Re: Build failure -- powerpc/boot: Add OPAL console to epapr wrappers

On 03/28/2017 02:35 AM, Michael Ellerman wrote:

Daniel Walker [off-list ref] writes:

quoted

On 03/24/2017 07:16 PM, Oliver O'Halloran wrote:

quoted

On Sat, Mar 25, 2017 at 4:00 AM, Daniel Walker [off-list ref] wrote:

quoted

I get this build failure,


In file included from arch/powerpc/boot/fdt.c:51:
../arch/powerpc/boot/libfdt_env.h:9: error: redefinition of typedef
'uint32_t'
../arch/powerpc/boot/types.h:20: note: previous declaration of 'uint32_t'
was here
../arch/powerpc/boot/libfdt_env.h:10: error: redefinition of typedef
'uint64_t'
../arch/powerpc/boot/types.h:21: note: previous declaration of 'uint64_t'
was here
make[2]: *** [arch/powerpc/boot/fdt.o] Error 1
make[1]: *** [uImage] Error 2
make[1]: Leaving directory `/nobackup/danielwa/linux/t1040'
make: *** [sub-make] Error 2


and it bisects to ,


commit 656ad58ef19e2a763fa5c938b20ae0f6b8d67242
Author: Oliver O'Halloran [off-list ref]
Date:   Fri Jul 1 00:34:37 2016 +1000

      powerpc/boot: Add OPAL console to epapr wrappers

      This patch adds an OPAL console backend to the powerpc boot wrapper so
      that decompression failures inside the wrapper can be reported to the
      user. This is important since it typically indicates data corruption in
      the firmware and other nasty things.

      Currently this only works when building a little endian kernel. When
      compiling a 64 bit BE kernel the wrapper is always build 32 bit to be
      compatible with some 32 bit firmwares. BE support will be added at a
      later date. Another limitation of this is that only the "raw" type of
      OPAL console is supported, however machines that provide a hvsi console
      also provide a raw console so this is not an issue in practice.

      Actually-written-by: Benjamin Herrenschmidt [off-list ref]
      Signed-off-by: Oliver O'Halloran [off-list ref]
      [mpe: Move #ifdef __powerpc64__ to avoid warnings on 32-bit]
      Signed-off-by: Michael Ellerman [off-list ref]


I can provide a config file if needed. My apologies if this was already
reported.

Thanks for the report, I don't think this is a known bug. mpe's build
testing is pretty thorough so I'm surprised this wasn't caught sooner.

A config file and the version of gcc that you're using would be useful.

Config attached , it's for a Fresecale t1042 machine. The GCC is custom
based on 4.4.1 .

o_O

So only an 8 year old compiler! :D

I can't reproduce here with 4.6.3 sorry, which is the oldest I have handy.

Can you debug it a bit further at your end? I assume your toolchain is
built with libc?

cheers

It appears to be fairly straight forward,


In file included from arch/powerpc/boot/fdt.c:51:
../arch/powerpc/boot/libfdt_env.h:9: error: redefinition of typedef 'uint32_t'
../arch/powerpc/boot/types.h:20: note: previous declaration of 'uint32_t' was here
../arch/powerpc/boot/libfdt_env.h:10: error: redefinition of typedef 'uint64_t'
../arch/powerpc/boot/types.h:21: note: previous declaration of 'uint64_t' was here


So in types.h you added a couple new types "unit32_t" and "uint64_t" , but alas libfdt_env.h is already defining those and includes types.h . So now libfdt_env.h ends up having two types with the same name. I suppose it's possible that newer compilers ignore this issue if the types are identical which in this case they are.

To fix this I removed the typedefs from libfdt_env.h , if you make a patch like that I can test it.

Daniel