Hi, I have a shiny new Intel Core Duo T2300 laptop with an nvidia geforce go 7400 graphics chipset. When I boot into Linux, there is a process in 'top' that is showing a 90% average cpu usage and it won't stop. I cannot kill it. There is an events/1 process that is always at 0%. I assume this is something to do with the processor and perhaps acpi, since there are two events processes and my processor has two cores. Any suggestions on what the problem may be? I have tried: 1. running with the basic provided graphics drivers for nvidia and no change 2. running with the nvidia provided graphics drivers and no change other than improved video performance Jon.
On Wed, 2006-04-19 at 11:30 +0100, Jon Black wrote:
Hi, I have a shiny new Intel Core Duo T2300 laptop with an nvidia geforce go 7400 graphics chipset. When I boot into Linux, there is a process in 'top' that is showing a 90% average cpu usage and it won't stop. I cannot kill it.
What is the process showing 90% utilization? What is it that leads you to believe it is related to the NVidia chip? What version of Linux are you running? Probably SUSE, but what release?
There is an events/1 process that is always at 0%. I assume this is something to do with the processor and perhaps acpi, since there are two events processes and my processor has two cores.
If you suspect the problem is multi-core specific, install the kernel-default rpm rather than the kernel-smp. That kernel will only use one processor (core). Also, if you suspect acpi, you could try the boot parameter "acpi=off". This is set if you boot using the failsafe option (from the bootloader). Thanks, Chad
On 19/04/06, Chad Groneman <linuxlistacct@gmail.com> wrote:
On Wed, 2006-04-19 at 11:30 +0100, Jon Black wrote:
Hi, I have a shiny new Intel Core Duo T2300 laptop with an nvidia geforce go 7400 graphics chipset. When I boot into Linux, there is a process in 'top' that is showing a 90% average cpu usage and it won't stop. I cannot kill it.
What is the process showing 90% utilization? What is it that leads you to believe it is related to the NVidia chip? What version of Linux are you running? Probably SUSE, but what release?
Sorry the information wasn't very concise. I am running SUSE 10.1 RC1. During install it was very, very slow, which may or may not be a related issue (worth mentioning in case). I do not think it is related to nvidia but again, wanted to add it to show what I had tried. The process that is taking up 90% of the CPU is called events/0. There is an events/1 process that has 0% of the CPU...I'm not sure what these processes do but I think they are processor related or perhaps acpi.
There is an events/1 process that is always at 0%. I assume this is something to do with the processor and perhaps acpi, since there are two events processes and my processor has two cores.
If you suspect the problem is multi-core specific, install the kernel-default rpm rather than the kernel-smp. That kernel will only use one processor (core).
Also, if you suspect acpi, you could try the boot parameter "acpi=off". This is set if you boot using the failsafe option (from the bootloader).
Thanks, Chad
--------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory-unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory-help@opensuse.org
On Wed, 2006-04-19 at 18:53 +0100, Jon Black wrote:
The process that is taking up 90% of the CPU is called events/0. There is an events/1 process that has 0% of the CPU...I'm not sure what these processes do but I think they are processor related or perhaps acpi.
I don't know exactly what the events/ processes are, but my guess pretty much the same as yours. My guess is interrupt handling. It could be ACPI interrupts, so you could try the failsafe method. I don't know the ramifications of doing that though - it might be that there's a lot of interrupts because something's not working properly (hardware). Chad
On Wednesday 19 April 2006 23:29, Chad Groneman wrote:
On Wed, 2006-04-19 at 18:53 +0100, Jon Black wrote:
The process that is taking up 90% of the CPU is called events/0. There is an events/1 process that has 0% of the CPU...I'm not sure what these processes do but I think they are processor related or perhaps acpi.
I don't know exactly what the events/ processes are, but my guess pretty much the same as yours. My guess is interrupt handling. It could be ACPI interrupts, so you could try the failsafe method. I don't know the ramifications of doing that though - it might be that there's a lot of interrupts because something's not working properly (hardware).
The output of dmesg and lspci might be useful. Maybe there are some evil ACPI or APIC errors who trouble the kernel :) Philipp
Sorry for the late reply. I have now installed 10.1 RC3 and the problem has spread so that both events/0 and events/1 are hogging the cpu. The output from top, dmesg and lspci can be found at this address: http://pastebin.com/697722 Thanks. On 20/04/06, Philipp Wollermann <philipp@wollermann.name> wrote:
On Wednesday 19 April 2006 23:29, Chad Groneman wrote:
On Wed, 2006-04-19 at 18:53 +0100, Jon Black wrote:
The process that is taking up 90% of the CPU is called events/0. There is an events/1 process that has 0% of the CPU...I'm not sure what these processes do but I think they are processor related or perhaps acpi.
I don't know exactly what the events/ processes are, but my guess pretty much the same as yours. My guess is interrupt handling. It could be ACPI interrupts, so you could try the failsafe method. I don't know the ramifications of doing that though - it might be that there's a lot of interrupts because something's not working properly (hardware).
The output of dmesg and lspci might be useful. Maybe there are some evil ACPI or APIC errors who trouble the kernel :)
Philipp
--------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory-unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory-help@opensuse.org
Jon Black escribió:
Sorry for the late reply. I have now installed 10.1 RC3 and the problem has spread so that both events/0 and events/1 are hogging the cpu. The output from top, dmesg and lspci can be found at this address:
Thanks.
did you tried booting with ACPI disabled ?
Firstly I tried to boot by using the failsafe mode. This was too slow because DMA is off (nodma). Removing this, it was still too slow to do anything, and even caused me to force a reboot :s So I booted normally just adding: acpi=off 3 So that it wouldn't start the xserver and looked at top. Sadly, the problem was still there. In fact, this time the events/1 and events/0 processes are averaging around 97% cpu usage, so it is slightly worse. I just wish there was some clue somewhere as to why. Was there anything suspect in my dmesg/lspci output? How might I find what is causing it? Jon. -----Original Message----- From: Cristian Rodriguez [mailto:judas_iscariote@imapmail.org] Sent: 04 May 2006 11:43 To: opensuse-factory@opensuse.org Subject: Re: [opensuse-factory] RC1 events/0 CPU usage Jon Black escribió:
Sorry for the late reply. I have now installed 10.1 RC3 and the problem has spread so that both events/0 and events/1 are hogging the cpu. The output from top, dmesg and lspci can be found at this address:
Thanks.
did you tried booting with ACPI disabled ? --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory-unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory-help@opensuse.org
participants (4)
-
Chad Groneman -
Cristian Rodriguez -
Jon Black -
Philipp Wollermann