linux-next: build failure after merge of the final tree (staging tree related)

linux-next: build failure after merge of the final tree (staging tree related)

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1929 bytes --]

Hi all,

After merging the final tree, today's linux-next build (powerpc
allyesconfig) failed like this:

drivers/staging/silicom/bypasslib/built-in.o: In function `.is_bypass':
(.text+0x3254): multiple definition of `.is_bypass'
drivers/staging/silicom/bpctl_mod.o:(.text+0x1d938): first defined here
drivers/staging/silicom/bypasslib/built-in.o: In function `is_bypass':
(.opd+0x480): multiple definition of `is_bypass'
drivers/staging/silicom/bpctl_mod.o:(.opd+0x15f0): first defined here

Caused by commit 7040e556c90a ("staging: add Silicom Bypass driver").

I added this patch to disable the driver:

From 884437b30775214e9f34696adb2be2f24f1e6df7 Mon Sep 17 00:00:00 2001
From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Mon, 10 Sep 2012 17:01:05 +1000
Subject: [PATCH] staging: disable Silicom Bypass driver

due to build breakage:

rivers/staging/silicom/bypasslib/built-in.o: In function `.is_bypass':
(.text+0x3254): multiple definition of `.is_bypass'
drivers/staging/silicom/bpctl_mod.o:(.text+0x1d938): first defined here
drivers/staging/silicom/bypasslib/built-in.o: In function `is_bypass':
(.opd+0x480): multiple definition of `is_bypass'
drivers/staging/silicom/bpctl_mod.o:(.opd+0x15f0): first defined here

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 drivers/staging/silicom/Kconfig |    1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/staging/silicom/Kconfig b/drivers/staging/silicom/Kconfig
index 3493ee8..1914844 100644
--- a/drivers/staging/silicom/Kconfig
+++ b/drivers/staging/silicom/Kconfig
@@ -6,6 +6,7 @@ config NET_VENDOR_SILICOM
 	bool "Silicom devices"
 	default y
 	depends on (SSB_POSSIBLE && HAS_DMA) || PCI 
+	depends on BROKEN
 	---help---
 	  If you have a network card (Ethernet) belonging to this class,
 	  say Y.
-- 
1.7.10.280.gaa39

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]

Re: linux-next: build failure after merge of the final tree (staging tree related)

[Greg KH]

On Mon, Sep 10, 2012 at 05:04:09PM +1000, Stephen Rothwell wrote:
> Hi all,
> 
> After merging the final tree, today's linux-next build (powerpc
> allyesconfig) failed like this:
> 
> drivers/staging/silicom/bypasslib/built-in.o: In function `.is_bypass':
> (.text+0x3254): multiple definition of `.is_bypass'
> drivers/staging/silicom/bpctl_mod.o:(.text+0x1d938): first defined here
> drivers/staging/silicom/bypasslib/built-in.o: In function `is_bypass':
> (.opd+0x480): multiple definition of `is_bypass'
> drivers/staging/silicom/bpctl_mod.o:(.opd+0x15f0): first defined here
> 
> Caused by commit 7040e556c90a ("staging: add Silicom Bypass driver").
> 
> I added this patch to disable the driver:
> 
> From 884437b30775214e9f34696adb2be2f24f1e6df7 Mon Sep 17 00:00:00 2001
> From: Stephen Rothwell <sfr@canb.auug.org.au>
> Date: Mon, 10 Sep 2012 17:01:05 +1000
> Subject: [PATCH] staging: disable Silicom Bypass driver
> 
> due to build breakage:
> 
> rivers/staging/silicom/bypasslib/built-in.o: In function `.is_bypass':
> (.text+0x3254): multiple definition of `.is_bypass'
> drivers/staging/silicom/bpctl_mod.o:(.text+0x1d938): first defined here
> drivers/staging/silicom/bypasslib/built-in.o: In function `is_bypass':
> (.opd+0x480): multiple definition of `is_bypass'
> drivers/staging/silicom/bpctl_mod.o:(.opd+0x15f0): first defined here
> 
> Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
> ---
>  drivers/staging/silicom/Kconfig |    1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/staging/silicom/Kconfig b/drivers/staging/silicom/Kconfig
> index 3493ee8..1914844 100644
> --- a/drivers/staging/silicom/Kconfig
> +++ b/drivers/staging/silicom/Kconfig
> @@ -6,6 +6,7 @@ config NET_VENDOR_SILICOM
>  	bool "Silicom devices"
>  	default y
>  	depends on (SSB_POSSIBLE && HAS_DMA) || PCI 
> +	depends on BROKEN

I think forcing this driver to be a module will resolve this issue.
It seems to be quite a common problem for most staging drivers when they
first land in the tree, sorry for not catching this one.

greg k-h

[PATCH 3.6rc5.linux-next 1/1] Force module

[DanielC]

Sorry about breaking the build, this should fix.

Staging: silicom: Depend on module until monolithic build fixed

Signed-off-by: Daniel Cotey <puff65537@bansheeslibrary.com>
---
 drivers/staging/silicom/Kconfig | 5 ++++-
 drivers/staging/silicom/TODO    | 1 +
 2 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/staging/silicom/Kconfig b/drivers/staging/silicom/Kconfig
index 1914844..045ab08 100644
--- a/drivers/staging/silicom/Kconfig
+++ b/drivers/staging/silicom/Kconfig
@@ -5,7 +5,7 @@
 config NET_VENDOR_SILICOM
 	bool "Silicom devices"
 	default y
-	depends on (SSB_POSSIBLE && HAS_DMA) || PCI 
+	depends on PCI 
 	depends on BROKEN
 	---help---
 	  If you have a network card (Ethernet) belonging to this class,
@@ -20,6 +20,8 @@ if NET_VENDOR_SILICOM
 
 config SBYPASS
 	tristate "Silicom BypassCTL library support"
+	depends on PCI
+	depends on m
 	---help---
 	  If you have a network (Ethernet) controller of this type, say Y
 
@@ -29,6 +31,7 @@ config SBYPASS
 config BPCTL
 	tristate "Silicom BypassCTL net support"
 	depends on PCI
+	depends on m
 	select SBYPASS
 	select NET_CORE
 	select MII
diff --git a/drivers/staging/silicom/TODO b/drivers/staging/silicom/TODO
index bf60714..a39821f 100644
--- a/drivers/staging/silicom/TODO
+++ b/drivers/staging/silicom/TODO
@@ -3,5 +3,6 @@ TODO:
 	- locking audit
 	- single module with all functionality
 	- userland 
+        - fix monolithic build.
 
 
-- 
1.7.11.4

Re: [PATCH 3.6rc5.linux-next 1/1] Force module

[Greg KH]

On Mon, Sep 10, 2012 at 06:43:30PM -0700, DanielC wrote:
> Sorry about breaking the build, this should fix.

This doesn't make sense in a changelog entry.

> 
> Staging: silicom: Depend on module until monolithic build fixed

This needs to be in the Subject: line of the patch, don't make me edit
it by hand to add it in.

> 
> Signed-off-by: Daniel Cotey <puff65537@bansheeslibrary.com>

You forgot to add:
	Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
to this list, as he is the one that found the problem.

Care to fix this up and resend the patch so I can apply it without
needing to edit it at all?

thanks,

greg k-h

[PATCHv2 3.6rc5.linux-next 1/1] Staging: silicom: Force depend on module

[DanielC]

Staging: silicom: Force depend on module until monolithic build fixed

Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Daniel Cotey <puff65537@bansheeslibrary.com>

---
 drivers/staging/silicom/Kconfig | 5 ++++-
 drivers/staging/silicom/TODO    | 1 +
 2 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/staging/silicom/Kconfig b/drivers/staging/silicom/Kconfig
index 1914844..045ab08 100644
--- a/drivers/staging/silicom/Kconfig
+++ b/drivers/staging/silicom/Kconfig
@@ -5,7 +5,7 @@
 config NET_VENDOR_SILICOM
 	bool "Silicom devices"
 	default y
-	depends on (SSB_POSSIBLE && HAS_DMA) || PCI 
+	depends on PCI 
 	depends on BROKEN
 	---help---
 	  If you have a network card (Ethernet) belonging to this class,
@@ -20,6 +20,8 @@ if NET_VENDOR_SILICOM
 
 config SBYPASS
 	tristate "Silicom BypassCTL library support"
+	depends on PCI
+	depends on m
 	---help---
 	  If you have a network (Ethernet) controller of this type, say Y
 
@@ -29,6 +31,7 @@ config SBYPASS
 config BPCTL
 	tristate "Silicom BypassCTL net support"
 	depends on PCI
+	depends on m
 	select SBYPASS
 	select NET_CORE
 	select MII
diff --git a/drivers/staging/silicom/TODO b/drivers/staging/silicom/TODO
index bf60714..a39821f 100644
--- a/drivers/staging/silicom/TODO
+++ b/drivers/staging/silicom/TODO
@@ -3,5 +3,6 @@ TODO:
 	- locking audit
 	- single module with all functionality
 	- userland 
+        - fix monolithic build.
 
 
-- 
1.7.11.4