netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH 2/3] dummy: make device count build-time configurable
@ 2016-01-12 11:57 Lubomir Rintel
  2016-01-12 18:42 ` Stephen Hemminger
  0 siblings, 1 reply; 3+ messages in thread
From: Lubomir Rintel @ 2016-01-12 11:57 UTC (permalink / raw)
  To: netdev; +Cc: linux-kernel, David S. Miller, Lubomir Rintel

The devices can be created at run-time for quite some time already and the
load-time device creation collides with attempts to create the device of
the same name:

  # rmmod dummy
  # ip link add dummy0 type dummy
  RTNETLINK answers: File exists

This is pretty much the same situation as was with the block loop devices
which was solved by adding a build-time configuration that the
distributions could use as they deem fit while keeping the default for
compatibility.

Let's do that here as well.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
---
 drivers/net/Kconfig | 15 +++++++++++++++
 drivers/net/dummy.c |  2 +-
 2 files changed, 16 insertions(+), 1 deletion(-)

diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
index 63535b4..f94af69 100644
--- a/drivers/net/Kconfig
+++ b/drivers/net/Kconfig
@@ -71,6 +71,21 @@ config DUMMY
 	  To compile this driver as a module, choose M here: the module
 	  will be called dummy.
 
+config DUMMY_COUNT
+        int "Number of dummy devices to pre-create at init time"
+        depends on DUMMY
+        default 1
+        help
+          Static number of dummy devices to be unconditionally pre-created
+          at init time.
+
+          This default value can be overwritten on the kernel command
+          line or with module-parameter dummy.numdummies.
+
+          The historic default is 1. If a mid-2007 version of iproute2
+          is used (v2.6.23 or later), it can be set to 0, since needed
+          dummy devices can be dynamically allocated via rtnetlink.
+
 config EQUALIZER
 	tristate "EQL (serial line load balancing) support"
 	---help---
diff --git a/drivers/net/dummy.c b/drivers/net/dummy.c
index 69fc840..e4506c9 100644
--- a/drivers/net/dummy.c
+++ b/drivers/net/dummy.c
@@ -41,7 +41,7 @@
 #define DRV_NAME	"dummy"
 #define DRV_VERSION	"1.0"
 
-static int numdummies = 1;
+static int numdummies = CONFIG_DUMMY_COUNT;
 
 /* fake multicast ability */
 static void set_multicast_list(struct net_device *dev)
-- 
2.5.0

^ permalink raw reply related	[flat|nested] 3+ messages in thread

* Re: [PATCH 2/3] dummy: make device count build-time configurable
  2016-01-12 11:57 [PATCH 2/3] dummy: make device count build-time configurable Lubomir Rintel
@ 2016-01-12 18:42 ` Stephen Hemminger
  2016-02-05 15:09   ` Lubomir Rintel
  0 siblings, 1 reply; 3+ messages in thread
From: Stephen Hemminger @ 2016-01-12 18:42 UTC (permalink / raw)
  To: Lubomir Rintel; +Cc: netdev, linux-kernel, David S. Miller

On Tue, 12 Jan 2016 12:57:33 +0100
Lubomir Rintel <lkundrak@v3.sk> wrote:

> The devices can be created at run-time for quite some time already and the
> load-time device creation collides with attempts to create the device of
> the same name:
> 
>   # rmmod dummy
>   # ip link add dummy0 type dummy
>   RTNETLINK answers: File exists
> 
> This is pretty much the same situation as was with the block loop devices
> which was solved by adding a build-time configuration that the
> distributions could use as they deem fit while keeping the default for
> compatibility.
> 
> Let's do that here as well.
> 
> Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>

There is already a module parameter for this, so making it a compile
time option adds nothing.

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH 2/3] dummy: make device count build-time configurable
  2016-01-12 18:42 ` Stephen Hemminger
@ 2016-02-05 15:09   ` Lubomir Rintel
  0 siblings, 0 replies; 3+ messages in thread
From: Lubomir Rintel @ 2016-02-05 15:09 UTC (permalink / raw)
  To: Stephen Hemminger; +Cc: netdev, linux-kernel, David S. Miller

Hi Stephen,

On Tue, 2016-01-12 at 10:42 -0800, Stephen Hemminger wrote:
> On Tue, 12 Jan 2016 12:57:33 +0100
> Lubomir Rintel <lkundrak@v3.sk> wrote:
> 
> > The devices can be created at run-time for quite some time already
> > and the
> > load-time device creation collides with attempts to create the
> > device of
> > the same name:
> > 
> >   # rmmod dummy
> >   # ip link add dummy0 type dummy
> >   RTNETLINK answers: File exists
> > 
> > This is pretty much the same situation as was with the block loop
> > devices
> > which was solved by adding a build-time configuration that the
> > distributions could use as they deem fit while keeping the default
> > for
> > compatibility.
> > 
> > Let's do that here as well.
> > 
> > Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
> 
> There is already a module parameter for this, so making it a compile
> time option adds nothing.

This option changes the defaults for the parameter.

When the module gets autoloaded, the user doesn't get a chance to
specify the module parameter and unwanted devices pop in.

Worse even, the automatically created devices are likely to collide
with what the user asked for.

Lubo

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2016-02-05 15:09 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2016-01-12 11:57 [PATCH 2/3] dummy: make device count build-time configurable Lubomir Rintel
2016-01-12 18:42 ` Stephen Hemminger
2016-02-05 15:09   ` Lubomir Rintel

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).