public inbox for systemtap@sourceware.org
 help / color / mirror / Atom feed
* [PATCH -mm 1/2] kprobes: indirectly call kprobe_target
@ 2008-06-06 17:58 Masami Hiramatsu
  2008-06-06 23:45 ` Ananth N Mavinakayanahalli
  0 siblings, 1 reply; 2+ messages in thread
From: Masami Hiramatsu @ 2008-06-06 17:58 UTC (permalink / raw)
  To: Andrew Morton
  Cc: Ananth N Mavinakayanahalli, Jim Keniston, David Miller,
	systemtap-ml, LKML

Call kprobe_target indirectly. This prevents gcc to unroll
a noinline function in caller function.

Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com>
CC: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
CC: Jim Keniston <jkenisto@us.ibm.com>
CC: David Miller <davem@davemloft.net>
---
I ported patches which had been discussed on
http://sources.redhat.com/bugzilla/show_bug.cgi?id=3542

 tests/test_kprobes.c |   21 ++++++---------------
 1 file changed, 6 insertions(+), 15 deletions(-)

Index: 2.6.26-rc2-mm1/tests/test_kprobes.c
===================================================================
--- 2.6.26-rc2-mm1.orig/tests/test_kprobes.c
+++ 2.6.26-rc2-mm1/tests/test_kprobes.c
@@ -22,21 +22,10 @@

 static u32 rand1, preh_val, posth_val, jph_val;
 static int errors, handler_errors, num_tests;
+static u32 (*target)(u32 value);

 static noinline u32 kprobe_target(u32 value)
 {
-	/*
-	 * gcc ignores noinline on some architectures unless we stuff
-	 * sufficient lard into the function. The get_kprobe() here is
-	 * just for that.
-	 *
-	 * NOTE: We aren't concerned about the correctness of get_kprobe()
-	 * here; hence, this call is neither under !preempt nor with the
-	 * kprobe_mutex held. This is fine(tm)
-	 */
-	if (get_kprobe((void *)0xdeadbeef))
-		printk(KERN_INFO "Kprobe smoke test: probe on 0xdeadbeef!\n");
-
 	return (value / div_factor);
 }

@@ -74,7 +63,7 @@ static int test_kprobe(void)
 		return ret;
 	}

-	ret = kprobe_target(rand1);
+	ret = target(rand1);
 	unregister_kprobe(&kp);

 	if (preh_val == 0) {
@@ -121,7 +110,7 @@ static int test_jprobe(void)
 		return ret;
 	}

-	ret = kprobe_target(rand1);
+	ret = target(rand1);
 	unregister_jprobe(&jp);
 	if (jph_val == 0) {
 		printk(KERN_ERR "Kprobe smoke test failed: "
@@ -177,7 +166,7 @@ static int test_kretprobe(void)
 		return ret;
 	}

-	ret = kprobe_target(rand1);
+	ret = target(rand1);
 	unregister_kretprobe(&rp);
 	if (krph_val != rand1) {
 		printk(KERN_ERR "Kprobe smoke test failed: "
@@ -193,6 +182,8 @@ int init_test_probes(void)
 {
 	int ret;

+	target = kprobe_target;
+
 	do {
 		rand1 = random32();
 	} while (rand1 <= div_factor);

-- 
Masami Hiramatsu

Software Engineer
Hitachi Computer Products (America) Inc.
Software Solutions Division

e-mail: mhiramat@redhat.com

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [PATCH -mm 1/2] kprobes: indirectly call kprobe_target
  2008-06-06 17:58 [PATCH -mm 1/2] kprobes: indirectly call kprobe_target Masami Hiramatsu
@ 2008-06-06 23:45 ` Ananth N Mavinakayanahalli
  0 siblings, 0 replies; 2+ messages in thread
From: Ananth N Mavinakayanahalli @ 2008-06-06 23:45 UTC (permalink / raw)
  To: Masami Hiramatsu
  Cc: Andrew Morton, Jim Keniston, David Miller, systemtap-ml, LKML

On Fri, Jun 06, 2008 at 10:08:08AM -0400, Masami Hiramatsu wrote:
> Call kprobe_target indirectly. This prevents gcc to unroll
> a noinline function in caller function.
> 
> Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com>

Acked-by: Ananth N Mavinakayanahalli <ananth@in.ibm.com>

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2008-06-06 14:21 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2008-06-06 17:58 [PATCH -mm 1/2] kprobes: indirectly call kprobe_target Masami Hiramatsu
2008-06-06 23:45 ` Ananth N Mavinakayanahalli

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 read-only IMAP folder(s) and NNTP newsgroup(s).