Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1576692 > unrolled thread

Re: [PATCH] serial: st-asc: Use new GPIOD API to obtain RTS pin

Started byAndy Shevchenko <andriy.shevchenko@linux.intel.com>
First post2017-02-08 17:40 +0100
Last post2017-02-08 22:30 +0100
Articles 4 — 3 participants

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.


Contents

  Re: [PATCH] serial: st-asc: Use new GPIOD API to obtain RTS pin Andy Shevchenko <andriy.shevchenko@linux.intel.com> - 2017-02-08 17:40 +0100
    Re: [PATCH] serial: st-asc: Use new GPIOD API to obtain RTS pin Greg KH <greg@kroah.com> - 2017-02-08 18:50 +0100
      Re: [PATCH] serial: st-asc: Use new GPIOD API to obtain RTS pin Andy Shevchenko <andriy.shevchenko@linux.intel.com> - 2017-02-08 20:50 +0100
        Re: [PATCH] serial: st-asc: Use new GPIOD API to obtain RTS pin Stephen Rothwell <sfr@canb.auug.org.au> - 2017-02-08 22:30 +0100

#1576692 — Re: [PATCH] serial: st-asc: Use new GPIOD API to obtain RTS pin

FromAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Date2017-02-08 17:40 +0100
SubjectRe: [PATCH] serial: st-asc: Use new GPIOD API to obtain RTS pin
Message-ID<t8DrI-2DS-11@gated-at.bofh.it>
On Wed, 2017-02-08 at 21:48 +0800, kbuild test robot wrote:
> Hi Lee,
> 
> [auto build test ERROR on tty/tty-testing]
> [cannot apply to v4.10-rc7 next-20170208]
> [if your patch is applied to the wrong git tree, please drop us a note
> to help improve the system]
> 
> url:    https://github.com/0day-ci/linux/commits/Lee-Jones/serial-st-a
> sc-Use-new-GPIOD-API-to-obtain-RTS-pin/20170208-180609
> base:   https://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty.git
>  tty-testing
> config: x86_64-allmodconfig (attached as .config)
> compiler: gcc-6 (Debian 6.2.0-3) 6.2.0 20160901
> reproduce:
>         # save the attached .config to linux build tree
>         make ARCH=x86_64 
> 

It requires to have immutable branch in one of the subsystem which the
other one can pull.

> All errors (new ones prefixed by >>):
> 
>    drivers/tty/serial/st-asc.c: In function 'asc_set_termios':
> > > drivers/tty/serial/st-asc.c:578:12: error: implicit declaration of
> > > function 'devm_fwnode_get_gpiod_from_child' [-Werror=implicit-
> > > function-declaration]
> 
>        gpiod = devm_fwnode_get_gpiod_from_child(port->dev,
>                ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>    drivers/tty/serial/st-asc.c:578:10: warning: assignment makes
> pointer from integer without a cast [-Wint-conversion]
>        gpiod = devm_fwnode_get_gpiod_from_child(port->dev,
>              ^
>    cc1: some warnings being treated as errors
> 
> vim +/devm_fwnode_get_gpiod_from_child +578 drivers/tty/serial/st-
> asc.c
> 
>    572		} else {
>    573			/* If flow-control disabled, it's safe
> to handle RTS manually */
>    574			if (!ascport->rts && ascport-
> >states[NO_HW_FLOWCTRL]) {
>    575				pinctrl_select_state(ascport-
> >pinctrl,
>    576						     ascport-
> >states[NO_HW_FLOWCTRL]);
>    577	
>  > 578				gpiod =
> devm_fwnode_get_gpiod_from_child(port->dev,
>    579									
>  "rts",
>    580									
>  &np->fwnode,
>    581									
>  GPIOD_OUT_LOW,
> 
> ---
> 0-DAY kernel test infrastructure                Open Source Technology
> Center
> https://lists.01.org/pipermail/kbuild-all                   Intel
> Corporation

-- 
Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Intel Finland Oy

[toc] | [next] | [standalone]


#1576778

FromGreg KH <greg@kroah.com>
Date2017-02-08 18:50 +0100
Message-ID<t8Exr-3hZ-3@gated-at.bofh.it>
In reply to#1576692
On Wed, Feb 08, 2017 at 06:31:10PM +0200, Andy Shevchenko wrote:
> On Wed, 2017-02-08 at 21:48 +0800, kbuild test robot wrote:
> > Hi Lee,
> > 
> > [auto build test ERROR on tty/tty-testing]
> > [cannot apply to v4.10-rc7 next-20170208]
> > [if your patch is applied to the wrong git tree, please drop us a note
> > to help improve the system]
> > 
> > url:    https://github.com/0day-ci/linux/commits/Lee-Jones/serial-st-a
> > sc-Use-new-GPIOD-API-to-obtain-RTS-pin/20170208-180609
> > base:   https://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty.git
> >  tty-testing
> > config: x86_64-allmodconfig (attached as .config)
> > compiler: gcc-6 (Debian 6.2.0-3) 6.2.0 20160901
> > reproduce:
> >         # save the attached .config to linux build tree
> >         make ARCH=x86_64 
> > 
> 
> It requires to have immutable branch in one of the subsystem which the
> other one can pull.

Which sucks, and is why you should not do api changes this way!

greg k-h

[toc] | [prev] | [next] | [standalone]


#1576986

FromAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Date2017-02-08 20:50 +0100
Message-ID<t8GpA-4sD-13@gated-at.bofh.it>
In reply to#1576778
On Wed, 2017-02-08 at 18:47 +0100, Greg KH wrote:
> On Wed, Feb 08, 2017 at 06:31:10PM +0200, Andy Shevchenko wrote:
> > On Wed, 2017-02-08 at 21:48 +0800, kbuild test robot wrote:
> > > Hi Lee,
> > > 
> > > [auto build test ERROR on tty/tty-testing]
> > > [cannot apply to v4.10-rc7 next-20170208]
> > > [if your patch is applied to the wrong git tree, please drop us a
> > > note
> > > to help improve the system]
> > > 
> > > url:    https://github.com/0day-ci/linux/commits/Lee-Jones/serial-
> > > st-a
> > > sc-Use-new-GPIOD-API-to-obtain-RTS-pin/20170208-180609
> > > base:   https://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty
> > > .git
> > >  tty-testing
> > > config: x86_64-allmodconfig (attached as .config)
> > > compiler: gcc-6 (Debian 6.2.0-3) 6.2.0 20160901
> > > reproduce:
> > >         # save the attached .config to linux build tree
> > >         make ARCH=x86_64 
> > > 
> > 
> > It requires to have immutable branch in one of the subsystem which
> > the
> > other one can pull.
> 
> Which sucks, and is why you should not do api changes this way!

Not only me :-)

If above will not work we may do something like below for this cycle:

static inline ... devm_get_gpiod_from_child()
{
 return devm_fwnode_get_gpiod_from_child(..., GPIO_AS_IS, "?");
}

in GPIO tree.

-- 
Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Intel Finland Oy

[toc] | [prev] | [next] | [standalone]


#1577081

FromStephen Rothwell <sfr@canb.auug.org.au>
Date2017-02-08 22:30 +0100
Message-ID<t8HYl-5wK-3@gated-at.bofh.it>
In reply to#1576986
Hi all,

On Wed, 08 Feb 2017 21:42:47 +0200 Andy Shevchenko <andriy.shevchenko@linux.intel.com> wrote:
>
> On Wed, 2017-02-08 at 18:47 +0100, Greg KH wrote:
> > On Wed, Feb 08, 2017 at 06:31:10PM +0200, Andy Shevchenko wrote:  
> > > 
> > > It requires to have immutable branch in one of the subsystem which
> > > the
> > > other one can pull.  
> > 
> > Which sucks, and is why you should not do api changes this way!  
> 
> Not only me :-)
> 
> If above will not work we may do something like below for this cycle:
> 
> static inline ... devm_get_gpiod_from_child()
> {
>  return devm_fwnode_get_gpiod_from_child(..., GPIO_AS_IS, "?");
> }
> 
> in GPIO tree.

I will use Lee's patch as a merge resolution when I merge the gpio
tree (as that is later in my list) from now on.  All that has to happen
now is that whichever tree is merged last by Linus (Torvalds) has to
have this same merge resolution applied.

In general, it is better if API changes can be done either as Greg
suggested or with a separate immutable topic branch merged into
whichever trees need it, but it doesn't happen that way very often, so
this is what we generally do.

-- 
Cheers,
Stephen Rothwell

[toc] | [prev] | [standalone]


Back to top | Article view | linux.kernel


csiph-web