Re: GAS patch for sh*-unknown-linux-gnu

[NIIBE Yutaka <gniibe at m17n dot org>]

Hans-Peter Nilsson wrote:
 > I'm not really sure about using "sh*eb-*-linux*" to denote a
 > GNU/Linux system using big endian code and data.  Is this triple
 > new or has it been used somewhere else?  I'm not sure I can
 > approve it if it's new.  Do we need it; do you know of any
 > GNU/Linux big endian variant?  It's not used in e.g. bfd.  Is
 > there an existing (non-SH-based) port where sh* would collide?
 > Can Ben Elliston, the config.* maintainer, shed some light?

It's not new (for me :-).  For GNU/Linux on SuperH Project
(http://www.m17n.org/linux-sh/), there're for targets:

	sh4-unknown-linux-gnu
	sh4eb-unknown-linux-gnu
	sh3-unknown-linux-gnu
	sh3eb-unknown-linux-gnu

Note that the kernel (uname on native machine) returns sh3, sh3eb,
sh4, or sh4eb.  IIRC, it was discussed around April 2000, and
implemented (in kernel) July 2000.  Before that, we only had cross
toolchain, and didn't care about native toolchain.  The kernel
returned "sh" for all of four targets.  We found the probelm when
we became to have native toolchain.

I've already send a changes to Ben, and all support of the four
targets is now included in config.{sub,guess}.  Thanks Ben for
maintaining them.

(I didn't know the maintainance of config.{sub,guess}, so it took time
since kernel change to toolchain change.

I think I have confgigury change for ld on native GNU/Linux system on
SuperH too, I will send later (after document addition of gas).

I think that the definition of CPU name is determined by kernel.
While I think "sh4el" could be candidate, kernel already has been use
the definition above for a year or so.
--