On Tue, Feb 09, 2021 at 02:01:37PM +0100, Greg Kroah-Hartman wrote:

On Tue, Feb 09, 2021 at 01:52:52PM +0100, Pavel Machek wrote:

quoted

Hi!

quoted

From: Arnd Bergmann <arnd@arndb.de>

commit 6e7b64b9dd6d96537d816ea07ec26b7dedd397b9 upstream.

kernel/elfcore.c only contains weak symbols, which triggers a bug with
clang in combination with recordmcount:

  Cannot find symbol for section 2: .text.
  kernel/elfcore.o: failed

Move the empty stubs into linux/elfcore.h as inline functions.  As only
two architectures use these, just use the architecture specific Kconfig
symbols to key off the declaration.

4.4 has this:

config BINFMT_ELF32
        bool
        default y if MIPS32_O32 || MIPS32_N32
                select ELFCORE

in arch/mips. So I believe we'll see problems in that
configuration...?

Hm, did I miss a patch to backport?  This is needed to keep 4.4 building
with newer versions of gcc.  Well, close to building, I'm still haveing
local issues with 10.2.0 and the 4.4.y tree right now...

Not a gcc issue, looks like a binutils issue on my side...

And the builders seem to have worked ok.

thanks,

greg k-h