public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Gautham R Shenoy <ego@in.ibm.com>
To: Andrew Morton <akpm@osdl.org>
Cc: ego@in.ibm.com, vatsa@in.ibm.com, linux-kernel@vger.kernel.org,
	mingo@elte.hu, torvalds@osdl.org, dipankar@in.ibm.com
Subject: Re: [PATCH] Handle per-subsystem mutexes for CONFIG_HOTPLUG_CPU not set.
Date: Fri, 24 Nov 2006 09:57:02 +0530	[thread overview]
Message-ID: <20061124042702.GA4666@in.ibm.com> (raw)
In-Reply-To: <20061123125446.3cd9ff0f.akpm@osdl.org>

Hi Andrew,

> >
> > Usage:
> > a) Each hotcpu aware subsystem defines the hotcpu_mutex as follows
> > #ifdef CONFIG_HOTPLUG_CPU
> > DEFINE_MUTEX(my_hotcpu_mutex);
> > #endif
> >
> > b) The hotcpu aware subsystem uses
> > cpuhotplug_mutex_lock(&my_hotcpu_mutex)
> > and
> > cpuhotplug_mutex_unlock(&my_hotcpu_mutex)
> > instead of the usual mutex_lock/mutex_unlock.
> >
> > Signed-off-by: Gautham R Shenoy <ego@in.ibm.com>
> >
> > ---
> >  include/linux/cpu.h |   15 +++++++++++++++
[snip]
> >  extern void unlock_cpu_hotplug(void);
> >  #define hotcpu_notifier(fn, pri) {				\
> > @@ -86,6 +98,9 @@ extern void unlock_cpu_hotplug(void);
> >  int cpu_down(unsigned int cpu);
> >  #define cpu_is_offline(cpu) unlikely(!cpu_online(cpu))
> >  #else
> > +#define cpuhotplug_mutex_lock(m)	do { } while (0)
> > +#define cpuhotplug_mutex_unlock(m)	do { } while (0)
> > +
> 
> But what to do about the now-unneeded mutex?
> 
> We can just leave it there if CONFIG_HOTPLUG_CPU=n but then we'll get
> unused variable warnings for statically-defined mutexes.

Why even leave it there?

Can't we do something as follows, which has already been suggested in
the patch description:

Each hotcpu aware subsystem defines the hotcpu_mutex as follows
#ifdef CONFIG_HOTPLUG_CPU
[static] DEFINE_MUTEX(my_hotcpu_mutex);
#endif

That way, we won't be having any unneeded mutexes at all.

> To fix that would require either
> 
> #ifdef CONFIG_HOTPLUG_CPU
> #define DEFINE_MUTEX_HOTPLUG_CPU(m) DEFINE_MUTEX(m)
> #else
> #define DEFINE_MUTEX_HOTPLUG_CPU(m)
> #endif
> 

Yup, this won't work. Wish we could cook up something like this, but
alas! Most of these mutexes are defined with the static keyword,
which causes compile errors with CONFIG_HOTPLUG_CPU=n.

> or
> 
> #define cpuhotplug_mutex_lock(m)	do { (void)(m); } while (0)
> 
> 
> Given that the former won't work, I'd suggest the latter ;)
> 

regards
gautham.
-- 
Gautham R Shenoy
Linux Technology Center
IBM India.
"Freedom comes with a price tag of responsibility, which is still a bargain,
because Freedom is priceless!"

  reply	other threads:[~2006-11-24  4:27 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-11-23 13:18 [PATCH] Handle per-subsystem mutexes for CONFIG_HOTPLUG_CPU not set Gautham R Shenoy
2006-11-23 20:54 ` Andrew Morton
2006-11-24  4:27   ` Gautham R Shenoy [this message]
2006-11-24 12:13     ` [PATCH][v2] " Gautham R Shenoy

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=20061124042702.GA4666@in.ibm.com \
    --to=ego@in.ibm.com \
    --cc=akpm@osdl.org \
    --cc=dipankar@in.ibm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=torvalds@osdl.org \
    --cc=vatsa@in.ibm.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox