'msoft-float' on 'powerpc-elf' tool chain

[Rohit Arul Raj]

Hi,

Looking at the multi-lib directories of ‘powerpc-elf’ tool chain,
there is no ‘nof’ subfolder.

$ powerpc-elf-gcc.exe --print-multi-lib
.;@mno-eabi at mstrict-align
64;@m64 at mno-eabi@mstrict-align

Does this mean ‘-msoft-float’ libraries are not available for 32-bit
as well as 64-bit targets?

Looking at other tool chains,

$ ./powerpc-linux-gnu-gcc.exe --print-multi-lib
.;@fPIC at mstrict-align
te500v1;@te500v1 at fPIC@mstrict-align
te500v2;@te500v2 at fPIC@mstrict-align
te600;@te600 at fPIC@mstrict-align
te500mc;@te500mc at fPIC@mstrict-align
64;@m64 at fPIC@mstrict-align
nof;@msoft-float at fPIC@mstrict-align

This tool chain too supports soft-float only for the target processor
603 (e.g. e300c2 core) that gets selected by default.
For e500mc and other 32-bit targets, soft-float libraries are not provided.

Same with 'powerpc-eabi’ tool chains, soft-float libraries provided
only for the target processor 603 (e.g. e300c2 core).

$ powerpc-eabi-gcc.exe --print-multi-lib
.;@mno-eabi at mstrict-align
te500v1;@te500v1 at mno-eabi@mstrict-align
te500v2;@te500v2 at mno-eabi@mstrict-align
te600;@te600 at mno-eabi@mstrict-align
nof;@msoft-float at mno-eabi@mstrict-align

Assuming that there is no ABI definition for soft-float target under
64bit, any reason why 32-bit targets like e500mc/e500v1/e500v2 doesn't
have 'soft-float' libraries?

Regards,
Rohit