From: Sam Ravnborg <sam@ravnborg.org>
To: Andrea Arcangeli <andrea@qumranet.com>
Cc: tvrtko.ursulin@sophos.com, linux-kbuild@vger.kernel.org
Subject: Re: kbuild variable $(src) broken in 2.6.23-rc3?
Date: Wed, 28 May 2008 21:09:07 +0200 [thread overview]
Message-ID: <20080528190907.GA6466@uranus.ravnborg.org> (raw)
In-Reply-To: <20080521194806.GJ22488@duo.random>
On Wed, May 21, 2008 at 09:48:08PM +0200, Andrea Arcangeli wrote:
> On Wed, May 21, 2008 at 09:04:53PM +0200, Sam Ravnborg wrote:
> > Andrea Arcangeli <andrea@qumranet.com> reported a similar issue.
> > I have not gotten around to look at it yet.
> > I minimal Makefile that exhibits the problem
> > would be helpfull!
>
> I don't have the minimal but I can easily provide you a way to
> reproduce:
>
> git clone git://git.kernel.org/pub/scm/virt/kvm/kvm-userspace.git
> cd kvm-userspace
> patch -p1 -R <../belowinlinedworkaround
> ./configure --kerneldir=/usr/src/anykernel
> cd kernel
> make
>
> The below has the benefit of making more robust and higher prio the
> inclusion of external-module-compat so it may not to be backed out
> after this 2.6.26-rc regression is fixed.
>
> The problem is that the $(src) in the EXTRA_CFLAGS below is expanded
> to "".
I have now tried to reproduce this bug with no luck.
- I tried to clone the git tree and follow your instructions.
But it failed to build due to some files missing in the
kernel directory so I gace up on that.
- Then I tried to create a few Makefiles for an external
module. I used a copy of the ext-2 filesystem and modifided
the Makefile like below:
#
# Makefile for the linux ext2-filesystem routines.
#
obj-m += ext2.o
ext2-y := balloc.o dir.o file.o fsync.o ialloc.o inode.o \
ioctl.o namei.o super.o symlink.o
EXTRA_CFLAGS := -I$(src)
Manually inspecting the command lines to gcc told me that $(src) was
properly expanded.
So something more subtle is going on.
Can I ask you to in an empty directory to create
a Makefile like this:
obj-m := foo.o
ccflags-y := -I$(src)
Another question. Do you use a -mm or -next kernel.
Or do you use a vanlla -linus kernel?
Because Andrew has some patches in -mm that may impact
how we handle -I but I do not have -mm available on this box.
Sam
next prev parent reply other threads:[~2008-05-28 19:08 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-05-21 14:42 kbuild variable $(src) broken in 2.6.23-rc3? tvrtko.ursulin
2008-05-21 19:04 ` Sam Ravnborg
2008-05-21 19:48 ` Andrea Arcangeli
2008-05-28 19:09 ` Sam Ravnborg [this message]
2008-05-28 19:13 ` Sam Ravnborg
2008-05-28 20:18 ` Andrea Arcangeli
2008-05-28 20:35 ` Sam Ravnborg
2008-05-29 19:36 ` Sam Ravnborg
2008-05-30 9:27 ` tvrtko.ursulin
2008-05-30 12:41 ` Andrea Arcangeli
2008-05-30 12:52 ` tvrtko.ursulin
2008-05-25 6:54 ` Sam Ravnborg
2008-05-26 15:15 ` Rafael J. Wysocki
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=20080528190907.GA6466@uranus.ravnborg.org \
--to=sam@ravnborg.org \
--cc=andrea@qumranet.com \
--cc=linux-kbuild@vger.kernel.org \
--cc=tvrtko.ursulin@sophos.com \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.