Re: Build regressions/improvements in v3.7-rc5

On 11/12/2012 03:02:46 PM, Geert Uytterhoeven wrote:

On Mon, Nov 12, 2012 at 9:58 PM, Geert Uytterhoeven
[off-list ref] wrote:

quoted

JFYI, when comparing v3.7-rc5 to v3.7-rc4[3], the summaries are:
  - build errors: +14/-4

=20
14 regressions:
  + drivers/virt/fsl_hypervisor.c: error: 'MSR_GS' undeclared (first
use in this function):  =3D> 799:93

These reports would be more useful if the actual .config were provided, =20
rather than just saying "powerpc randconfig".

I doubt this particular one is a regression; it looks like the =20
dependencies have always been too weak (patch sent) and you're just =20
comparing different randomly-generated configs.

powerpc-randconfig
=20
  + error: relocation truncated to fit: R_PPC64_REL24 against symbol
`._mcount' defined in .text section in arch/powerpc/kernel/entry_64.o:
(.text+0x1ff9eb8) =3D> (.text+0x1ffa274), (.text+0x1ff7840)
=20
powerpc-allyesconfig

It's been over 3 years since powerpc-allyesconfig built (it's just too =20
big for the relocation model the kernel uses) -- how is this a =20
regression?

-Scott=