Re: [PATCH 08/11] iwlwifi: call apm stop on exit

[Marcel Holtmann <holtmann@linux.intel.com>]

Hi Tomas,

> >> > > This patch calls apm stop on exit and suspend and sets STATUS_EXIT_PENDING
> >> > > accordingly. Without this patch hardware consumes power even after driver
> >> > > is removed or suspended.
> >> >
> >> > is this a regression to 2.6.26? That is always the important question
> >> > here since that will determine if the patch should go in. If the 2.6.26
> >> > version did this better (even if not prefect), then this patch should go
> >> > on, otherwise it could wait until next merge window.
> >>
> >> AFAIR people have been complaining about this behaviour _forever_, so
> >> it's hardly a regression.
> >
> > I think so too. However the question is still if 2.6.26 maybe did better
> > and we are now doing worse because of other patches. This needs to be in
> > the commit message. Otherwise this falls for me in the next merge window
> > category.
> >
> Regression or not, this is an important bug fix.  With millions of
> laptops sold this is significant contribution to global worming.

I fully agree with you here (no questions asked), but that is besides
the point.

The whole idea is to speed up the release cycle and every fix that is
not a regression fix might have impacts. Also if it is not a regression,
I can wait until the next merge window. I think that Linus made it
pretty clear.

So while I personally would vote for including this patch, it doesn't
really fit the criteria for being submitted after -rc4. In the end this
is up to Dave if he feels it is worth to explain this to Linus. There is
a gray area.

If you wanna play the "being green" card, then include numbers on how
much power this would save :)

Regards

Marcel