Re: [PATCH v5 0/4] Fix MTRR bug for intel-lpss-pci

[Andy Shevchenko]

On Thu, Oct 17, 2019 at 09:22:01AM +0100, Lee Jones wrote:
> On Thu, 17 Oct 2019, Andy Shevchenko wrote:
> > On Thu, Oct 17, 2019 at 08:31:16AM +0100, Lee Jones wrote:
> > > On Thu, 17 Oct 2019, Andy Shevchenko wrote:
> > > > On Wed, Oct 16, 2019 at 03:06:25PM -0600, Tuowen Zhao wrote:
> > > > > Some BIOS erroneously specifies write-combining BAR for intel-lpss-pci
> > > > > in MTRR. This will cause the system to hang during boot. If possible,
> > > > > this bug could be corrected with a firmware update.
> > > > > 
> > > > > Previous version: https://lkml.org/lkml/2019/10/14/575
> > > > > 
> > > > > Changes from previous version:
> > > > > 
> > > > >  * implement ioremap_uc for sparc64
> > > > >  * split docs changes to not CC stable (doc location moved since 5.3)
> > > > > 
> > > > 
> > > > It forgot to explicitly mention through which tree is supposed to go.
> > > > I think it's MFD one, correct?
> > > 
> > > To be fair, that's not really up to the submitter to decide.
> > 
> > Submitter still can share their view, no?
> 
> Preferences can be voiced, if held, and will always be taken into
> consideration.  The final decision will always be made by the people
> managing the trees.
> 
> The comment above implies a requirement to specify which tree is
> preferred, which is not the case.  In almost all cases, it's best not
> to specify.

In my practice I had been asked several times to specify and express my
understanding which subsystem should take series. It's not so unequivocally.

-- 
With Best Regards,
Andy Shevchenko