Mailing-List: contact dev-help@stdcxx.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@stdcxx.apache.org
Received-SPF: pass (athena.apache.org: domain of msebor@gmail.com designates
 209.85.200.169 as permitted sender)
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=message-id:date:from:organization:user-agent:mime-version:to:subject:references:in-reply-to:content-type:content-transfer-encoding:sender;
        b=sBR91558EFoqcxebuagTEXzFGwd81Kn3PYUtX7Z2s3zbjI8gjs+u1mF8dpYPzDjLZV4vqq9GEybPN7680sD3pIrGMcGa/rpKjWMvP9UuPNh2XOdbE/azZakoWRIqvurcu6XQfD8qmGuq2KqPBwupAPFyyMzF2px8pwzNg6S8eyE=
Message-ID: <482E0D7F.6060203@roguewave.com>
Date: Fri, 16 May 2008 16:41:03 -0600
From: Martin Sebor <sebor@roguewave.com>
Organization: Rogue Wave Software, Inc.
User-Agent: Thunderbird 2.0.0.14 (X11/20080501)
MIME-Version: 1.0
To: dev@stdcxx.apache.org
Subject: Re: XLC++ build failures
References: <482B8461.9090402@roguewave.com>
 <CFFDD219128FD94FB4F92B99F52D0A49010A37BE@exchmail01.Blue.Roguewave.Com>
 <482C6A92.8090103@roguewave.com>
 <CFFDD219128FD94FB4F92B99F52D0A49010A37C2@exchmail01.Blue.Roguewave.Com>
 <482C943F.7080202@roguewave.com>
 <CFFDD219128FD94FB4F92B99F52D0A499AA855@exchmail01.Blue.Roguewave.Com>
In-Reply-To: 
 <CFFDD219128FD94FB4F92B99F52D0A499AA855@exchmail01.Blue.Roguewave.Com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Sender: Martin Sebor <msebor@gmail.com>

Travis Vitek wrote:
>  
> 
>> Martin Sebor wrote:
>>> Eric Lemings wrote:
>>>
>>>> Martin Sebor wrote:
>>>>
>>>> The page is fine (the presentation could be better), but there
>>>> is a problem with all the AIX builds. They've all been failing
>>>> due to undefined library symbols. I've been assuming it had to
>>>> do with all the recent infrastructure screwups that Scott noted
>>>> the other day but upon closer inspection I noticed that we're
>>>> using -qtempinc when we should be using -qtemplateregistry. I
>>>> wonder if this change might be responsible:
>>>>    http://svn.apache.org/viewvc?view=rev&revision=652954
>>> Uh yeah, quite possible.  The build directory, but not necessarily
>>> the source directory, by definition has to be writable.  I assume
>>> that's where configuration tests are built and executed also,
>>> unless I'm missing something.
>> Most but not all of them. The change above is to code that runs
>> with CWD=$TOPDIR. My local builds run fine so I'm puzzled as to
>> why it's failing. Travis, any ideas?
>>
> 
> No, I've got nothing. I noticed the failures on Wednesday. I also did a
> manual test at that time and had no problems. I was hoping that it was
> some kind of glitch in the nightly build system.

I must have missed the error below in the logs. I see what
the problem is now: a missing '$' in the fix I committed.
It works when TMPDIR is not defined, which it isn't in the
default environment, but our bin/buildntest script sets it
to /tmp/stdcxx-tmp.<PID> and that's when things break. I
just checked in a fix for it:
     http://svn.apache.org/viewvc?rev=657235&view=rev

### gmake  builddir   BUILDMODE=,archive,optimized,narrow 
CONFIG=vacpp.config 2>&1 | sed -e 
"s/\/build2\/batman\/5.0.0\/builds\/34761602\/source-buildspace\/build/\$(BUILDDIR)/g;s/\/build2\/batman\/5.0.0\/builds\/34761602\/source-buildspace/\$(TOPDIR)/g" 
| tee /tmp/build.1016008.log:
/bin/sh: /tmp/stdcxx-tmp.1016008/tmp:  not found
rm: foo.1138728.c: No such file or directory
creating BUILDDIR=$(BUILDDIR)
generating $(BUILDDIR)/makefile.in from $(TOPDIR)/etc/config/vacpp.config
/bin/sh: /tmp/stdcxx-tmp.1016008/tmp:  not found
rm: foo.1138762.c: No such file or directory
build directory created: BUILDDIR=$(BUILDDIR)

Martin