lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date:	Wed, 3 Feb 2016 12:46:34 -0600
From:	Bjorn Helgaas <helgaas@...nel.org>
To:	Felipe Balbi <balbi@...nel.org>
Cc:	Bjorn Helgaas <bhelgaas@...gle.com>, Felipe Balbi <balbi@...com>,
	Alexandre Courbot <gnurou@...il.com>,
	Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
	Linus Walleij <linus.walleij@...aro.org>,
	linux-usb@...r.kernel.org, Liam Girdwood <lgirdwood@...il.com>,
	linux-kernel@...r.kernel.org, linux-gpio@...r.kernel.org
Subject: Re: [PATCH 1/3] usb: phy: generic: use forward declarations instead
 of #includes

[-cc Mark]

On Wed, Feb 03, 2016 at 08:18:02PM +0200, Felipe Balbi wrote:
> Bjorn Helgaas <bhelgaas@...gle.com> writes:
> > In include/linux/usb/usb_phy_generic.h, use a forward declaration for
> > struct gpio_desc instead of including linux/gpio/consumer.h.
> >
> > Of the files that include usb_phy_generic.h, only
> > drivers/usb/phy/phy-generic.c uses the gpiod_*() interfaces from
> > linux/gpio/consumer.h, so include consumer.h explicitly there.
> >
> > This is a little more efficient and ensures that users of the gpiod
> > interfaces include linux/gpio/consumer.h directly rather than getting it
> > accidentally via linux/usb/usb_phy_generic.h.
> >
> > Signed-off-by: Bjorn Helgaas <bhelgaas@...gle.com>
> > ---
> >  drivers/usb/phy/phy-generic.c       |    1 +
> >  include/linux/usb/usb_phy_generic.h |    3 ++-
> >  2 files changed, 3 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/usb/phy/phy-generic.c b/drivers/usb/phy/phy-generic.c
> > index 5320cb8..2c1349e 100644
> > --- a/drivers/usb/phy/phy-generic.c
> > +++ b/drivers/usb/phy/phy-generic.c
> > @@ -36,6 +36,7 @@
> >  #include <linux/of.h>
> >  #include <linux/of_gpio.h>
> >  #include <linux/gpio.h>
> > +#include <linux/gpio/consumer.h>
> >  #include <linux/delay.h>
> >  
> >  #include "phy-generic.h"
> > diff --git a/include/linux/usb/usb_phy_generic.h b/include/linux/usb/usb_phy_generic.h
> > index c13632d5..45c20060 100644
> > --- a/include/linux/usb/usb_phy_generic.h
> > +++ b/include/linux/usb/usb_phy_generic.h
> > @@ -2,7 +2,8 @@
> >  #define __LINUX_USB_NOP_XCEIV_H
> >  
> >  #include <linux/usb/otg.h>
> > -#include <linux/gpio/consumer.h>
> > +
> > +struct gpio_desc;
> 
> then someone decides to rename struct gpio_desc and this just becomes
> yet another place to fix. Why is this any better than just including the
> header ?

You're right.  The advantage is that the includers of
usb_phy_generic.h that don't use gpiod_*() functions no longer need to
include linux/gpio/consumer.h, so they build faster.  It's a minor
advantage in this case because not very many files include
usb_phy_generic.h.

Bjorn

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ