[Pardus-kullanicilari] Pardus 2008 has issues with 82Q33 Express Integrated Graphics

[Baran]

This is a request of Mustafa Akgül to help the Pardus users if they somehow
encounter this problem...
 
It looks like diary...but will be useful the ones who encounter this
problem...
 
I had an image of XP SP3 and Pardus Linux 2008 (not hyaena, the version
before
hyaena)HP DC5800DC with Acronis snap deploy...There is no problem while
imaging...We had first try with Norton Ghost...but no use...Acronis is the
only
solution...
 
After the deployment, the weird problem with Linux Pardus is:
 
When i start the machine, while the services are loading, it may come up
with
an error when i decrease init level...So, the error is the black screen
freezes
with _ char and cannot load KDE...
 
But, THIS OCCURS RANDOMLY...Sometimes, the black is passed perfectly and the
KDE is loaded...Another thing is resolution is 1440x900...It's a weird
solution
for Intel i guess...
 
I think the problem is with the Intel 82Q33 Express Integrated Graphics
Controller...
 
You may say: Check the repository for XORG-Server...I did...No use...
You may say: Change XORG.conf under /etc/X11, give some values to it...
You may say: Change the kernel parameter from:
root=LABEL=PARDUS_ROOT vga=791 splash=silent quiet resume=/dev/sda7
 
to xorg=safe...Please follow my steps that i am executing:
 
1) xorg=prob->error chains->_ char
2) xorg=safe->error chains->ok (I can see GUI but resolution is 640x480, I
need
1440x900)
3) xorg=off->error chains->startx->no init 5->fails
4) I installed  intel updates from test repository:
http://paketler.pardus.org.tr/pardus-2008-test/xorg-video-7.3-52-15.pisi
http://paketler.pardus.org.tr/pardus-2008-test/xorg-server-7.3-46-13.pisi
->error chains->_ (no safe or no off or no probe set)
5) xorg.conf-> Option "FramebufferCompression" "false"->error chains->_ char
6) step 4 + step5.+xorg=off->error chains->startx->fails
 In each machine there is the _ error randomly...After restart or close,
then
start operations, the problem still exists...
 
Step 2 the error log i got:
eth0: Mac 5, PHY: 6, PBA No: 1042ff-0ff
agpgart: Detected an Intel Q33 Chipset.
agpgart: AGP architecture is 256M @ 0xe0000000
rtc_coms 00:03: rtc core: registered rtc_cmos as rtc0
ACPI Exception (video-1613): AE_NOT_FOUND, Evaluating _DOD [20070126]
input: Video Bus as
/devices/LNXSYSTM:00/device:00/PNP0A08:00/device:1e/input/input5
ACPI: Video Device [GFX0] (multi-head: no rom: yes post: no)
ACPI: PCI Interrupt 0000:00:1b.0[A] -> GSI 21 (level, low) ->IRQ 21
PCI: Setting latency timer of device 0000:00:1b.0 to 64
 
Now, after that i try to use 915resolution and this time i got this:
 
Intel chipset detected.  However, 915resolution was unable to determine the
chipset type.
Chipset Id: 29d08086
 
****************************************************************************
**
 
ok...let me tell behavior of machines:
 
For example: I opened today 6 machines to probe 1/6 fail...
A month ago when i try for 8 machines 4/8 fail...
 
When a machine fails to open Linux, he has a tendency of failing again...For
that one machine, for 5 try, 3 fail...So, if he has a tendency to fail once,
he
fail again definitely...I am now sending a XORG.conf of 3/5 failing machine
as
snippet:
 
As you see for screen section, in the snippet, there is no resolution
set...I
had set that before to 1440x900 but again fails...
 
So, another i am now probing is:
commenting out this "BusId   "PCI:0:2:0"" from your /etc/X11/xorg.conf file,
or
play around with different values. like PCI:0:0:1 etc.
 
I deleted the line with busid but again in 3 machines, it failed and _ char
is
seen...
 
no use...
 
**************************************************************************
 
The interesting thing before i am changing the line, i select the machines
that
failed for the first time...After i edit, they worked well for couple
startups...But interesting thing is they failed at the same try...
 
Weird...
 
*************************************************************************
 
I tried ACPI=off
 
no use...
*************************************************************************
I tried several combination noapic nolapic acpi=off pci=noacpi under
grub.conf...
 
Each time a machine, random machine fails...out of 3 machines...
 
I am getting insane...
 
The error starts with acpi id 1 already in use...
**************************************************************************
 
BIOS bug IO#ACPI 0 id 1 is already used.
Fixing upto 2.
 
**************************************************************************
This is getting insane!
pci=bios
 
a small white pixel/dot in the right part of the screen, only one dot...it's
the weirdest thing i have ever seen in my life!!
***************************************************************************
 
pci=bios kills 4 machines...96 left...
acpi=off noapic -> no use
 
I am using
http://lxr.linux.no/linux+v2.6.24.7/Documentation/kernel-parameters.txt#L119
4
 
***************************************************************************
ok...i solved the problem but please check the issue:
 
1) adding xorg=safe as kernel param...This will add some lines to
xorg.conf...Reboot..
2) Removing xorg=safe and adding the 1440x900 resolution to
xorg.conf...Reboot...This will cause the display card to have only one
resolution as 1440x900 that will suffice my problem so the students cannot
change the resolution to absurd values...
 
Voila!
 
I will send you more files to enlighten the issue further....

 

-------------- sonraki bölüm --------------
Bir HTML eklentisi temizlendi...
URL: http://liste.pardus.org.tr/pardus-kullanicilari/attachments/20081103/89dbd808/attachment-0001.htm