[PATCH] sh: Fix address calculation of Initrd

[PATCH] sh: Fix address calculation of Initrd

[Nobuhiro Iwamatsu]

From: Nobuhiro Iwamatsu <iwamatsu.nobuhiro@renesas.com>

Signed-off-by: Nobuhiro Iwamatsu <nobuhiro.iwamatsu.yj@renesas.com>
---
 arch/sh/kernel/setup.c |    7 ++++---
 1 files changed, 4 insertions(+), 3 deletions(-)

diff --git a/arch/sh/kernel/setup.c b/arch/sh/kernel/setup.c
index f24d0ea..4e27846 100644
--- a/arch/sh/kernel/setup.c
+++ b/arch/sh/kernel/setup.c
@@ -135,8 +135,9 @@ void __init check_for_initrd(void)
 		goto disable;
 	}
 
-	if (unlikely(start < PAGE_OFFSET)) {
-		pr_err("initrd start < PAGE_OFFSET\n");
+	if (unlikely(start < __MEMORY_START)) {
+		pr_err("initrd start (%08lx) < __MEMORY_START(%x)\n",
+			start, __MEMORY_START);
 		goto disable;
 	}
 
@@ -157,7 +158,7 @@ void __init check_for_initrd(void)
 	/*
 	 * Address sanitization
 	 */
-	initrd_start = (unsigned long)__va(__pa(start));
+	initrd_start = (unsigned long)__va(start);
 	initrd_end = initrd_start + INITRD_SIZE;
 
 	memblock_reserve(__pa(initrd_start), INITRD_SIZE);
-- 
1.7.1

Re: [PATCH] sh: Fix address calculation of Initrd

[Paul Mundt]

On Tue, Oct 05, 2010 at 05:16:08PM +0900, Nobuhiro Iwamatsu wrote:
> From: Nobuhiro Iwamatsu <iwamatsu.nobuhiro@renesas.com>
> 
> Signed-off-by: Nobuhiro Iwamatsu <nobuhiro.iwamatsu.yj@renesas.com>

Can you describe what exactly the issue is? Is this broken on 29 bit or
32 bit? Did you test both?

Re: [PATCH] sh: Fix address calculation of Initrd

[Nobuhiro Iwamatsu]

Hi,

2010/10/5 Paul Mundt <lethal@linux-sh.org>:
> On Tue, Oct 05, 2010 at 05:16:08PM +0900, Nobuhiro Iwamatsu wrote:
>> From: Nobuhiro Iwamatsu <iwamatsu.nobuhiro@renesas.com>
>>
>> Signed-off-by: Nobuhiro Iwamatsu <nobuhiro.iwamatsu.yj@renesas.com>
>
> Can you describe what exactly the issue is?

Because there is it on physics memory, start value fails in the check
by PAGE_OFFSET.
And __pa() is not necessary for the same reason.

> Is this broken on 29 bit or
> 32 bit? Did you test both?

Yes, I tested on both.

Best regards,
  Nobuhiro

Re: [PATCH] sh: Fix address calculation of Initrd

[Paul Mundt]

On Tue, Oct 05, 2010 at 05:34:29PM +0900, Nobuhiro Iwamatsu wrote:
> Hi,
> 
> 2010/10/5 Paul Mundt <lethal@linux-sh.org>:
> > On Tue, Oct 05, 2010 at 05:16:08PM +0900, Nobuhiro Iwamatsu wrote:
> >> From: Nobuhiro Iwamatsu <iwamatsu.nobuhiro@renesas.com>
> >>
> >> Signed-off-by: Nobuhiro Iwamatsu <nobuhiro.iwamatsu.yj@renesas.com>
> >
> > Can you describe what exactly the issue is?
> 
> Because there is it on physics memory, start value fails in the check
> by PAGE_OFFSET.
> And __pa() is not necessary for the same reason.
> 
> > Is this broken on 29 bit or
> > 32 bit? Did you test both?
> 
> Yes, I tested on both.
> 
Ok, applied, thanks.