[PATCH] ppc: Marvell mv64x60 drivers for ppc platforms fix

[PATCH] ppc: Marvell mv64x60 drivers for ppc platforms fix

[Yuri Tikhonov]

 This patch fixes the issue of erroneously involving the 
arch/powerpc/sysdev/mv64x60*.c files into the compilation 
process for ppc- (not powerpc-) based platforms (e.g., 
P3M750).

 Signed-off-by: Yuri Tikhonov <yur@emcraft.com>

--

diff --git a/arch/powerpc/sysdev/Makefile b/arch/powerpc/sysdev/Makefile
index c3ce0bd..25eaa14 100644
--- a/arch/powerpc/sysdev/Makefile
+++ b/arch/powerpc/sysdev/Makefile
@@ -17,7 +17,9 @@ obj-$(CONFIG_FSL_PCIE)		+= fsl_pcie.o
 obj-$(CONFIG_TSI108_BRIDGE)	+= tsi108_pci.o tsi108_dev.o
 obj-$(CONFIG_QUICC_ENGINE)	+= qe_lib/
 mv64x60-$(CONFIG_PCI)		+= mv64x60_pci.o
+ifeq ($(CONFIG_PPC_MERGE),y)
 obj-$(CONFIG_MV64X60)		+= $(mv64x60-y) mv64x60_pic.o mv64x60_dev.o
+endif
 
 # contains only the suspend handler for time
 obj-$(CONFIG_PM)		+= timer.o

Re: [PATCH] ppc: Marvell mv64x60 drivers for ppc platforms fix

[Dale Farnsworth]

In article <200706241552.31483.yur@emcraft.com> you write:
>  This patch fixes the issue of erroneously involving the 
> arch/powerpc/sysdev/mv64x60*.c files into the compilation 
> process for ppc- (not powerpc-) based platforms (e.g., 
> P3M750).
> 
>  Signed-off-by: Yuri Tikhonov <yur@emcraft.com>
> 
> --
> 
> diff --git a/arch/powerpc/sysdev/Makefile b/arch/powerpc/sysdev/Makefile
> index c3ce0bd..25eaa14 100644
> --- a/arch/powerpc/sysdev/Makefile
> +++ b/arch/powerpc/sysdev/Makefile
> @@ -17,7 +17,9 @@ obj-$(CONFIG_FSL_PCIE)		+= fsl_pcie.o
>  obj-$(CONFIG_TSI108_BRIDGE)	+= tsi108_pci.o tsi108_dev.o
>  obj-$(CONFIG_QUICC_ENGINE)	+= qe_lib/
>  mv64x60-$(CONFIG_PCI)		+= mv64x60_pci.o

Shouldn't the above line also be placed inside the of following ifeq ?

> +ifeq ($(CONFIG_PPC_MERGE),y)
>  obj-$(CONFIG_MV64X60)		+= $(mv64x60-y) mv64x60_pic.o mv64x60_dev.o
> +endif
>  
>  # contains only the suspend handler for time
>  obj-$(CONFIG_PM)		+= timer.o

Thanks,

-Dale

Re: [PATCH] ppc: Marvell mv64x60 drivers for ppc platforms fix

[Kumar Gala]

On Jun 24, 2007, at 5:46 PM, Dale Farnsworth wrote:

> In article <200706241552.31483.yur@emcraft.com> you write:
>>  This patch fixes the issue of erroneously involving the
>> arch/powerpc/sysdev/mv64x60*.c files into the compilation
>> process for ppc- (not powerpc-) based platforms (e.g.,
>> P3M750).
>>
>>  Signed-off-by: Yuri Tikhonov <yur@emcraft.com>
>>
>> --
>>
>> diff --git a/arch/powerpc/sysdev/Makefile b/arch/powerpc/sysdev/ 
>> Makefile
>> index c3ce0bd..25eaa14 100644
>> --- a/arch/powerpc/sysdev/Makefile
>> +++ b/arch/powerpc/sysdev/Makefile
>> @@ -17,7 +17,9 @@ obj-$(CONFIG_FSL_PCIE)		+= fsl_pcie.o
>>  obj-$(CONFIG_TSI108_BRIDGE)	+= tsi108_pci.o tsi108_dev.o
>>  obj-$(CONFIG_QUICC_ENGINE)	+= qe_lib/
>>  mv64x60-$(CONFIG_PCI)		+= mv64x60_pci.o
>
> Shouldn't the above line also be placed inside the of following ifeq ?
>
>> +ifeq ($(CONFIG_PPC_MERGE),y)
>>  obj-$(CONFIG_MV64X60)		+= $(mv64x60-y) mv64x60_pic.o mv64x60_dev.o
>> +endif
>>
>>  # contains only the suspend handler for time
>>  obj-$(CONFIG_PM)		+= timer.o

Also, lets just move these into the existing ifeq  
($CONFIG_PPC_MERGE,y) block in that makefile rather than introducing  
a new one.

- k

Re: [PATCH] ppc: Marvell mv64x60 drivers for ppc platforms fix

[Yuri Tikhonov]

 Hi Dale, Kumar,

 Thank you for your comments. All they make sense. Here is an update.

 --

diff --git a/arch/powerpc/sysdev/Makefile b/arch/powerpc/sysdev/Makefile
index c3ce0bd..37c37a9 100644
--- a/arch/powerpc/sysdev/Makefile
+++ b/arch/powerpc/sysdev/Makefile
@@ -16,8 +16,6 @@ obj-$(CONFIG_FSL_SOC)         += fsl_soc.o
 obj-$(CONFIG_FSL_PCIE)         += fsl_pcie.o
 obj-$(CONFIG_TSI108_BRIDGE)    += tsi108_pci.o tsi108_dev.o
 obj-$(CONFIG_QUICC_ENGINE)     += qe_lib/
-mv64x60-$(CONFIG_PCI)          += mv64x60_pci.o
-obj-$(CONFIG_MV64X60)          += $(mv64x60-y) mv64x60_pic.o mv64x60_dev.o

 # contains only the suspend handler for time
 obj-$(CONFIG_PM)               += timer.o
@@ -26,6 +24,8 @@ ifeq ($(CONFIG_PPC_MERGE),y)
 obj-$(CONFIG_PPC_I8259)                += i8259.o
 obj-$(CONFIG_PPC_83xx)         += ipic.o
 obj-$(CONFIG_4xx)              += uic.o
+mv64x60-$(CONFIG_PCI)          += mv64x60_pci.o
+obj-$(CONFIG_MV64X60)          += $(mv64x60-y) mv64x60_pic.o mv64x60_dev.o
 endif

 # Temporary hack until we have migrated to asm-powerpc

Re: [PATCH] ppc: Marvell mv64x60 drivers for ppc platforms fix

[Segher Boessenkool]

> +mv64x60-$(CONFIG_PCI)          += mv64x60_pci.o
> +obj-$(CONFIG_MV64X60)          += $(mv64x60-y) mv64x60_pic.o 
> mv64x60_dev.o

How about

mv64x60-pci-$(CONFIG_PCI)      += mv64x60_pci.o
mv64x60-objs                   := $(mv64x60-pci-y) mv64x60_pic.o 
mv64x60_dev.o
obj-$(CONFIG_MV64X60)          += mv64x60.o

so you can build it is a module, too (although other things
might be preventing that right now)?


Segher

Re: [PATCH] ppc: Marvell mv64x60 drivers for ppc platforms fix

[Yuri Tikhonov]

 Hi Segher,

 The mv64x60 driver has no support for compiling as a module. Built-in only. 
And I think there is no much sense in making the driver which provide you 
with such a base stuff as a boot console, ethernet, etc. to be a LKM.

 Regards, Yuri.

On Monday 25 June 2007 13:11, you wrote:
> > +mv64x60-$(CONFIG_PCI)          += mv64x60_pci.o
> > +obj-$(CONFIG_MV64X60)          += $(mv64x60-y) mv64x60_pic.o
> > mv64x60_dev.o
>
> How about
>
> mv64x60-pci-$(CONFIG_PCI)      += mv64x60_pci.o
> mv64x60-objs                   := $(mv64x60-pci-y) mv64x60_pic.o
> mv64x60_dev.o
> obj-$(CONFIG_MV64X60)          += mv64x60.o
>
> so you can build it is a module, too (although other things
> might be preventing that right now)?
>
>
> Segher

Re: [PATCH] ppc: Marvell mv64x60 drivers for ppc platforms fix

[Segher Boessenkool]

>  The mv64x60 driver has no support for compiling as a module. Built-in 
> only.

Yeah I expected that that is the case right now.  Could be
fixed later though.

> And I think there is no much sense in making the driver which provide 
> you
> with such a base stuff as a boot console, ethernet, etc. to be a LKM.

That's orthogonal to expressing it as one "module" in Kbuild,
which is really just about grouping highly interdependent
stuff together.

But anyway, your patch looks good, I was just suggesting to
clean it up some more.  You don't have to do it now if you
don't want to, I'm sure it will happen later :-)

[Oh, and don't top-post.]

>>> +mv64x60-$(CONFIG_PCI)          += mv64x60_pci.o
>>> +obj-$(CONFIG_MV64X60)          += $(mv64x60-y) mv64x60_pic.o
>>> mv64x60_dev.o
>>
>> How about
>>
>> mv64x60-pci-$(CONFIG_PCI)      += mv64x60_pci.o
>> mv64x60-objs                   := $(mv64x60-pci-y) mv64x60_pic.o
>> mv64x60_dev.o
>> obj-$(CONFIG_MV64X60)          += mv64x60.o
>>
>> so you can build it is a module, too (although other things
>> might be preventing that right now)?


Segher