[RESEND][PATCH 1/2] stacktrace: add saved stack traces to backtrace self-test

[RESEND][PATCH 1/2] stacktrace: add saved stack traces to backtrace self-test

[Vegard Nossum]

Hi,

I'm not really sure to whom this goes. It seems that the original code
came through the x86 tree, although the code itself seems to be
entirely generic.

(The last time I sent it, it was to Andrew only, though he seemed to
be on holidays at the time. Now I have a patch #2 as well!)

Arjan, are there any objections to this?

(This is not intended for 2.6.26. In case that wasn't obvious, heh.)


Vegard


From: Vegard Nossum <vegard.nossum@gmail.com>
Date: Thu, 26 Jun 2008 13:27:35 +0200
Subject: [PATCH] stacktrace: add saved stack traces to backtrace self-test

This patch adds saved stack-traces to the backtrace suite of self-tests.

Note that we don't depend on or unconditionally enable CONFIG_STACKTRACE
because not all architectures may have it (and we still want to enable the
other tests for those architectures).

Cc: Arjan van de Ven <arjan@infradead.org>
Signed-off-by: Vegard Nossum <vegard.nossum@gmail.com>
---
 kernel/backtracetest.c |   30 +++++++++++++++++++++++++++++-
 lib/Kconfig.debug      |    3 +++
 2 files changed, 32 insertions(+), 1 deletions(-)

diff --git a/kernel/backtracetest.c b/kernel/backtracetest.c
index d1a7605..50f7abd 100644
--- a/kernel/backtracetest.c
+++ b/kernel/backtracetest.c
@@ -10,9 +10,10 @@
  * of the License.
  */
 
+#include <linux/delay.h>
 #include <linux/module.h>
 #include <linux/sched.h>
-#include <linux/delay.h>
+#include <linux/stacktrace.h>
 
 static struct timer_list backtrace_timer;
 
@@ -22,6 +23,31 @@ static void backtrace_test_timer(unsigned long data)
 	printk("The following trace is a kernel self test and not a bug!\n");
 	dump_stack();
 }
+
+#ifdef CONFIG_STACKTRACE
+static void backtrace_test_saved(void)
+{
+	struct stack_trace trace;
+	unsigned long entries[8];
+
+	printk("Testing a saved backtrace.\n");
+	printk("The following trace is a kernel self test and not a bug!\n");
+
+	trace.nr_entries = 0;
+	trace.max_entries = ARRAY_SIZE(entries);
+	trace.entries = entries;
+	trace.skip = 0;
+
+	save_stack_trace(&trace);
+	print_stack_trace(&trace, 0);
+}
+#else
+static void backtrace_test_saved(void)
+{
+	printk("Saved backtrace test skipped.\n");
+}
+#endif
+
 static int backtrace_regression_test(void)
 {
 	printk("====[ backtrace testing ]===========\n");
@@ -29,6 +55,8 @@ static int backtrace_regression_test(void)
 	printk("The following trace is a kernel self test and not a bug!\n");
 	dump_stack();
 
+	backtrace_test_saved();
+
 	init_timer(&backtrace_timer);
 	backtrace_timer.function = backtrace_test_timer;
 	mod_timer(&backtrace_timer, jiffies + 10);
diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
index d2099f4..4dac3c7 100644
--- a/lib/Kconfig.debug
+++ b/lib/Kconfig.debug
@@ -563,6 +563,9 @@ config BACKTRACE_SELF_TEST
 	  for distributions or general kernels, but only for kernel
 	  developers working on architecture code.
 
+	  Note that if you want to also test saved backtraces, you will
+	  have to enable STACKTRACE as well.
+
 	  Say N if you are unsure.
 
 config LKDTM
-- 
1.5.4.1

Re: [RESEND][PATCH 1/2] stacktrace: add saved stack traces to backtrace self-test

[Ingo Molnar]

* Vegard Nossum <vegard.nossum@gmail.com> wrote:

> Hi,
> 
> I'm not really sure to whom this goes. It seems that the original code 
> came through the x86 tree, although the code itself seems to be 
> entirely generic.

given that i've written the generic stacktrace code for lockdep, i guess 
it comes to me - the only question would be into which branch ;-)

fortunately there's already a tip/core/stacktrace topic, hosting a 
change for you as well:

 Ingo Molnar (1):
       debugging: make stacktrace independent from DEBUG_KERNEL

 Johannes Berg (1):
       stacktrace: don't crash on invalid stack trace structs

 Vegard Nossum (1):
       stacktrace: print_stack_trace() cleanup

i'll queue this one up there.

	Ingo

Re: [RESEND][PATCH 1/2] stacktrace: add saved stack traces to backtrace self-test

[Ingo Molnar]

FYI, it needed the fixlet below.

	Ingo

------------->
commit 7bcebd2e3ee8435e2ff0f674a409f46961fda3cc
Author: Ingo Molnar <mingo@elte.hu>
Date:   Fri Jun 27 21:20:17 2008 +0200

    stacktrace: export print_stack_trace and save_stack_trace
    
    fix:
    
    ERROR: "print_stack_trace" [kernel/backtracetest.ko] undefined!
    ERROR: "save_stack_trace" [kernel/backtracetest.ko] undefined!
    
    Signed-off-by: Ingo Molnar <mingo@elte.hu>
    
    Conflicts:
    
    Signed-off-by: Ingo Molnar <mingo@elte.hu>

diff --git a/arch/x86/kernel/stacktrace.c b/arch/x86/kernel/stacktrace.c
index c28c342..b7b483b 100644
--- a/arch/x86/kernel/stacktrace.c
+++ b/arch/x86/kernel/stacktrace.c
@@ -74,10 +74,4 @@ void save_stack_trace(struct stack_trace *trace)
 	if (trace->nr_entries < trace->max_entries)
 		trace->entries[trace->nr_entries++] = ULONG_MAX;
 }
-
-void save_stack_trace_tsk(struct task_struct *tsk, struct stack_trace *trace)
-{
-	dump_trace(tsk, NULL, NULL, 0, &save_stack_ops_nosched, trace);
-	if (trace->nr_entries < trace->max_entries)
-		trace->entries[trace->nr_entries++] = ULONG_MAX;
-}
+EXPORT_SYMBOL_GPL(save_stack_trace);
diff --git a/kernel/stacktrace.c b/kernel/stacktrace.c
index 7eaea9d..eda2f4c 100644
--- a/kernel/stacktrace.c
+++ b/kernel/stacktrace.c
@@ -21,4 +21,5 @@ void print_stack_trace(struct stack_trace *trace, int spaces)
 		print_ip_sym(trace->entries[i]);
 	}
 }
+EXPORT_SYMBOL_GPL(print_stack_trace);
 

Re: [RESEND][PATCH 1/2] stacktrace: add saved stack traces to backtrace self-test

[Vegard Nossum]

On Fri, Jun 27, 2008 at 9:20 PM, Ingo Molnar <mingo@elte.hu> wrote:
>
> FYI, it needed the fixlet below.
>
>        Ingo
>

Oh. I'm sorry :-(

(I'll admit that I've never worked with modules before. I'll try to
remember now.)

> ------------->
> commit 7bcebd2e3ee8435e2ff0f674a409f46961fda3cc
> Author: Ingo Molnar <mingo@elte.hu>
> Date:   Fri Jun 27 21:20:17 2008 +0200
>
>    stacktrace: export print_stack_trace and save_stack_trace
>
>    fix:
>
>    ERROR: "print_stack_trace" [kernel/backtracetest.ko] undefined!
>    ERROR: "save_stack_trace" [kernel/backtracetest.ko] undefined!
>
>    Signed-off-by: Ingo Molnar <mingo@elte.hu>
>
>    Conflicts:
>
>    Signed-off-by: Ingo Molnar <mingo@elte.hu>
>
> diff --git a/arch/x86/kernel/stacktrace.c b/arch/x86/kernel/stacktrace.c
> index c28c342..b7b483b 100644
> --- a/arch/x86/kernel/stacktrace.c
> +++ b/arch/x86/kernel/stacktrace.c
> @@ -74,10 +74,4 @@ void save_stack_trace(struct stack_trace *trace)
>        if (trace->nr_entries < trace->max_entries)
>                trace->entries[trace->nr_entries++] = ULONG_MAX;
>  }
> -
> -void save_stack_trace_tsk(struct task_struct *tsk, struct stack_trace *trace)
> -{
> -       dump_trace(tsk, NULL, NULL, 0, &save_stack_ops_nosched, trace);
> -       if (trace->nr_entries < trace->max_entries)
> -               trace->entries[trace->nr_entries++] = ULONG_MAX;
> -}

Why is this removed, though? (It seems unrelated, and that's a useful function.)


Vegard

-- 
"The animistic metaphor of the bug that maliciously sneaked in while
the programmer was not looking is intellectually dishonest as it
disguises that the error is the programmer's own creation."
	-- E. W. Dijkstra, EWD1036

Re: [RESEND][PATCH 1/2] stacktrace: add saved stack traces to backtrace self-test

[Ingo Molnar]

* Vegard Nossum <vegard.nossum@gmail.com> wrote:

> On Fri, Jun 27, 2008 at 9:20 PM, Ingo Molnar <mingo@elte.hu> wrote:
> >
> > FYI, it needed the fixlet below.
> >
> >        Ingo
> >
> 
> Oh. I'm sorry :-(
> 
> (I'll admit that I've never worked with modules before. I'll try to 
> remember now.)

i too work with bzImages most of the time. They are convenient 
single-file objects that are easy to distribute and easy to use, and 
they also have no Linux-distribution-flavor dependencies. (module load 
order can be influenced by things like /etc/modprobe.conf and other 
details - make reproduction harder and testing less conclusive, etc.)

> > ------------->
> > commit 7bcebd2e3ee8435e2ff0f674a409f46961fda3cc
> > Author: Ingo Molnar <mingo@elte.hu>
> > Date:   Fri Jun 27 21:20:17 2008 +0200
> >
> >    stacktrace: export print_stack_trace and save_stack_trace
> >
> >    fix:
> >
> >    ERROR: "print_stack_trace" [kernel/backtracetest.ko] undefined!
> >    ERROR: "save_stack_trace" [kernel/backtracetest.ko] undefined!
> >
> >    Signed-off-by: Ingo Molnar <mingo@elte.hu>
> >
> >    Conflicts:
> >
> >    Signed-off-by: Ingo Molnar <mingo@elte.hu>
> >
> > diff --git a/arch/x86/kernel/stacktrace.c b/arch/x86/kernel/stacktrace.c
> > index c28c342..b7b483b 100644
> > --- a/arch/x86/kernel/stacktrace.c
> > +++ b/arch/x86/kernel/stacktrace.c
> > @@ -74,10 +74,4 @@ void save_stack_trace(struct stack_trace *trace)
> >        if (trace->nr_entries < trace->max_entries)
> >                trace->entries[trace->nr_entries++] = ULONG_MAX;
> >  }
> > -
> > -void save_stack_trace_tsk(struct task_struct *tsk, struct stack_trace *trace)
> > -{
> > -       dump_trace(tsk, NULL, NULL, 0, &save_stack_ops_nosched, trace);
> > -       if (trace->nr_entries < trace->max_entries)
> > -               trace->entries[trace->nr_entries++] = ULONG_MAX;
> > -}
> 
> Why is this removed, though? (It seems unrelated, and that's a useful 
> function.)

that was an oversight - thanks - fixed it - updated commit below.

	Ingo

---------------->
commit a73d1be0974856741a258928966b8a03ae6bb74f
Author: Ingo Molnar <mingo@elte.hu>
Date:   Fri Jun 27 21:20:17 2008 +0200

    stacktrace: fix modular build, export print_stack_trace and save_stack_trace
    
    fix:
    
    ERROR: "print_stack_trace" [kernel/backtracetest.ko] undefined!
    ERROR: "save_stack_trace" [kernel/backtracetest.ko] undefined!
    
    Signed-off-by: Ingo Molnar <mingo@elte.hu>
    
    and fix:
    
      Building modules, stage 2.
      MODPOST 376 modules
    ERROR: "print_stack_trace" [kernel/backtracetest.ko] undefined!
    make[1]: *** [__modpost] Error 1
    
    Signed-off-by: Ingo Molnar <mingo@elte.hu>

diff --git a/arch/x86/kernel/stacktrace.c b/arch/x86/kernel/stacktrace.c
index c28c342..a03e7f6 100644
--- a/arch/x86/kernel/stacktrace.c
+++ b/arch/x86/kernel/stacktrace.c
@@ -74,6 +74,7 @@ void save_stack_trace(struct stack_trace *trace)
 	if (trace->nr_entries < trace->max_entries)
 		trace->entries[trace->nr_entries++] = ULONG_MAX;
 }
+EXPORT_SYMBOL_GPL(save_stack_trace);
 
 void save_stack_trace_tsk(struct task_struct *tsk, struct stack_trace *trace)
 {
@@ -81,3 +82,4 @@ void save_stack_trace_tsk(struct task_struct *tsk, struct stack_trace *trace)
 	if (trace->nr_entries < trace->max_entries)
 		trace->entries[trace->nr_entries++] = ULONG_MAX;
 }
+EXPORT_SYMBOL_GPL(save_stack_trace_tsk);
diff --git a/kernel/stacktrace.c b/kernel/stacktrace.c
index 7eaea9d..94b527e 100644
--- a/kernel/stacktrace.c
+++ b/kernel/stacktrace.c
@@ -6,6 +6,7 @@
  *  Copyright (C) 2006 Red Hat, Inc., Ingo Molnar <mingo@redhat.com>
  */
 #include <linux/sched.h>
+#include <linux/module.h>
 #include <linux/kallsyms.h>
 #include <linux/stacktrace.h>
 
@@ -21,4 +22,5 @@ void print_stack_trace(struct stack_trace *trace, int spaces)
 		print_ip_sym(trace->entries[i]);
 	}
 }
+EXPORT_SYMBOL_GPL(print_stack_trace);