1. need to add acpi support in s2881 LinuxBIOS support to get it. Anyway 8111 acpi support is already there in Serengeti_leopard, so it will be easy to get that.
YH
-----Original Message----- From: linuxbios-bounces@linuxbios.org [mailto:linuxbios-bounces@linuxbios.org] On Behalf Of Ward Vandewege Sent: Monday, April 10, 2006 3:43 PM To: linuxbios@linuxbios.org Subject: [LinuxBIOS] some more tyan/s2881 questions
So the board is stable now. That's great.
Some observations:
1. 'halt' no longer powers the machine off, which it does with the proprietary bios. 'halt -p' does the job though. LinuxBIOS oddity?
2. reboots seem to cause problems sometimes. I'm not sure yet what causes this, but I've had problems with the machine hanging after
ACPI: Unable to locate RSDP
It normally sits there for a few seconds, but I had the machine in such a state that it simply hung at that point.
This may have had something to do with the IPMI card, which I'm experimenting with. I had to unplug the power before this problem went away, a cold restart didn't help. I will do more testing to see if I can pinpoint this.
Thanks for all your help so far! Ward.
On Mon, Apr 10, 2006 at 04:00:27PM -0700, Lu, Yinghai wrote:
- need to add acpi support in s2881 LinuxBIOS support to get it.
Anyway 8111 acpi support is already there in Serengeti_leopard, so it will be easy to get that.
Thanks, I see that halt now powers off the machine properly (svn 2254)!
Still has the delays after
ACPI: Unable to locate RSDP
on boot.
Thanks, Ward.
-----Original Message----- From: linuxbios-bounces@linuxbios.org [mailto:linuxbios-bounces@linuxbios.org] On Behalf Of Ward Vandewege Sent: Monday, April 10, 2006 3:43 PM To: linuxbios@linuxbios.org Subject: [LinuxBIOS] some more tyan/s2881 questions
So the board is stable now. That's great.
Some observations:
- 'halt' no longer powers the machine off, which it does with the
proprietary bios. 'halt -p' does the job though. LinuxBIOS oddity?
- reboots seem to cause problems sometimes. I'm not sure yet what
causes this, but I've had problems with the machine hanging after
ACPI: Unable to locate RSDP
It normally sits there for a few seconds, but I had the machine in such a state that it simply hung at that point.
This may have had something to do with the IPMI card, which I'm experimenting with. I had to unplug the power before this problem went away, a cold restart didn't help. I will do more testing to see if I can pinpoint this.
Thanks for all your help so far! Ward.
-- Ward Vandewege ward@fsf.org Free Software Foundation - Senior System Administrator
-- linuxbios mailing list linuxbios@linuxbios.org http://www.openbios.org/mailman/listinfo/linuxbios
-- linuxbios mailing list linuxbios@linuxbios.org http://www.openbios.org/mailman/listinfo/linuxbios
!DSPAM:443ae48e228441410093335!
On Mon, Apr 10, 2006 at 04:00:27PM -0700, Lu, Yinghai wrote:
- need to add acpi support in s2881 LinuxBIOS support to get it.
Anyway 8111 acpi support is already there in Serengeti_leopard, so it will be easy to get that.
So.... if there is no ACPI support yet for this board in LinuxBIOS, does that mean that there is no automatic fan control in hardware? In other words, that the CPUs will cook if they get too hot, because the fans won't start blowing harder automatically?
As a point of interest, somewhere in the process of upgrading this box to LinuxBIOS, the fans got stuck at maximum speed (which is pretty loud). I'm not sure where it happened, but in any case booting with the proprietary bios didn't change a thing - they were still stuck at 17K RPM.
I've now figured out how to monitor and set fan speed through lm_sensors, but I want to know if I need to run fancontrold (daemon that keeps an eye on cpu temperature and adjusts fan speed as necessary) to prevent hardware overheating?
Thanks, Ward.
On 4/20/06, Ward Vandewege ward@gnu.org wrote:
On Mon, Apr 10, 2006 at 04:00:27PM -0700, Lu, Yinghai wrote:
- need to add acpi support in s2881 LinuxBIOS support to get it.
Anyway 8111 acpi support is already there in Serengeti_leopard, so it will be easy to get that.
So.... if there is no ACPI support yet for this board in LinuxBIOS, does that mean that there is no automatic fan control in hardware? In other words, that the CPUs will cook if they get too hot, because the fans won't start blowing harder automatically?
If they do then the hardware designers failed miserably. Generally what I find is that when you first power on the machine the fans come on full. Then the bios takes over. What happens in the case when the power supply is going flaky and machine glitches on startup and doesn't fully boot?
If you pull the bios chip and then power up the board and the fans don't come on full then I would find a different manufacturer since that one is just waiting to cook itself in a crash.
-- Richard A. Smith
On Thu, Apr 20, 2006 at 06:25:29PM -0500, Richard Smith wrote:
On 4/20/06, Ward Vandewege ward@gnu.org wrote:
On Mon, Apr 10, 2006 at 04:00:27PM -0700, Lu, Yinghai wrote:
- need to add acpi support in s2881 LinuxBIOS support to get it.
Anyway 8111 acpi support is already there in Serengeti_leopard, so it will be easy to get that.
So.... if there is no ACPI support yet for this board in LinuxBIOS, does that mean that there is no automatic fan control in hardware? In other words, that the CPUs will cook if they get too hot, because the fans won't start blowing harder automatically?
If they do then the hardware designers failed miserably. Generally what I find is that when you first power on the machine the fans come on full. Then the bios takes over. What happens in the case when the power supply is going flaky and machine glitches on startup and doesn't fully boot?
If you pull the bios chip and then power up the board and the fans don't come on full then I would find a different manufacturer since that one is just waiting to cook itself in a crash.
So far I'm really happy with Tyan and this board in particular. I'll see if I can try that tomorrow. I'd be surprised if the fans don't do the right thing.
Ward.
So far I'm really happy with Tyan and this board in particular. I'll see if I can try that tomorrow. I'd be surprised if the fans don't do the right thing.
I'm sure they will. Tyan is good stuff.
One thing many years of embedded developement has taught me is that if all your safeguards are software then its not a matter of "if" but "when" for system failure in a some catastrophic method.
If it were not for watchdog timers I'm sure 990% of the embedded systems we rely on would stop on a regular basis.
-- Richard A. Smith
On Thu, Apr 20, 2006 at 06:25:29PM -0500, Richard Smith wrote:
On 4/20/06, Ward Vandewege ward@gnu.org wrote:
On Mon, Apr 10, 2006 at 04:00:27PM -0700, Lu, Yinghai wrote:
- need to add acpi support in s2881 LinuxBIOS support to get it.
Anyway 8111 acpi support is already there in Serengeti_leopard, so it will be easy to get that.
So.... if there is no ACPI support yet for this board in LinuxBIOS, does that mean that there is no automatic fan control in hardware? In other words, that the CPUs will cook if they get too hot, because the fans won't start blowing harder automatically?
If they do then the hardware designers failed miserably. Generally what I find is that when you first power on the machine the fans come on full. Then the bios takes over.
I can confirm that this is what happens when the s2881 is first plugged in. The fans go full speed, regardless of the BIOS.
Ward.
If they do then the hardware designers failed miserably. Generally what I find is that when you first power on the machine the fans come on full. Then the bios takes over.
I can confirm that this is what happens when the s2881 is first plugged in. The fans go full speed, regardless of the BIOS.
Soif you set the fan speed really low does the system override you when you run something like cpuburn?
-- Richard A. Smith
-
Lu, Yinghai -
Richard Smith -
Ward Vandewege