[OpenAFS] Need SYS_NAME_ID_i386_nbsd_16

Derrick J Brashear shadow@dementia.org
Mon, 9 Feb 2004 16:06:18 -0500 (EST)


On Mon, 9 Feb 2004, John Hascall wrote:

>
> > of course if you look at the cvs head, 2600 is used. you're not doing
> > something dumb like porting 1.2 are you?
>
> Yes.  I tried openafs-1.3.52 and openafs-snap-2004-01-23
> both of which seem not at all ready for prime time...
>
> For example, snap-2004-01-23 (apparently the latest)
> didn't have a configure, and trying to make one was
> not good:

./regen.sh at the top level. Didn't read README.DEVEL, did you?

> rhovanion: {559} autoconf
> configure.in:2: error: possibly undefined macro: AM_INIT_AUTOMAKE
>       If this token and others are legitimate, please use m4_pattern_allow.
>       See the Autoconf documentation.

See above.

> 1.3.52 had a configure script at least, but the make takes a crap early:
>
>    ...
> cd src && cd pinstall && make all
> cc  -O -I/home/john/openafs-1.3.52/src/config -I. -I. -I/home/john/openafs-1.3.52/include -I/home/john/openafs-1.3.52/include/afs -I/home/john/openafs-1.3.52/include/rx -I/home/john/openafs-1.3.52 -I/home/john/openafs-1.3.52/src -I/home/john/openafs-1.3.52/src -O2 -pipe ./install.c -o pinstall
> install.c: In function `quickStrip':
> install.c:446: error: structure has no member named `a_magic'
> *** Error code 1

and fixing problems you see with a new port is entirely unexpected.

> So I figured I might as well work with 1.2.11 because that's what we're
> using elsewhere in our cell.  At least autoconf worked in that tree.

Patches won't apply to the head, when you're done.