linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Rusty Russell <rusty@rustcorp.com.au>
To: Steven Rostedt <rostedt@goodmis.org>, Ingo Molnar <mingo@elte.hu>
Cc: LKML <linux-kernel@vger.kernel.org>,
	Andrew Morton <akpm@linux-foundation.org>,
	Frederic Weisbecker <fweisbec@gmail.com>,
	Li Zefan <lizf@cn.fujitsu.com>
Subject: Re: [RFC][PATCH] tracing/module: Move tracepoint out of module.h
Date: Wed, 01 Feb 2012 17:18:18 +1030	[thread overview]
Message-ID: <87y5snhwwd.fsf@rustcorp.com.au> (raw)
In-Reply-To: <1328014230.5882.29.camel@gandalf.stny.rr.com>

On Tue, 31 Jan 2012 07:50:30 -0500, Steven Rostedt <rostedt@goodmis.org> wrote:
> On Tue, 2012-01-31 at 13:20 +0100, Ingo Molnar wrote:
> 
> > > Saves only 400 bytes of text here, but I don't do preempt or 
> > > tracepoints.
> > 
> > Most distro kernels do tracepoints so I guess that's where the 
> > size delta comes from :-) In any case:
> 
> Yes, having tracepoints inlined causes a lot of bloat. But I still did
> get a 1799 bytes savings between moving the tracepoint out of line but
> keeping the if, and totally moving the entire functions out of line. Not
> sure what the discrepancy was there.
> 
> My last set of numbers came from the default 3.0.0 Debian config, which
> probably adds more things that call these functions into the kernel
> proper?

Here's the results I get, using the Ubuntu 3.0.0 config (minus
CONFIG_DEBUG_INFO, and rest with Enter held down).

Applied,
Rusty.

From: Steven Rostedt <rostedt@goodmis.org>
Subject: module: move __module_get and try_module_get() out of line.

With the preempt, tracepoint and everything, it's getting a bit
chubby.  For an Ubuntu-based config:

Before:
	$ size -t `find * -name '*.ko'` | grep TOTAL
	56199906        3870760	1606616	61677282	3ad1ee2	(TOTALS)
	$ size vmlinux
	   text	   data	    bss	    dec	    hex	filename
	8509342	 850368	3358720	12718430	 c2115e	vmlinux

After:
	$ size -t `find * -name '*.ko'` | grep TOTAL
	56183760	3867892	1606616	61658268	3acd49c	(TOTALS)
	$ size vmlinux
	   text	   data	    bss	    dec	    hex	filename
	8501842	 849088	3358720	12709650	 c1ef12	vmlinux

Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> (made all out-of-line)

diff --git a/include/linux/module.h b/include/linux/module.h
--- a/include/linux/module.h
+++ b/include/linux/module.h
@@ -21,8 +21,6 @@
 #include <linux/percpu.h>
 #include <asm/module.h>
 
-#include <trace/events/module.h>
-
 /* Not Yet Implemented */
 #define MODULE_SUPPORTED_DEVICE(name)
 
@@ -452,33 +450,11 @@ void symbol_put_addr(void *addr);
 
 /* Sometimes we know we already have a refcount, and it's easier not
    to handle the error case (which only happens with rmmod --wait). */
-static inline void __module_get(struct module *module)
-{
-	if (module) {
-		preempt_disable();
-		__this_cpu_inc(module->refptr->incs);
-		trace_module_get(module, _THIS_IP_);
-		preempt_enable();
-	}
-}
+extern void __module_get(struct module *module);
 
-static inline int try_module_get(struct module *module)
-{
-	int ret = 1;
-
-	if (module) {
-		preempt_disable();
-
-		if (likely(module_is_live(module))) {
-			__this_cpu_inc(module->refptr->incs);
-			trace_module_get(module, _THIS_IP_);
-		} else
-			ret = 0;
-
-		preempt_enable();
-	}
-	return ret;
-}
+/* This is the Right Way to get a module: if it fails, it's being removed,
+ * so pretend it's not there. */
+extern bool try_module_get(struct module *module);
 
 extern void module_put(struct module *module);
 
diff --git a/kernel/module.c b/kernel/module.c
--- a/kernel/module.c
+++ b/kernel/module.c
@@ -903,6 +903,36 @@ static ssize_t show_refcnt(struct module
 static struct module_attribute modinfo_refcnt =
 	__ATTR(refcnt, 0444, show_refcnt, NULL);
 
+void __module_get(struct module *module)
+{
+	if (module) {
+		preempt_disable();
+		__this_cpu_inc(module->refptr->incs);
+		trace_module_get(module, _RET_IP_);
+		preempt_enable();
+	}
+}
+EXPORT_SYMBOL(__module_get);
+
+bool try_module_get(struct module *module)
+{
+	bool ret = true;
+
+	if (module) {
+		preempt_disable();
+
+		if (likely(module_is_live(module))) {
+			__this_cpu_inc(module->refptr->incs);
+			trace_module_get(module, _RET_IP_);
+		} else
+			ret = false;
+
+		preempt_enable();
+	}
+	return ret;
+}
+EXPORT_SYMBOL(try_module_get);
+
 void module_put(struct module *module)
 {
 	if (module) {

  reply	other threads:[~2012-02-01  7:06 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-01-16 22:57 [PATCH][GIT PULL][v3.3] tracing: Add header wrappers event_headers_start.h and event_headers_end.h Steven Rostedt
2012-01-17  9:54 ` Ingo Molnar
2012-01-17 13:32   ` Steven Rostedt
2012-01-18 12:07     ` Ingo Molnar
2012-01-18 17:56       ` Steven Rostedt
2012-01-22 22:59         ` Rusty Russell
2012-01-26  2:41           ` [RFC][PATCH] tracing/module: Move tracepoint out of module.h Steven Rostedt
2012-01-26  2:45             ` Steven Rostedt
2012-01-26 10:28             ` Ingo Molnar
2012-01-26 13:52               ` Steven Rostedt
2012-01-26 13:55                 ` Ingo Molnar
2012-01-26 14:04                   ` Steven Rostedt
2012-01-26 14:07                     ` Steven Rostedt
2012-01-26 14:36                     ` Steven Rostedt
2012-01-26 18:39                       ` Ingo Molnar
2012-01-27  3:02                         ` Rusty Russell
2012-01-30 11:52                           ` Steven Rostedt
2012-01-30 17:28                             ` Steven Rostedt
2012-01-31  3:58                             ` Rusty Russell
2012-01-31 12:20                               ` Ingo Molnar
2012-01-31 12:50                                 ` Steven Rostedt
2012-02-01  6:48                                   ` Rusty Russell [this message]
2012-02-01 13:27                                     ` Steven Rostedt
2012-02-01 13:49                                     ` Ingo Molnar
2012-02-01 14:25                                       ` Steven Rostedt
2012-03-29  4:22                                     ` Eric Dumazet
2012-03-29  5:24                                       ` Rusty Russell
2012-02-01  1:10                                 ` Rusty Russell
2012-02-01  7:09                                   ` Ingo Molnar
2012-01-30  6:40                       ` Li Zefan
2012-02-17 13:46       ` [tip:perf/core] tracing/softirq: Move __raise_softirq_irqoff() out of header tip-bot for Steven Rostedt

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=87y5snhwwd.fsf@rustcorp.com.au \
    --to=rusty@rustcorp.com.au \
    --cc=akpm@linux-foundation.org \
    --cc=fweisbec@gmail.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=lizf@cn.fujitsu.com \
    --cc=mingo@elte.hu \
    --cc=rostedt@goodmis.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).