Re: [Suggestion] ARM:S5pv210: compiling issue for s5pv210 by using randconfig

[Arnd Bergmann <arnd@arndb.de>]

On Wednesday 17 April 2013, Will Deacon wrote:
> > arm-linux-gnu-gcc -Wp,-MD,arch/arm/vfp/.vfphw.o.d  -nostdinc -isystem /usr/lib/gcc/arm-linux-gnueabi/4.7.1/include -I/root/linux-next/arch/arm/include -Iarch/arm/include/generated  -Iinclude -I/root/linux-next/arch/arm/include/uapi -Iarch/arm/include/generated/uapi -I/root/linux-next/include/uapi -Iinclude/generated/uapi -include /root/linux-next/include/linux/kconfig.h -D__KERNEL__ -mlittle-endian -Iarch/arm/mach-s5pv210/include -Iarch/arm/plat-samsung/include  -D__ASSEMBLY__ -mabi=apcs-gnu -mno-thumb-interwork -marm -D__LINUX_ARM_ARCH__=4 -march=armv4t -mtune=arm9tdmi -include asm/unified.h -Wa,-mfpu=softvfp+vfp -mfloat-abi=soft        -c -o arch/arm/vfp/vfphw.o arch/arm/vfp/vfphw.S
> > 
> > 
> >   compiling err:
> > 
> > arch/arm/vfp/vfphw.S: Assembler messages:
> > arch/arm/vfp/vfphw.S:295: Error: selected processor does not support ARM mode `mrrc p11,3,r0,r1,c0'
> > arch/arm/vfp/vfphw.S:295: Error: selected processor does not support ARM mode `mrrc p11,3,r0,r1,c1'
> 
> The problem here is that you've ended up targetting a platform (s5pv210)
> that selects CPU_V7. VFP is then subsequently selected, but CONFIG_MMU=n, so
> 7TDMI and 9TDMI (v4 CPUs, no VFP) are selectable. Selecting either of those,
> causes these warnings.
> 
> Unfortunately, I'm not sure how best to fix this. Most of the !MMU CPUs are
> tied to a particular board (lots of `if ARCH_INTEGRATOR' predicates), but we
> don't want to do that for 7tdmi.
> 
> If we could enforce the strict exclusion of {<= ARMv5} and {ARMv6+} in the
> Kconfig, that would solve your problem.

I have not tried to get no-MMU kernels to build in general. I think the way
it should be done is to not offer any user-selectable CPU types at all but
always select the CPU from the board.

For randconfig tests, I would recommend turning on CONFIG_MMU unconditionally
using an appropriate KCONFIG_ALLCONFIG= file.

The alternative is to use the patch below, but it may be incomplete: I could
not find anything other than AT91x40 in the kernel that actually has an
ARM7TDMI or ARM9TDMI.

	Arnd

diff --git a/arch/arm/mach-at91/Kconfig.non_dt b/arch/arm/mach-at91/Kconfig.non_dt
index 6c24985..dc972e1 100644
--- a/arch/arm/mach-at91/Kconfig.non_dt
+++ b/arch/arm/mach-at91/Kconfig.non_dt
@@ -47,6 +47,7 @@ config ARCH_AT91X40
 	select ARCH_USES_GETTIMEOFFSET
 	select MULTI_IRQ_HANDLER
 	select SPARSE_IRQ
+	select CPU_ARM7TDMI
 
 endchoice
 
diff --git a/arch/arm/mm/Kconfig b/arch/arm/mm/Kconfig
index 84af266..c5e4ef0 100644
--- a/arch/arm/mm/Kconfig
+++ b/arch/arm/mm/Kconfig
@@ -6,7 +6,6 @@ comment "Processor Type"
 
 # ARM7TDMI
 config CPU_ARM7TDMI
-	bool "Support ARM7TDMI processor"
 	depends on !MMU
 	select CPU_32v4T
 	select CPU_ABRT_LV4T
@@ -56,7 +55,6 @@ config CPU_ARM740T
 
 # ARM9TDMI
 config CPU_ARM9TDMI
-	bool "Support ARM9TDMI processor"
 	depends on !MMU
 	select CPU_32v4T
 	select CPU_ABRT_NOMMU