This is the mail archive of the
mailing list for the GCC project.
Re: final install problems
- To: David Edelsohn <dje at watson dot ibm dot com>
- Subject: Re: final install problems
- From: Zack Weinberg <zack at codesourcery dot com>
- Date: Fri, 2 Nov 2001 14:04:37 -0800
- Cc: Joe Buck <jbuck at synopsys dot com>, Phil Edwards <pedwards at disaster dot jaj dot com>,David <david_fly113 at yahoo dot com>, gcc at gcc dot gnu dot org
- References: <firstname.lastname@example.org> <200111021934.LAA19167@atrus.synopsys.com> <200111022137.QAA26104@makai.watson.ibm.com>
On Fri, Nov 02, 2001 at 04:36:54PM -0500, David Edelsohn wrote:
> >>>>> Joe Buck writes:
> Joe> 1) if a known-good system, see if install exists in the standard
> Joe> place (/usr/ucb/install for Solaris and AIX, /usr/bin/install for
> Joe> BSD or GNU/Linux), if so assume it's OK. The full path should be
> Joe> used, to avoid using the wrong install.
> AIX /usr/ucb/install does meet GNU install requirements. /usr/ucb
> is not good enough.
I was thinking along the lines of "for each install found in the user's
PATH, test to see if it works and understands the command line arguments
we need." More reliable, and more in the spirit of autoconf.
(Do we use any switches other than -c and -m?)