On 11.02.2010 19:00, Dimitrios Apostolou wrote: > On Wed, 10 Feb 2010, Arjan van de Ven wrote: >> On Wed, 10 Feb 2010 22:51:38 +0200 (EET) >> Dimitrios Apostolou wrote: >>> >>> As I understand, in his case the C3 state is unstable and exits >>> immediately. I have asked him to post the dmidecode output so you can >>> put him on the exception list too. However I now believe that more >>> and more users will be facing the same problem, it's not something >>> you find easily, especially on desktop machines! What do you think? >> >> if C3 does not work, this needs to be fixed in the code that implements >> C3, not in the code that selects C3. >> >> >> Modern systems should have working C3; if one does not it needs to be >> investigated as to why it's not working. One cause could be a PME that >> we're not handling (I've seen that a few times in our lab), lspci -vvv >> will show that. >> >> But regardless, it's not the task of the code that selects a C state to >> deal with.... > > Wojo (CC'd) can you run as root lspci -vvv and attach the output, so the > experts can have a look? > > Arjan, in this case a bisection was not performed but the symptoms are > exactly the same as mine: > * powertop showing thousands of interrups but showing no specific > process causing them > * The situation is caused only when the "processor" module is inserted > and after a message about "marking TSC as unstable due to halts in > idle", exactly like my case > > Hmmm actually a difference is that in my case the system used the > acpi_pm clocksource, but in Wojo's case it used hpet. > > If I understand correctly what you said, this is a bug in another piece > of code, and I assume that the previous behaviour of the governor was > hiding it, avoiding C3 state completely, right? Sure Dimitris. You can find lspci -vvv output from that unfortunate laptop in attachments. -- Sincerely Yours, Wojciech 'Wojo' Płoskonka