This is the mail archive of the
binutils@sources.redhat.com
mailing list for the binutils project.
Re: GAS patch for sh*-unknown-linux-gnu
- To: NIIBE Yutaka <gniibe at m17n dot org>
- Subject: Re: GAS patch for sh*-unknown-linux-gnu
- From: Ralf Corsepius <corsepiu at faw dot uni-ulm dot de>
- Date: 03 Oct 2001 13:22:19 +0200
- Cc: binutils at sources dot redhat dot com
- References: <200110030934.f939YfR13613@mule.m17n.org>
Am Mit, 2001-10-03 um 11:34 Uhr schrieb NIIBE Yutaka:
> Hi there,
>
> Here's a patch for gas for the target sh*-unknown-linux-gnu.
>
> It does some clean-up for endian handling. Removing the variable
> shl and use target_big_endian directly, let configure define
> TARGET_BYTES_BIG_ENDIAN.
>
> Besides, we'd like to set default as little endian for
> sh-unknown-linux-gnu, because it's more populer for SH-3 and SH-4.
>
> While only little endian is supported for sh-*-pe, we need
> support of big endian, thus, the option -big.
Well, I am not sure if this patch is a good idea.
IMHO, it introduces compatibility problems and complications at doubtful
benefits:
* backward compatiblity problems. Users are used to sh-binutils to
default to big endianness, all gcc version up to now have pre-supposed
using bigendian as default and packages might be relying on it.
* -big would be redundant to -little and actually is of little use for
all sh users but linux-gnu users.
Further, I don't understand the need for this patch. There are other
means to provide "convenience-defaults" to users (rpm, gcc/specs, etc.).
Additionally, I also do not understand why a new variable
"target_big_endian" shall be used instead of "shl". AFAIU, it breaks
sourcecode backward compatibility at no real functional benefits.
Ralf