Re: [PATCH 1/3] arch-x86: Add x86-x32 to MACHINEOVERRIDES

[Richard Purdie <richard.purdie@linuxfoundation.org>]

On Tue, 2017-09-19 at 11:27 -0700, Saul Wold wrote:
> This is needed as an x32 more generic x32 override later in the
> OVERRIDES, currently linux-gnux32 is the first override, but we
> need a stronger (later in the list) x32 override to deal with some
> needed x32 dependency overrides.
> 
> Signed-off-by: Saul Wold <sgw@linux.intel.com>
> ---
>  meta/conf/machine/include/x86/arch-x86.inc | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/meta/conf/machine/include/x86/arch-x86.inc b/meta/conf/machine/include/x86/arch-x86.inc
> index e51d595f74..31d30b3304 100644
> --- a/meta/conf/machine/include/x86/arch-x86.inc
> +++ b/meta/conf/machine/include/x86/arch-x86.inc
> @@ -26,6 +26,7 @@ TUNE_LDARGS += "${@bb.utils.contains('TUNE_FEATURES', 'mx32', '-m elf32_x86_64',
>  TUNE_ASARGS += "${@bb.utils.contains('TUNE_FEATURES', 'mx32', '-x32', '', d)}"
>  # user mode qemu doesn't support x32
>  MACHINE_FEATURES_BACKFILL_CONSIDERED_append = " ${@bb.utils.contains('TUNE_FEATURES', 'mx32', 'qemu-usermode', '', d)}"
> +MACHINEOVERRIDES =. "${@bb.utils.contains('TUNE_FEATURES', 'mx32', 'x86_x32:', '' ,d)}"
>  
>  # ELF64 ABI
>  TUNEVALID[m64] = "IA32e (x86_64) ELF64 standard ABI"

I was ok with this until I realised the patch does not do what it says
in the commit message, it adds "x86_x32", not "x86-x32". Since "_" is
the override modifier, I worry about how this reacts with the rest of
the system and I suspect its a bad idea. Is there a reason you didn't
use "x86-x32" (following the example of x86-64)?

Cheers,

Richard