[PATCH 0/2] x86,nmi: Section header fixes

[PATCH 0/2] x86,nmi: Section header fixes

[Don Zickus]

An earlier attempt at a fix for the problem described in the second patch was
committed to Ingo's tree.  Later a smaller/cleaner patch was posted.  This
series just reverts the original patch and then applies the second patch.

The second patch looks a lot cleaner this way instead of just provding the
delta patch on top.

Ingo, let me know if this is not what you want.

Cheers,
Don

Don Zickus (1):
  Revert "x86/nmi: Fix section mismatch warnings on 32-bit"

Li Zhong (1):
  x86,nmi: Fix section mismatch warnings on 32-bit

 arch/x86/include/asm/nmi.h     |   20 +++-----------------
 arch/x86/kernel/nmi_selftest.c |    7 ++++---
 2 files changed, 7 insertions(+), 20 deletions(-)

-- 
1.7.7.6

[PATCH 1/2] Revert "x86/nmi: Fix section mismatch warnings on 32-bit"

[Don Zickus]

This reverts commit eeaaa96a3a2134a174100afd129bb0891d05f4b2.

There is a cleaner/simpler way to achieve this.

Signed-off-by: Don Zickus <dzickus@redhat.com>
---
 arch/x86/include/asm/nmi.h     |   14 --------------
 arch/x86/kernel/nmi_selftest.c |    4 ++--
 2 files changed, 2 insertions(+), 16 deletions(-)

diff --git a/arch/x86/include/asm/nmi.h b/arch/x86/include/asm/nmi.h
index dc580c4..0e3793b 100644
--- a/arch/x86/include/asm/nmi.h
+++ b/arch/x86/include/asm/nmi.h
@@ -54,20 +54,6 @@ struct nmiaction {
 	__register_nmi_handler((t), &fn##_na);	\
 })
 
-/*
- * For special handlers that register/unregister in the
- * init section only.  This should be considered rare.
- */
-#define register_nmi_handler_initonly(t, fn, fg, n)		\
-({							\
-	static struct nmiaction fn##_na __initdata = {		\
-		.handler = (fn),			\
-		.name = (n),				\
-		.flags = (fg),				\
-	};						\
-	__register_nmi_handler((t), &fn##_na);	\
-})
-
 int __register_nmi_handler(unsigned int, struct nmiaction *);
 
 void unregister_nmi_handler(unsigned int, const char *);
diff --git a/arch/x86/kernel/nmi_selftest.c b/arch/x86/kernel/nmi_selftest.c
index 149b8d9..e31bf8d 100644
--- a/arch/x86/kernel/nmi_selftest.c
+++ b/arch/x86/kernel/nmi_selftest.c
@@ -42,7 +42,7 @@ static int __init nmi_unk_cb(unsigned int val, struct pt_regs *regs)
 static void __init init_nmi_testsuite(void)
 {
 	/* trap all the unknown NMIs we may generate */
-	register_nmi_handler_initonly(NMI_UNKNOWN, nmi_unk_cb, 0, "nmi_selftest_unk");
+	register_nmi_handler(NMI_UNKNOWN, nmi_unk_cb, 0, "nmi_selftest_unk");
 }
 
 static void __init cleanup_nmi_testsuite(void)
@@ -64,7 +64,7 @@ static void __init test_nmi_ipi(struct cpumask *mask)
 {
 	unsigned long timeout;
 
-	if (register_nmi_handler_initonly(NMI_LOCAL, test_nmi_ipi_callback,
+	if (register_nmi_handler(NMI_LOCAL, test_nmi_ipi_callback,
 				 NMI_FLAG_FIRST, "nmi_selftest")) {
 		nmi_fail = FAILURE;
 		return;
-- 
1.7.7.6

[PATCH 2/2] x86,nmi: Fix section mismatch warnings on 32-bit

[Don Zickus]

From: Li Zhong <zhong@linux.vnet.ibm.com>

It was reported that compiling for 32-bit caused a bunch of section
mismatch warnings:

 VDSOSYM arch/x86/vdso/vdso32-syms.lds
  LD      arch/x86/vdso/built-in.o
  LD      arch/x86/built-in.o
WARNING: arch/x86/built-in.o(.data+0x5af0): Section mismatch in reference from the variable
test_nmi_ipi_callback_na.10451 to the function .init.text:test_nmi_ipi_callback()
The variable test_nmi_ipi_callback_na.10451 references
the function __init test_nmi_ipi_callback()
If the reference is valid then annotate the
variable with __init* or __refdata (see linux/init.h) or name the variable:
*_template, *_timer, *_sht, *_ops, *_probe, *_probe_one, *_console

WARNING: arch/x86/built-in.o(.data+0x5b04): Section mismatch in reference from the variable
nmi_unk_cb_na.10399 to the function .init.text:nmi_unk_cb()
The variable nmi_unk_cb_na.10399 references
the function __init nmi_unk_cb()
If the reference is valid then annotate the
variable with __init* or __refdata (see linux/init.h) or name the variable:
*_template, *_timer, *_sht, *_ops, *_probe, *_probe_one, *_console

Both of these are attributed to the internal representation of the nmiaction
struct created during register_nmi_handler.  The reason for this is that those
structs are not defined in the init section whereas the rest of the code in
nmi_selftest.c is.

This is resolved by passing in another parameter that sets the structure as
initdata or not.

[ Big thanks to Jan Beulich for decoding this for me as I didn't have a clue
what was going on. ]

Reported-by: Witold Baryluk <baryluk@smp.if.uj.edu.pl>
Cc: Jan Beulich <JBeulich@suse.com>
Signed-off-by: Li Zhong <zhong@linux.vnet.ibm.com>
Signed-off-by: Don Zickus <dzickus@redhat.com>
---
 arch/x86/include/asm/nmi.h     |    6 +++---
 arch/x86/kernel/nmi_selftest.c |    5 +++--
 2 files changed, 6 insertions(+), 5 deletions(-)

diff --git a/arch/x86/include/asm/nmi.h b/arch/x86/include/asm/nmi.h
index 0e3793b..c0fa356 100644
--- a/arch/x86/include/asm/nmi.h
+++ b/arch/x86/include/asm/nmi.h
@@ -44,14 +44,14 @@ struct nmiaction {
 	const char		*name;
 };
 
-#define register_nmi_handler(t, fn, fg, n)		\
+#define register_nmi_handler(t, fn, fg, n, init...)	\
 ({							\
-	static struct nmiaction fn##_na = {		\
+	static struct nmiaction init fn##_na = {	\
 		.handler = (fn),			\
 		.name = (n),				\
 		.flags = (fg),				\
 	};						\
-	__register_nmi_handler((t), &fn##_na);	\
+	__register_nmi_handler((t), &fn##_na);		\
 })
 
 int __register_nmi_handler(unsigned int, struct nmiaction *);
diff --git a/arch/x86/kernel/nmi_selftest.c b/arch/x86/kernel/nmi_selftest.c
index e31bf8d..6d9582e 100644
--- a/arch/x86/kernel/nmi_selftest.c
+++ b/arch/x86/kernel/nmi_selftest.c
@@ -42,7 +42,8 @@ static int __init nmi_unk_cb(unsigned int val, struct pt_regs *regs)
 static void __init init_nmi_testsuite(void)
 {
 	/* trap all the unknown NMIs we may generate */
-	register_nmi_handler(NMI_UNKNOWN, nmi_unk_cb, 0, "nmi_selftest_unk");
+	register_nmi_handler(NMI_UNKNOWN, nmi_unk_cb, 0, "nmi_selftest_unk",
+			__initdata);
 }
 
 static void __init cleanup_nmi_testsuite(void)
@@ -65,7 +66,7 @@ static void __init test_nmi_ipi(struct cpumask *mask)
 	unsigned long timeout;
 
 	if (register_nmi_handler(NMI_LOCAL, test_nmi_ipi_callback,
-				 NMI_FLAG_FIRST, "nmi_selftest")) {
+				 NMI_FLAG_FIRST, "nmi_selftest", __initdata)) {
 		nmi_fail = FAILURE;
 		return;
 	}
-- 
1.7.7.6

Re: [PATCH 0/2] x86,nmi: Section header fixes

[Ingo Molnar]

* Don Zickus <dzickus@redhat.com> wrote:

> An earlier attempt at a fix for the problem described in the second patch was
> committed to Ingo's tree.  Later a smaller/cleaner patch was posted.  This
> series just reverts the original patch and then applies the second patch.
> 
> The second patch looks a lot cleaner this way instead of just provding the
> delta patch on top.
> 
> Ingo, let me know if this is not what you want.

Wouldnt this be nicer as a delta enhancement, instead of a 
revert+fix-done-differently?

That way we'd should that the original one was good as well, 
it's just that this one is easier to maintain.

Thanks,

	Ingo

Re: [PATCH 0/2] x86,nmi: Section header fixes

[Don Zickus]

On Mon, Jun 18, 2012 at 11:07:31AM +0200, Ingo Molnar wrote:
> 
> * Don Zickus <dzickus@redhat.com> wrote:
> 
> > An earlier attempt at a fix for the problem described in the second patch was
> > committed to Ingo's tree.  Later a smaller/cleaner patch was posted.  This
> > series just reverts the original patch and then applies the second patch.
> > 
> > The second patch looks a lot cleaner this way instead of just provding the
> > delta patch on top.
> > 
> > Ingo, let me know if this is not what you want.
> 
> Wouldnt this be nicer as a delta enhancement, instead of a 
> revert+fix-done-differently?
> 
> That way we'd should that the original one was good as well, 
> it's just that this one is easier to maintain.

Ok.  I originally had a delta diff, but it looked ugly and hard to
understand the new changes.  That is why I did it this way.  I'll send the
delta enhancement and let you decide.

Cheers,
Don