linux-next: driver-core tree build warning

linux-next: driver-core tree build warning

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 465 bytes --]

Hi Greg,

Today's linux-next build (x86_64 allmodconfig) produced this warning:

drivers/firewire/core-device.c: In function 'init_fw_attribute_group':
drivers/firewire/core-device.c:315: warning: assignment from incompatible pointer type

Introduced by commit 025a6a26b70e5dc12ab9333ea43d70c735f459ac ("driver
model: constify attribute groups").

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]

Re: linux-next: driver-core tree build warning

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 659 bytes --]

Hi Greg,

On Mon, 13 Jul 2009 17:26:46 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> Today's linux-next build (x86_64 allmodconfig) produced this warning:
> 
> drivers/firewire/core-device.c: In function 'init_fw_attribute_group':
> drivers/firewire/core-device.c:315: warning: assignment from incompatible pointer type
> 
> Introduced by commit 025a6a26b70e5dc12ab9333ea43d70c735f459ac ("driver
> model: constify attribute groups").

Also:

drivers/mtd/mtdcore.c:227: warning: initialization from incompatible pointer type

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]

Re: linux-next: driver-core tree build warning

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 898 bytes --]

Hi Greg,

On Mon, 13 Jul 2009 17:28:31 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> On Mon, 13 Jul 2009 17:26:46 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> >
> > Today's linux-next build (x86_64 allmodconfig) produced this warning:
> > 
> > drivers/firewire/core-device.c: In function 'init_fw_attribute_group':
> > drivers/firewire/core-device.c:315: warning: assignment from incompatible pointer type
> > 
> > Introduced by commit 025a6a26b70e5dc12ab9333ea43d70c735f459ac ("driver
> > model: constify attribute groups").
> 
> Also:
> 
> drivers/mtd/mtdcore.c:227: warning: initialization from incompatible pointer type

And from a PowerPC allyesconfig build:

drivers/block/cciss.c:582: warning: initialization from incompatible pointer type

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]

Re: linux-next: driver-core tree build warning

[Greg KH]

On Mon, Jul 13, 2009 at 05:51:06PM +1000, Stephen Rothwell wrote:
> Hi Greg,
> 
> On Mon, 13 Jul 2009 17:28:31 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> >
> > On Mon, 13 Jul 2009 17:26:46 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> > >
> > > Today's linux-next build (x86_64 allmodconfig) produced this warning:
> > > 
> > > drivers/firewire/core-device.c: In function 'init_fw_attribute_group':
> > > drivers/firewire/core-device.c:315: warning: assignment from incompatible pointer type
> > > 
> > > Introduced by commit 025a6a26b70e5dc12ab9333ea43d70c735f459ac ("driver
> > > model: constify attribute groups").
> > 
> > Also:
> > 
> > drivers/mtd/mtdcore.c:227: warning: initialization from incompatible pointer type
> 
> And from a PowerPC allyesconfig build:
> 
> drivers/block/cciss.c:582: warning: initialization from incompatible pointer type

David, care to send patches fixing these warnings?

thanks,

greg k-h

Re: linux-next: driver-core tree build warning

[David Brownell]

On Monday 13 July 2009, Greg KH wrote:
> David, care to send patches fixing these warnings?

Here's a single patch:

=================== CUT HERE
From: David Brownell <dbrownell@users.sourceforge.net>

Some more attribute group pointers need to become const-friendly:

drivers/firewire/core-device.c:315: warning: assignment from incompatible pointer type
drivers/mtd/mtdcore.c:227: warning: initialization from incompatible pointer type
drivers/block/cciss.c:582: warning: initialization from incompatible pointer type

Two of these post-date the original constification patch; not sure
how I missed Firewire the first time.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
---
 drivers/block/cciss.c          |    2 +-
 drivers/firewire/core-device.c |    2 +-
 drivers/mtd/mtdcore.c          |    2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)

--- a/drivers/block/cciss.c
+++ b/drivers/block/cciss.c
@@ -572,7 +572,7 @@ static struct attribute_group cciss_dev_
 	.attrs = cciss_dev_attrs,
 };
 
-static struct attribute_group *cciss_dev_attr_groups[] = {
+static const struct attribute_group *cciss_dev_attr_groups[] = {
 	&cciss_dev_attr_group,
 	NULL
 };
--- a/drivers/firewire/core-device.c
+++ b/drivers/firewire/core-device.c
@@ -312,7 +312,7 @@ static void init_fw_attribute_group(stru
 	group->groups[0] = &group->group;
 	group->groups[1] = NULL;
 	group->group.attrs = group->attrs;
-	dev->groups = group->groups;
+	dev->groups = (const struct attribute_group **) group->groups;
 }
 
 static ssize_t modalias_show(struct device *dev,
--- a/drivers/mtd/mtdcore.c
+++ b/drivers/mtd/mtdcore.c
@@ -217,7 +217,7 @@ struct attribute_group mtd_group = {
 	.attrs		= mtd_attrs,
 };
 
-struct attribute_group *mtd_groups[] = {
+struct const attribute_group *mtd_groups[] = {
 	&mtd_group,
 	NULL,
 };

Re: linux-next: driver-core tree build warning

[Greg KH]

On Mon, Jul 13, 2009 at 02:23:59PM -0700, David Brownell wrote:
> On Monday 13 July 2009, Greg KH wrote:
> > David, care to send patches fixing these warnings?
> 
> Here's a single patch:

Thanks, I'll merge this with the original one.

greg k-h

Re: linux-next: driver-core tree build warning

[Greg KH]

On Mon, Jul 13, 2009 at 02:23:59PM -0700, David Brownell wrote:
> On Monday 13 July 2009, Greg KH wrote:
> > David, care to send patches fixing these warnings?
> 
> Here's a single patch:

Now merged and pushed out.  Stephen, this should be now resolved in your
builds for the next -next tree.

thanks,

greg k-h

Re: linux-next: driver-core tree build warning

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 513 bytes --]

On Mon, 13 Jul 2009 14:36:14 -0700 Greg KH <greg@kroah.com> wrote:
>
> On Mon, Jul 13, 2009 at 02:23:59PM -0700, David Brownell wrote:
> > On Monday 13 July 2009, Greg KH wrote:
> > > David, care to send patches fixing these warnings?
> > 
> > Here's a single patch:
> 
> Now merged and pushed out.  Stephen, this should be now resolved in your
> builds for the next -next tree.

Thanks all.

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]

linux-next: driver-core tree build warning

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 457 bytes --]

Hi Greg,

Today's linux-next build (powerpc ppc64_defconfig) produced this warning:

drivers/base/cpu.c:264: warning: initialization from incompatible pointer type
drivers/base/cpu.c:265: warning: initialization from incompatible pointer type

Introduced by commit 380567d3dc35a62f5cc0ec640a45dfc238238417 ("sysdev:
Convert node driver")?

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]

Re: linux-next: driver-core tree build warning

[Greg KH]

On Tue, Jan 19, 2010 at 05:27:41PM +1100, Stephen Rothwell wrote:
> Hi Greg,
> 
> Today's linux-next build (powerpc ppc64_defconfig) produced this warning:
> 
> drivers/base/cpu.c:264: warning: initialization from incompatible pointer type
> drivers/base/cpu.c:265: warning: initialization from incompatible pointer type
> 
> Introduced by commit 380567d3dc35a62f5cc0ec640a45dfc238238417 ("sysdev:
> Convert node driver")?

No, I don't think that patch causes the warning, I can't duplicate it
here.

I don't know what patch causes that problem, as I can't duplicate it in
my tree at all.  Maybe some 'const' patch snuck in from some other tree?

strange,

greg k-h

Re: linux-next: driver-core tree build warning

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 957 bytes --]

Hi Greg,

On Tue, 19 Jan 2010 13:04:17 -0800 Greg KH <greg@kroah.com> wrote:
>
> On Tue, Jan 19, 2010 at 05:27:41PM +1100, Stephen Rothwell wrote:
> > Hi Greg,
> > 
> > Today's linux-next build (powerpc ppc64_defconfig) produced this warning:
> > 
> > drivers/base/cpu.c:264: warning: initialization from incompatible pointer type
> > drivers/base/cpu.c:265: warning: initialization from incompatible pointer type
> > 
> > Introduced by commit 380567d3dc35a62f5cc0ec640a45dfc238238417 ("sysdev:
> > Convert node driver")?
> 
> No, I don't think that patch causes the warning, I can't duplicate it
> here.
> 
> I don't know what patch causes that problem, as I can't duplicate it in
> my tree at all.  Maybe some 'const' patch snuck in from some other tree?
> 
> strange,

OK, if it is still there today, I will look deeper.

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]

Re: linux-next: driver-core tree build warning

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1662 bytes --]

Hi Greg,

On Wed, 20 Jan 2010 10:45:25 +1100 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> On Tue, 19 Jan 2010 13:04:17 -0800 Greg KH <greg@kroah.com> wrote:
> >
> > On Tue, Jan 19, 2010 at 05:27:41PM +1100, Stephen Rothwell wrote:
> > > 
> > > Today's linux-next build (powerpc ppc64_defconfig) produced this warning:
> > > 
> > > drivers/base/cpu.c:264: warning: initialization from incompatible pointer type
> > > drivers/base/cpu.c:265: warning: initialization from incompatible pointer type
> > > 
> > > Introduced by commit 380567d3dc35a62f5cc0ec640a45dfc238238417 ("sysdev:
> > > Convert node driver")?
> > 
> > No, I don't think that patch causes the warning, I can't duplicate it
> > here.
> > 
> > I don't know what patch causes that problem, as I can't duplicate it in
> > my tree at all.  Maybe some 'const' patch snuck in from some other tree?
> > 
> > strange,
> 
> OK, if it is still there today, I will look deeper.

I still get those warnings.  The two referenced lines are added by the commit above.  The relevant parts of the file (in the version I have) are:

(lines 98 and 99)
static CLASS_ATTR(probe, S_IWUSR, NULL, cpu_probe_store);
static CLASS_ATTR(release, S_IWUSR, NULL, cpu_release_store);

...

(lines 262-266)
static struct sysdev_class_attribute *cpu_sysdev_class_attrs[] = {
#ifdef CONFIG_ARCH_CPU_PROBE_RELEASE
        &class_attr_probe.attr,
        &class_attr_release.attr,
#endif


class_attr_probe.attr is a "struct attribute" not a "struct
sysdev_class_attribute".

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]

Re: linux-next: driver-core tree build warning

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 565 bytes --]

Hi Greg,

On Wed, 20 Jan 2010 17:25:43 +1100 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> (lines 262-266)
> static struct sysdev_class_attribute *cpu_sysdev_class_attrs[] = {
> #ifdef CONFIG_ARCH_CPU_PROBE_RELEASE
>         &class_attr_probe.attr,
>         &class_attr_release.attr,
> #endif

The reason you may not see this is because CONFIG_ARCH_CPU_PROBE_RELEASE
is not set for x86_64 allmodconfig builds (and probably many others).

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]

Re: linux-next: driver-core tree build warning

[Greg KH]

On Wed, Jan 20, 2010 at 05:44:41PM +1100, Stephen Rothwell wrote:
> Hi Greg,
> 
> On Wed, 20 Jan 2010 17:25:43 +1100 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> >
> > (lines 262-266)
> > static struct sysdev_class_attribute *cpu_sysdev_class_attrs[] = {
> > #ifdef CONFIG_ARCH_CPU_PROBE_RELEASE
> >         &class_attr_probe.attr,
> >         &class_attr_release.attr,
> > #endif
> 
> The reason you may not see this is because CONFIG_ARCH_CPU_PROBE_RELEASE
> is not set for x86_64 allmodconfig builds (and probably many others).

Ah, you are right, I was focusing on the lines after this.

Here's a patch that I just applied to my tree to hopefully resolve this
issue.

thanks,

greg k-h

----------

From: Greg Kroah-Hartman <gregkh@suse.de>
Subject: sysdev: fix up the probe/release attributes

These should be sysdev attributes, not class attributes.  This patch
should resolve the problem.

Thanks to Stephen Rothwell for pointing out the problem.

Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Andi Kleen <andi@firstfloor.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/base/cpu.c |   16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

--- a/drivers/base/cpu.c
+++ b/drivers/base/cpu.c
@@ -79,24 +79,24 @@ void unregister_cpu(struct cpu *cpu)
 }
 
 #ifdef CONFIG_ARCH_CPU_PROBE_RELEASE
-static ssize_t cpu_probe_store(struct class *class,
-				struct class_attribute *attr,
+static ssize_t cpu_probe_store(struct sys_device *dev,
+				struct sysdev_attribute *attr,
 				const char *buf,
 			       size_t count)
 {
 	return arch_cpu_probe(buf, count);
 }
 
-static ssize_t cpu_release_store(struct class *class,
-				struct class_attribute *attr,
+static ssize_t cpu_release_store(struct sys_device *dev,
+				struct sysdev_attribute *attr,
 				const char *buf,
 				 size_t count)
 {
 	return arch_cpu_release(buf, count);
 }
 
-static CLASS_ATTR(probe, S_IWUSR, NULL, cpu_probe_store);
-static CLASS_ATTR(release, S_IWUSR, NULL, cpu_release_store);
+static SYSDEV_ATTR(probe, S_IWUSR, NULL, cpu_probe_store);
+static SYSDEV_ATTR(release, S_IWUSR, NULL, cpu_release_store);
 #endif /* CONFIG_ARCH_CPU_PROBE_RELEASE */
 
 #else /* ... !CONFIG_HOTPLUG_CPU */
@@ -261,8 +261,8 @@ int __init cpu_dev_init(void)
 
 static struct sysdev_class_attribute *cpu_sysdev_class_attrs[] = {
 #ifdef CONFIG_ARCH_CPU_PROBE_RELEASE
-	&class_attr_probe.attr,
-	&class_attr_release.attr,
+	&attr_probe.attr,
+	&attr_release.attr,
 #endif
 	&cpu_attrs[0].attr,
 	&cpu_attrs[1].attr,

Re: linux-next: driver-core tree build warning

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 354 bytes --]

Hi Greg,

On Thu, 21 Jan 2010 15:10:10 -0800 Greg KH <greg@kroah.com> wrote:
>
> Ah, you are right, I was focusing on the lines after this.
> 
> Here's a patch that I just applied to my tree to hopefully resolve this
> issue.

Thanks.

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]

Re: linux-next: driver-core tree build warning

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1159 bytes --]

Hi Greg,

On Thu, 21 Jan 2010 15:10:10 -0800 Greg KH <greg@kroah.com> wrote:
>
> Here's a patch that I just applied to my tree to hopefully resolve this
> issue.

I still get the warnings.

> -static CLASS_ATTR(probe, S_IWUSR, NULL, cpu_probe_store);
> -static CLASS_ATTR(release, S_IWUSR, NULL, cpu_release_store);
> +static SYSDEV_ATTR(probe, S_IWUSR, NULL, cpu_probe_store);

This defines attr_probe as a "struct sysdev_attribute" whose attr element
is a "struct attribute".

> +static SYSDEV_ATTR(release, S_IWUSR, NULL, cpu_release_store);
>  #endif /* CONFIG_ARCH_CPU_PROBE_RELEASE */
>  
>  #else /* ... !CONFIG_HOTPLUG_CPU */
> @@ -261,8 +261,8 @@ int __init cpu_dev_init(void)
>  
>  static struct sysdev_class_attribute *cpu_sysdev_class_attrs[] = {
>  #ifdef CONFIG_ARCH_CPU_PROBE_RELEASE
> -	&class_attr_probe.attr,
> -	&class_attr_release.attr,
> +	&attr_probe.attr,

so here we are initialising a "struct sysdev_class_attribute *" with a
"struct attribute *".  Thus the warning.

> +	&attr_release.attr,

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]

Re: linux-next: driver-core tree build warning

[Greg KH]

On Fri, Jan 29, 2010 at 06:21:15PM +1100, Stephen Rothwell wrote:
> Hi Greg,
> 
> On Thu, 21 Jan 2010 15:10:10 -0800 Greg KH <greg@kroah.com> wrote:
> >
> > Here's a patch that I just applied to my tree to hopefully resolve this
> > issue.
> 
> I still get the warnings.
> 
> > -static CLASS_ATTR(probe, S_IWUSR, NULL, cpu_probe_store);
> > -static CLASS_ATTR(release, S_IWUSR, NULL, cpu_release_store);
> > +static SYSDEV_ATTR(probe, S_IWUSR, NULL, cpu_probe_store);
> 
> This defines attr_probe as a "struct sysdev_attribute" whose attr element
> is a "struct attribute".
> 
> > +static SYSDEV_ATTR(release, S_IWUSR, NULL, cpu_release_store);
> >  #endif /* CONFIG_ARCH_CPU_PROBE_RELEASE */
> >  
> >  #else /* ... !CONFIG_HOTPLUG_CPU */
> > @@ -261,8 +261,8 @@ int __init cpu_dev_init(void)
> >  
> >  static struct sysdev_class_attribute *cpu_sysdev_class_attrs[] = {
> >  #ifdef CONFIG_ARCH_CPU_PROBE_RELEASE
> > -	&class_attr_probe.attr,
> > -	&class_attr_release.attr,
> > +	&attr_probe.attr,
> 
> so here we are initialising a "struct sysdev_class_attribute *" with a
> "struct attribute *".  Thus the warning.

Doh, sorry about that.  I've fixed it up now and pushed it out.

thanks,

greg k-h

Re: linux-next: driver-core tree build warning

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 276 bytes --]

Hi Greg,

On Tue, 2 Feb 2010 18:22:47 -0800 Greg KH <greg@kroah.com> wrote:
>
> Doh, sorry about that.  I've fixed it up now and pushed it out.

Great, thanks.

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]

Re: linux-next: driver-core tree build warning

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1414 bytes --]

Hi Greg,

On Tue, 2 Feb 2010 18:22:47 -0800 Greg KH <greg@kroah.com> wrote:
>
> On Fri, Jan 29, 2010 at 06:21:15PM +1100, Stephen Rothwell wrote:
> > 
> > > -static CLASS_ATTR(probe, S_IWUSR, NULL, cpu_probe_store);
> > > -static CLASS_ATTR(release, S_IWUSR, NULL, cpu_release_store);
> > > +static SYSDEV_ATTR(probe, S_IWUSR, NULL, cpu_probe_store);
> > 
> > This defines attr_probe as a "struct sysdev_attribute" whose attr element
> > is a "struct attribute".
> > 
> > > +static SYSDEV_ATTR(release, S_IWUSR, NULL, cpu_release_store);
> > >  #endif /* CONFIG_ARCH_CPU_PROBE_RELEASE */
> > >  
> > >  #else /* ... !CONFIG_HOTPLUG_CPU */
> > > @@ -261,8 +261,8 @@ int __init cpu_dev_init(void)
> > >  
> > >  static struct sysdev_class_attribute *cpu_sysdev_class_attrs[] = {
> > >  #ifdef CONFIG_ARCH_CPU_PROBE_RELEASE
> > > -	&class_attr_probe.attr,
> > > -	&class_attr_release.attr,
> > > +	&attr_probe.attr,
> > 
> > so here we are initialising a "struct sysdev_class_attribute *" with a
> > "struct attribute *".  Thus the warning.
> 
> Doh, sorry about that.  I've fixed it up now and pushed it out.

Still not right :-( &attr_probe and &attr_release are "struct
sysdev_attribute *" after this patch, but are being assigned to "struct
sysdev_class_attribute *" ...

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]