[PATCH 1/2] Revert "ipv4: Make INET_LRO a bool instead of tristate."

[PATCH 1/2] Revert "ipv4: Make INET_LRO a bool instead of tristate."

[Ben Hutchings]

This reverts commit e81963b180ac502fda0326edf059b1e29cdef1a2.

LRO is now deprecated in favour of GRO, and only a few drivers use it,
so it is desirable to build it as a module in distribution kernels.

The original change to prevent building it as a module was made in an
attempt to avoid the case where some dependents are set to y and some
to m, and INET_LRO can be set to m rather than y.  However, the
Kconfig system will reliably set INET_LRO=y in this case.

Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
---
Dave,

You made the change I want to revert in response to
<http://article.gmane.org/gmane.linux.kernel/825646>.  The real problem
with its configuration is actually that CONFIG_INET is not set but
CONFIG_INET_LRO=m, and the fix is to make CONFIG_PASEMI_MAC depend on
CONFIG_INET.

Ben.

 net/ipv4/Kconfig |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/net/ipv4/Kconfig b/net/ipv4/Kconfig
index 94e0b51..704a0cf 100644
--- a/net/ipv4/Kconfig
+++ b/net/ipv4/Kconfig
@@ -420,7 +420,7 @@ config INET_XFRM_MODE_BEET
 	  If unsure, say Y.
 
 config INET_LRO
-	bool "Large Receive Offload (ipv4/tcp)"
+	tristate "Large Receive Offload (ipv4/tcp)"
 	default y
 	---help---
 	  Support for Large Receive Offload (ipv4/tcp).
-- 
1.7.1

[PATCH 2/2] netdev: Depend on INET before selecting INET_LRO

[Ben Hutchings]

Since 'select' ignores dependencies, drivers that select INET_LRO must
depend on INET.  This fixes the broken configuration reported in
<http://article.gmane.org/gmane.linux.kernel/825646>.

Reported-by: Subrata Modak <subrata@linux.vnet.ibm.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
---
 drivers/net/Kconfig |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
index ef683a9..13d01f3 100644
--- a/drivers/net/Kconfig
+++ b/drivers/net/Kconfig
@@ -2428,7 +2428,7 @@ config UGETH_TX_ON_DEMAND
 
 config MV643XX_ETH
 	tristate "Marvell Discovery (643XX) and Orion ethernet support"
-	depends on MV64X60 || PPC32 || PLAT_ORION
+	depends on (MV64X60 || PPC32 || PLAT_ORION) && INET
 	select INET_LRO
 	select PHYLIB
 	help
@@ -2815,7 +2815,7 @@ config NIU
 
 config PASEMI_MAC
 	tristate "PA Semi 1/10Gbit MAC"
-	depends on PPC_PASEMI && PCI
+	depends on PPC_PASEMI && PCI && INET
 	select PHYLIB
 	select INET_LRO
 	help
-- 
1.7.1

Re: [PATCH 1/2] Revert "ipv4: Make INET_LRO a bool instead of tristate."

[David Miller]

From: Ben Hutchings <ben@decadent.org.uk>
Date: Mon, 04 Oct 2010 02:37:42 +0100

> You made the change I want to revert in response to
> <http://article.gmane.org/gmane.linux.kernel/825646>.  The real problem
> with its configuration is actually that CONFIG_INET is not set but
> CONFIG_INET_LRO=m, and the fix is to make CONFIG_PASEMI_MAC depend on
> CONFIG_INET.

Ben, you can't just revert this by itself.

That knowingly breaks the build.

If you want the tristate back, you must do it after or at the
same time as fixing the driver deps.

Thanks.

Re: [PATCH 1/2] Revert "ipv4: Make INET_LRO a bool instead of tristate."

[Ben Hutchings]

[-- Attachment #1: Type: text/plain, Size: 942 bytes --]

On Sun, 2010-10-03 at 19:56 -0700, David Miller wrote:
> From: Ben Hutchings <ben@decadent.org.uk>
> Date: Mon, 04 Oct 2010 02:37:42 +0100
> 
> > You made the change I want to revert in response to
> > <http://article.gmane.org/gmane.linux.kernel/825646>.  The real problem
> > with its configuration is actually that CONFIG_INET is not set but
> > CONFIG_INET_LRO=m, and the fix is to make CONFIG_PASEMI_MAC depend on
> > CONFIG_INET.
> 
> Ben, you can't just revert this by itself.
> 
> That knowingly breaks the build.
> 
> If you want the tristate back, you must do it after or at the
> same time as fixing the driver deps.

The fact that the driver dependencies are broken has nothing to do with
whether CONFIG_INET_LRO is boolean or tristate.  You fixed a problem
that didn't exist rather than the problem that did.

Ben.

-- 
Ben Hutchings
Once a job is fouled up, anything done to improve it makes it worse.

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 828 bytes --]

Re: [PATCH 1/2] Revert "ipv4: Make INET_LRO a bool instead of tristate."

[David Miller]

From: Ben Hutchings <ben@decadent.org.uk>
Date: Mon, 04 Oct 2010 04:54:55 +0100

> On Sun, 2010-10-03 at 19:56 -0700, David Miller wrote:
>> From: Ben Hutchings <ben@decadent.org.uk>
>> Date: Mon, 04 Oct 2010 02:37:42 +0100
>> 
>> > You made the change I want to revert in response to
>> > <http://article.gmane.org/gmane.linux.kernel/825646>.  The real problem
>> > with its configuration is actually that CONFIG_INET is not set but
>> > CONFIG_INET_LRO=m, and the fix is to make CONFIG_PASEMI_MAC depend on
>> > CONFIG_INET.
>> 
>> Ben, you can't just revert this by itself.
>> 
>> That knowingly breaks the build.
>> 
>> If you want the tristate back, you must do it after or at the
>> same time as fixing the driver deps.
> 
> The fact that the driver dependencies are broken has nothing to do with
> whether CONFIG_INET_LRO is boolean or tristate.  You fixed a problem
> that didn't exist rather than the problem that did.

Oh, I see, ok I'll apply your patches thanks for explaining Ben.