Em Fri, 13 Sep 2019 08:56:30 -0400
Matthew Wilcox <[email protected]> escreveu:
> It's easy enough to move the kernel-doc warnings out from under W=1. I only
> out them there to avoid overwhelming us with new warnings. If they're
> mostly fixed now, let's make checking them the default.
Didn't try doing it kernelwide, but for media we do use W=1 by default,
on our CI instance.
There's a few warnings at EDAC, but they all seem easy enough to be
fixed.
So, from my side, I'm all to make W=1 default.
Regards,
Mauro
>
> On Thu., Sep. 12, 2019, 16:01 Bart Van Assche, <[email protected]> wrote:
>
> > On 9/12/19 8:34 AM, Joe Perches wrote:
> > > On Thu, 2019-09-12 at 14:31 +0100, Bart Van Assche wrote:
> > >> On 9/11/19 5:40 PM, Martin K. Petersen wrote:
> > >>> * The patch must compile without warnings (make C=1
> > CF="-D__CHECK_ENDIAN__")
> > >>> and does not incur any zeroday test robot complaints.
> > >>
> > >> How about adding W=1 to that make command?
> > >
> > > That's rather too compiler version dependent and new
> > > warnings frequently get introduced by new compiler versions.
> >
> > I've never observed this myself. If a new compiler warning is added to
> > gcc and if it produces warnings that are not useful for kernel code
> > usually Linus or someone else is quick to suppress that warning.
> >
> > Another argument in favor of W=1 is that the formatting of kernel-doc
> > headers is checked only if W=1 is passed to make.
> >
> > Bart.
> >
> > _______________________________________________
> > Ksummit-discuss mailing list
> > [email protected]
> > https://lists.linuxfoundation.org/mailman/listinfo/ksummit-discuss
> >
Thanks,
Mauro
On Fri, Sep 13, 2019 at 10:54:46AM -0300, Mauro Carvalho Chehab wrote:
> Em Fri, 13 Sep 2019 08:56:30 -0400
> Matthew Wilcox <[email protected]> escreveu:
>
> > It's easy enough to move the kernel-doc warnings out from under W=1. I only
> > out them there to avoid overwhelming us with new warnings. If they're
> > mostly fixed now, let's make checking them the default.
>
> Didn't try doing it kernelwide, but for media we do use W=1 by default,
> on our CI instance.
>
I used to do that as well, but gave up on it since it resulted in lots
of warnings from generic kernel include files. I have not tried recently,
so maybe that is no longer the case.
> There's a few warnings at EDAC, but they all seem easy enough to be
> fixed.
>
Acceptance depends on the maintainer, really. I had patches rejected
when trying to fix W=1 warnings, so I no longer do it.
> So, from my side, I'm all to make W=1 default.
>
Seems to me that would require a common agreement that maintainers
are expected to accept fixes for problems reported with W=1.
Guenter
> Regards,
> Mauro
>
> >
> > On Thu., Sep. 12, 2019, 16:01 Bart Van Assche, <[email protected]> wrote:
> >
> > > On 9/12/19 8:34 AM, Joe Perches wrote:
> > > > On Thu, 2019-09-12 at 14:31 +0100, Bart Van Assche wrote:
> > > >> On 9/11/19 5:40 PM, Martin K. Petersen wrote:
> > > >>> * The patch must compile without warnings (make C=1
> > > CF="-D__CHECK_ENDIAN__")
> > > >>> and does not incur any zeroday test robot complaints.
> > > >>
> > > >> How about adding W=1 to that make command?
> > > >
> > > > That's rather too compiler version dependent and new
> > > > warnings frequently get introduced by new compiler versions.
> > >
> > > I've never observed this myself. If a new compiler warning is added to
> > > gcc and if it produces warnings that are not useful for kernel code
> > > usually Linus or someone else is quick to suppress that warning.
> > >
> > > Another argument in favor of W=1 is that the formatting of kernel-doc
> > > headers is checked only if W=1 is passed to make.
> > >
> > > Bart.
> > >
> > > _______________________________________________
> > > Ksummit-discuss mailing list
> > > [email protected]
> > > https://lists.linuxfoundation.org/mailman/listinfo/ksummit-discuss
> > >
>
>
>
> Thanks,
> Mauro
> _______________________________________________
> Ksummit-discuss mailing list
> [email protected]
> https://lists.linuxfoundation.org/mailman/listinfo/ksummit-discuss