public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Takashi Iwai <tiwai@suse.de>
To: Jaswinder Singh Rajput <jaswinder@kernel.org>
Cc: Arnd Bergmann <arnd@arndb.de>, Sam Ravnborg <sam@ravnborg.org>,
	alsa-devel@alsa-project.org, Ingo Molnar <mingo@elte.hu>,
	LKML <linux-kernel@vger.kernel.org>
Subject: Re: usr/include/linux/soundcard.h warnings for 'make headers_check'
Date: Wed, 04 Feb 2009 10:03:06 +0100	[thread overview]
Message-ID: <s5h4ozak20l.wl%tiwai@suse.de> (raw)
In-Reply-To: <1233736494.3198.15.camel@localhost.localdomain>

At Wed, 04 Feb 2009 14:04:54 +0530,
Jaswinder Singh Rajput wrote:
> 
> On Wed, 2009-02-04 at 07:39 +0100, Takashi Iwai wrote:
> > At Wed, 04 Feb 2009 09:34:12 +0530,
> > Jaswinder Singh Rajput wrote:
> > > 
> > > On Tue, 2009-02-03 at 21:27 +0100, Arnd Bergmann wrote:
> > > > On Wednesday 21 January 2009, Sam Ravnborg wrote:
> > > > > 
> > > > > > No, these are declarations for OSS-lib.
> > > > > > It doesn't belong to kernel, but it's better to keep it for
> > > > > > compatibility in some way.
> > > > > 
> > > > > But not in the kernel for sure.
> > > > > I just checked and for example: OSS_init is not used anywhere in the kernel.
> > > > > The kernel headers are not a "dump all your stuff" ground anymore.
> > > > > 
> > > > > So I encourage you to find a better home for the user space library definitons.
> > > > > We can let the definiton stay for a while - but eventually they have to
> > > > > be dropped from the kernel.
> > > > 
> > > > The last free version of OSSlib that matches this header is from ~1997, and
> > > > no current distro appears to be shipping it. All new OSSlib versions (the
> > > > oldest I found was from ~2003) contain in their documentation:
> > > > 
> > > > "Older versions of some OSS include files are distributed
> > > >  with various operating systems. There is no danger in
> > > >  using them but applications written for more recent OSS
> > > >  versions will not compile with older header files.
> > > > 
> > > >  The latest versions of these include files (such as
> > > >  soundcard.h) are distributed in ../include/sys. Use
> > > >  the -I/usr/lib/oss/include switch when compiling
> > > >  programs. Alternatively copy these files to /usr/include/sys
> > > >  so that they replace the original ones."
> > > > 
> > > > I think it's safe enough to assume that everyone that builds against
> > > > OSSlib by now also has the correct version of the header installed.
> > > > Consequently, we should just drop the #ifndef __KERNEL__ section
> > > > of soundcard.h (all the macros in there depend on the extern
> > > > declarations), and maybe add an
> > > > 
> > > > #ifdef OSSLIB
> > > > #error need to use <sys/soundcard.h> from libOSSlib
> > > > #endif
> > > > 
> > > 
> > > sys/soundcard.h is also pointing to linux/soundcard.h, we need to move
> > > all this stuff to sys/soundcard.h.
> > 
> > All which stuff?
> > If you are referring to OSS-lib stuff in soundcard.h, then I'm for
> > it.
> > 
> 
> We need to fix things in such a way that we can replace:
> #if (!defined(__KERNEL__) && !defined(KERNEL) && !defined(INKERNEL) && !defined(_KERNEL)) || defined(USE_SEQ_MACROS)
> with
> #ifdef __KERNEL__
> and only keep those things in this file which we need in kernel and move
> userspace declarations which are not required for kernel to
> sys/soundcard.h

Which means to remove only the OSS-lib part.  Some other macros are
used in sound/oss/*.c.  And, there are cases that apps want to use
SEQ* macros without oss-lib (which are actually most cases).


> > Also giving an error with an ifdef OSSLIB like Arnd suggested would be
> > nice, too.  But maybe we need one more ifdef such as
> > 
> > #ifdef OSSLIB
> > #ifndef _SYS_SOUNDCARD_H
> > #error ...
> > #endif
> > #endif
> > 
> 
> hmm, so you want to confirm that we only come through sys/soundcard.h
> for OSSLIB

Yes.


Takashi

  reply	other threads:[~2009-02-04  9:03 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-01-20 15:38 usr/include/linux/soundcard.h warnings for 'make headers_check' Jaswinder Singh Rajput
2009-01-21  0:08 ` Takashi Iwai
2009-01-21  0:25   ` Jaswinder Singh Rajput
2009-01-21  5:41   ` Sam Ravnborg
2009-02-03 20:27     ` Arnd Bergmann
2009-02-04  4:04       ` Jaswinder Singh Rajput
2009-02-04  6:39         ` Takashi Iwai
2009-02-04  8:34           ` Jaswinder Singh Rajput
2009-02-04  9:03             ` Takashi Iwai [this message]
2009-02-04 10:37               ` Jaswinder Singh Rajput
2009-02-04 10:52                 ` Takashi Iwai
2009-02-04 11:31                   ` Jaswinder Singh Rajput
2009-02-04 11:43                     ` Takashi Iwai
2009-02-04 12:51                       ` Arnd Bergmann
2009-02-04 12:58                         ` Takashi Iwai
2009-02-04 13:44                         ` Jaswinder Singh Rajput
2009-02-04 13:49                           ` Takashi Iwai
2009-02-04 14:02                           ` Arnd Bergmann
2009-02-09 23:06                             ` Takashi Iwai
2009-01-24  6:50   ` Jaswinder Singh Rajput

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=s5h4ozak20l.wl%tiwai@suse.de \
    --to=tiwai@suse.de \
    --cc=alsa-devel@alsa-project.org \
    --cc=arnd@arndb.de \
    --cc=jaswinder@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=sam@ravnborg.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