[PATCH] x86: Fix reboot issue for Dell Optiplex 990 V2

[PATCH] x86: Fix reboot issue for Dell Optiplex 990 V2

[Thomas Renninger]

This was the only (reboot=p) option that worked for this desktop.
Default does not reboot as well.

V2: Remove copy and paste left over from another machine

CC: Ismail Donmez <idoenmez@suse.de>
CC: stable@kernel.org
Signed-off-by: Thomas Renninger <trenn@suse.de>
---
 arch/x86/kernel/reboot.c |    9 +++++++++
 1 files changed, 9 insertions(+), 0 deletions(-)

diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c
index e334be1..42f8cd7 100644
--- a/arch/x86/kernel/reboot.c
+++ b/arch/x86/kernel/reboot.c
@@ -443,6 +443,15 @@ static struct dmi_system_id __initdata pci_reboot_dmi_table[] = {
 			DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E6420"),
 		},
 	},
+	{
+		.callback = set_pci_reboot,
+		.ident = "Dell Optiplex 990",
+		.matches = {
+			/* Found in BIOS version A06 */
+			DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 990"),
+		},
+	},
 	{ }
 };
 
-- 
1.7.6.1

Re: [PATCH] x86: Fix reboot issue for Dell Optiplex 990 V2

[Ismail Donmez]

Hi;

On 11/04/2011 03:29 AM, Thomas Renninger wrote:
> This was the only (reboot=p) option that worked for this desktop.
> Default does not reboot as well.
>
> V2: Remove copy and paste left over from another machine
>
> CC: Ismail Donmez<idoenmez@suse.de>
> CC: stable@kernel.org
> Signed-off-by: Thomas Renninger<trenn@suse.de>
> ---
>   arch/x86/kernel/reboot.c |    9 +++++++++
>   1 files changed, 9 insertions(+), 0 deletions(-)
>
> diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c
> index e334be1..42f8cd7 100644
> --- a/arch/x86/kernel/reboot.c
> +++ b/arch/x86/kernel/reboot.c
> @@ -443,6 +443,15 @@ static struct dmi_system_id __initdata pci_reboot_dmi_table[] = {
>   			DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E6420"),
>   		},
>   	},
> +	{
> +		.callback = set_pci_reboot,
> +		.ident = "Dell Optiplex 990",
> +		.matches = {
> +			/* Found in BIOS version A06 */
> +			DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
> +			DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 990"),
> +		},
> +	},
>   	{ }
>   };

I have this machine and can confirm reboot=pci is only way to get it to 
reboot. But also Dell's very recent Latitude E6520 laptop needs the same 
workaround. It would be nice to send a patch for it too.

Regards.

-- 
İsmail Dönmez - openSUSE Booster
SUSE LINUX Products GmbH
Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer, HRB 16746 (AG Nürnberg)

[PATCH] X86: Solve Dell Latitudes do not reboot on x86_64 more generally

[Thomas Renninger]

There are quite some reports that those do not reboot:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/833705
->E6520, E6220
http://www.linux-club.de/viewtopic.php?f=29&t=114459&start=20
->E5520
and some more already are blacklisted

This patch improves the blacklist for all E6xxx and E5xxx Latitudes.

Strange is that on above Ubuntu bug it was mentioned that 32 bit
reboots just fine, which should use the same mechanism than 64 bit
by default (ACPI).

While this should solve the problem on short term for quite some users,
hopefully a more generic way is found sooner or later...


Signed-off-by: Thomas Renninger <trenn@suse.de>
CC: stable@kernel.org
---
 arch/x86/kernel/reboot.c |   20 ++++++--------------
 1 files changed, 6 insertions(+), 14 deletions(-)

diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c
index 42f8cd7..7fcb20a 100644
--- a/arch/x86/kernel/reboot.c
+++ b/arch/x86/kernel/reboot.c
@@ -419,28 +419,20 @@ static struct dmi_system_id __initdata pci_reboot_dmi_table[] = {
 			DMI_MATCH(DMI_PRODUCT_NAME, "iMac9,1"),
 		},
 	},
-	{	/* Handle problems with rebooting on the Latitude E6320. */
-		.callback = set_pci_reboot,
-		.ident = "Dell Latitude E6320",
-		.matches = {
-			DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
-			DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E6320"),
-		},
-	},
-	{	/* Handle problems with rebooting on the Latitude E5420. */
+	{
 		.callback = set_pci_reboot,
-		.ident = "Dell Latitude E5420",
+		.ident = "Dell Latitude E5xxx",
 		.matches = {
 			DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
-			DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E5420"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E5"),
 		},
 	},
-	{	/* Handle problems with rebooting on the Latitude E6420. */
+	{
 		.callback = set_pci_reboot,
-		.ident = "Dell Latitude E6420",
+		.ident = "Dell Latitude E6xxx",
 		.matches = {
 			DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
-			DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E6420"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E6"),
 		},
 	},
 	{
-- 
1.7.6.1

Re: [PATCH] X86: Solve Dell Latitudes do not reboot on x86_64 more generally

[Len Brown]

On 11/04/2011 08:44 PM, Thomas Renninger wrote:

> There are quite some reports that those do not reboot:
> https://bugs.launchpad.net/ubuntu/+source/linux/+bug/833705
> ->E6520, E6220
> http://www.linux-club.de/viewtopic.php?f=29&t=114459&start=20
> ->E5520
> and some more already are blacklisted
> 
> This patch improves the blacklist for all E6xxx and E5xxx Latitudes.
> 
> Strange is that on above Ubuntu bug it was mentioned that 32 bit
> reboots just fine, which should use the same mechanism than 64 bit
> by default (ACPI).
> 
> While this should solve the problem on short term for quite some users,
> hopefully a more generic way is found sooner or later...
> 
> 
> Signed-off-by: Thomas Renninger <trenn@suse.de>
> CC: stable@kernel.org


Acked-by: Len Brown <len.brown@intel.com>

Although this may be an issue with ACPI,
looks like this patch should go though TIP
because of the file changed.

thanks,
-Len


> ---
>  arch/x86/kernel/reboot.c |   20 ++++++--------------
>  1 files changed, 6 insertions(+), 14 deletions(-)
> 
> diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c
> index 42f8cd7..7fcb20a 100644
> --- a/arch/x86/kernel/reboot.c
> +++ b/arch/x86/kernel/reboot.c
> @@ -419,28 +419,20 @@ static struct dmi_system_id __initdata pci_reboot_dmi_table[] = {
>  			DMI_MATCH(DMI_PRODUCT_NAME, "iMac9,1"),
>  		},
>  	},
> -	{	/* Handle problems with rebooting on the Latitude E6320. */
> -		.callback = set_pci_reboot,
> -		.ident = "Dell Latitude E6320",
> -		.matches = {
> -			DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
> -			DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E6320"),
> -		},
> -	},
> -	{	/* Handle problems with rebooting on the Latitude E5420. */
> +	{
>  		.callback = set_pci_reboot,
> -		.ident = "Dell Latitude E5420",
> +		.ident = "Dell Latitude E5xxx",
>  		.matches = {
>  			DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
> -			DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E5420"),
> +			DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E5"),
>  		},
>  	},
> -	{	/* Handle problems with rebooting on the Latitude E6420. */
> +	{
>  		.callback = set_pci_reboot,
> -		.ident = "Dell Latitude E6420",
> +		.ident = "Dell Latitude E6xxx",
>  		.matches = {
>  			DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
> -			DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E6420"),
> +			DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E6"),
>  		},
>  	},
>  	{

Re: [PATCH] X86: Solve Dell Latitudes do not reboot on x86_64 more generally

[Matthew Garrett]

On Sat, Nov 05, 2011 at 01:44:03AM +0100, Thomas Renninger wrote:
> There are quite some reports that those do not reboot:
> https://bugs.launchpad.net/ubuntu/+source/linux/+bug/833705
> ->E6520, E6220
> http://www.linux-club.de/viewtopic.php?f=29&t=114459&start=20
> ->E5520
> and some more already are blacklisted

It's a bug in the SMM code on these machines, and it goes away if you 
disable VT-D. Let's try to actually fix it rather than adding to the 
blacklist, otherwise we'll just end up with another batch of broken 
Dells later on.

-- 
Matthew Garrett | mjg59@srcf.ucam.org

Re: [PATCH] X86: Solve Dell Latitudes do not reboot on x86_64 more generally

[Thomas Renninger]

On Monday, November 07, 2011 02:59:24 PM Matthew Garrett wrote:
> On Sat, Nov 05, 2011 at 01:44:03AM +0100, Thomas Renninger wrote:
> > There are quite some reports that those do not reboot:
> > https://bugs.launchpad.net/ubuntu/+source/linux/+bug/833705
> > ->E6520, E6220
> > http://www.linux-club.de/viewtopic.php?f=29&t=114459&start=20
> > ->E5520
> > and some more already are blacklisted
> 
> It's a bug in the SMM code on these machines, and it goes away if you 
> disable VT-D. Let's try to actually fix it
How do you intend to fix that?
>From description above it sounds as if this problem is not
fixable from OS side.

   Thomas

> rather than adding to the 
> blacklist, otherwise we'll just end up with another batch of broken 
> Dells later on.

Re: [PATCH] X86: Solve Dell Latitudes do not reboot on x86_64 more generally

[Matthew Garrett]

On Mon, Nov 07, 2011 at 03:25:13PM +0100, Thomas Renninger wrote:
> On Monday, November 07, 2011 02:59:24 PM Matthew Garrett wrote:
> > On Sat, Nov 05, 2011 at 01:44:03AM +0100, Thomas Renninger wrote:
> > > There are quite some reports that those do not reboot:
> > > https://bugs.launchpad.net/ubuntu/+source/linux/+bug/833705
> > > ->E6520, E6220
> > > http://www.linux-club.de/viewtopic.php?f=29&t=114459&start=20
> > > ->E5520
> > > and some more already are blacklisted
> > 
> > It's a bug in the SMM code on these machines, and it goes away if you 
> > disable VT-D. Let's try to actually fix it
> How do you intend to fix that?
> From description above it sounds as if this problem is not
> fixable from OS side.

Tear down VT-D state before reboot.

-- 
Matthew Garrett | mjg59@srcf.ucam.org

Re: [PATCH] X86: Solve Dell Latitudes do not reboot on x86_64 more generally

[Thomas Renninger]

On Monday, November 07, 2011 03:27:40 PM Matthew Garrett wrote:
> On Mon, Nov 07, 2011 at 03:25:13PM +0100, Thomas Renninger wrote:
> > On Monday, November 07, 2011 02:59:24 PM Matthew Garrett wrote:
> > > On Sat, Nov 05, 2011 at 01:44:03AM +0100, Thomas Renninger wrote:
> > > > There are quite some reports that those do not reboot:
> > > > https://bugs.launchpad.net/ubuntu/+source/linux/+bug/833705
> > > > ->E6520, E6220
> > > > http://www.linux-club.de/viewtopic.php?f=29&t=114459&start=20
> > > > ->E5520
> > > > and some more already are blacklisted
> > > 
> > > It's a bug in the SMM code on these machines, and it goes away if you 
> > > disable VT-D. Let's try to actually fix it
> > How do you intend to fix that?
> > From description above it sounds as if this problem is not
> > fixable from OS side.
> 
> Tear down VT-D state before reboot.
But doing this for all X86 machines because of a handful of
Dells with broken SMM code sounds wrong as well.

Doesn't Dell ship with Ubuntu pre-loaded?
Can't someone point them to this to get this fixed in BIOS?

   Thomas

Re: [PATCH] X86: Solve Dell Latitudes do not reboot on x86_64 more generally

[Matthew Garrett]

On Mon, Nov 07, 2011 at 04:09:54PM +0100, Thomas Renninger wrote:
> On Monday, November 07, 2011 03:27:40 PM Matthew Garrett wrote:
> > Tear down VT-D state before reboot.
> But doing this for all X86 machines because of a handful of
> Dells with broken SMM code sounds wrong as well.

We've already torn everything down at this point. What's the downside?

> Doesn't Dell ship with Ubuntu pre-loaded?
> Can't someone point them to this to get this fixed in BIOS?

I've been trying.

-- 
Matthew Garrett | mjg59@srcf.ucam.org

Re: [PATCH] X86: Solve Dell Latitudes do not reboot on x86_64 more generally

[Len Brown]

The SMM bugs relating to CPU0 pre-date the existence of VT-D

It is the same old story...
Linux is on thin-ice exercising FW that has been validated only
under (some version of) Windows.
-- 
Len Brown, Intel Open Source Technology Center

Re: [PATCH] x86: Fix reboot issue for Dell Optiplex 990 V2

[Valdis.Kletnieks]

[-- Attachment #1: Type: text/plain, Size: 741 bytes --]

On Fri, 04 Nov 2011 08:16:09 BST, Ismail Donmez said:
> On 11/04/2011 03:29 AM, Thomas Renninger wrote:
> > This was the only (reboot=p) option that worked for this desktop.
> > Default does not reboot as well.

> >   			DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E6420"),

> > +			DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 990"),

> I have this machine and can confirm reboot=pci is only way to get it to 
> reboot. But also Dell's very recent Latitude E6520 laptop needs the same 
> workaround. It would be nice to send a patch for it too.

Just as another data point, my Latitude E6500 does *not* need this, as it's perfectly
happy using the default reboot method.  So please avoid the temptation to do
a match on "Latitude E6" ;)

[-- Attachment #2: Type: application/pgp-signature, Size: 227 bytes --]