From: Fam Zheng <famz@redhat.com>
To: Paolo Bonzini <pbonzini@redhat.com>
Cc: Peter Maydell <peter.maydell@linaro.org>,
agraf@suse.de, Michael Tokarev <mjt@tls.msk.ru>,
qemu-devel@nongnu.org, Stefan Weil <sw@weilnetz.de>
Subject: Re: [Qemu-devel] [PATCH] rules.mak: Force CFLAGS for all objects in DSO
Date: Wed, 6 May 2015 23:01:49 +0800 [thread overview]
Message-ID: <20150506150149.GC1535@ad.nay.redhat.com> (raw)
In-Reply-To: <554A26EE.7050108@redhat.com>
On Wed, 05/06 16:36, Paolo Bonzini wrote:
>
>
> On 06/05/2015 16:23, Fam Zheng wrote:
> >>> > > + $(eval $($v:%.mo=%$(DSOSUF)) $($v) $(foreach o,$($v),$($o-objs)) .PHONY: CFLAGS += -fPIC -DBUILD_DSO)
> > ^ ^ ^ ^
> > | | | |
> > | | | `- In case all others are empty.
> > | | |
> > | | `- Expansion of all %.mo-objs so it's a %.o list.
> > | | For $v=block-obj-m, this will contain curl.o, iscsi.o, ...
> > | |
> > | `- %.mo list
> > | For $v=block-obj-m, this will contain curl.mo, iscsi.mo, ...
> > |
> > `- %.so list
> > For $v=block-obj-m, this will contain curl.so, iscsi.so, .
>
> Great. :) You should have used Unicode drawing characters to be
> consistent with rules.mak!
Well, I'd have to find and read the vim plugin doc to do that again!
>
> But hopefully the ASCII art is not needed at all. Do we need all of those?
>
> - foo.mo is a prerequisite of foo.so. You cannot use foo.so (this is
> the bug you are fixing) but foo.so doesn't use CFLAGS in its rule, and
> you should be able to remove it.
>
> - can you use foo.mo without hitting the original bug? If so, could you
> simply use:
>
> -%$(DSOSUF): CFLAGS += -fPIC -DBUILD_DSO
> +%.mo: CFLAGS += -fPIC -DBUILD_DSO
I believe this will propagate the flags correctly. However that will affect
non-module build, so I didn't want to do it unconditionally.
>
> (and if so move the rule down, above "%.mo:")? If you cannot use
> foo.mo, foo.mo also doesn't use CFLAGS in its rules, and you should be
> able to remove foo.mo too.
>
> - so the other possibility is to just use $(foreach o,$($v),$($o-objs)).
> In this case there's already a convenient $(foreach) just above, and
> you can just add another $(eval) inside it.
>
> - and if that is true, you do not need .PHONY either because you have
> $(foreach ... $(eval)) instead of $(eval $(foreach)).
>
OK.
It's been a long day for me, so you can see that I'm going defensive with the
code I'm writing, but yes, your last "if" looks about right, except we still
need the "$(if $(CONFIG_MODULES),...)".
Sanity check: is -fPIC only ever needed in compiling, but not in (partial)
linking, right?
Fam
next prev parent reply other threads:[~2015-05-06 15:02 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-05-06 13:46 [Qemu-devel] [PATCH] rules.mak: Force CFLAGS for all objects in DSO Fam Zheng
2015-05-06 14:07 ` Paolo Bonzini
2015-05-06 14:23 ` Fam Zheng
2015-05-06 14:36 ` Paolo Bonzini
2015-05-06 15:01 ` Fam Zheng [this message]
2015-05-06 15:03 ` Paolo Bonzini
2015-05-06 23:25 ` Alexander Graf
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=20150506150149.GC1535@ad.nay.redhat.com \
--to=famz@redhat.com \
--cc=agraf@suse.de \
--cc=mjt@tls.msk.ru \
--cc=pbonzini@redhat.com \
--cc=peter.maydell@linaro.org \
--cc=qemu-devel@nongnu.org \
--cc=sw@weilnetz.de \
/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).