C2 state used much more than C3 when computer seems to be idling

Olaus · 11-02-2009, 02:18 PM

Quote:

Originally Posted by nimnull22

I have upload it to http://www.filefactory.com/file/a081...file372053_zip

To open it type where necessary "$pcitw$", without " ".
Copy it to /usr/bin and try to execute: under root type "pcitweak -l", if it works it shows all PCI devices.
Give me some time to find out how to look what we need.

Code:

./pcitweak -l
PCI: Probing config type using method 1
PCI: Config type is 1
PCI: PCI scan (all values are in hex)
PCI: 00:00:0: chip 8086,2a40 card 1025,0253 rev 07 class 06,00,00 hdr 00
PCI: 00:01:0: chip 8086,2a41 card 0000,0000 rev 07 class 06,04,00 hdr 01
PCI: 00:1a:0: chip 8086,2937 card 1025,0253 rev 03 class 0c,03,00 hdr 80
PCI: 00:1a:1: chip 8086,2938 card 1025,0253 rev 03 class 0c,03,00 hdr 00
PCI: 00:1a:7: chip 8086,293c card 1025,0253 rev 03 class 0c,03,20 hdr 00
PCI: 00:1b:0: chip 8086,293e card 1025,0253 rev 03 class 04,03,00 hdr 00
PCI: 00:1c:0: chip 8086,2940 card 0000,0000 rev 03 class 06,04,00 hdr 81
PCI: 00:1c:3: chip 8086,2946 card 0000,0000 rev 03 class 06,04,00 hdr 81
PCI: 00:1c:5: chip 8086,294a card 0000,0000 rev 03 class 06,04,00 hdr 81
PCI: 00:1d:0: chip 8086,2934 card 1025,0253 rev 03 class 0c,03,00 hdr 80
PCI: 00:1d:1: chip 8086,2935 card 1025,0253 rev 03 class 0c,03,00 hdr 00
PCI: 00:1d:2: chip 8086,2936 card 1025,0253 rev 03 class 0c,03,00 hdr 00
PCI: 00:1d:3: chip 8086,2939 card 1025,0253 rev 03 class 0c,03,00 hdr 00
PCI: 00:1d:7: chip 8086,293a card 1025,0253 rev 03 class 0c,03,20 hdr 00
PCI: 00:1e:0: chip 8086,2448 card 0000,0000 rev 93 class 06,04,01 hdr 01
PCI: 00:1f:0: chip 8086,2919 card 1025,0253 rev 03 class 06,01,00 hdr 80
PCI: 00:1f:2: chip 8086,2929 card 1025,0253 rev 03 class 01,06,01 hdr 00
PCI: 00:1f:3: chip 8086,2930 card 1025,0253 rev 03 class 0c,05,00 hdr 00
PCI: 01:00:0: chip 10de,06ec card 1025,0253 rev a1 class 03,00,00 hdr 00
PCI: 07:00:0: chip 168c,002a card 105b,e006 rev 01 class 02,80,00 hdr 00
PCI: 09:00:0: chip 1969,1063 card 1025,0253 rev c0 class 02,00,00 hdr 00
PCI: 0f:16:0: chip ffff,ffff card ffff,ffff rev ff class ff,ff,ff hdr 00
PCI: d6:17:0: chip ffff,ffff card ffff,ffff rev ff class ff,ff,ff hdr 00
PCI: End of PCI scan

nimnull22 · 11-02-2009, 02:58 PM

Can you post output of the: "pcitweak -r 00:1f:0 -b 0xa9" - We need to know what is there before change.

Olaus · 11-03-2009, 02:07 AM

Quote:

Originally Posted by nimnull22

Can you post output of the: "pcitweak -r 00:1f:0 -b 0xa9" - We need to know what is there before change.

Code:

pcitweak -r 00:1f:0 -b 0xa9
0x1c

nimnull22 · 11-03-2009, 09:45 AM

"1c" - it means that configuration is OK, mine was "10".
Other thing you can check as well is /sys/module/usbhid/parameters/mousepoll, I have "0" in this file.

nimnull22 · 11-03-2009, 09:49 AM

And as well /sys/module/usbcore/parameters/autosuspend

Olaus · 11-03-2009, 10:30 AM

Quote:

Originally Posted by nimnull22

And as well /sys/module/usbcore/parameters/autosuspend

I guess the problem might be my external USB hub, that I have connected the mouse and keyboard to. If I connect them directly to the computer I am fine (at least this applies to the mouse) regarding the computer sleeping, even if they are not suspended. If I suspend both of them, then the hub also go to sleep and I get a lot of time in C3-state. The only problem is that I sometimes lose some of the keyboard input when that wakes up.
Thank you for taking your time to help.