/raw/2XbEiz4E

[root@home usr]# ldd /usr/bin/glxinfo
	linux-vdso.so.1 (0x00007fff961c6000)
	libGL.so.1 => /usr/lib/libGL.so.1 (0x00007fa3d0621000)
	libX11.so.6 => /usr/lib/libX11.so.6 (0x00007fa3d04de000)
	libc.so.6 => /usr/lib/libc.so.6 (0x00007fa3d02f7000)
	libGLdispatch.so.0 => /usr/lib/libGLdispatch.so.0 (0x00007fa3d023f000)
	libGLX.so.0 => /usr/lib/libGLX.so.0 (0x00007fa3d020d000)
	libxcb.so.1 => /usr/lib/libxcb.so.1 (0x00007fa3d01e2000)
	/lib64/ld-linux-x86-64.so.2 => /usr/lib64/ld-linux-x86-64.so.2 (0x00007fa3d06d2000)
	libXau.so.6 => /usr/lib/libXau.so.6 (0x00007fa3d01db000)
	libXdmcp.so.6 => /usr/lib/libXdmcp.so.6 (0x00007fa3d01d3000)
[root@home usr]# ldd /usr/bin/glxinfo32
	linux-gate.so.1 (0xf7f8d000)
	libGL.so.1 => /usr/lib32/libGL.so.1 (0xf7ef3000)
	libX11.so.6 => /usr/lib32/libX11.so.6 (0xf7da1000)
	libc.so.6 => /usr/lib32/libc.so.6 (0xf7a00000)
	libGLdispatch.so.0 => /usr/lib32/libGLdispatch.so.0 (0xf7d25000)
	libGLX.so.0 => /usr/lib32/libGLX.so.0 (0xf7cea000)
	libxcb.so.1 => /usr/lib32/libxcb.so.1 (0xf7cbe000)
	/lib/ld-linux.so.2 => /usr/lib/ld-linux.so.2 (0xf7f8f000)
	libXau.so.6 => /usr/lib32/libXau.so.6 (0xf7cb9000)
	libXdmcp.so.6 => /usr/lib32/libXdmcp.so.6 (0xf7cb2000)
[root@home usr]#

freeartist · 2023-01-31 22:37:42

>You would have to completely uninstall all nvidia packages
Removed nvidia 340.108, what now?

>But if you blacklist nvidia, unblacklist nouveau, make sure the xorg config doesn't reference nvidia and remove /usr/lib*/nvidia/ from the ld path (what should™ be achieved by removing those ld.so.conf.d files) you can keep the packages installed.
remove /etc/X11//xorg.conf.d/20-nvidia.conf

now working nouveau

Last edited by freeartist (2023-01-31 22:47:07)

seth · 2023-01-31 22:50:44

So the nvidia libraries are now gone, but you're (again? still?) runing on the binary nvidia driver.
But we're lookcing for backtraces on nouveau (which you claimed to not work either)

freeartist · 2023-01-31 22:57:24

bt coredump at nouveau

sdl
https://pastebin.com/raw/SCV2FsLW

opengl
https://pastebin.com/raw/GEyZUKuk

Last edited by freeartist (2023-01-31 23:02:50)

Arch Linux

#51 2023-01-28 22:31:09

Re: [SOLVED] UT99 not start

#52 2023-01-29 00:58:25

Re: [SOLVED] UT99 not start

#53 2023-01-29 07:51:57

Re: [SOLVED] UT99 not start

#54 2023-01-29 19:06:27

Re: [SOLVED] UT99 not start

#55 2023-01-29 19:12:01

Re: [SOLVED] UT99 not start

#56 2023-01-29 19:14:25

Re: [SOLVED] UT99 not start

#57 2023-01-29 19:32:25

Re: [SOLVED] UT99 not start

#58 2023-01-29 23:37:10

Re: [SOLVED] UT99 not start

#59 2023-01-30 08:06:05

Re: [SOLVED] UT99 not start

#60 2023-01-30 12:31:50

Re: [SOLVED] UT99 not start

#61 2023-01-30 14:35:28

Re: [SOLVED] UT99 not start

#62 2023-01-30 20:52:18

Re: [SOLVED] UT99 not start

#63 2023-01-30 21:05:39

Re: [SOLVED] UT99 not start

#64 2023-01-31 00:15:59

Re: [SOLVED] UT99 not start

#65 2023-01-31 09:06:05

Re: [SOLVED] UT99 not start

#66 2023-01-31 21:21:47

Re: [SOLVED] UT99 not start

#67 2023-01-31 21:39:49

Re: [SOLVED] UT99 not start

#68 2023-01-31 21:59:56

Re: [SOLVED] UT99 not start

#69 2023-01-31 22:01:47

Re: [SOLVED] UT99 not start

#70 2023-01-31 22:09:48

Re: [SOLVED] UT99 not start

#71 2023-01-31 22:22:00

Re: [SOLVED] UT99 not start

#72 2023-01-31 22:30:51

Re: [SOLVED] UT99 not start

#73 2023-01-31 22:37:42

Re: [SOLVED] UT99 not start

#74 2023-01-31 22:50:44

Re: [SOLVED] UT99 not start

#75 2023-01-31 22:57:24

Re: [SOLVED] UT99 not start

Board footer