[PATCH] x86, efi: Convert efi_phys_get_time() args to physical addresses

[PATCH] x86, efi: Convert efi_phys_get_time() args to physical addresses

[Matt Fleming]

From: Maurice Ma <maurice.ma@intel.com>

Because callers of efi_phys_get_time() pass virtual stack addresses as
arguments, we need to find their corresponding physical addresses and
when calling GetTime() in physical mode.

Without this patch the following line is printed on boot,

	"Oops: efitime: can't read time!"

Cc: Matthew Garrett <mjg@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: x86@kernel.org
Cc: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Maurice Ma <maurice.ma@intel.com>
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
---
 arch/x86/platform/efi/efi.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/arch/x86/platform/efi/efi.c b/arch/x86/platform/efi/efi.c
index 3ae4128..d182bba 100644
--- a/arch/x86/platform/efi/efi.c
+++ b/arch/x86/platform/efi/efi.c
@@ -237,7 +237,8 @@ static efi_status_t __init phys_efi_get_time(efi_time_t *tm,
 
 	spin_lock_irqsave(&rtc_lock, flags);
 	efi_call_phys_prelog();
-	status = efi_call_phys2(efi_phys.get_time, tm, tc);
+	status = efi_call_phys2(efi_phys.get_time, virt_to_phys(tm),
+				virt_to_phys(tc));
 	efi_call_phys_epilog();
 	spin_unlock_irqrestore(&rtc_lock, flags);
 	return status;
-- 
1.7.4.4

Re: [PATCH] x86, efi: Convert efi_phys_get_time() args to physical addresses

[Matt Fleming]

On Tue, 2011-10-11 at 11:52 +0100, Matt Fleming wrote:
> From: Maurice Ma <maurice.ma@intel.com>
> 
> Because callers of efi_phys_get_time() pass virtual stack addresses as
> arguments, we need to find their corresponding physical addresses and
> when calling GetTime() in physical mode.
> 
> Without this patch the following line is printed on boot,
> 
> 	"Oops: efitime: can't read time!"
> 
> Cc: Matthew Garrett <mjg@redhat.com>
> Cc: Thomas Gleixner <tglx@linutronix.de>
> Cc: Ingo Molnar <mingo@elte.hu>
> Cc: "H. Peter Anvin" <hpa@zytor.com>
> Cc: x86@kernel.org
> Cc: Zhang Rui <rui.zhang@intel.com>
> Signed-off-by: Maurice Ma <maurice.ma@intel.com>
> Signed-off-by: Matt Fleming <matt.fleming@intel.com>
> ---
>  arch/x86/platform/efi/efi.c |    3 ++-
>  1 files changed, 2 insertions(+), 1 deletions(-)

Ping?

Re: [PATCH] x86, efi: Convert efi_phys_get_time() args to physical addresses

[H. Peter Anvin]

On 11/07/2011 03:31 AM, Matt Fleming wrote:
> 
> Ping?
> 

Is this -stable material?

	-hpa

Re: [PATCH] x86, efi: Convert efi_phys_get_time() args to physical addresses

[Matt Fleming]

On Mon, 2011-11-07 at 11:08 -0800, H. Peter Anvin wrote:
> On 11/07/2011 03:31 AM, Matt Fleming wrote:
> > 
> > Ping?
> > 
> 
> Is this -stable material?

I don't think so. It only affects things early in boot (before
efi_enter_virtual_mode()) is called. Once we've called
SetVirtualAddressMap() we call virt_efi_get_time() to get the time and
everything works fine.

I don't think this bug is even visible to the user apart from the "Oops:
efitime: can't read time!" message displayed on boot. It certainly
doesn't cause a crash or anything like that.

-- 
Matt Fleming, Intel Open Source Technology Center

Re: [PATCH] x86, efi: Convert efi_phys_get_time() args to physical addresses

[Matt Fleming]

On Mon, 2011-11-07 at 11:08 -0800, H. Peter Anvin wrote:
> On 11/07/2011 03:31 AM, Matt Fleming wrote:
> > 
> > Ping?
> > 
> 
> Is this -stable material?

Are you picking this up, Peter?

-- 
Matt Fleming, Intel Open Source Technology Center

[tip:x86/asm] x86, efi: Convert efi_phys_get_time() args to physical addresses

[tip-bot for Maurice Ma]

Commit-ID:  e9a9eca517d4cd94e816538efc400257e34bc63e
Gitweb:     http://git.kernel.org/tip/e9a9eca517d4cd94e816538efc400257e34bc63e
Author:     Maurice Ma <maurice.ma@intel.com>
AuthorDate: Tue, 11 Oct 2011 11:52:13 +0100
Committer:  Ingo Molnar <mingo@elte.hu>
CommitDate: Mon, 5 Dec 2011 12:45:14 +0100

x86, efi: Convert efi_phys_get_time() args to physical addresses

Because callers of efi_phys_get_time() pass virtual stack
addresses as arguments, we need to find their corresponding
physical addresses and when calling GetTime() in physical mode.

Without this patch the following line is printed on boot,

	"Oops: efitime: can't read time!"

Signed-off-by: Maurice Ma <maurice.ma@intel.com>
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
Cc: Zhang Rui <rui.zhang@intel.com>
Cc: Matthew Garrett <mjg@redhat.com>
Link: http://lkml.kernel.org/r/1318330333-4617-1-git-send-email-matt@console-pimps.org
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 arch/x86/platform/efi/efi.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/arch/x86/platform/efi/efi.c b/arch/x86/platform/efi/efi.c
index 37718f0..d2376eb 100644
--- a/arch/x86/platform/efi/efi.c
+++ b/arch/x86/platform/efi/efi.c
@@ -238,7 +238,8 @@ static efi_status_t __init phys_efi_get_time(efi_time_t *tm,
 
 	spin_lock_irqsave(&rtc_lock, flags);
 	efi_call_phys_prelog();
-	status = efi_call_phys2(efi_phys.get_time, tm, tc);
+	status = efi_call_phys2(efi_phys.get_time, virt_to_phys(tm),
+				virt_to_phys(tc));
 	efi_call_phys_epilog();
 	spin_unlock_irqrestore(&rtc_lock, flags);
 	return status;