[PATCH] Support kernel/hotplug sysctl variable when !CONFIG_NET

[PATCH] Support kernel/hotplug sysctl variable when !CONFIG_NET

[Ian Abbott]

From: Ian Abbott <abbotti@mev.co.uk>

The kernel/hotplug sysctl variable (/proc/sys/kernel/hotplug file) was
made conditional on CONFIG_NET by commit
f743ca5e10f4145e0b3e6d11b9b46171e16af7ce (applied in 2.6.18) to fix
problems with undefined references in 2.6.16 when CONFIG_HOTPLUG=y &&
!CONFIG_NET, but this restriction is no longer needed.

This patch makes the kernel/hotplug sysctl variable depend only on
CONFIG_HOTPLUG.

Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
---
 kernel/sysctl.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/kernel/sysctl.c b/kernel/sysctl.c
index 702709b..70bc796 100644
--- a/kernel/sysctl.c
+++ b/kernel/sysctl.c
@@ -567,7 +567,7 @@ static struct ctl_table kern_table[] = {
 		.extra2		= &one,
 	},
 #endif
-#if defined(CONFIG_HOTPLUG) && defined(CONFIG_NET)
+#ifdef CONFIG_HOTPLUG
 	{
 		.procname	= "hotplug",
 		.data		= &uevent_helper,
-- 
1.7.1

Re: [PATCH] Support kernel/hotplug sysctl variable when !CONFIG_NET

[Ian Abbott]

On 07/06/10 12:57, Ian Abbott wrote:
> The kernel/hotplug sysctl variable (/proc/sys/kernel/hotplug file) was
      ^^^^^^^^^^^^^^

That should read "kernel.hotplug" of course!

-- 
-=( Ian Abbott @ MEV Ltd.    E-mail: <abbotti@mev.co.uk>        )=-
-=( Tel: +44 (0)161 477 1898   FAX: +44 (0)161 718 3587         )=-

Re: [PATCH] Support kernel/hotplug sysctl variable when !CONFIG_NET

[Greg KH]

On Mon, Jun 07, 2010 at 12:57:12PM +0100, Ian Abbott wrote:
> From: Ian Abbott <abbotti@mev.co.uk>
> 
> The kernel/hotplug sysctl variable (/proc/sys/kernel/hotplug file) was
> made conditional on CONFIG_NET by commit
> f743ca5e10f4145e0b3e6d11b9b46171e16af7ce (applied in 2.6.18) to fix
> problems with undefined references in 2.6.16 when CONFIG_HOTPLUG=y &&
> !CONFIG_NET, but this restriction is no longer needed.

Why is this restriction no longer needed?  What changed?

> This patch makes the kernel/hotplug sysctl variable depend only on
> CONFIG_HOTPLUG.

Why is this needed?  Are you really running without a network and still
want to change the default /sbin/hotplug binary?

confused,

greg k-h

Re: [PATCH] Support kernel/hotplug sysctl variable when !CONFIG_NET

[Randy Dunlap]

On Mon, 7 Jun 2010 10:43:50 -0700 Greg KH wrote:

> On Mon, Jun 07, 2010 at 12:57:12PM +0100, Ian Abbott wrote:
> > From: Ian Abbott <abbotti@mev.co.uk>
> > 
> > The kernel/hotplug sysctl variable (/proc/sys/kernel/hotplug file) was
> > made conditional on CONFIG_NET by commit
> > f743ca5e10f4145e0b3e6d11b9b46171e16af7ce (applied in 2.6.18) to fix
> > problems with undefined references in 2.6.16 when CONFIG_HOTPLUG=y &&
> > !CONFIG_NET, but this restriction is no longer needed.

Ack, builds for me on x86_64 when CONFIG_NET is not enabled.

> Why is this restriction no longer needed?  What changed?


commit cd3772e6898c6386f21d2958346d6dd57d4204f5
Author: Ming Lei <tom.leiming@gmail.com>
Date:   Sun Nov 16 18:22:09 2008 +0800

    kernel/ksysfs.c:fix dependence on CONFIG_NET
    
    Access to uevent_seqnum and uevent_helper does not need to
    depend on CONFIG_NET, so remove it.
    
    Signed-off-by: Ming Lei <tom.leiming@gmail.com>
    Cc: Kay Sievers <kay.sievers@vrfy.org>
    Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>


> > This patch makes the kernel/hotplug sysctl variable depend only on
> > CONFIG_HOTPLUG.
> 
> Why is this needed?  Are you really running without a network and still
> want to change the default /sbin/hotplug binary?
> 
> confused,


---
~Randy
*** Remember to use Documentation/SubmitChecklist when testing your code ***

Re: [PATCH] Support kernel/hotplug sysctl variable when !CONFIG_NET

[Greg KH]

On Mon, Jun 07, 2010 at 11:11:45AM -0700, Randy Dunlap wrote:
> On Mon, 7 Jun 2010 10:43:50 -0700 Greg KH wrote:
> 
> > On Mon, Jun 07, 2010 at 12:57:12PM +0100, Ian Abbott wrote:
> > > From: Ian Abbott <abbotti@mev.co.uk>
> > > 
> > > The kernel/hotplug sysctl variable (/proc/sys/kernel/hotplug file) was
> > > made conditional on CONFIG_NET by commit
> > > f743ca5e10f4145e0b3e6d11b9b46171e16af7ce (applied in 2.6.18) to fix
> > > problems with undefined references in 2.6.16 when CONFIG_HOTPLUG=y &&
> > > !CONFIG_NET, but this restriction is no longer needed.
> 
> Ack, builds for me on x86_64 when CONFIG_NET is not enabled.
> 
> > Why is this restriction no longer needed?  What changed?
> 
> 
> commit cd3772e6898c6386f21d2958346d6dd57d4204f5
> Author: Ming Lei <tom.leiming@gmail.com>
> Date:   Sun Nov 16 18:22:09 2008 +0800
> 
>     kernel/ksysfs.c:fix dependence on CONFIG_NET
>     
>     Access to uevent_seqnum and uevent_helper does not need to
>     depend on CONFIG_NET, so remove it.
>     
>     Signed-off-by: Ming Lei <tom.leiming@gmail.com>
>     Cc: Kay Sievers <kay.sievers@vrfy.org>
>     Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

So something back in 2008 changed which requires a change now?

I'm still confused, what use case calls for the original patch here
being made?  Who wants to modify this option with CONFIG_NET disabled on
their machine?

thanks,

greg k-h

Re: [PATCH] Support kernel/hotplug sysctl variable when !CONFIG_NET

[Randy Dunlap]

On 06/07/10 11:16, Greg KH wrote:
> On Mon, Jun 07, 2010 at 11:11:45AM -0700, Randy Dunlap wrote:
>> On Mon, 7 Jun 2010 10:43:50 -0700 Greg KH wrote:
>>
>>> On Mon, Jun 07, 2010 at 12:57:12PM +0100, Ian Abbott wrote:
>>>> From: Ian Abbott <abbotti@mev.co.uk>
>>>>
>>>> The kernel/hotplug sysctl variable (/proc/sys/kernel/hotplug file) was
>>>> made conditional on CONFIG_NET by commit
>>>> f743ca5e10f4145e0b3e6d11b9b46171e16af7ce (applied in 2.6.18) to fix
>>>> problems with undefined references in 2.6.16 when CONFIG_HOTPLUG=y &&
>>>> !CONFIG_NET, but this restriction is no longer needed.
>>
>> Ack, builds for me on x86_64 when CONFIG_NET is not enabled.
>>
>>> Why is this restriction no longer needed?  What changed?
>>
>>
>> commit cd3772e6898c6386f21d2958346d6dd57d4204f5
>> Author: Ming Lei <tom.leiming@gmail.com>
>> Date:   Sun Nov 16 18:22:09 2008 +0800
>>
>>     kernel/ksysfs.c:fix dependence on CONFIG_NET
>>     
>>     Access to uevent_seqnum and uevent_helper does not need to
>>     depend on CONFIG_NET, so remove it.
>>     
>>     Signed-off-by: Ming Lei <tom.leiming@gmail.com>
>>     Cc: Kay Sievers <kay.sievers@vrfy.org>
>>     Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
> 
> So something back in 2008 changed which requires a change now?

"requires" is too strong of a word to use here.  "allows" a change.

> I'm still confused, what use case calls for the original patch here
> being made?  Who wants to modify this option with CONFIG_NET disabled on
> their machine?


Are you trying to slow down on kernel patches for some reason?

Using CONFIG_NET is more restrictive than is needed here, so remove it.

-- 
~Randy
*** Remember to use Documentation/SubmitChecklist when testing your code ***

Re: [PATCH] Support kernel/hotplug sysctl variable when !CONFIG_NET

[Greg KH]

On Mon, Jun 07, 2010 at 11:26:58AM -0700, Randy Dunlap wrote:
> On 06/07/10 11:16, Greg KH wrote:
> > On Mon, Jun 07, 2010 at 11:11:45AM -0700, Randy Dunlap wrote:
> >> On Mon, 7 Jun 2010 10:43:50 -0700 Greg KH wrote:
> >>
> >>> On Mon, Jun 07, 2010 at 12:57:12PM +0100, Ian Abbott wrote:
> >>>> From: Ian Abbott <abbotti@mev.co.uk>
> >>>>
> >>>> The kernel/hotplug sysctl variable (/proc/sys/kernel/hotplug file) was
> >>>> made conditional on CONFIG_NET by commit
> >>>> f743ca5e10f4145e0b3e6d11b9b46171e16af7ce (applied in 2.6.18) to fix
> >>>> problems with undefined references in 2.6.16 when CONFIG_HOTPLUG=y &&
> >>>> !CONFIG_NET, but this restriction is no longer needed.
> >>
> >> Ack, builds for me on x86_64 when CONFIG_NET is not enabled.
> >>
> >>> Why is this restriction no longer needed?  What changed?
> >>
> >>
> >> commit cd3772e6898c6386f21d2958346d6dd57d4204f5
> >> Author: Ming Lei <tom.leiming@gmail.com>
> >> Date:   Sun Nov 16 18:22:09 2008 +0800
> >>
> >>     kernel/ksysfs.c:fix dependence on CONFIG_NET
> >>     
> >>     Access to uevent_seqnum and uevent_helper does not need to
> >>     depend on CONFIG_NET, so remove it.
> >>     
> >>     Signed-off-by: Ming Lei <tom.leiming@gmail.com>
> >>     Cc: Kay Sievers <kay.sievers@vrfy.org>
> >>     Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
> > 
> > So something back in 2008 changed which requires a change now?
> 
> "requires" is too strong of a word to use here.  "allows" a change.
> 
> > I'm still confused, what use case calls for the original patch here
> > being made?  Who wants to modify this option with CONFIG_NET disabled on
> > their machine?
> 
> 
> Are you trying to slow down on kernel patches for some reason?

Heh, no, I'm trying to see where to put this, in my "for .35" queue, or
my "for .36" queue.

> Using CONFIG_NET is more restrictive than is needed here, so remove it.

Fair enough, I'll queue it up.

thanks,

greg k-h