From: Andreas Rottmann <a.rottmann@gmx.at>
To: ludo@gnu.org (Ludovic Courtès)
Cc: guile-devel@gnu.org
Subject: Re: Build failure (stable-2.0, "make distcheck")
Date: Sun, 06 Jan 2013 01:13:42 +0100 [thread overview]
Message-ID: <877gnopx9i.fsf@delenn.home.rotty.xx.vu> (raw)
In-Reply-To: 87bod7bjhy.fsf@gnu.org
ludo@gnu.org (Ludovic Courtès) writes:
> Andreas Rottmann <a.rottmann@gmx.at> skribis:
>
>> It is the case; guile.info is present, but it seems that it's not newer
>> that one of the sources it's built from:
>>
>> % tar --utc -tvf _build/guile-2.0.7.13-b5d83.tar.gz | grep -E '\.(info|texi)$' | awk '{ print $4 " " $5 " " $6 }' | sort | tail -n2
>> 2012-12-26 13:59 guile-2.0.7.13-b5d83/doc/ref/guile.info
>> 2012-12-26 13:59 guile-2.0.7.13-b5d83/doc/ref/standard-library.texi
>>
>> Probably standard-library.texi is created just a tiny bit earlier than
>> guile.info (in the same second). That might be the reason the build
>> tries to run makeinfo inside the unpacked "dist" tarball?
>
> Can you unpack the tarball and check that hypothesis?
>
It turned out my hypothesis was wrong, the guile.info actually has a
newer timestamp. I've now run a build using "make distcheck --debug=v"
and attached the log for future reference. It seems that problem is the
following:
,----
| Prerequisite `effective-version.texi' is newer than target `../../../doc/ref/guile.info'.
| Must remake target `../../../doc/ref/guile.info'.
| make[5]: Entering directory `/home/rotty/src/guile/_build/guile-2.0.7.12-c23fb/_build/doc/ref'
| [... actual failure happens here ... ]
`----
While "doc/ref/effective-version.texi", as contained in the tarball, is
indeed older than guile.info (A), it seems a build-directory version of
"effective-version.texi" is generated by configure (B, C), and that copy
is then used by make for calculating whether guile.info needs to be
(re-)built, as indicated by the above log snippet.
The below commands are executed inside the extracted copy of the
tarball, in which the build fails
(i.e. /home/rotty/src/guile/_build/guile-2.0.7.12-c23fb/):
(A)
% ls --full-time doc/ref/guile.info doc/ref/effective-version.texi
-r--r--r-- 1 rotty rotty 27 2012-12-26 17:35:08.000000000 +0100 doc/ref/effective-version.texi
-r--r--r-- 1 rotty rotty 29821 2012-12-26 17:35:16.000000000 +0100 doc/ref/guile.info
(B)
% ls --full-time _build/doc/ref/effective-version.texi
-rw-r--r-- 1 rotty rotty 27 2013-01-05 16:32:37.150982801 +0100 _build/doc/ref/effective-version.texi
(C)
,----
| ...
| config.status: creating doc/ref/effective-version.texi
| ...
`----
At least that's what I read out of the logs. I just wonder why nobody
else seems to have run into this issue, which now looks like a genuine
buildsystem bug to me. Thoughts?
Kind Regards, Rotty
--
Andreas Rottmann -- <http://rotty.xx.vu/>
prev parent reply other threads:[~2013-01-06 0:13 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-12-26 17:34 Build failure (stable-2.0, "make distcheck") Andreas Rottmann
2013-01-02 11:35 ` Ludovic Courtès
2013-01-02 12:07 ` Andreas Rottmann
2013-01-02 17:08 ` Ludovic Courtès
2013-01-06 0:13 ` Andreas Rottmann [this message]
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
List information: https://www.gnu.org/software/guile/
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=877gnopx9i.fsf@delenn.home.rotty.xx.vu \
--to=a.rottmann@gmx.at \
--cc=guile-devel@gnu.org \
--cc=ludo@gnu.org \
/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).