On 6/10/07, Wit <[EMAIL PROTECTED]> wrote: > Tijnema wrote: > > On 6/9/07, Wit <[EMAIL PROTECTED]> wrote: > >> Tijnema wrote: > >> > On 6/9/07, Wit <[EMAIL PROTECTED]> wrote: > >> >> Tijnema wrote: > >> >> ><snip> > > >> >> I suggest doing your kernel with CONFIG_ACPI=y and > >> CONFIG_ACPI_THERMAL=y > >> >> if you mobo support acpi. Then you can cat > >> >> /proc/acpi/thermal_zone/THRM/temperature > >> >> in a loop to get a more accurate reading. > >> > I've compiled kernel (2.6.17.11) with both options, and my motherboard > >> > does support ACPI, (I can see in dmesg log too), but the folder > >> > /proc/acpi/thermal_zone is empty for me. > >> > >> ?? Is /proc mounted? I *think* that unless something like acpi=off is on > >> the kernel command line, that ought to be available when /proc is > >> mounted. However, I am a "casual" user and maybe more is needed. There > >> were quite a few parms in the .config file related to acpi and I > >> selected only two obvious ones to mention. > > My LFS is shutdown ATM, but my CentOS 4.5 EL is 2.6.9 ATM and my CentOS > 5.0 is 2.6.18-8. Anyway, I showed the ACPI kernel settings for those > below. Since the mobo does support ACPI, is it enabled? I know it seems > dumb to ask such basic questions, but my experience is that I tend to > overlook the obvious too often.
When i look into the boot logs, it shows some ACPI things, so i guess it's enabled... /# dmesg | grep ACPI acpiphp: ACPI Hot Plug PCI Controller Driver version: 0.5 ACPI: PCI Interrupt Link [LNKA] enabled at IRQ 10 ACPI: PCI Interrupt 0000:01:00.0[A] -> Link [LNKA] -> GSI 10 (level, low) -> IRQ 10 ACPI: Power Button (FF) [PWRF] ACPI: Power Button (CM) [PWRB] ACPI: Sleep Button (CM) [SLPB] ACPI: CPU0 (power states: C1[C1] C2[C2]) ACPI: Processor [CPU0] (supports 2 throttling states) ACPI: PCI Interrupt 0000:00:0c.0[A] -> Link [LNKA] -> GSI 10 (level, low) -> IRQ 10 ACPI: PCI Interrupt Link [LNKD] enabled at IRQ 11 ACPI: PCI Interrupt 0000:00:07.2[D] -> Link [LNKD] -> GSI 11 (level, low) -> IRQ 11 ACPI: PCI Interrupt 0000:00:07.3[D] -> Link [LNKD] -> GSI 11 (level, low) -> IRQ 11 ACPI: PCI Interrupt Link [LNKC] enabled at IRQ 5 ACPI: PCI Interrupt 0000:00:07.5[C] -> Link [LNKC] -> GSI 5 (level, low) -> IRQ 5 ACPI: PCI Interrupt 0000:00:0a.0[A] -> Link [LNKC] -> GSI 5 (level, low) -> IRQ 5 ACPI: PCI Interrupt 0000:00:0a.1[A] -> Link [LNKC] -> GSI 5 (level, low) -> IRQ 5 ACPI: PCI interrupt for device 0000:00:0a.1 disabled > > > > > Yes, of course /proc is mounted :P > > > > The directory's /proc/acpi/thermal_zone exist, but that one is empty. > > Please look at this: > > cat config-2.6.17.11 | grep ACPI > > # Power management options (ACPI, APM) > > # ACPI (Advanced Configuration and Power Interface) Support > > CONFIG_ACPI=y > > CONFIG_ACPI_AC=y > > CONFIG_ACPI_BATTERY=y > > CONFIG_ACPI_BUTTON=y > > CONFIG_ACPI_VIDEO=y > > # CONFIG_ACPI_HOTKEY is not set > > CONFIG_ACPI_FAN=y > > CONFIG_ACPI_PROCESSOR=y > > CONFIG_ACPI_THERMAL=y > > # CONFIG_ACPI_ASUS is not set > > # CONFIG_ACPI_IBM is not set > > # CONFIG_ACPI_TOSHIBA is not set > > CONFIG_ACPI_BLACKLIST_YEAR=0 > > # CONFIG_ACPI_DEBUG is not set > > CONFIG_ACPI_EC=y > > CONFIG_ACPI_POWER=y > > CONFIG_ACPI_SYSTEM=y > > # CONFIG_ACPI_CONTAINER is not set > > CONFIG_HOTPLUG_PCI_ACPI=y > > CONFIG_HOTPLUG_PCI_ACPI_IBM=y > > CONFIG_PNPACPI=y > > > > Seems that i've enabled enough ACPI options right? > > Yes. But maybe you need this too? > > CONFIG_X86_MCE_P4THERMAL=y I have that one enabled too, any more suggestions? > > Just in case, I'll show my other acpi-related params from both my CentOS > systems. > > My CentOS 4.5 has these. > > CONFIG_ACPI=y > CONFIG_ACPI_BOOT=y > CONFIG_ACPI_INTERPRETER=y > CONFIG_ACPI_SLEEP=y > CONFIG_ACPI_SLEEP_PROC_FS=y > CONFIG_ACPI_AC=m > CONFIG_ACPI_BATTERY=m > CONFIG_ACPI_BUTTON=m > CONFIG_ACPI_FAN=y > CONFIG_ACPI_PROCESSOR=y > CONFIG_ACPI_THERMAL=y > CONFIG_ACPI_ASUS=m > CONFIG_ACPI_TOSHIBA=m > CONFIG_ACPI_BLACKLIST_YEAR=2001 > # CONFIG_ACPI_DEBUG is not set > CONFIG_ACPI_BUS=y > CONFIG_ACPI_EC=y > CONFIG_ACPI_POWER=y > CONFIG_ACPI_PCI=y > CONFIG_ACPI_SYSTEM=y > > This system has an SiS chipset, IIRC and seems to not need a special > module, as I see none in the lsmod output. > > And my 5.0 has this > > CONFIG_ACPI=y > CONFIG_ACPI_SLEEP=y > CONFIG_ACPI_SLEEP_PROC_FS=y > # CONFIG_ACPI_SLEEP_PROC_SLEEP is not set > CONFIG_ACPI_AC=m > CONFIG_ACPI_BATTERY=m > CONFIG_ACPI_BUTTON=m > CONFIG_ACPI_VIDEO=m > # CONFIG_ACPI_HOTKEY is not set > CONFIG_ACPI_FAN=y > CONFIG_ACPI_DOCK=y > CONFIG_ACPI_PROCESSOR=y > CONFIG_ACPI_HOTPLUG_CPU=y > CONFIG_ACPI_THERMAL=y > CONFIG_ACPI_ASUS=m > CONFIG_ACPI_IBM=m > CONFIG_ACPI_TOSHIBA=m > CONFIG_ACPI_BLACKLIST_YEAR=1999 > # CONFIG_ACPI_DEBUG is not set > CONFIG_ACPI_EC=y > CONFIG_ACPI_POWER=y > CONFIG_ACPI_SYSTEM=y > CONFIG_ACPI_CONTAINER=y > CONFIG_ACPI_SBS=m > > Also, the 5.0 has an inserted module, asus_acpi. I don't think every > system needs a special driver though, whether module or compiled in. But > you may need to check extended explanations in kernel config and see if > you might need the IBM/ASUS/TOSHIBA driver enabled? I have an Biostart M7VKD mobo, with 2 VIA chipsets on it, so I guess I don't need anything built in/ module. > > >> > >> >> <snip> > >> > >> >> Another test: remove everything you can and see if boot succeeds > >> at full > >> >> clocking. Then add items back in, one at time. The power requirements > >> >> listed on various devices may let you calculate the maximum load with > >> >> doing all this remove/re-insert. > >> > > >> > Are you sure you read the whole thread? I said that when I set my CPU > >> > speed to the original speed, the whole PC wouldn't start, nor the fans > >> > started spinning... > >> > I tried to boot the PC with only CPU+Motherboard+RAM+PS, but even then > >> > the fans didn't start to spin. > >> > >> Yes, I read the whole thread. Please forgive my faulty memory of all the > >> details. > > > > Sure i will :P > >> > >> >> <snip> > >> > >> > > >> > When downclocked, booting isn't a problem. It's just in a regular > >> > >> That's one of the things that reminds me that PS may be involved. Higher > >> frequency = higher amperage needs. Of course, there could be other > >> things, like timing issues, that would also be affected by the frequency > >> selection. > > > > And why does it stop working after some time? > > Asking for theories here? Or just testing them? Regardless, the power > supply is affected by heat. The capacitors are so affected. Everything > on the board that is electrical will suffer some degradation as they > heat up. Overheated CPU/memory/... are the most common victims of > overheat, but a power supply or capacitor or circuit (with a "cold > solder" joint) can begin to fail as things get warm. Ok, just to make sure the CPU cooling was good enough, I placed 3 fans on it, all 3 in right direction, so that it got ultra cooled :) Still, it didn't pass the GCC testsuite, as it stopped after maybe 30-60 seconds. > > Experience indicates that when things fail at seemingly random times > after running for a short (long) while, and does it consistently, > something is affected by the heat. It may be not too hot, it may be a > week circuit that develops an open/short when it gets warmed up. > > As power supply heats up, its efficiency drops. If it has a bad circuit, > it may open/short. > > It all *seems* somewhat random to us because we cannot see the variables > that are in play. > > That's wehn the aggregate experience of the list may become helpful. > Sometimes others (in large numbers) have faced this sort of problem > before and found "solutions". Some of the folks are even hardware > "experts" with EE degrees, etc. I'm not, so I rely on others and my own > experience over many decades. > > Regardless, it's all "a shot in the dark" when talking about a system > you can't instrument. Tomorrow I will try the system with a Asus 350W PSU, which now runs an AMD X2 4400+ with 2 HDDs and a GeForce 7600GT..., should be enough for my old system :) > > >> <snip> > > >> > >> > > >> > I've built my own case, it's more like a wooden block :P No covers or > >> > such, but I can get front and back off it, both are bigger than a > >> > regular cover :P > >> > >> What are you using for chassis ground? I'm no elec whiz, but I seem to > >> recall stuff in the past that talks about Vcc floating around and having > >> nasty effects if good ground is not available. Seems like a weak > >> possibility though. > > > > There are just a few pins screwed into the wood, and well, that's it... > > I don't have ground at the place my server is, so It doesn't matter if > > I connect it to the ground of the PSU... > > Does your plug have 3 prongs? If so, a "ground" is provided via the > household wiring. Nope, and a dutch plug does never. We have different plugs than else around the world, but I've already opened the power-point a few times, and there are only 2 wires inside.... > > But it is important that the mobo and PS see the same Vcc (IIRC) so that > there is no ground potential difference. Run a wire from one of the mobo > mounting pads that has metal under it to a convenient screw on the PS > case. That should assure both are at same ground potential, avoid static > charge buildup on the mobo (another possible cause of "random" lockup) > and is a good thing even if not related to your problem. Be sure the > wire is insulated except at the very ends. Don't let it touch anything > hot or be placed such that is could suffer abrasion damage. Hmm, when I'm replacing the PSU tomorrow, i'll see if i can add another wire in the case :) > > A wire from the PS external screw to any metal cold water pipe, metal > radiator, ... should provide a good ground and bleed off any static > buildup. A wooden case may tend to encourage static buildup. I'm not sure. That's the problem, the room where my pc (=server) is located is in a room, where there is only some wood and a stone wall... If I really need it, I would need to place a cable of about 10-15 meters to get to some metal, but then I could better connect it to a real ground i think :) > > > <snip> > > > > > Tijnema > > > > Well, I'm out of thoughts. Again, good luck with it. > > -- > Wit. > You've already brought up a lot of points that could cause the problem :) I'll try them all one by one... :) Really thanks! Tijnema -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
