Re: [PATCH 6/7] Cpuidle: Deal with timer expiring in the past

2014-03-21 Thread Len Brown
Tuukka, I've reproduced this negative on a 48 thread 2-socket Xeon during boot (seen it only once, so far). expected_us gets calculated to be -1, which is truthful, since the next timer return value was about 500ns in the past and our math truncates. This, in turn, confuses the heck out of menu's

Re: [PATCH 6/7] Cpuidle: Deal with timer expiring in the past

2014-03-10 Thread Tuukka Tikkanen
Hi, On 6 March 2014 09:41, Len Brown wrote: > > On Mon, Feb 24, 2014 at 1:29 AM, Tuukka Tikkanen > wrote: > > Sometimes (fairly often) when the cpuidle menu governor is making a decision > > about idle state to enter the next timer for the cpu appears to expire in > > the past. The menu governor

Re: [PATCH 6/7] Cpuidle: Deal with timer expiring in the past

2014-03-06 Thread Len Brown
NAK tick_nohz_get_sleep_length() has exactly 1 consumer -- cpuidle. Rather than have cpuidle second guess what is returned, tick_nohz_get_sleep_length() should be updated to return a value that makes sense to cpuidle. thanks, Len Brown, Intel Open Source Technology Cent -- To unsubscribe from th

Re: [PATCH 6/7] Cpuidle: Deal with timer expiring in the past

2014-03-05 Thread Len Brown
On Mon, Feb 24, 2014 at 1:29 AM, Tuukka Tikkanen wrote: > Sometimes (fairly often) when the cpuidle menu governor is making a decision > about idle state to enter the next timer for the cpu appears to expire in > the past. The menu governor expects the expiry to always be in the future > and in fa

Re: [PATCH 6/7] Cpuidle: Deal with timer expiring in the past

2014-02-24 Thread Rafael J. Wysocki
On Monday, February 24, 2014 12:05:39 PM Nicolas Pitre wrote: > On Mon, 24 Feb 2014, Tuukka Tikkanen wrote: > > > Sometimes (fairly often) when the cpuidle menu governor is making a decision > > about idle state to enter the next timer for the cpu appears to expire in > > the past. The menu govern

Re: [PATCH 6/7] Cpuidle: Deal with timer expiring in the past

2014-02-24 Thread Nicolas Pitre
On Mon, 24 Feb 2014, Tuukka Tikkanen wrote: > Sometimes (fairly often) when the cpuidle menu governor is making a decision > about idle state to enter the next timer for the cpu appears to expire in > the past. The menu governor expects the expiry to always be in the future > and in fact stores th

[PATCH 6/7] Cpuidle: Deal with timer expiring in the past

2014-02-23 Thread Tuukka Tikkanen
Sometimes (fairly often) when the cpuidle menu governor is making a decision about idle state to enter the next timer for the cpu appears to expire in the past. The menu governor expects the expiry to always be in the future and in fact stores the time delta in an unsigned variable. However, when t