CoreDuo Bus-Master-Activity on Core1 ffffffff when idle, no C3
Hi List!
i recently got a Lenovo X60T, installed Gentoo Linux, and since then i am trying to figure out why i miss 2h of battery-life compared to Windows-Users. After a while i noticed that the Bus-Master-Activity on CPU1 is ffffffff when its idle and only drops if i give it some work to do (like a kernel compile with make -j3)
When i put some load on both cores the Bus Master Activity of CPU1 is normalizing itself and varies according to the load.
When i leave the Laptop idle for a while it gets actually hotter (60C) then when i use it (44C).
The CPU is not able to stay in C3, it tries but comes back to C2 within the second, i assume that is because of the high Bus Master Activity of CPU1?
I have already looked for help about this on various forums and mailing lists and i talked to other people with CoreDuo's and they do not have this issue with the Bus Master Activity.
I tried alot of different kernels (2.6.19 and 2.6.20) and setups or used preconfigured Kernels or even Live-Distros.
I also booted a *very* minimal Gentoo-Linux setup, no WLAN, no USB, no X etc., and played with the BIOS-Settings (disabling as much as possible), and now i am at a point where i really dont know what to do anymore and i fear that im not getting any answers anywhere.
If anyone has any advice, information or would like to see some logs or configs, i would be very gratefull.
Greetings from Germany,
Sascha
Len Brown
reviewerQuoted 27 lines · Sascha Heid
Hi List!
i recently got a Lenovo X60T, installed Gentoo Linux, and since then i am trying to figure out why i miss 2h of battery-life compared to Windows-Users. After a while i noticed that the Bus-Master-Activity on CPU1 is ffffffff when its idle and only drops if i give it some work to do (like a kernel compile with make -j3)
When i put some load on both cores the Bus Master Activity of CPU1 is normalizing itself and varies according to the load.
When i leave the Laptop idle for a while it gets actually hotter (60C) then when i use it (44C).
The CPU is not able to stay in C3, it tries but comes back to C2 within the second, i assume that is because of the high Bus Master Activity of CPU1?I have already looked for help about this on various forums and mailing lists and i talked to other people with CoreDuo's and they do not have this issue with the Bus Master Activity.
I tried alot of different kernels (2.6.19 and 2.6.20) and setups or used preconfigured Kernels or even Live-Distros.
I also booted a *very* minimal Gentoo-Linux setup, no WLAN, no USB, no X etc., and played with the BIOS-Settings (disabling as much as possible), and now i am at a point where i really dont know what to do anymore and i fear that im not getting any answers anywhere.
Quoted 5 lines
If anyone has any advice, information or would like to see some logs or configs, i would be very gratefull.
Greetings from Germany,
Sascha
Try disconnecting usb devices until lsusb shows nothing.
cheers from Boston,
-Len
Quoted 1 line · Len Brown
Try disconnecting usb devices until lsusb shows nothing.
On some ThinkPads, this means removing uhci_hcd and ehci_hcd, because they have internal USB devices (bluetooth, fingerprint reader, etc).
Signature · 4 lines
"One disk to rule them all, One disk to find them. One disk to bring them all and in the darkness grind them. In the Land of Redmond where the shadows lie." -- The Silicon Valley Tarot Henrique Holschuh
Am Freitag, den 09.02.2007, 12:29 -0200 schrieb Henrique de Moraes Holschuh:
Quoted 6 lines · Len Brown
Quoted 1 line
Try disconnecting usb devices until lsusb shows nothing.
On some ThinkPads, this means removing
uhci_hcdandehci_hcd, because they have internal USB devices (bluetooth, fingerprint reader, etc).
Thats it, added 1h Battery-Life immediately.
Thank you very much!
Quoted 9 lines · Sascha Heid
Am Freitag, den 09.02.2007, 12:29 -0200 schrieb Henrique de Moraes Holschuh:
Quoted 5 lines · Len Brown
Quoted 1 line
Try disconnecting usb devices until lsusb shows nothing.
On some ThinkPads, this means removing
uhci_hcdandehci_hcd, because they have internal USB devices (bluetooth, fingerprint reader, etc).Thats it, added 1h Battery-Life immediately.
Now, for the golden question: Should uhci_hcd be causing that much bus-mastering activity? And why it causes that *only* when the CPU is idle? That is extremely weird.
I think there indeed is a real bug at work here.
Signature · 4 lines
"One disk to rule them all, One disk to find them. One disk to bring them all and in the darkness grind them. In the Land of Redmond where the shadows lie." -- The Silicon Valley Tarot Henrique Holschuh