Don't pass the stack pointer to zImage's start() function

At present, the zImage entry code passes a copy of the stack pointer
to the start() function.  There's no real reason for this; the only
thing start() does with it is print it out.  It appears to be a
leftover debugging hack, so, this patch removes it.

Signed-off-by: David Gibson <redacted>
---

As requested, a cleanup for this wart.  Please apply.

Index: working-2.6/arch/powerpc/boot/crt0.S
===================================================================

--- working-2.6.orig/arch/powerpc/boot/crt0.S	2007-03-15 14:04:14.000000000 +1100
+++ working-2.6/arch/powerpc/boot/crt0.S	2007-03-15 14:04:23.000000000 +1100

@@ -86,5 +86,4 @@ _zimage_start:
 	bl	platform_init
 
 	/* Call start */
-	mr	r3,r1
 	b	start

Index: working-2.6/arch/powerpc/boot/main.c
===================================================================

--- working-2.6.orig/arch/powerpc/boot/main.c	2007-03-15 14:03:09.000000000 +1100
+++ working-2.6/arch/powerpc/boot/main.c	2007-03-15 14:07:05.000000000 +1100

@@ -256,7 +256,7 @@ struct dt_ops dt_ops;
 struct console_ops console_ops;
 struct loader_info loader_info;
 
-void start(void *sp)
+void start(void)
 {
 	struct addr_range vmlinux, initrd;
 	kernel_entry_t kentry;

@@ -268,8 +268,7 @@ void start(void *sp)
 	if (platform_ops.fixups)
 		platform_ops.fixups();
 
-	printf("\n\rzImage starting: loaded at 0x%p (sp: 0x%p)\n\r",
-	       _start, sp);
+	printf("\n\rzImage starting: loaded at 0x%p (sp: 0x%p)\n\r", _start);
 
 	vmlinux = prep_kernel();
 	initrd = prep_initrd(vmlinux, loader_info.initrd_addr,

Index: working-2.6/arch/powerpc/boot/ops.h
===================================================================

--- working-2.6.orig/arch/powerpc/boot/ops.h	2007-03-15 14:07:13.000000000 +1100
+++ working-2.6/arch/powerpc/boot/ops.h	2007-03-15 14:07:17.000000000 +1100

@@ -66,7 +66,7 @@ struct loader_info {
 };
 extern struct loader_info loader_info;
 
-void start(void *sp);
+void start(void);
 int ft_init(void *dt_blob, unsigned int max_size, unsigned int max_find_device);
 int serial_console_init(void);
 int ns16550_console_init(void *devp, struct serial_console_data *scdp);

-- 
David Gibson			| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you.  NOT _the_ _other_
				| _way_ _around_!
http://www.ozlabs.org/~dgibson

On Thu, Mar 15, 2007 at 03:29:04PM +1100, David Gibson wrote:

At present, the zImage entry code passes a copy of the stack pointer
to the start() function.  There's no real reason for this; the only
thing start() does with it is print it out.  It appears to be a
leftover debugging hack, so, this patch removes it.

<snip>

Shouldn't that be:

+	printf("\n\rzImage starting: loaded at 0x%p\n\r", _start);

or similar?

Yours Tony

  linux.conf.au        http://linux.conf.au/ || http://lca2008.linux.org.au/
  Jan 28 - Feb 02 2008 The Australian Linux Technical Conference!

On Thu, Mar 15, 2007 at 03:35:46PM +1100, Tony Breeds wrote:

On Thu, Mar 15, 2007 at 03:29:04PM +1100, David Gibson wrote:

quoted

At present, the zImage entry code passes a copy of the stack pointer
to the start() function.  There's no real reason for this; the only
thing start() does with it is print it out.  It appears to be a
leftover debugging hack, so, this patch removes it.

<snip>

quoted

--- working-2.6.orig/arch/powerpc/boot/main.c	2007-03-15 14:03:09.000000000 +1100
+++ working-2.6/arch/powerpc/boot/main.c	2007-03-15 14:07:05.000000000 +1100

@@ -256,7 +256,7 @@ struct dt_ops dt_ops;
 struct console_ops console_ops;
 struct loader_info loader_info;
 
-void start(void *sp)
+void start(void)
 {
 	struct addr_range vmlinux, initrd;
 	kernel_entry_t kentry;

@@ -268,8 +268,7 @@ void start(void *sp)
 	if (platform_ops.fixups)
 		platform_ops.fixups();
 
-	printf("\n\rzImage starting: loaded at 0x%p (sp: 0x%p)\n\r",
-	       _start, sp);
+	printf("\n\rzImage starting: loaded at 0x%p (sp: 0x%p)\n\r", _start);

Shouldn't that be:

+	printf("\n\rzImage starting: loaded at 0x%p\n\r", _start);

or similar?

Crap, yes, I suck.

Fixed version coming, plus a patch to turn on gcc printf format
warnings for the zImage.

-- 
David Gibson			| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you.  NOT _the_ _other_
				| _way_ _around_!
http://www.ozlabs.org/~dgibson

On Thu, Mar 15, David Gibson wrote:

At present, the zImage entry code passes a copy of the stack pointer
to the start() function.  There's no real reason for this; the only
thing start() does with it is print it out.  It appears to be a
leftover debugging hack, so, this patch removes it.

Signed-off-by: David Gibson <redacted>

NACK.
It helps to determine the firmware memory layout.

On Fri, 2007-03-16 at 19:37 +0100, Olaf Hering wrote:

On Thu, Mar 15, David Gibson wrote:

quoted

At present, the zImage entry code passes a copy of the stack pointer
to the start() function.  There's no real reason for this; the only
thing start() does with it is print it out.  It appears to be a
leftover debugging hack, so, this patch removes it.

Signed-off-by: David Gibson <redacted>

NACK.
It helps to determine the firmware memory layout.

How, why is that useful, and where is that ever used?

josh

On Fri, Mar 16, Josh Boyer wrote:

How, why is that useful, and where is that ever used?

prom_init runs with the stack passed from the firmware (unless that
changed in since 2.6.13). Its better to know where the stack is when
looking at early boot bugreports.