Bug 216451

Created attachment 103110 [details]
Installation YaST logs

The needed installation YaST logs.

I selected Major, because due to this bug, I didn't sound, printers, well, I didn't any of the hardware that should be configured in the last part of the installation.

It also lead to errors about installation after the system "finished the install".

I though that this problems were enough for Major. So I guess Major means blocking the installation, preventing nay use of the computer. I will remember that.

*** Bug 217368 has been marked as a duplicate of this bug. ***

I'm sorry but I cannot reproduce this. I did the following maybe you can
repeat the following steps on your machine as user root:

   PATH=$PATH:/usr/lib/YaST2/bin
   cd /usr/share/YaST2/clients

   y2base test_proposal '("hardware")' qt -geometry 800x600

I selected the aspect setup and canceled it... nothing serious happens   

With beta2 it got worst. Know I cannot even configure the monitor. If I try to vonfigure the monitor in the last stage of the installation, after having selected the monitor, I select "Test configuration", then I accept the configuration, and the YAST craches and skips the hardware configuration.

Next boot starts last instalation phase again.

So I still wasn't able to test what you sai in your last comment, since I still wasn't able to complete the openSUSE 10.2 beta2 installation properly. I will change the monitor without testing to see If I can complete the install.

Ok, I skipped the test part and was able to complete the install part (by the way, selecting OK on the aspect ratio causes no problem, only cancel). Also remember that I have "Aspect Unconfigured" when the dialog appears.

Running what you told me crashes YaST with the following output:

 
Uncaught exception from user code:
        TypeError in method 'SaXManipulateDesktop_setDisplayRatioAndTraversal', argument 2 of type 'double'
 at /usr/share/YaST2/modules/XLib.pm line 195
        XLib::setDisplaySize('XLib', 'ARRAY(0x14e4be0)') called at -e line 0
Can't undef active subroutine during global destruction.

I'm sorry but I cannot reproduce your problem. It works as it should
on the current beta version tested on:

    cat /etc/SuSE-release 
    openSUSE 10.2 (i586) Beta2
    VERSION = 10.2
 

I don't know why I was linked to this report as a duplicate, but Marcus, could you take a look:
https://bugzilla.novell.com/show_bug.cgi?id=217368

It is not about error messages -- there, simply yast "accepts" cancel.

Guys

My 00.02 on this. Beta2 fresh installation on a Thinkpad R40.

At the final Hardware screen I usually click on the check configuration for the detected video and resolution, just to make sure. When I said the config was OK it suddenly started to do the last bit of startup and next thing I was at the KDE login screen. It seemed to miss the final done button.

Anyways after logging in I went to configure something that required the
audio components and they were not configured. I started YaST and chose
the Quick method to configure the sound card (the hardware had been
detected OK) and it started to work.

I think there might be something a bit broken with the final screen of
the installation and the Sax control panel...

Cheers
Jim

Looks to me like the problem is screen configuration doesn't know how to exit properly. I tested my configuration after doing everything else in final configuration. Test never returned me to where I could click on finish, but instead sent me directly to KDM. Then on next boot, YAST2 started itself up again claiming the installation had been aborted.

I thought it could be something related to the xorg open-source NVIDIA driver, but all the issues still happend even after having the NVIDIA binary only driver.

Trying the test written on Comment #4 crashes when cancelling the aspect ration, crashes when canceling the "Test Configuration", and crashes when accepting ("save") on "Test configuration".

This is very annoying, and from other people comments it seems to be a real bug that affects not just me. So shouldn't the bug be reopened, even if it works for you, Marcus? At least people would know that it is really a bug but that it is not being worked on.

yes I think you are right. I tested what you mentioned and all this
works as expected if tested remotely but I was able to reproduce a
crash when I manually called

   /usr/lib/YaST2/startup/YaST2.Second-Stage

if another X-Server is running ( I have a Matrox card and clicked on test )
the fbdev server crashed if you save the configuration. The X log file from
the fbdev server said:

   Backtrace:
0: Xorg(xf86SigHandler+0x81) [0x80c00b1]
1: [0xb7f7a420]
2: /usr/lib/xorg/modules/input//mouse_drv.so [0xb7c23072]
3: Xorg [0x80c022a]
4: Xorg [0x80aa441]
5: [0xb7f7a420]
6: Xorg(xf86UnblockSIGIO+0x57) [0x80a9e57]
7: Xorg [0x80b49dd]
8: Xorg(xf86EnterServerState+0x1f) [0x80b4a0f]
9: Xorg(xf86Wakeup+0x3de) [0x80c17ee]
10: Xorg(WakeupHandler+0x59) [0x808a279]
11: Xorg(WaitForSomething+0x1b9) [0x819aa29]
12: Xorg(Dispatch+0x82) [0x80861c2]
13: Xorg(main+0x495) [0x806e705]
14: /lib/libc.so.6(__libc_start_main+0xdc) [0xb7d0ef9c]
15: Xorg(FontFileCompleteXLFD+0x1e1) [0x806da21]

Fatal server error:
Caught signal 11.  Server aborting

I will attach the complete log

Created attachment 105068 [details]
Xorg.0.log

*** This bug has been marked as a duplicate of bug 219946 ***