From: Jan Kiszka <jan.kiszka@siemens.com>
To: Juan Quintela <quintela@trasno.org>
Cc: qemu-devel@nongnu.org, Nathan Froyd <froydnj@codesourcery.com>
Subject: [Qemu-devel] Re: [PATCH] configure: fix breakage of --extra-{cflags, ldflags}
Date: Thu, 23 Jul 2009 18:32:06 +0200 [thread overview]
Message-ID: <4A689086.6080705@siemens.com> (raw)
In-Reply-To: <m33a8n5slq.fsf@neno.mitica>
Juan Quintela wrote:
> Nathan Froyd <froydnj@codesourcery.com> wrote:
>> On Thu, Jul 23, 2009 at 02:43:26PM +0200, Juan Quintela wrote:
>>> Jan Kiszka <jan.kiszka@siemens.com> wrote:
>>>>> -CFLAGS="$CFLAGS -Wall -Wundef -Wendif-labels -Wwrite-strings -Wmissing-prototypes -Wstrict-prototypes -Wredundant-decls"
>>>>> -LDFLAGS="$LDFLAGS -g"
>>>>> +CFLAGS="$CFLAGS $EXTRA_CFLAGS -Wall -Wundef -Wendif-labels -Wwrite-strings -Wmissing-prototypes -Wstrict-prototypes -Wredundant-decls"
>>>>> +LDFLAGS="$LDFLAGS $EXTRA_LDFLAGS -g"
>>>> ...but now we add the extra flags twice. So please remove the injection
>>>> into the flags for $config_mak.
>>> It is still broken. Thinking of a solution.
>>> People wanted EXTRA_CFLAGS to be added at the end of CFLAGS, not in the
>>> middle. Trying to find a way that is consistent for everything.
>>>
>>> Could you told in what test do you need extra_cflags?
>> The particular instance was for detecting zlib in a cross
>> configuration. But that's not necessarily the only test where
>> --extra-{cflags,ldflags} would be useful: every test that needs external
>> libraries is subject to this bug (zlib, TLS, SASL, SDL, VDE libraries,
>> BrlAPI?, Xen?, bluez?, etc.).
>>
>> If adding EXTRA_{CFLAGS,LDFLAGS} to all such tests is necessary and
>> people feel that's the best solution, then I can do that.
>
> I think that the solution is the other way around :)
>
> call configure with:
>
> CFLAGS="foo bar" ./configure
>
> and use CFLAGS for all configure script
>
> That was what I had in mind.
>
> If we do always in configure:
>
> CFLAGS="baz goo $CFLAGS"
>
> We also maintain the propierte that whatever you pass to CFLAGS is the
> last option in CFLAG (i.e. can overwrote whatever found by configure.
>
> Then we can:
> - kill extra cflags,
> - win your feature,
> - and have user wrote at the end of CFLAGS variable
>
> What do you think?
No concerns. My only additional requirement is that configure should not
forget about the provided CFLAGS when going through automatic reconfig,
just like it's currently the case with --extra-* - and like autoconf
behaves, too.
Jan
--
Siemens AG, Corporate Technology, CT SE 2
Corporate Competence Center Embedded Linux
next prev parent reply other threads:[~2009-07-23 16:32 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-07-23 11:40 [Qemu-devel] [PATCH] configure: fix breakage of --extra-{cflags, ldflags} Nathan Froyd
2009-07-23 14:13 ` [Qemu-devel] " Jan Kiszka
[not found] ` <m3bpnb6041.fsf@neno.mitica>
2009-07-23 14:20 ` Nathan Froyd
[not found] ` <m33a8n5slq.fsf@neno.mitica>
2009-07-23 16:32 ` Jan Kiszka [this message]
2009-07-23 14:50 ` Jan Kiszka
2009-07-23 16:58 ` [Qemu-devel] " Nathan Froyd
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
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4A689086.6080705@siemens.com \
--to=jan.kiszka@siemens.com \
--cc=froydnj@codesourcery.com \
--cc=qemu-devel@nongnu.org \
--cc=quintela@trasno.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 NNTP newsgroup(s).