 |
|
| Linux Forum Index » Linux - Suse Forum » Linux BIOS Challenge |
|
Page 1 of 1 |
|
| Author |
Message |
| Guest |
 Posted: Fri Mar 25, 2005 10:11 am |
|
|
|
|
HI
I'd like to conduct a survey on how many Suse Linux users found that
their BIOS was no longer detecting their hard drives, and what BIOS
they were using.
This week I have found 3 users who had such difficulty from the
moment they installed that O/S. It's difficult to rationalise, but
then it always is until you have the answer.
Last night I installed XP on a 4Gb had drive that had been acting as a
paper weight on my desk, and as the only hard drive with nothing else
on it, it is running like silk. A sewing machine couldn't be steadier.
Yet the day before, when Linux was present, it was falling all over
the place.
If Linux is responsible how can it affect the BIOS before Linux is
loaded? I just don't know. Could it be affecting the BIOS on log off
and leaving the fault as some sort of legacy? There has to be an
explanation.
Colin |
|
|
| Back to top |
|
|
|
| Stephan |
| Posted: Fri Mar 25, 2005 10:30 am |
|
|
|
Guest
|
me@privacy.net wrote:
Quote: If Linux is responsible how can it affect the BIOS before Linux is
loaded? I just don't know. Could it be affecting the BIOS on log off
and leaving the fault as some sort of legacy? There has to be an
explanation.
Linux can't affect the Bios, but it uses different ways to work with your
hardware. Explanation from the Kernel docs:
On PCI systems, the BIOS can be used to detect the PCI devices and
determine their configuration. However, some old PCI motherboards
have BIOS bugs and may crash if this is done. Also, some embedded
PCI-based systems don't have any BIOS at all. Linux can also try to
detect the PCI hardware directly without using the BIOS.
With this option, you can specify how Linux should detect the
PCI devices. If you choose "BIOS", the BIOS will be used,
if you choose "Direct", the BIOS won't be used, and if you
choose "MMConfig", then PCI Express MMCONFIG will be used.
If you choose "Any", the kernel will try MMCONFIG, then the
direct access method and falls back to the BIOS if that doesn't
work. If unsure, go with the default, which is "Any". |
|
|
| Back to top |
|
|
|
| Guest |
| Posted: Fri Mar 25, 2005 2:24 pm |
|
|
|
|
On Fri, 25 Mar 2005 16:30:08 +0100, Stephan <Spaceman26@gmx.net>
wrote:
Quote: me@privacy.net wrote:
If Linux is responsible how can it affect the BIOS before Linux is
loaded? I just don't know. Could it be affecting the BIOS on log off
and leaving the fault as some sort of legacy? There has to be an
explanation.
Linux can't affect the Bios, but it uses different ways to work with your
hardware. Explanation from the Kernel docs:
On PCI systems, the BIOS can be used to detect the PCI devices and
determine their configuration. However, some old PCI motherboards
have BIOS bugs and may crash if this is done. Also, some embedded
PCI-based systems don't have any BIOS at all. Linux can also try to
detect the PCI hardware directly without using the BIOS.
With this option, you can specify how Linux should detect the
PCI devices. If you choose "BIOS", the BIOS will be used,
if you choose "Direct", the BIOS won't be used, and if you
choose "MMConfig", then PCI Express MMCONFIG will be used.
If you choose "Any", the kernel will try MMCONFIG, then the
direct access method and falls back to the BIOS if that doesn't
work. If unsure, go with the default, which is "Any".
Good stuff, but I think it's gone beyond "it can't possibly happen".
It already has happened. We should find out how widespread it is and
explain it.
Colin |
|
|
| Back to top |
|
|
|
| Dave M |
| Posted: Fri Mar 25, 2005 4:31 pm |
|
|
|
Guest
|
me@privacy.net wrote:
Quote: On Fri, 25 Mar 2005 16:30:08 +0100, Stephan <Spaceman26@gmx.net
wrote:
me@privacy.net wrote:
If Linux is responsible how can it affect the BIOS before Linux is
loaded? I just don't know. Could it be affecting the BIOS on log off
and leaving the fault as some sort of legacy? There has to be an
explanation.
Linux can't affect the Bios, but it uses different ways to work with your
hardware. Explanation from the Kernel docs:
On PCI systems, the BIOS can be used to detect the PCI devices and
determine their configuration. However, some old PCI motherboards
have BIOS bugs and may crash if this is done. Also, some embedded
PCI-based systems don't have any BIOS at all. Linux can also try to
detect the PCI hardware directly without using the BIOS.
With this option, you can specify how Linux should detect the
PCI devices. If you choose "BIOS", the BIOS will be used,
if you choose "Direct", the BIOS won't be used, and if you
choose "MMConfig", then PCI Express MMCONFIG will be used.
If you choose "Any", the kernel will try MMCONFIG, then the
direct access method and falls back to the BIOS if that doesn't
work. If unsure, go with the default, which is "Any".
Good stuff, but I think it's gone beyond "it can't possibly happen".
It already has happened. We should find out how widespread it is and
explain it.
Colin
Your linux-based OS installation is not doing anything to the firmware of
the BIOS. Software doesn't just go around flashing BIOS ROMs. And if
somehow your BIOS information was modified, it certainly wouldn't just
revert back to a previous state when the software was removed.
I believe you may have some issue, but I think part of it is with diagnosis
of the problem.
--Dave M |
|
|
| Back to top |
|
|
|
| Colin Bearfield |
| Posted: Fri Mar 25, 2005 5:30 pm |
|
|
|
Guest
|
On Fri, 25 Mar 2005 16:31:50 -0500, Dave M <dave@landlordhat.com>
wrote:
Quote: me@privacy.net wrote:
On Fri, 25 Mar 2005 16:30:08 +0100, Stephan <Spaceman26@gmx.net
wrote:
me@privacy.net wrote:
If Linux is responsible how can it affect the BIOS before Linux is
loaded? I just don't know. Could it be affecting the BIOS on log off
and leaving the fault as some sort of legacy? There has to be an
explanation.
Linux can't affect the Bios, but it uses different ways to work with your
hardware. Explanation from the Kernel docs:
On PCI systems, the BIOS can be used to detect the PCI devices and
determine their configuration. However, some old PCI motherboards
have BIOS bugs and may crash if this is done. Also, some embedded
PCI-based systems don't have any BIOS at all. Linux can also try to
detect the PCI hardware directly without using the BIOS.
With this option, you can specify how Linux should detect the
PCI devices. If you choose "BIOS", the BIOS will be used,
if you choose "Direct", the BIOS won't be used, and if you
choose "MMConfig", then PCI Express MMCONFIG will be used.
If you choose "Any", the kernel will try MMCONFIG, then the
direct access method and falls back to the BIOS if that doesn't
work. If unsure, go with the default, which is "Any".
Good stuff, but I think it's gone beyond "it can't possibly happen".
It already has happened. We should find out how widespread it is and
explain it.
Colin
Your linux-based OS installation is not doing anything to the firmware of
the BIOS. Software doesn't just go around flashing BIOS ROMs. And if
somehow your BIOS information was modified, it certainly wouldn't just
revert back to a previous state when the software was removed.
I believe you may have some issue, but I think part of it is with diagnosis
of the problem.
--Dave M
Wel, yes. Er.. we can't diagnose it because we don't yet have the
explanation, but it only happens in the presence of Linux. No one has
explained that away. |
|
|
| Back to top |
|
|
|
| George Peatty |
| Posted: Fri Mar 25, 2005 7:42 pm |
|
|
|
Guest
|
On Fri, 25 Mar 2005 15:11:17 +0000, me wrote:
Quote: HI
I'd like to conduct a survey on how many Suse Linux users found that
their BIOS was no longer detecting their hard drives, and what BIOS
they were using.
[snip]
I have unresolved problems with SuSE 9.2 failing to detect a slave drive.
When I exit out of Linux, the slave drive no longer appears in the BIOS.
The only way to get it back is to cold boot, but that does recover the
drive. The drive is read/writeable in Windows 98, and even read/writeable
from a networked PC also running SuSE 9.2, so I have to believe it's not
the hard drive. I have an email into SuSE support. Since YaST didn't
detect the drive during install, I should get an answer. If I hear
anything useful, I'll post here. |
|
|
| Back to top |
|
|
|
| Larry I Smith |
| Posted: Fri Mar 25, 2005 9:13 pm |
|
|
|
Guest
|
George Peatty wrote:
Quote: On Fri, 25 Mar 2005 15:11:17 +0000, me wrote:
HI
I'd like to conduct a survey on how many Suse Linux users found that
their BIOS was no longer detecting their hard drives, and what BIOS
they were using.
[snip]
I have unresolved problems with SuSE 9.2 failing to detect a slave drive.
When I exit out of Linux, the slave drive no longer appears in the BIOS.
The only way to get it back is to cold boot, but that does recover the
drive. The drive is read/writeable in Windows 98, and even read/writeable
from a networked PC also running SuSE 9.2, so I have to believe it's not
the hard drive. I have an email into SuSE support. Since YaST didn't
detect the drive during install, I should get an answer. If I hear
anything useful, I'll post here.
I had that problem once, a long time ago...
It turned out that the 2 drives (master and slave)
were not jumpered correctly.
Regards,
Larry
--
Anti-spam address, change each 'X' to '.' to reply directly. |
|
|
| Back to top |
|
|
|
| Kevin Miller |
| Posted: Fri Mar 25, 2005 10:42 pm |
|
|
|
Guest
|
me@privacy.net wrote:
Quote: HI
I'd like to conduct a survey on how many Suse Linux users found that
their BIOS was no longer detecting their hard drives, and what BIOS
they were using.
This week I have found 3 users who had such difficulty from the
moment they installed that O/S. It's difficult to rationalise, but
then it always is until you have the answer.
Last night I installed XP on a 4Gb had drive that had been acting as a
paper weight on my desk, and as the only hard drive with nothing else
on it, it is running like silk. A sewing machine couldn't be steadier.
Yet the day before, when Linux was present, it was falling all over
the place.
If Linux is responsible how can it affect the BIOS before Linux is
loaded? I just don't know. Could it be affecting the BIOS on log off
and leaving the fault as some sort of legacy? There has to be an
explanation.
Colin
Can you be more specific? You say when it isn't detected. You mean at
boot time? It flashes by rapidly, but usually the bios lists the drives
it finds when going through the POST process. It isn't listed? (You
may have to turn off the splash screen that so many vendors put on
computers these days to see it.) What happens if you go into the BIOS?
Does it show up in there? I presume we're talking an IDE drive. Is
it set to auto-recognize the attached device?
When you say the day before when Linux was present, do you mean just
Linux or along w/XP? Any other devices on the IDE cable? What BIOS are
you using? What hardware? Did you try reseating the cables?
If you want to troubleshoot this you need to be systematic. Strip out
everything that isn't involved like a 2nd drive or OS, put the drive on
it's own IDE cable, not shared w/the CD ROM, DVD, etc. Check the cable
select jumper on the drive to make sure it's set appropriately, etc.
Your description is just a bit vague to really expect us to try to
pinpoint what's going on. I haven't seen anything like that w/a hard
drive but I had a similar issue w/a DVD drive once. When I first
powered on the computer, that is from a cold start by toggling the power
supply switch, I couldn't see the drive. If I then pressed the power
button on the front or did a warm boot with Ctl-Alt-Del the DVD drive
would be accessible. It turned out to be a bad DVD drive. I got a
replacement and haven't had a problem since. You description really
smacks of a loose cable or something that is now inserted fully when you
promoted the HD from paperweight status and (re)installed it into the
computer. Or maybe I'm misreading that? You have to be clear in your
write up because we can't see the computer from here...
....Kevin
--
Kevin Miller
http://www.alaska.net/~atftb
Juneau, Alaska |
|
|
| Back to top |
|
|
|
| Guest |
| Posted: Sat Mar 26, 2005 1:41 am |
|
|
|
|
On Fri, 25 Mar 2005 18:42:09 -0900, Kevin Miller
<millerboys@acsalaska.net> wrote:
Quote: me@privacy.net wrote:
HI
I'd like to conduct a survey on how many Suse Linux users found that
their BIOS was no longer detecting their hard drives, and what BIOS
they were using.
This week I have found 3 users who had such difficulty from the
moment they installed that O/S. It's difficult to rationalise, but
then it always is until you have the answer.
Last night I installed XP on a 4Gb had drive that had been acting as a
paper weight on my desk, and as the only hard drive with nothing else
on it, it is running like silk. A sewing machine couldn't be steadier.
Yet the day before, when Linux was present, it was falling all over
the place.
If Linux is responsible how can it affect the BIOS before Linux is
loaded? I just don't know. Could it be affecting the BIOS on log off
and leaving the fault as some sort of legacy? There has to be an
explanation.
Colin
Can you be more specific? You say when it isn't detected. You mean at
boot time? It flashes by rapidly, but usually the bios lists the drives
it finds when going through the POST process. It isn't listed? (You
may have to turn off the splash screen that so many vendors put on
computers these days to see it.) What happens if you go into the BIOS?
Does it show up in there? I presume we're talking an IDE drive. Is
it set to auto-recognize the attached device?
When you say the day before when Linux was present, do you mean just
Linux or along w/XP? Any other devices on the IDE cable? What BIOS are
you using? What hardware? Did you try reseating the cables?
If you want to troubleshoot this you need to be systematic. Strip out
everything that isn't involved like a 2nd drive or OS, put the drive on
it's own IDE cable, not shared w/the CD ROM, DVD, etc. Check the cable
select jumper on the drive to make sure it's set appropriately, etc.
Your description is just a bit vague to really expect us to try to
pinpoint what's going on. I haven't seen anything like that w/a hard
drive but I had a similar issue w/a DVD drive once. When I first
powered on the computer, that is from a cold start by toggling the power
supply switch, I couldn't see the drive. If I then pressed the power
button on the front or did a warm boot with Ctl-Alt-Del the DVD drive
would be accessible. It turned out to be a bad DVD drive. I got a
replacement and haven't had a problem since. You description really
smacks of a loose cable or something that is now inserted fully when you
promoted the HD from paperweight status and (re)installed it into the
computer. Or maybe I'm misreading that? You have to be clear in your
write up because we can't see the computer from here...
...Kevin
I'm very happy to give you more info. I didn't want to scare people
off with loads of un interesting detail.
I know a lot more now than I did 3 months ago.
The system is my old machine running two hard drives. HD0 is 18Gb
which 5½ years ago was a staggering size. HD1 is 20Gb. Both are
entirely FAT32. Until the advent of Linux I ran XP, 98SE (twice), and
DOS 6.22, and multibooting was performed by BootitNG. The system was
rock steady during the time and was networked through a Linksys router
to my "best" computer which ran only XP. The processor is an AMD
550Ghz. The BIOS is Award and the mobo MS-6167.
I installed Linux and when finished I booted up and it failed. I
noticed that the error message said "disk failure" and referred to the
CD-ROM. I assumed that the BIOS was still pointing to the CD-ROM as
the first boot up device and entered the BIOS to change it. It was in
good order and pointing to the floppy drive.
Linux had been installed in 12Gb of a second partition behind XP.
Overtime I realised that the POST message had flown past and that this
was happening:
At boot up the BIOS found no floppy in the first device and had failed
to detect both hard drives (usually) and had finished up pointing at
an empty 3rd device - the CD-ROM.
I found that very often the solution was to do a warm boot whereupon
the BIOS successfully detected both drives and ran the Suse boot menu.
Recently I flashed an update (1999) to the BIOS to cover that option.
A few days ago I decided to buy a new hard drive. The decision meant
that I had decided to support the old machine and maintain it for the
near future.
I had an old 4Gb hard drive acting as a paper weight on my desk and
decide to do a dry run. I installed XP and it went like a dream. It
created my internet connection and found my network. It booted up
perfectly every time. Even my wife used it for the web. BIOS
aberration is a thing of the past.
I had disconnected the data cable serving my 2 hard drives at mobo
level and used a spare cable to connect to the old drive (which is
just resting on the floor of my case for the moment).
If it helps I could run this system for a period of time, say a week
or a month, and then move back to just HD0 with XP and Linux, or both
hard drives and monitor any BIOS non-detection, having first ensured
proper seating of cables.
Thank you for your time and interest. I acknowledge that my premise
might be false, but in the past the impossible has been found to be
possible (just think how the flat earth people feel today) and I am
glad that this issue is being considered.
Colin |
|
|
| Back to top |
|
|
|
| Joost van der Waa |
| Posted: Sat Mar 26, 2005 3:30 am |
|
|
|
Guest
|
Dave M wrote:
Quote: me@privacy.net wrote:
On Fri, 25 Mar 2005 16:30:08 +0100, Stephan <Spaceman26@gmx.net
wrote:
me@privacy.net wrote:
If Linux is responsible how can it affect the BIOS before Linux is
loaded? I just don't know. Could it be affecting the BIOS on log off
and leaving the fault as some sort of legacy? There has to be an
explanation.
Linux can't affect the Bios, but it uses different ways to work with your
hardware. Explanation from the Kernel docs:
On PCI systems, the BIOS can be used to detect the PCI devices and
determine their configuration. However, some old PCI motherboards
have BIOS bugs and may crash if this is done. Also, some embedded
PCI-based systems don't have any BIOS at all. Linux can also try to
detect the PCI hardware directly without using the BIOS.
With this option, you can specify how Linux should detect the
PCI devices. If you choose "BIOS", the BIOS will be used,
if you choose "Direct", the BIOS won't be used, and if you
choose "MMConfig", then PCI Express MMCONFIG will be used.
If you choose "Any", the kernel will try MMCONFIG, then the
direct access method and falls back to the BIOS if that doesn't
work. If unsure, go with the default, which is "Any".
Good stuff, but I think it's gone beyond "it can't possibly happen".
It already has happened. We should find out how widespread it is and
explain it.
Colin
Your linux-based OS installation is not doing anything to the firmware of
the BIOS. Software doesn't just go around flashing BIOS ROMs. And if
somehow your BIOS information was modified, it certainly wouldn't just
revert back to a previous state when the software was removed.
I believe you may have some issue, but I think part of it is with diagnosis
of the problem.
--Dave M
Couldn't agree more
Joost |
|
|
| Back to top |
|
|
|
| Guest |
| Posted: Sat Mar 26, 2005 4:06 am |
|
|
|
|
On Sat, 26 Mar 2005 09:30:50 +0100, Joost van der Waa
<Joost.vander.waa@whipethishccnet.nl> wrote:
Quote: Dave M wrote:
me@privacy.net wrote:
On Fri, 25 Mar 2005 16:30:08 +0100, Stephan <Spaceman26@gmx.net
wrote:
me@privacy.net wrote:
If Linux is responsible how can it affect the BIOS before Linux is
loaded? I just don't know. Could it be affecting the BIOS on log off
and leaving the fault as some sort of legacy? There has to be an
explanation.
Linux can't affect the Bios, but it uses different ways to work with your
hardware. Explanation from the Kernel docs:
On PCI systems, the BIOS can be used to detect the PCI devices and
determine their configuration. However, some old PCI motherboards
have BIOS bugs and may crash if this is done. Also, some embedded
PCI-based systems don't have any BIOS at all. Linux can also try to
detect the PCI hardware directly without using the BIOS.
With this option, you can specify how Linux should detect the
PCI devices. If you choose "BIOS", the BIOS will be used,
if you choose "Direct", the BIOS won't be used, and if you
choose "MMConfig", then PCI Express MMCONFIG will be used.
If you choose "Any", the kernel will try MMCONFIG, then the
direct access method and falls back to the BIOS if that doesn't
work. If unsure, go with the default, which is "Any".
Good stuff, but I think it's gone beyond "it can't possibly happen".
It already has happened. We should find out how widespread it is and
explain it.
Colin
Your linux-based OS installation is not doing anything to the firmware of
the BIOS. Software doesn't just go around flashing BIOS ROMs. And if
somehow your BIOS information was modified, it certainly wouldn't just
revert back to a previous state when the software was removed.
I believe you may have some issue, but I think part of it is with diagnosis
of the problem.
--Dave M
Couldn't agree more
Joost
That's what someone standing next to the flat earther said! I think
that Kevin Miller's mind is more open.
BTW
When I installed Linux there was interference with the cables or the
hardware of any kind. Seems to rule out loose connections! The fact
stubbornly remains - without Linux, total detection, with Linux less
than 50%. And the issue repeats across time and space.
Colin |
|
|
| Back to top |
|
|
|
| James Knott |
| Posted: Sat Mar 26, 2005 6:08 pm |
|
|
|
Guest
|
Kevin Miller wrote:
Quote: Isn't that bizarre? And in 5 more years we'll wonder how we ever got by
w/only a 160 GB drive!
And around that time, Linux will be ready for 128 bit CPUs, about two years
ahead of Windows!  |
|
|
| Back to top |
|
|
|
|
|
All times are GMT - 5 Hours
The time now is Wed Mar 17, 2010 10:24 pm
|
|