Catherine Moore <clm@codesourcery.com> writes:
This patch adds the -mips16e option to gas enabling backwards
compatibility with the SDE toolchain.
FWIW, we did discuss this with MTI a few years back and "agreed"
(as I remember it ;-)) that the -mips16e option wasn't really
worthwhile, and that FSF compatibility wasn't necessary.
-mips16 is effectively a mode selection that flips between 16-bit
and 32-bit instruction encodings, and as such there isn't any
need for a separate option to say that the MIPS16 encoding on
certain architectures supports more instructions than on other
architectures. There's no non-MIPS32 or non-MIPS64 archicture
that supports the full MIPS16e encoding and no MIPS32 or MIPS64
architecture that supports only the "old" MIPS16 encoding.
Adding -mips16e risks confusing people about what -mips16
actually means.