Please review: JDK-8011348 "use of which in common/autoconf/autogen.sh is not portable"
David Katleman
david.katleman at oracle.com
Thu Apr 4 02:14:27 UTC 2013
On 4/2/2013 7:42 PM, Tim Bell wrote:
> All -
>
> I regret the fact that I did not test the fix for 8009988 (see below)
> when I reviewed it. The script as written there assumes that the
> 'which' utility will return nothing if the command is not found. On
> some platforms (Solaris and Windows/Cygwin to name two), the return is
> more like 'no autoconf-2.67 in $PATH'
>
> Please review the fix for 8011348, which restores autogen.sh to
> platform independence.
> 46 AUTOCONF=$(which autoconf 2> /dev/null);
> 47 if test $? -ne 0; then
> 48 unset AUTOCONF
> 49 fi
The above still won't work on older Solaris releases, most S10 and earlier.
As both success and failure will result in $? being 0, if means AUTOCONF
will be set to either
/PATH/autoconf
or
no autoconf in AAA BBB CCC
I would borrow what we just came up with for hgforest.sh:
AUTOCONF="`which autoconf 2> /dev/null | grep -v '^no autoconf in'`"
You don't need the trailing test check then.
Thanks
Dave
>
> The bug report should be visible soon at:
> http://bugs.sun.com/view_bug.do?bug_id=8011348
>
> Webrev is here:
> http://cr.openjdk.java.net/~tbell/8011348/webrev.00/
>
> Thanks in advance for any reviews. Feedback welcome-
>
> Tim
>
>
> -------- Original Message --------
> Subject: hg: jdk8/build: 8009988: build-infra: Fix configure
> output for zip debuginfo check
> Date: Tue, 02 Apr 2013 12:59:43 +0000
> From: ahughes at redhat.com
> To: jdk8-changes at openjdk.java.net, build-dev at openjdk.java.net
>
>
>
> Changeset: 15c1642967c9
> Author: andrew
> Date: 2013-04-02 13:59 +0100
> URL: http://hg.openjdk.java.net/jdk8/build/rev/15c1642967c9
>
> 8009988: build-infra: Fix configure output for zip debuginfo check
> Summary: No output from zip debuginfo option when default is used.
> Reviewed-by: tbell
>
> ! common/autoconf/autogen.sh
> ! common/autoconf/generated-configure.sh
> ! common/autoconf/jdk-options.m4
>
>
>
>
More information about the build-dev
mailing list