Path: csiph.com!xmission!news.snarked.org!news.linkpendium.com!news.linkpendium.com!panix!usenet.stanford.edu!not-for-mail
From: Bruce Korb <bruce.korb@gmail.com>
Newsgroups: gnu.utils.bug
Subject: Re: [sharutils] Does not compile with -fno-common
Date: Tue, 28 Jan 2020 18:06:53 -0800
Lines: 27
Approved: bug-gnu-utils@gnu.org
Message-ID: <mailman.1.1580263625.2384.bug-gnu-utils@gnu.org>
References: <dc59915b-005b-12fd-d990-96a483cdb6b4@suse.de> <20200123151904.GA16884@dhcp-0-146.brq.redhat.com> <90678e4f-ff76-213e-f96c-28830197148b@gmail.com>
NNTP-Posting-Host: lists.gnu.org
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
To: bug-gnu-utils@gnu.org
Envelope-to: bug-gnu-utils@gnu.org
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101 Thunderbird/68.2.1
In-Reply-To: <20200123151904.GA16884@dhcp-0-146.brq.redhat.com>
Content-Language: en-US
Precedence: list
Xref: csiph.com gnu.utils.bug:2274

On 1/23/20 7:19 AM, Petr Pisar wrote:
> On Mon, Jan 20, 2020 at 07:47:15PM +0100, Christian Voegl wrote:
>> Trying to build sharutils 4.15.2 with  -fno-common fails with
>>
>>    [   30s]
>>    /usr/lib64/gcc/x86_64-suse-linux/9/../../../../x86_64-suse-linux/bin/ld:
>>    shar-opts.o:(.data.rel.ro.local+0x0): multiple definition of
>>    `program_name'; shar.o:(.rodata+0x10): first defined here
>>    [   30s]
>>    /usr/lib64/gcc/x86_64-suse-linux/9/../../../../x86_64-suse-linux/bin/ld:
>>    ../lib/libgnu.a(md5.o): in function `md5_stream':
>>    [   30s] md5.c:(.text+0x15d): multiple definition of `md5_stream';
>>    shar.o:shar.c:(.text+0x28): first defined here
>>
>>
>> As -fno-common is going to be the default, I think this should be fixed.
>>
> Please try the two patches attached. The first one should fix the program_name
> symbol. The second the md5_stream symbol. I was able to reproduce the first
> one, but not the second one. (Depiste tha fact I do not link to openssl.)
> 
> -- Petr

Thank you guys. I'll queue this up for when I can figure out how to
re-run "autoreconf" once again. Meanwhile, use Petr's patch.