Bug Report
----------
To:
Linux Kernel Mailing List
Cc:
Reid [email protected]
Kenneth [email protected]
Roman Drahtm?ller [email protected]
I would send a copy of this to ABIT but they aren't very reachable. I don't
know a point of contact for the macro-society known as Microsoft.
The following is a collection of information and data accumulated after playing
with a specific motherboard and a variety of operating systems (for 18 hours).
The failure to install, run and operate some of the OS' on the board or rather
operate them at the meant to be CPU speed.
The report regards:
Abit BX6 Rev. 2 also knows as
Abit BX6 Rev. 2.0 also knows as
Abit BX6-2.0
and contains the following sections:
1. GENERAL
2. HARDWARE SETUP
3. PROBLEM SITUATION
4. SUMMARY
5. WEB RESEARCH
6. MY ANALYSIS
7. THANK YOU
8. KERNEL OOPSES
9. LINKS
GENERAL
-------
The board was purchased in December 1999 (built into a machine) and came from
the manufacturer with the BIOS 'BXR_MN.BIN'.
The latest BIOS is 'BXR_QR.BIN'.
The BIOS history can be found here:
http://fae.abit.com.tw/eng/download/bios/bx620.htm
The machine is filled with a Pentium III CPU 550 MHz. I tested this CPU using
the Intel test utility and this CPU is neither overclocked or been tampered
with.
This board has been specified to work with Linux at:
http://lhd.zdnet.com/db/dispproduct.php3?DISP?334
Throughout my research I have been in contact with BX6-2.0 users who had no
trouble and others who had the same trouble as I did.
This board has been known to be used by a variety of overclockers. In my setup
no overclocked CPU has been used or in other words my CPU was never intended to
be running above its meant to be speed. The target speed of the Pentium III 550
MHz was always 550MHz.
During my testing I have removed all unnecessary hardware from the setup trying
to find a hardware component that caused the problems. I could not however.
HARDWARE SETUP
--------------
Abit BX6-2.0 (SLOT 1 mobo)
Pentium III 550 Mhz
Kingston Valueram 256 MB PC100
LG CD-RW CED-8042B
NVIDIA RIVA TNT2 32 MB AGP
IBM 30.7 GB 75GXP DTLA-307030 (brand new so its got nothing to do with the
drive; I know this drive has been having its own bad calls in the news but it
is not the issue here)
Lite-ON Monitor
Over the course of my testing I also tried another stick of memory, a Trident
3Dimage 4MB AGP card and a Toshiba 24x CD-Rom. The same problem I am about to
discuss occurred.
PROBLEM SITUATION
-----------------
After the purchase in 1999 I installed SuSE 6.2 on the machine. I don't
remember having any trouble at it and I am pretty sure it was at the CPU speed
of 550MHz. The machine came with Win98 and has been running for 2 years with
Win98. The SuSE install was only a test run back then and I had wiped it off
the disk afterwards.
2002: My latest goal was installing Windows XP on this new drive in the
machine. (The drive before was a DTLA-305030). I had updated the BIOS two
months ago to QR and Win98 ran fine.
For the next few paragraphs the CPU is running at 550MHz!!!
I put in the XP CD and it booted and I formatted a 6 Gig partition in NTFS. In
the install procedure it copies over a bunch of files and then reboots to then
drive the rest of the installation from harddisk. On this reboot I got a Blue
Screen Of Death. It reported that there was a Stop error when a process tried
to write to read-only memory. I therefore removed all hardware and turned of
caching and shadowing in the BIOS to diagnose the problem. The problem
persisted and I get either BSOD's or stalls. The BSOD's report
(DRIVER_)IRQL_NOT_LESS_OR_EQUAL errors now. Microsoft support pages give the
same advice I already applied but the OS won't load.
I then went back and installed Win98 to check for sure something wasn't out of
order. It installed without hitch and booted and ran fine.
I went back to installing XP with the same problem occuring.
I went back to older BIOS. I tried MN, KU, HJ. It didn't help.
I pulled out a set of Mandrake 8.1 CDs and it booted off CD but then stalled or
crashed. I pulled out a old set of SuSE 6.4 CDs and it booted but as soon as
the kernel loaded it oopsed.
All while doing this I was searching the web for help and gathered tips from a
couple of mailing lists.
Accidentally I set the CPU to a lower speed than 550MHz and booted the machine.
The BIOS complained that the CPU was misconfigured and I should change the BIOS
settings but it also gave me the option to go ahead (F1) and boot anyhow.
The CPU is at 366 MHz (underclocked)!!!!!
I went ahead and suddenly XP booted.
Here I went and pulled out my SuSE 6.4 CD and now it wouldn't oops anymore. I
went ahead and installed SuSE 6.4 on a second partition. I rebooted and it
booted just fine.
I went back and set the CPU back to 550Mhz and booting XP or SuSE became
impossible again. SuSE kernel dumped a large oops onto my screen.
Today I got a set of RedHat 7.2 CDs and it will not install at 550MHz. It
stalls or suddenly reboots. I set the CPU to 366 MHz again and installed RH
onto another partition. My setup is now:
/dev/hda1 XP
/dev/hda2 SuSE 6.4
/dev/hda3 swap
/dev/hda4 RH 7.2
After having installed RH I rebooted and it boots fine at 366 MHz. If I set the
CPU to 550Mhz the RH kernel oopses.
Now many people suggested to turn off about everything in the BIOS and work my
way upwards. I did this and only left IDE controller enabled. It booted fine at
366 Mhz. As soon as I switched to 550 Mhz even with L2 Cache and L1 Cache
disabled and any shadowing turned off too it would stall.
SUMMARY
-------
Original motherboard allowed for Linux installation (SuSE 6.2) and runs Win98
fine at speed 550MHz Pentium III CPU. No problems for two years and to this
date with Win98.
The BIOS was updated in late 2001 to QR. The machine still runs Win98 fine at
550MHz. Operating systems such as XP, RH, SuSE must be installed and run at 366
MHz now. UNDERCLOCKED!!!
WEB RESEARCH
------------
I have dug into google and found quite some information. This problem has been
known but occurs very rarely and in only special cases and I think I know why.
If I ran XP at 550 Mhz and chose Safe Mode it would run down a list of modules
loaded and suddenly stall at "driver/agp440.sys".
I searched google and a ton of people using 2K or XP have trouble with the
agp440.sys problem. This doesn't mean AGP is the cause here since XP doesn't
dump out its real module loading log when loading. agp440.sys is just the last
module mentioned before it stalls.
The agp440.sys problem occurs also for many other boards and configurations
(e.g. Athlon, P4 and other NVIDIA cards)
I figured this was not the right direction.
Having remembered that I once ran Linux on this board two years ago I figured
its either the CPU or the mobo. I remembered I flashed the BIOS recently. I
looked up what I could find and voila people did the same as me and had the
same trouble:
Reid reports that all his mobo can run is Win98 and neither 2K, Linux or BSDs
runs:
http://groups.google.com/groups?hl=en&frame=right&th=b5b6671ee8e4a3ad&seekm=q2vZ
6.77281%24ru2.22650002%40typhoon.southeast.rr.com#link1
A reverting to the old BIOS does not help. I contacted Reid and he still has no
fix for the situation.
I got another example which is exactly like mine and Reid's:
http://groups.google.com/groups?q=bx6+kernel+panic&hl=en&selm=37B6EE47.3D9A3AE5%
40ucs.orst.edu&rnum=3
Kenneth ran NT on his BX6-2. It worked wonderfully. Then he flashed his BIOS
and it oopsed from thereon in. He went back to use his old BIOS but his OS is
still unstable. He tried Mandrake on it then and it won't succeed but oopses.
MY ANALYSIS
-----------
The board as it was manufactured was able to run all types of OS. Manypeople
have assured me that is what they are doing.
I and Reid though cannot run anything but Win9x. The thing that makes our case
special is that we both flashed the BIOS and then tried to install XP or Linux.
I do wonder if the BIOS installed at the plant was in some way special and any
BIOS flashed at home and downloaded from the web are containing inconsistencies
which break the mobo from working with NT or Linux kernels. The fact that
others run Linux and Nt fine on their mobo but my flashed and revertedly
flashed mobo cannot makes me wonder.
THANK YOU
---------
I would love to contact Abit about this but they don't leave a single point of
contact on their web pages.
I appreciate any comments in regards to this issue. I will now report what the
kernel errored on and it may help the hackers in building an opinion about this
issue.
Thank you for reading this lengthy document! I could care less if Ican run a
certain OS on this specific by today's standard's outdatedbox but it bugs me
that there seems to have creaped in a bug intoABIT's board which makes it
incompatible with a advanced kernel like the Linux 2.2 or 2.4 kernel and the NT
kernel as well but gives it theability to handle the functionality of the Win9x
kernel.
I wish somebody would invest the time in researching this topic in detail.
I know ABIT has their own Linux distro. I do not have a copy of it but I would
be damn curious to know if their distro runs on their ownboard after a BIOS
flash.
Thanks for your time and if you need more info contact me,
Marcel
KERNEL OOPSES
-------------
RH 7.2 default i686 kernel (I think its 2.4.7):
EXT3-fs: mounted filesystem with ordered data mode.
Freeing unused kernel memory: 220k freed.
CPU 0: Machine Check Exception: 0000000000000004
Bank 3: b20000000002010a
Kernel panic: CPU context corrupt
In interrupt handler - not syncing
SuSE 6.4 default 2.2.14 kernel:
Unable to handle kernel paging request at virtual address 05efa769.2%
current->tss.cr3 = 00101000, %cr3 = 00101000
*pde = 00000000
Oops: 0002
CPU: 0
EIP: 0010:[<c011825>]
EFLAGS: 00010013
eax: 000000ff ebx: c0289f24 ecx: 00008c7e edx: c0289f24
esi: 20000001 edi: 00000000 ebp: c0289e6c esp: c0289e74
ds: 0018 es: 0018 ss: 0018
Process swapper (pid: 0, process nr: 0, stackpage=c0289000)
Stack: c0289f24 24000001 0000000e c0289f24 c026ad58 00000001 c0289eacc01d43b4
00000001 00000695............
Call Trace: ........................
Code: 28 8b 45 08 c7 45 f4 00 00 00 00 c7 45 f0 00 00 00 00 89 45
Aiee, killing interrupt handler
Kernel panic: Attempted to kill idle task!
In swapper task - not syncing.
Kenneth's Mandrake 6.0 Oops:
Kernel panic: LRU block list corrupted
Unable to handle kernul NULL pointer dereference at virtual address
00000028
current->tss.cr3 = 00101000, %cr3 = 00101000
*pde = 00000000
Oops: 0000
CPU: 0
.
.
.
LINKS
-----
Abit Support:
http://fae.abit.com.tw/eng/
Abit BX6 Rev. 2:
http://fae.abit.com.tw/eng/download/bios/bx620.htm
I just got mail from Reid and he said he now slowed down his 450 CPu to a
lower speed and it now boots Linux as well. I guess its a work around but
there must be a bug to catch either in the kernel or in the BIOS or the
engineering of the board itsxelf.
I guess one can only find the bug by getting two original boards and flashing
the BIOS of one of them and then comparing their functioning.
mk
> > Bug Report >
---------- > > To: > Linux Kernel Mailing List
>
> Cc:
> Reid [email protected]
> Kenneth [email protected]
> Roman Drahtm?ller [email protected]
>
> I would send a copy of this to ABIT but they aren't very reachable. I don't
> know a point of contact for the macro-society known as Microsoft.
>
>
>
> The following is a collection of information and data accumulated after playin
g
> with a specific motherboard and a variety of operating systems (for 18 hours).
> The failure to install, run and operate some of the OS' on the board or rather
> operate them at the meant to be CPU speed.
>
>
>
> The report regards:
>
> Abit BX6 Rev. 2 also knows as
> Abit BX6 Rev. 2.0 also knows as
> Abit BX6-2.0
>
>
> and contains the following sections:
>
> 1. GENERAL
> 2. HARDWARE SETUP
> 3. PROBLEM SITUATION
> 4. SUMMARY
> 5. WEB RESEARCH
> 6. MY ANALYSIS
> 7. THANK YOU
> 8. KERNEL OOPSES
> 9. LINKS
>
>
>
> GENERAL
> -------
>
> The board was purchased in December 1999 (built into a machine) and came from
> the manufacturer with the BIOS 'BXR_MN.BIN'.
>
> The latest BIOS is 'BXR_QR.BIN'.
>
> The BIOS history can be found here:
> http://fae.abit.com.tw/eng/download/bios/bx620.htm
>
> The machine is filled with a Pentium III CPU 550 MHz. I tested this CPU using
> the Intel test utility and this CPU is neither overclocked or been tampered
> with.
>
> This board has been specified to work with Linux at:
>
> http://lhd.zdnet.com/db/dispproduct.php3?DISP?334
>
> Throughout my research I have been in contact with BX6-2.0 users who had no
> trouble and others who had the same trouble as I did.
>
> This board has been known to be used by a variety of overclockers. In my setup
> no overclocked CPU has been used or in other words my CPU was never intended t
o
> be running above its meant to be speed. The target speed of the Pentium III 55
0
> MHz was always 550MHz.
>
>
> During my testing I have removed all unnecessary hardware from the setup tryin
g
> to find a hardware component that caused the problems. I could not however.
>
>
> HARDWARE SETUP
> --------------
>
> Abit BX6-2.0 (SLOT 1 mobo)
> Pentium III 550 Mhz
> Kingston Valueram 256 MB PC100
> LG CD-RW CED-8042B
> NVIDIA RIVA TNT2 32 MB AGP
> IBM 30.7 GB 75GXP DTLA-307030 (brand new so its got nothing to do with the
> drive; I know this drive has been having its own bad calls in the news but it
> is not the issue here)
> Lite-ON Monitor
>
> Over the course of my testing I also tried another stick of memory, a Trident
> 3Dimage 4MB AGP card and a Toshiba 24x CD-Rom. The same problem I am about to
> discuss occurred.
>
>
> PROBLEM SITUATION
> -----------------
>
> After the purchase in 1999 I installed SuSE 6.2 on the machine. I don't
> remember having any trouble at it and I am pretty sure it was at the CPU speed
> of 550MHz. The machine came with Win98 and has been running for 2 years with
> Win98. The SuSE install was only a test run back then and I had wiped it off
> the disk afterwards.
>
> 2002: My latest goal was installing Windows XP on this new drive in the
> machine. (The drive before was a DTLA-305030). I had updated the BIOS two
> months ago to QR and Win98 ran fine.
>
> For the next few paragraphs the CPU is running at 550MHz!!!
>
> I put in the XP CD and it booted and I formatted a 6 Gig partition in NTFS. In
> the install procedure it copies over a bunch of files and then reboots to then
> drive the rest of the installation from harddisk. On this reboot I got a Blue
> Screen Of Death. It reported that there was a Stop error when a process tried
> to write to read-only memory. I therefore removed all hardware and turned of
> caching and shadowing in the BIOS to diagnose the problem. The problem
> persisted and I get either BSOD's or stalls. The BSOD's report
> (DRIVER_)IRQL_NOT_LESS_OR_EQUAL errors now. Microsoft support pages give the
> same advice I already applied but the OS won't load.
>
> I then went back and installed Win98 to check for sure something wasn't out of
> order. It installed without hitch and booted and ran fine.
>
> I went back to installing XP with the same problem occuring.
>
> I went back to older BIOS. I tried MN, KU, HJ. It didn't help.
>
> I pulled out a set of Mandrake 8.1 CDs and it booted off CD but then stalled o
r
> crashed. I pulled out a old set of SuSE 6.4 CDs and it booted but as soon as
> the kernel loaded it oopsed.
>
> All while doing this I was searching the web for help and gathered tips from a
> couple of mailing lists.
>
> Accidentally I set the CPU to a lower speed than 550MHz and booted the machine
.
> The BIOS complained that the CPU was misconfigured and I should change the BIO
S
> settings but it also gave me the option to go ahead (F1) and boot anyhow.
>
> The CPU is at 366 MHz (underclocked)!!!!!
>
> I went ahead and suddenly XP booted.
>
> Here I went and pulled out my SuSE 6.4 CD and now it wouldn't oops anymore. I
> went ahead and installed SuSE 6.4 on a second partition. I rebooted and it
> booted just fine.
>
> I went back and set the CPU back to 550Mhz and booting XP or SuSE became
> impossible again. SuSE kernel dumped a large oops onto my screen.
>
> Today I got a set of RedHat 7.2 CDs and it will not install at 550MHz. It
> stalls or suddenly reboots. I set the CPU to 366 MHz again and installed RH
> onto another partition. My setup is now:
>
> /dev/hda1 XP
> /dev/hda2 SuSE 6.4
> /dev/hda3 swap
> /dev/hda4 RH 7.2
>
> After having installed RH I rebooted and it boots fine at 366 MHz. If I set th
e
> CPU to 550Mhz the RH kernel oopses.
>
>
> Now many people suggested to turn off about everything in the BIOS and work my
> way upwards. I did this and only left IDE controller enabled. It booted fine a
t
> 366 Mhz. As soon as I switched to 550 Mhz even with L2 Cache and L1 Cache
> disabled and any shadowing turned off too it would stall.
>
>
> SUMMARY
> -------
>
> Original motherboard allowed for Linux installation (SuSE 6.2) and runs Win98
> fine at speed 550MHz Pentium III CPU. No problems for two years and to this
> date with Win98.
>
> The BIOS was updated in late 2001 to QR. The machine still runs Win98 fine at
> 550MHz. Operating systems such as XP, RH, SuSE must be installed and run at 36
6
> MHz now. UNDERCLOCKED!!!
>
>
> WEB RESEARCH
> ------------
>
> I have dug into google and found quite some information. This problem has been
> known but occurs very rarely and in only special cases and I think I know why.
>
> If I ran XP at 550 Mhz and chose Safe Mode it would run down a list of modules
> loaded and suddenly stall at "driver/agp440.sys".
>
> I searched google and a ton of people using 2K or XP have trouble with the
> agp440.sys problem. This doesn't mean AGP is the cause here since XP doesn't
> dump out its real module loading log when loading. agp440.sys is just the last
> module mentioned before it stalls.
>
> The agp440.sys problem occurs also for many other boards and configurations
> (e.g. Athlon, P4 and other NVIDIA cards)
>
> I figured this was not the right direction.
>
> Having remembered that I once ran Linux on this board two years ago I figured
> its either the CPU or the mobo. I remembered I flashed the BIOS recently. I
> looked up what I could find and voila people did the same as me and had the
> same trouble:
>
> Reid reports that all his mobo can run is Win98 and neither 2K, Linux or BSDs
> runs:
>
> http://groups.google.com/groups?hl=en&frame=right&th=b5b6671ee8e4a3ad&seekm=q2
vZ
> 6.77281%24ru2.22650002%40typhoon.southeast.rr.com#link1
>
> A reverting to the old BIOS does not help. I contacted Reid and he still has n
o
> fix for the situation.
>
> I got another example which is exactly like mine and Reid's:
>
> http://groups.google.com/groups?q=bx6+kernel+panic&hl=en&selm=37B6EE47.3D9A3AE
5%
> 40ucs.orst.edu&rnum=3
>
> Kenneth ran NT on his BX6-2. It worked wonderfully. Then he flashed his BIOS
> and it oopsed from thereon in. He went back to use his old BIOS but his OS is
> still unstable. He tried Mandrake on it then and it won't succeed but oopses.
>
>
> MY ANALYSIS
> -----------
>
> The board as it was manufactured was able to run all types of OS. Manypeople
> have assured me that is what they are doing.
>
> I and Reid though cannot run anything but Win9x. The thing that makes our case
> special is that we both flashed the BIOS and then tried to install XP or Linux
.
>
>
> I do wonder if the BIOS installed at the plant was in some way special and any
> BIOS flashed at home and downloaded from the web are containing inconsistencie
s
> which break the mobo from working with NT or Linux kernels. The fact that
> others run Linux and Nt fine on their mobo but my flashed and revertedly
> flashed mobo cannot makes me wonder.
>
>
> THANK YOU
> ---------
>
> I would love to contact Abit about this but they don't leave a single point of
> contact on their web pages.
>
> I appreciate any comments in regards to this issue. I will now report what the
> kernel errored on and it may help the hackers in building an opinion about thi
s
> issue.
>
> Thank you for reading this lengthy document! I could care less if Ican run a
> certain OS on this specific by today's standard's outdatedbox but it bugs me
> that there seems to have creaped in a bug intoABIT's board which makes it
> incompatible with a advanced kernel like the Linux 2.2 or 2.4 kernel and the N
T
> kernel as well but gives it theability to handle the functionality of the Win9
x
> kernel.
>
> I wish somebody would invest the time in researching this topic in detail.
>
> I know ABIT has their own Linux distro. I do not have a copy of it but I would
> be damn curious to know if their distro runs on their ownboard after a BIOS
> flash.
>
> Thanks for your time and if you need more info contact me,
>
> Marcel
>
>
> KERNEL OOPSES
> -------------
>
> RH 7.2 default i686 kernel (I think its 2.4.7):
>
> EXT3-fs: mounted filesystem with ordered data mode.
> Freeing unused kernel memory: 220k freed.
> CPU 0: Machine Check Exception: 0000000000000004
> Bank 3: b20000000002010a
> Kernel panic: CPU context corrupt
> In interrupt handler - not syncing
>
>
> SuSE 6.4 default 2.2.14 kernel:
>
> Unable to handle kernel paging request at virtual address 05efa769.2%
> current->tss.cr3 = 00101000, %cr3 = 00101000
> *pde = 00000000
> Oops: 0002
> CPU: 0
> EIP: 0010:[<c011825>]
> EFLAGS: 00010013
> eax: 000000ff ebx: c0289f24 ecx: 00008c7e edx: c0289f24
> esi: 20000001 edi: 00000000 ebp: c0289e6c esp: c0289e74
> ds: 0018 es: 0018 ss: 0018
> Process swapper (pid: 0, process nr: 0, stackpage=c0289000)
> Stack: c0289f24 24000001 0000000e c0289f24 c026ad58 00000001 c0289eacc01d43b4
> 00000001 00000695............
> Call Trace: ........................
> Code: 28 8b 45 08 c7 45 f4 00 00 00 00 c7 45 f0 00 00 00 00 89 45
> Aiee, killing interrupt handler
> Kernel panic: Attempted to kill idle task!
> In swapper task - not syncing.
>
>
>
>
> Kenneth's Mandrake 6.0 Oops:
>
> Kernel panic: LRU block list corrupted
> Unable to handle kernul NULL pointer dereference at virtual address
> 00000028
> current->tss.cr3 = 00101000, %cr3 = 00101000
> *pde = 00000000
> Oops: 0000
> CPU: 0
> .
> .
> .
>
>
> LINKS
> -----
>
> Abit Support:
> http://fae.abit.com.tw/eng/
>
> Abit BX6 Rev. 2:
> http://fae.abit.com.tw/eng/download/bios/bx620.htm
>
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
On Fri, 25 Jan 2002, Marcel Kunath wrote:
> I just got mail from Reid and he said he now slowed down his 450 CPu to a
> lower speed and it now boots Linux as well. I guess its a work around but
> there must be a bug to catch either in the kernel or in the BIOS or the
> engineering of the board itsxelf.
>
> I guess one can only find the bug by getting two original boards and flashing
> the BIOS of one of them and then comparing their functioning.
>
> mk
>
I use a BX6 at home, running Win98SE fine. When I made the mistake of
installing Windows 2000, it loaded and ran fine for me. Had to dump it
because of the lack of dual monitor support for a lot of cards that worked
fine in Win98SE.
As well, the machine ran Debian 2.2r3 with both 2.2.19 and early 2.4
kernels.
BX6r2, QR bios, Celeron 466 @ 466, 512 meg (2 x 256 meg Micron PC133 cas3)
Quantum LM 30 gig, Quantum KX 13.6, Advansys 960U scsi card
Creative Annihilator Geforce2 GTS(something like that, marketing sucks! :D)
ATI Mach64 based secondary video, SB Live!, DFE-530TX network card
My only problems is when the USB webcam(3com's high end one, don't
remember the name) is running in netmeeting, my sound breaks up. Very
very frustrating, I've spent a lot of time trying to fix it, it's almost
to the point where i find my old sb32awe.
It sounds like something's not set right with your CMOS settings.
Something at 100mhz is causing problems that goes away at 66. Try
underclocking to 456(5.5*83). If that don't work, try 412(5.5*75).
Otherwise, try setting your cmos settings to default, and shut off ACPI,
cause I've ran into problems when that is enabled.
These boards can be tempermental, but stable once you get them setup
right. I have maybe a dozen of them at the office running 95/98, and
haven't had problems, compared to some of the VIA chipset crap here that
ends up being bought in the interest of saving money.
mike
> I would send a copy of this to ABIT but they aren't very reachable. I don't
[email protected] is their general address
For the US: [email protected]
If you go to http://www.abit.com.tw or http://www.abit-usa.com and select
Compy, and then Contacts, you'll find their tech contact page. As well,
1-510-492-0968 is their Tehcnical support line, 1-510-623-1092 is their
fax.
hope this helps.
mike
Hi,
Just a thought. Have you tried clearing the cmos completely by shorting
the jumper on the mb?
skd
On Fri, Jan 25, 2002 at 07:43:12AM -0500, Marcel Kunath wrote:
> I just got mail from Reid and he said he now slowed down his 450 CPu to a
> lower speed and it now boots Linux as well. I guess its a work around but
> there must be a bug to catch either in the kernel or in the BIOS or the
> engineering of the board itsxelf.
>
> I guess one can only find the bug by getting two original boards and flashing
> the BIOS of one of them and then comparing their functioning.
>
> mk
>
> > > Bug Report >
> ---------- > > To: > Linux Kernel Mailing List
> >
> > Cc:
> > Reid [email protected]
> > Kenneth [email protected]
> > Roman Drahtm?ller [email protected]
> >
> > I would send a copy of this to ABIT but they aren't very reachable. I don't
> > know a point of contact for the macro-society known as Microsoft.
> >
> >
> >
> > The following is a collection of information and data accumulated after playin
> g
> > with a specific motherboard and a variety of operating systems (for 18 hours).
>
> > The failure to install, run and operate some of the OS' on the board or rather
>
> > operate them at the meant to be CPU speed.
> >
> >
> >
> > The report regards:
> >
> > Abit BX6 Rev. 2 also knows as
> > Abit BX6 Rev. 2.0 also knows as
> > Abit BX6-2.0
> >
> >
> > and contains the following sections:
> >
> > 1. GENERAL
> > 2. HARDWARE SETUP
> > 3. PROBLEM SITUATION
> > 4. SUMMARY
> > 5. WEB RESEARCH
> > 6. MY ANALYSIS
> > 7. THANK YOU
> > 8. KERNEL OOPSES
> > 9. LINKS
> >
> >
> >
> > GENERAL
> > -------
> >
> > The board was purchased in December 1999 (built into a machine) and came from
> > the manufacturer with the BIOS 'BXR_MN.BIN'.
> >
> > The latest BIOS is 'BXR_QR.BIN'.
> >
> > The BIOS history can be found here:
> > http://fae.abit.com.tw/eng/download/bios/bx620.htm
> >
> > The machine is filled with a Pentium III CPU 550 MHz. I tested this CPU using
> > the Intel test utility and this CPU is neither overclocked or been tampered
> > with.
> >
> > This board has been specified to work with Linux at:
> >
> > http://lhd.zdnet.com/db/dispproduct.php3?DISP?334
> >
> > Throughout my research I have been in contact with BX6-2.0 users who had no
> > trouble and others who had the same trouble as I did.
> >
> > This board has been known to be used by a variety of overclockers. In my setup
>
> > no overclocked CPU has been used or in other words my CPU was never intended t
> o
> > be running above its meant to be speed. The target speed of the Pentium III 55
> 0
> > MHz was always 550MHz.
> >
> >
> > During my testing I have removed all unnecessary hardware from the setup tryin
> g
> > to find a hardware component that caused the problems. I could not however.
> >
> >
> > HARDWARE SETUP
> > --------------
> >
> > Abit BX6-2.0 (SLOT 1 mobo)
> > Pentium III 550 Mhz
> > Kingston Valueram 256 MB PC100
> > LG CD-RW CED-8042B
> > NVIDIA RIVA TNT2 32 MB AGP
> > IBM 30.7 GB 75GXP DTLA-307030 (brand new so its got nothing to do with the
> > drive; I know this drive has been having its own bad calls in the news but it
> > is not the issue here)
> > Lite-ON Monitor
> >
> > Over the course of my testing I also tried another stick of memory, a Trident
> > 3Dimage 4MB AGP card and a Toshiba 24x CD-Rom. The same problem I am about to
> > discuss occurred.
> >
> >
> > PROBLEM SITUATION
> > -----------------
> >
> > After the purchase in 1999 I installed SuSE 6.2 on the machine. I don't
> > remember having any trouble at it and I am pretty sure it was at the CPU speed
>
> > of 550MHz. The machine came with Win98 and has been running for 2 years with
> > Win98. The SuSE install was only a test run back then and I had wiped it off
> > the disk afterwards.
> >
> > 2002: My latest goal was installing Windows XP on this new drive in the
> > machine. (The drive before was a DTLA-305030). I had updated the BIOS two
> > months ago to QR and Win98 ran fine.
> >
> > For the next few paragraphs the CPU is running at 550MHz!!!
> >
> > I put in the XP CD and it booted and I formatted a 6 Gig partition in NTFS. In
>
> > the install procedure it copies over a bunch of files and then reboots to then
>
> > drive the rest of the installation from harddisk. On this reboot I got a Blue
> > Screen Of Death. It reported that there was a Stop error when a process tried
> > to write to read-only memory. I therefore removed all hardware and turned of
> > caching and shadowing in the BIOS to diagnose the problem. The problem
> > persisted and I get either BSOD's or stalls. The BSOD's report
> > (DRIVER_)IRQL_NOT_LESS_OR_EQUAL errors now. Microsoft support pages give the
> > same advice I already applied but the OS won't load.
> >
> > I then went back and installed Win98 to check for sure something wasn't out of
>
> > order. It installed without hitch and booted and ran fine.
> >
> > I went back to installing XP with the same problem occuring.
> >
> > I went back to older BIOS. I tried MN, KU, HJ. It didn't help.
> >
> > I pulled out a set of Mandrake 8.1 CDs and it booted off CD but then stalled o
> r
> > crashed. I pulled out a old set of SuSE 6.4 CDs and it booted but as soon as
> > the kernel loaded it oopsed.
> >
> > All while doing this I was searching the web for help and gathered tips from a
>
> > couple of mailing lists.
> >
> > Accidentally I set the CPU to a lower speed than 550MHz and booted the machine
> .
> > The BIOS complained that the CPU was misconfigured and I should change the BIO
> S
> > settings but it also gave me the option to go ahead (F1) and boot anyhow.
> >
> > The CPU is at 366 MHz (underclocked)!!!!!
> >
> > I went ahead and suddenly XP booted.
> >
> > Here I went and pulled out my SuSE 6.4 CD and now it wouldn't oops anymore. I
> > went ahead and installed SuSE 6.4 on a second partition. I rebooted and it
> > booted just fine.
> >
> > I went back and set the CPU back to 550Mhz and booting XP or SuSE became
> > impossible again. SuSE kernel dumped a large oops onto my screen.
> >
> > Today I got a set of RedHat 7.2 CDs and it will not install at 550MHz. It
> > stalls or suddenly reboots. I set the CPU to 366 MHz again and installed RH
> > onto another partition. My setup is now:
> >
> > /dev/hda1 XP
> > /dev/hda2 SuSE 6.4
> > /dev/hda3 swap
> > /dev/hda4 RH 7.2
> >
> > After having installed RH I rebooted and it boots fine at 366 MHz. If I set th
> e
> > CPU to 550Mhz the RH kernel oopses.
> >
> >
> > Now many people suggested to turn off about everything in the BIOS and work my
>
> > way upwards. I did this and only left IDE controller enabled. It booted fine a
> t
> > 366 Mhz. As soon as I switched to 550 Mhz even with L2 Cache and L1 Cache
> > disabled and any shadowing turned off too it would stall.
> >
> >
> > SUMMARY
> > -------
> >
> > Original motherboard allowed for Linux installation (SuSE 6.2) and runs Win98
> > fine at speed 550MHz Pentium III CPU. No problems for two years and to this
> > date with Win98.
> >
> > The BIOS was updated in late 2001 to QR. The machine still runs Win98 fine at
> > 550MHz. Operating systems such as XP, RH, SuSE must be installed and run at 36
> 6
> > MHz now. UNDERCLOCKED!!!
> >
> >
> > WEB RESEARCH
> > ------------
> >
> > I have dug into google and found quite some information. This problem has been
>
> > known but occurs very rarely and in only special cases and I think I know why.
>
> >
> > If I ran XP at 550 Mhz and chose Safe Mode it would run down a list of modules
>
> > loaded and suddenly stall at "driver/agp440.sys".
> >
> > I searched google and a ton of people using 2K or XP have trouble with the
> > agp440.sys problem. This doesn't mean AGP is the cause here since XP doesn't
> > dump out its real module loading log when loading. agp440.sys is just the last
>
> > module mentioned before it stalls.
> >
> > The agp440.sys problem occurs also for many other boards and configurations
> > (e.g. Athlon, P4 and other NVIDIA cards)
> >
> > I figured this was not the right direction.
> >
> > Having remembered that I once ran Linux on this board two years ago I figured
> > its either the CPU or the mobo. I remembered I flashed the BIOS recently. I
> > looked up what I could find and voila people did the same as me and had the
> > same trouble:
> >
> > Reid reports that all his mobo can run is Win98 and neither 2K, Linux or BSDs
> > runs:
> >
> > http://groups.google.com/groups?hl=en&frame=right&th=b5b6671ee8e4a3ad&seekm=q2
> vZ
> > 6.77281%24ru2.22650002%40typhoon.southeast.rr.com#link1
> >
> > A reverting to the old BIOS does not help. I contacted Reid and he still has n
> o
> > fix for the situation.
> >
> > I got another example which is exactly like mine and Reid's:
> >
> > http://groups.google.com/groups?q=bx6+kernel+panic&hl=en&selm=37B6EE47.3D9A3AE
> 5%
> > 40ucs.orst.edu&rnum=3
> >
> > Kenneth ran NT on his BX6-2. It worked wonderfully. Then he flashed his BIOS
> > and it oopsed from thereon in. He went back to use his old BIOS but his OS is
> > still unstable. He tried Mandrake on it then and it won't succeed but oopses.
> >
> >
> > MY ANALYSIS
> > -----------
> >
> > The board as it was manufactured was able to run all types of OS. Manypeople
> > have assured me that is what they are doing.
> >
> > I and Reid though cannot run anything but Win9x. The thing that makes our case
>
> > special is that we both flashed the BIOS and then tried to install XP or Linux
> .
> >
> >
> > I do wonder if the BIOS installed at the plant was in some way special and any
>
> > BIOS flashed at home and downloaded from the web are containing inconsistencie
> s
> > which break the mobo from working with NT or Linux kernels. The fact that
> > others run Linux and Nt fine on their mobo but my flashed and revertedly
> > flashed mobo cannot makes me wonder.
> >
> >
> > THANK YOU
> > ---------
> >
> > I would love to contact Abit about this but they don't leave a single point of
>
> > contact on their web pages.
> >
> > I appreciate any comments in regards to this issue. I will now report what the
>
> > kernel errored on and it may help the hackers in building an opinion about thi
> s
> > issue.
> >
> > Thank you for reading this lengthy document! I could care less if Ican run a
> > certain OS on this specific by today's standard's outdatedbox but it bugs me
> > that there seems to have creaped in a bug intoABIT's board which makes it
> > incompatible with a advanced kernel like the Linux 2.2 or 2.4 kernel and the N
> T
> > kernel as well but gives it theability to handle the functionality of the Win9
> x
> > kernel.
> >
> > I wish somebody would invest the time in researching this topic in detail.
> >
> > I know ABIT has their own Linux distro. I do not have a copy of it but I would
>
> > be damn curious to know if their distro runs on their ownboard after a BIOS
> > flash.
> >
> > Thanks for your time and if you need more info contact me,
> >
> > Marcel
> >
> >
> > KERNEL OOPSES
> > -------------
> >
> > RH 7.2 default i686 kernel (I think its 2.4.7):
> >
> > EXT3-fs: mounted filesystem with ordered data mode.
> > Freeing unused kernel memory: 220k freed.
> > CPU 0: Machine Check Exception: 0000000000000004
> > Bank 3: b20000000002010a
> > Kernel panic: CPU context corrupt
> > In interrupt handler - not syncing
> >
> >
> > SuSE 6.4 default 2.2.14 kernel:
> >
> > Unable to handle kernel paging request at virtual address 05efa769.2%
> > current->tss.cr3 = 00101000, %cr3 = 00101000
> > *pde = 00000000
> > Oops: 0002
> > CPU: 0
> > EIP: 0010:[<c011825>]
> > EFLAGS: 00010013
> > eax: 000000ff ebx: c0289f24 ecx: 00008c7e edx: c0289f24
> > esi: 20000001 edi: 00000000 ebp: c0289e6c esp: c0289e74
> > ds: 0018 es: 0018 ss: 0018
> > Process swapper (pid: 0, process nr: 0, stackpage=c0289000)
> > Stack: c0289f24 24000001 0000000e c0289f24 c026ad58 00000001 c0289eacc01d43b4
> > 00000001 00000695............
> > Call Trace: ........................
> > Code: 28 8b 45 08 c7 45 f4 00 00 00 00 c7 45 f0 00 00 00 00 89 45
> > Aiee, killing interrupt handler
> > Kernel panic: Attempted to kill idle task!
> > In swapper task - not syncing.
> >
> >
> >
> >
> > Kenneth's Mandrake 6.0 Oops:
> >
> > Kernel panic: LRU block list corrupted
> > Unable to handle kernul NULL pointer dereference at virtual address
> > 00000028
> > current->tss.cr3 = 00101000, %cr3 = 00101000
> > *pde = 00000000
> > Oops: 0000
> > CPU: 0
> > .
> > .
> > .
> >
> >
> > LINKS
> > -----
> >
> > Abit Support:
> > http://fae.abit.com.tw/eng/
> >
> > Abit BX6 Rev. 2:
> > http://fae.abit.com.tw/eng/download/bios/bx620.htm
> >
> >
> > -
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to [email protected]
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at http://www.tux.org/lkml/
> >
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/