public inbox for gcc@gcc.gnu.org
 help / color / mirror / Atom feed
From: Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
To: law@cygnus.com
Cc: egcs@cygnus.com
Subject: Re: install stuff
Date: Mon, 01 Dec 1997 07:38:00 -0000	[thread overview]
Message-ID: <199712011518.HAA21890@cygnus.com> (raw)
In-Reply-To: <10137.880986874@hurl.cygnus.com>

At 15:34 01.12.97 , Jeffrey A Law wrote:
>
>  In message <199712010928.BAA13558@cygnus.com>you write:
>  > configure:
>  > Make sure cc (gcc is _not_ enough, eventually make a symlink from gcc to
>  > cc) is in your PATH or start configure with "CC=/path/to/your/ccompiler
>  > configure", otherwise config.guess will guess wrong for some native
>  > platforms (notably alpha and powerpc), but configure will success
cause it
>  > finds gcc later.
>Can you have examples of problems this causes?  I've had no problems with
>this on the alpha (either osf or linux).

Ok, eg. for powerpc-unknown-linux-gnulibc1:
- configure calls config.guess before determing the cc to use
- config.guess defaults to ${CC-cc} for the compiler
- if CC is not set and cc is not in the path (gcc only system, egcs also
does not install any symlink from gcc to cc), all compile attempts in
config.guess will fail and some default values will be used.

Examples:
- powerpc-unknown-linux-gnu <-> powerpc-unknown-linux-gnulibc1
- alphaev5 <-> alphaev56 (if I remember correctly)
- and every other check in config.guess that relies on ${CC-cc}

Simple solution:
check for gcc/cc/${CC} in configure before calling config.guess and bail
out if nothing valid is found.

This reminds me of a post-release wish:
- support for install_root= (like glibc) during make install to ease
package building

>  > powerpc-*-linux-gnu*:
>  > You will need binutils-2.8.1.0.17 available from
>  > < ftp://ftp.yggdrasil.com/private/hjl > for a working egcs. It is strongly
>  > recommended to recompile binutils with egcs if you initially built it
with
>  > gcc-2.7.2.*.
>Is this really correct?? -- I know Meissner's been working with a ppc/linux
>box, but I don't know if he's using the new binutils.  Exactly what problems
>does this binutils fix?

Hmm, 2.8.1.0.15 is definitely to old. .17 includes gas-971105 + some
additional patches for Linux + some patches for Linux/PPC from Geoff, which
have not been accepted/included (?) in the gas snapshots in /private/gas.

Bye,
Franz.


  reply	other threads:[~1997-12-01  7:38 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <199712010928.BAA13558@cygnus.com>
1997-12-01  7:38 ` Jeffrey A Law
1997-12-01  7:38   ` Franz Sirl [this message]
1997-12-01 11:19 meissner
1997-12-01 11:32 ` Joern Rennecke
     [not found] <199712011518.HAA21889@cygnus.com>
1997-12-01 11:08 ` Jeffrey A Law
1997-12-01 11:33   ` Franz Sirl
  -- strict thread matches above, loose matches on Subject: below --
1997-12-01  9:03 meissner
1997-12-01 10:45 ` The Unknown User-ID
1997-11-30 13:50 Jeffrey A Law
1997-12-01  1:28 ` Franz Sirl

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=199712011518.HAA21890@cygnus.com \
    --to=franz.sirl-kernel@lauterbach.com \
    --cc=egcs@cygnus.com \
    --cc=law@cygnus.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).