[PATCH] check_signature depends on CONFIG_CHECK_SIGNATURE

[PATCH] check_signature depends on CONFIG_CHECK_SIGNATURE

[Jean Delvare]

check_signature() is only available if CONFIG_CHECK_SIGNATURE has been
selected. Don't declare the function if it won't be available at link
time. That way, any failure to select CONFIG_CHECK_SIGNATURE as needed
will show up immediately at build time.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
---
 include/linux/io.h |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

--- linux-2.6.40-rc0.orig/include/linux/io.h	2011-05-20 10:41:02.000000000 +0200
+++ linux-2.6.40-rc0/include/linux/io.h	2011-05-20 16:00:31.000000000 +0200
@@ -63,8 +63,11 @@ void __iomem *devm_ioremap(struct device
 void __iomem *devm_ioremap_nocache(struct device *dev, resource_size_t offset,
 				    unsigned long size);
 void devm_iounmap(struct device *dev, void __iomem *addr);
+void devm_ioremap_release(struct device *dev, void *res);
+
+#ifdef CONFIG_CHECK_SIGNATURE
 int check_signature(const volatile void __iomem *io_addr,
 			const unsigned char *signature, int length);
-void devm_ioremap_release(struct device *dev, void *res);
+#endif
 
 #endif /* _LINUX_IO_H */


-- 
Jean Delvare

Re: [PATCH] check_signature depends on CONFIG_CHECK_SIGNATURE

[Andrew Morton]

On Fri, 20 May 2011 16:14:48 +0200
Jean Delvare <khali@linux-fr.org> wrote:

> check_signature() is only available if CONFIG_CHECK_SIGNATURE has been
> selected. Don't declare the function if it won't be available at link
> time. That way, any failure to select CONFIG_CHECK_SIGNATURE as needed
> will show up immediately at build time.
> 
> Signed-off-by: Jean Delvare <khali@linux-fr.org>
> ---
>  include/linux/io.h |    5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
> 
> --- linux-2.6.40-rc0.orig/include/linux/io.h	2011-05-20 10:41:02.000000000 +0200
> +++ linux-2.6.40-rc0/include/linux/io.h	2011-05-20 16:00:31.000000000 +0200
> @@ -63,8 +63,11 @@ void __iomem *devm_ioremap(struct device
>  void __iomem *devm_ioremap_nocache(struct device *dev, resource_size_t offset,
>  				    unsigned long size);
>  void devm_iounmap(struct device *dev, void __iomem *addr);
> +void devm_ioremap_release(struct device *dev, void *res);
> +
> +#ifdef CONFIG_CHECK_SIGNATURE
>  int check_signature(const volatile void __iomem *io_addr,
>  			const unsigned char *signature, int length);
> -void devm_ioremap_release(struct device *dev, void *res);
> +#endif

We usually avoid doing this.  Detecting the error at link time is
sufficient.

Yes, that will very occasionally cost a small number of people a little
additional time.  But we must balance that against permanently messing
up the header files.

Re: [PATCH] check_signature depends on CONFIG_CHECK_SIGNATURE

[Jean Delvare]

Hi Andrew,

On Fri, 20 May 2011 15:11:13 -0700, Andrew Morton wrote:
> On Fri, 20 May 2011 16:14:48 +0200
> Jean Delvare <khali@linux-fr.org> wrote:
> 
> > check_signature() is only available if CONFIG_CHECK_SIGNATURE has been
> > selected. Don't declare the function if it won't be available at link
> > time. That way, any failure to select CONFIG_CHECK_SIGNATURE as needed
> > will show up immediately at build time.
> > 
> > Signed-off-by: Jean Delvare <khali@linux-fr.org>
> > ---
> >  include/linux/io.h |    5 ++++-
> >  1 file changed, 4 insertions(+), 1 deletion(-)
> > 
> > --- linux-2.6.40-rc0.orig/include/linux/io.h	2011-05-20 10:41:02.000000000 +0200
> > +++ linux-2.6.40-rc0/include/linux/io.h	2011-05-20 16:00:31.000000000 +0200
> > @@ -63,8 +63,11 @@ void __iomem *devm_ioremap(struct device
> >  void __iomem *devm_ioremap_nocache(struct device *dev, resource_size_t offset,
> >  				    unsigned long size);
> >  void devm_iounmap(struct device *dev, void __iomem *addr);
> > +void devm_ioremap_release(struct device *dev, void *res);
> > +
> > +#ifdef CONFIG_CHECK_SIGNATURE
> >  int check_signature(const volatile void __iomem *io_addr,
> >  			const unsigned char *signature, int length);
> > -void devm_ioremap_release(struct device *dev, void *res);
> > +#endif
> 
> We usually avoid doing this.  Detecting the error at link time is
> sufficient.
> 
> Yes, that will very occasionally cost a small number of people a little
> additional time.  But we must balance that against permanently messing
> up the header files.

OK, fair enough. Patch deleted.

-- 
Jean Delvare