C2, C3 and bus master activity
Hello
My CPU does not enter C3 if i run the stock configuration. Way too much bus master activity. I want to find out what is causing it. Any hints how to do that ?
I found one possible cause for some of the activity. Without X i have a 50:50 mix between C2 and C3. Using the vesa-driver the mix stays at 50:50. Using the radeon-driver (dri disabled) the CPU stops using C3.
the facts in case somebody already has a idea:
- ibm R51
- PentiumM 725 (1.6GHz)
- Intel 855PM chipset
- radeon 7500 mobil
- xorg 6.8.2
- kernel 2.6.11-1-686 (debian stock-kernel)
thanks.
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
A message that is not in the archive (sent off-list or not yet received): [email protected]
Hi
In a word USB. I don't fully understand why, but if I disable USB then both my laptops will go into C3 or C4, with USB enabled stays in C2.
Laptops are Compaq EVO 800c and IBM T42p both running SUSE 9.3 with a stock SuSE kernel 2.6.11-suse modified
Bob
Quoted 30 lines · pumpkin
Hello
My CPU does not enter C3 if i run the stock configuration. Way too much bus master activity. I want to find out what is causing it. Any hints how to do that ?
I found one possible cause for some of the activity. Without X i have a 50:50 mix between C2 and C3. Using the vesa-driver the mix stays at 50:50. Using the radeon-driver (dri disabled) the CPU stops using C3.
the facts in case somebody already has a idea:
- ibm R51
- PentiumM 725 (1.6GHz)
- Intel 855PM chipset
- radeon 7500 mobil
- xorg 6.8.2
- kernel 2.6.11-1-686 (debian stock-kernel)thanks.
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
_______________________________________________
Acpi-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/acpi-devel
Signature · 16 lines
V +44 (0) 1296 747667 F +44 (0) 1296 747557 C +44 (0) 7860 406093 Diamond Consulting Services Ltd Dinton Aylesbury Bucks, HP17 8UG England ------------------------------------------------------- SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
A message that is not in the archive (sent off-list or not yet received): 200507192118.37528.bob-l+PWtdWbHAsG2Il/[email protected]
Hello
I tested this by removing kernel/driver/usb/* and rebuilding initrd (just to be sure). It did not change anything for console (still 50:50 between C2 and C3) and made Xorg crash pretty hard.
Do you think removing all the other external buses (Firewire, ethernet, WLAN ,..) would change something ?
Could cpufreqd be part of the problem ?
Thanks,
Quoted 5 lines · Bob Lees
Hi
In a word USB. I don't fully understand why, but if I disable USB then both my laptops will go into C3 or C4, with USB enabled stays in C2.
Quoted 36 lines
Laptops are Compaq EVO 800c and IBM T42p both running SUSE 9.3 with a stock SuSE kernel 2.6.11-suse modified
Bob
Quoted 30 lines · pumpkin
Hello
My CPU does not enter C3 if i run the stock configuration. Way too much bus master activity. I want to find out what is causing it. Any hints how to do that ?
I found one possible cause for some of the activity. Without X i have a 50:50 mix between C2 and C3. Using the vesa-driver the mix stays at 50:50. Using the radeon-driver (dri disabled) the CPU stops using C3.
the facts in case somebody already has a idea:
- ibm R51
- PentiumM 725 (1.6GHz)
- Intel 855PM chipset
- radeon 7500 mobil
- xorg 6.8.2
- kernel 2.6.11-1-686 (debian stock-kernel)thanks.
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
_______________________________________________
Acpi-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/acpi-devel
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
A message that is not in the archive (sent off-list or not yet received): [email protected]
Quoted 6 lines · pumpkin
Hello
I tested this by removing
kernel/driver/usb/*and rebuilding initrd (just to be sure). It did not change anything for console (still 50:50 between C2 and C3) and made Xorg crash pretty hard.Do you think removing all the other external buses (Firewire, ethernet, WLAN ,..) would change something ?
Sometimes an active WLAN driver causes bus master activity... You also should not use an USB Mouse...
And by the way, what's wrong with 50:50 C2 C3 ?
This seems pretty normal to me.
Quoted 16 lines
Could cpufreqd be part of the problem ?
Thanks,
Quoted 10 lines · Bob Lees
Hi
In a word USB. I don't fully understand why, but if I disable USB then both my laptops will go into C3 or C4, with USB enabled stays in C2.
Laptops are Compaq EVO 800c and IBM T42p both running SUSE 9.3 with a stock SuSE kernel 2.6.11-suse modified
Bob
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
A message that is not in the archive (sent off-list or not yet received): [email protected]
Hello
Quoted 1 line
Sometimes an active WLAN driver causes bus master activity...
Inserted USB, removed WLAN and ieee1394. C2/C3-ratio dropps to 10:90 in console-mode. That's better. I don't mind removing ieee1349, but without networking the laptop is more or less useless. So i'm f***ed.
Quoted 1 line
You also should not use an USB Mouse...
No USB in use except the 4 root-hubs (the trackpoint is using PS/2).
Quoted 2 lines
And by the way, what's wrong with 50:50 C2 C3 ?
This seems pretty normal to me.
I have seen 10:90 on AthlonXP-systems running XFree. That's what i want, even if it is a long way to go.
So i only have 2 problems left:
- Is it possible to reconfigure WLAN to cause less Bus Master Activity ? (i will ask the ipw2200 guys and grok the code)
- and why is the radeon driver causing ffffffff-Activity ?
Running x without radeon driver is not an option. The radeon driver enables powerplay, which in turn saves quite a bit.
Thanks,
Quoted 11 lines
Quoted 10 lines · Bob Lees
Quoted 9 lines
Hi
In a word USB. I don't fully understand why, but if I disable USB then both my laptops will go into C3 or C4, with USB enabled stays in C2.
Laptops are Compaq EVO 800c and IBM T42p both running SUSE 9.3 with a stock SuSE kernel 2.6.11-suse modified
Bob
Quoted 9 lines
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
_______________________________________________
Acpi-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/acpi-devel
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
A message that is not in the archive (sent off-list or not yet received): [email protected]
Hello
i have made some experiments and now i'm scared shitless:
Removing and reinserting ieee1394, eth1394 and ohic1394 stops all bus master activity, even under X. USB and WLAN are active. Now the CPU stays at C4, a mode which didn't even show up before.
Now one battery charge should be enough for 4 hours. IBM says 3 and a half hours. That's why i love linux.
Thanks,
Quoted 52 lines · pumpkin
Hello
Quoted 1 line
Sometimes an active WLAN driver causes bus master activity...
Inserted USB, removed WLAN and ieee1394. C2/C3-ratio dropps to 10:90 in console-mode. That's better. I don't mind removing ieee1349, but without networking the laptop is more or less useless. So i'm f***ed.
Quoted 1 line
You also should not use an USB Mouse...
No USB in use except the 4 root-hubs (the trackpoint is using PS/2).
Quoted 2 lines
And by the way, what's wrong with 50:50 C2 C3 ?
This seems pretty normal to me.I have seen 10:90 on AthlonXP-systems running XFree. That's what i want, even if it is a long way to go.
So i only have 2 problems left:
- Is it possible to reconfigure WLAN to cause less Bus Master Activity ? (i will ask the ipw2200 guys and grok the code)
- and why is the radeon driver causing ffffffff-Activity ?
Running x without radeon driver is not an option. The radeon driver enables powerplay, which in turn saves quite a bit.
Thanks,
Quoted 11 lines
Quoted 10 lines · Bob Lees
Quoted 9 lines
Hi
In a word USB. I don't fully understand why, but if I disable USB then both my laptops will go into C3 or C4, with USB enabled stays in C2.
Laptops are Compaq EVO 800c and IBM T42p both running SUSE 9.3 with a stock SuSE kernel 2.6.11-suse modified
Bob
Quoted 9 lines
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
_______________________________________________
Acpi-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/acpi-devel
Quoted 9 lines
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
_______________________________________________
Acpi-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/acpi-devel
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
A message that is not in the archive (sent off-list or not yet received): [email protected]
A message that is not in the archive (sent off-list or not yet received): [email protected]
Hello
Quoted 2 lines
This is interesting, I have a theory why the IBM laptops keep crashing when you use the the deeper sleep states.
Well,... i don't really understand what is going on here.
From time-to-time Xorg tries to load and crashs even before starting the new logfile. But if it gets past that point everything works fine. And even with the stock kernel and every module (including ieee1394...) loaded the CPU enters C4. The BIOS/CMOS/Laptop... whatever has changed it's behavior after my first tests. Maybe it has something to do with running of AC or battery.
Once i have the time to an intensive test with all combinations (w/ and w/o battery, w/ and w/o AC, w/ and w/o 1394 at startup ... ) i will send you my results. Any logs/files that are important to you ?
Quoted 2 lines
Can you try out something for
me ?
Sure, but not today. I will recompile the kernel once again but with debugging enabled.
cu
Quoted 10 lines
First try to load or do something, that would cause your laptop to disable the C4 state. If this hapens, you shuld see in your logs (as long as you compiled your kernel with debug messages) something like \_SB.CPU0,
0x81. This means, that your laptop has changed the available C-States. (this should also appear, if you unload eee1394, eth1394 and ohic1394)I saw some places in the ACPI code, where the C-States are not correctly deleted after this event, and my theory is that this actually crashes some IBM laptops.
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_idt77&alloc_id492&op=click
Quoted 1 line
- and why is the radeon driver causing ffffffff-Activity ?
Disable dri (comment out the 'Load "dri"' in xorg.conf). Then the radeon driver will no longer make pointless bus master transfers. I have no idea why dri causes the bus master activity, but it should not.
Volker
PS: My CPU switches to C3/C4 even with ipw2200 running (2915a/b/g card).
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
Quoted 2 lines · Bob Lees
In a word USB. I don't fully understand why, but if I disable USB then both my laptops will go into C3 or C4, with USB enabled stays in C2.
On my laptop (Acer Travelmate 4001WLMi[1]) it's just the uhci-hcd driver. Unloading it enables C3 again.
-- Johan
[1] http://www.squirrel.nl/people/jvromans/articles/TM4001WLMi/
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
A message that is not in the archive (sent off-list or not yet received): m23bq98fir.fsf-KjnUIgV0B0bak1Ioo/c9IoRWq/[email protected]
I agree with Johan, I think it was the uhci-hcd driver which caused the problem. I must admit that I don't have wireless running when I shut everything off so that may also be a problem.
Unfortunately its the uhci-hcd driver which allows me to use a usb mouse so I would like to know how the usb driver could be used and still have C3/4 (C4 on the IBM)
Bob
Quoted 12 lines · Johan Vromans
Quoted 2 lines · Bob Lees
In a word USB. I don't fully understand why, but if I disable USB then both my laptops will go into C3 or C4, with USB enabled stays in C2.
On my laptop (Acer Travelmate 4001WLMi[1]) it's just the uhci-hcd driver. Unloading it enables C3 again.
-- Johan
[1] http://www.squirrel.nl/people/jvromans/articles/TM4001WLMi/
Quoted 9 lines
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
_______________________________________________
Acpi-devel mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/acpi-devel
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
A message that is not in the archive (sent off-list or not yet received): 200507202031.38488.bob-l+PWtdWbHAsG2Il/[email protected]
Quoted 3 lines · Bob Lees
I agree with Johan, I think it was the uhci-hcd driver which caused the problem. I must admit that I don't have wireless running when I shut everything off so that may also be a problem.
Unfortunately its the uhci-hcd driver which allows me to use a usb mouse so I would like to know how the usb driver could be used and still have C3/4 (C4 on the IBM)
Actually ff you use an USB device, you can never use the C3 or C4 mode of your CPU. USB devices just cause busmaster activity, and as long as the kernel detecs busmaster activity, it stops taking the CPU in anything behind C2.
If you want to know more please read ACPIspec3.0
8.1.4 Processor Power State C3
on page 253 (the number in the document / the pdfreader says it's page 273)
Janosch
Quoted 6 lines
Bob
Quoted 1 line · Johan Vromans
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
Hi!
Quoted 2 lines
In a word USB. I don't fully understand why, but if I disable USB then both my laptops will go into C3 or C4, with USB enabled stays in C2.
Misdesigned USB controllers do DMA all the time... I think it is uhci-specific problem.
Pavel
Signature · 8 lines
teflon -- maybe it is a trademark, but it should not be. ------------------------------------------------------- SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click