Re: host endianness

[Iain Hibbert <plunky%rya-online.net@localhost>]

Well, in summary, you guys are pretty clever.

On Mon, 8 Feb 2010, Izumi Tsutsui wrote:
> With nbtool_config.h like src/sbin/disklabel/bswap.h?
>
> ---
> #include <sys/types.h>
>
> #if HAVE_NBTOOL_CONFIG_H
> #ifndef BYTE_ORDER
> #ifdef WORDS_BIGENDIAN
> #define BYTE_ORDER            BIG_ENDIAN
> #else
> #define BYTE_ORDER            LITTLE_ENDIAN
> #endif
> #endif
> #endif
>
> #define HOST_BYTE_ORDER               BYTE_ORDER

This however, requires modifying the pcc dist for a NetBSD-ism, which I'm
reluctant to do (the native pcc build just uses GNU configure to provide
this definition for each use case)

On Mon, 8 Feb 2010, Christos Zoulas wrote:
> compile and run:
>
> #include <stdio.h>
> union {
>         int x;
>         char c[4];
> } z = {
>         .x = ('4' << 24) | ('3' << 16) | ('2' << 8) | '1'
> };
>
> int
> main(void)
> {
>         printf("%4.4s\n", z.c);
>         return 0;
> }

which makes some sense but as noted, it does require building and running
a host program in a strange place (that Makefile.inc will be included
several times)

Then, various contributors come up with:
> printf "%s" "C!" | od -t x2 | sed 's/[^1234]*//;s/2143/4321/p;d'

which Alan Barrett doesn't particularly like:
> src/tools/compat already has a configure script that sets BYTE_ORDER,
> and I'd prefer it if we could use that definition, or add another mini C
> program to the configure script, rather than relying on yet another host
> tool.

and I wonder if it would be simpler to use my original idea just extended
with a quick copy/paste from bsd.own.mk to guarantee we know the host
architecture, like so?

Index: bsd.endian.mk
===================================================================
RCS file: /cvsroot/src/share/mk/bsd.endian.mk,v
retrieving revision 1.15
diff -u -r1.15 bsd.endian.mk
--- bsd.endian.mk       9 Jan 2008 11:26:14 -0000       1.15
+++ bsd.endian.mk       10 Feb 2010 09:13:21 -0000
@@ -24,4 +24,31 @@
 TARGET_ENDIANNESS=     4321
 .endif

+# For _HOST_ARCH, if uname -p fails, or prints "unknown", or prints
+# something that does not look like an identifier, then use uname -m.
+_HOST_ARCH!=   uname -p 2>/dev/null
+_HOST_ARCH:=   ${HOST_ARCH:tW:C/.*[^-_A-Za-z0-9].*//:S/unknown//}
+.if empty(_HOST_ARCH)
+_HOST_ARCH!=   uname -m
+.endif
+
+.if ${_HOST_ARCH} == "alpha" || \
+    ${_HOST_ARCH} == "arm" || \
+    ${_HOST_ARCH} == "i386" || \
+    ${_HOST_ARCH} == "ia64" || \
+    ${_HOST_ARCH} == "vax" || \
+    ${_HOST_ARCH} == "x86_64" || \
+    ${_HOST_ARCH:C/^.*el$/el/} == "el"
+HOST_ENDIANNESS=       1234
+.elif ${_HOST_ARCH} == "hppa" || \
+      ${_HOST_ARCH} == "m68000" || \
+      ${_HOST_ARCH} == "m68k" || \
+      ${_HOST_ARCH} == "powerpc" || \
+      ${_HOST_ARCH} == "powerpc64" || \
+      ${_HOST_ARCH} == "sparc" || \
+      ${_HOST_ARCH} == "sparc64" || \
+      ${_HOST_ARCH:C/^.*eb$/eb/} == "eb"
+HOST_ENDIANNESS=       4321
+.endif
+
 .endif  # !defined(_BSD_ENDIAN_MK_)

regards,
iain