This is the mail archive of the
binutils@sourceware.org
mailing list for the binutils project.
Re: PATCH: PR binutils/12283: bfd/doc doesn't support parallel build
- From: Steve Ellcey <sje at cup dot hp dot com>
- To: Ralf Wildenhues <Ralf dot Wildenhues at gmx dot de>
- Cc: binutils at sourceware dot org, "H.J. Lu" <hjl dot tools at gmail dot com>
- Date: Mon, 31 Jan 2011 10:33:01 -0800
- Subject: Re: PATCH: PR binutils/12283: bfd/doc doesn't support parallel build
- References: <AANLkTikAd7jWXN+2bi-OeR_-cQq7ivpSyVz+4165trVS@mail.gmail.com> <201101282332.p0SNWFT04949@lucas.cup.hp.com> <20110129094232.GD11288@gmx.de>
- Reply-to: sje at cup dot hp dot com
On Sat, 2011-01-29 at 10:42 +0100, Ralf Wildenhues wrote:
> * Steve Ellcey wrote on Sat, Jan 29, 2011 at 12:32:15AM CET:
> > This patch is intended to be a secondary fix/workaround for the problem
> > that can happen when building bfd/doc using a parallel make. Ralf
> > Wildenhues has a fix to move-if-changed that should fix the problem for
> > any target where compiling a file multiple times will always result in
> > exactly the same binary, but I have at least one platform where this is
> > not the case: hppa*-*-hpux*. So I would also like to have this fix as a
> > backup to Ralf's patch (which I think should also go in).
>
> Note that with the changed move-if-change from gnulib, then instead of
> your patch, you could just use
> export CMPPROG=cmp --ignore-initial=16'
> make -jN
>
> for a successful build.
>
> Cheers,
> Ralf
I tried that method but it did not work. When I use 'cmp
--ignore-initial=16' on object files (.o's) it works fine, but when I
use it to compare the chew binary I get a difference beyond the first 16
bytes (at byte 111 or 112 depending on the binary).
Steve Ellcey
sje@cup.hp.com