Re: [PATCH] cpufreq: Stop BUGing the system
From: Rafael J. Wysocki
Date: Sat Dec 27 2014 - 15:15:26 EST
On Wednesday, December 24, 2014 09:06:05 AM Nishanth Menon wrote:
> On 12/18/2014 08:08 PM, Rafael J. Wysocki wrote:
> > On Friday, December 19, 2014 07:11:19 AM Viresh Kumar wrote:
> >> On 18 December 2014 at 20:19, Nishanth Menon <nm@xxxxxx> wrote:
> >>> I can add "could be unstable" -> the point being there can be psuedo
> >>> errors reported in the system - example - clock framework bugs. Dont
> >>> just stop the boot. example: what if cpufreq was a driver module - it
> >>> would not have rescued the system because cpufreq had'nt detected the
> >>> logic - if we are going to force this on the system, we should probably
> >>> not do this in cpufreq code, instead should be somewhere generic.
> >>> While I do empathise (and had infact advocated in the past) of not
> >>> favouring system attempting to continue at an invalid configuration and
> >>> our attempt to rescue has failed - given that we cannot provide a
> >>> consistent behavior (it is not a core system behavior) and potential of
> >>> a false-postive (example clk framework or underlying bug), it should be
> >>> good enough to "enhance" WARN to be "severe sounding enough" to
> >>> flag it for developer and continue while keeping the system alive as
> >>> much as possible.
> >> There is no way out for the kernel to know if its a false positive or a real
> >> bug. And in the worst case, it can screw up a platform completely.
> >> I am still not sure if changing it to a WARN would be good idea.
> >> @Rafael: Thoughts ?
> > I'm a bit divided here. On the one hand I don't like BUG_ON() as a rule and it
> > is used in too many places where it doesn't have to be used.
> > On the other hand, in this particular case, I'm not sure if allowing the system
> > to run without cpufreq when it might rely on it for CPU cooling, for one example,
> > is a good idea.
> but then, CPUFReq is not a mandatory feature - we could as well do the
> same with CPU_FREQ disabled.
Some platforms pretty much require CPU_FREQ and will always have it set, but
with the $subject patch they may end up not using it.
So this isn't a valid argument to me, sorry.
I speak only for myself.
Rafael J. Wysocki, Intel Open Source Technology Center.
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/