public inbox for linux-acpi@vger.kernel.org
 help / color / mirror / Atom feed
* AER/hotplug _OSC wierdness?
@ 2009-12-07  3:55 Robert Hancock
  2009-12-07  9:26 ` Kenji Kaneshige
  2009-12-07 10:55 ` Kenji Kaneshige
  0 siblings, 2 replies; 5+ messages in thread
From: Robert Hancock @ 2009-12-07  3:55 UTC (permalink / raw)
  To: linux-acpi; +Cc: kaneshige.kenji

I've been looking into why I get these messages on bootup on an Asus 
P7P55D PRO motherboard:

Firmware did not grant requested _OSC control
aer 0000:00:03.0:pcie02: AER service couldn't init device: no _OSC support
pci_hotplug: PCI Hot Plug PCI Core version: 0.5
Firmware did not grant requested _OSC control
Firmware did not grant requested _OSC control
Firmware did not grant requested _OSC control
Firmware did not grant requested _OSC control

The AML code for _OSC in the DSDT (below) is almost identical to the 
example code in the ACPI spec. From looking at the _OSC method, I don't 
see why it should reject any requests for PCI Express hotplug or AER 
control.

I can't quite make sense of the ACPI code that is supposed to handle 
this. When the AER driver calls acpi_pci_osc_control_set, it first does 
this:

	/* Need to query controls first before requesting them */
	if (!root->osc_queried) {
		status = acpi_pci_query_osc(root, root->osc_support_set);
		if (ACPI_FAILURE(status))
			goto out;
	}
	if ((root->osc_control_qry & control_req) != control_req) {
		printk(KERN_DEBUG
		       "Firmware did not grant requested _OSC control\n");
		status = AE_SUPPORT;
		goto out;
	}

Inside acpi_pci_query_osc:

	/* do _OSC query for all possible controls */
	support_set = root->osc_support_set | (flags & OSC_SUPPORT_MASKS);
	capbuf[OSC_QUERY_TYPE] = OSC_QUERY_ENABLE;
	capbuf[OSC_SUPPORT_TYPE] = support_set;
	capbuf[OSC_CONTROL_TYPE] = OSC_CONTROL_MASKS;

	status = acpi_pci_run_osc(root->device->handle, capbuf, &result);
	if (ACPI_SUCCESS(status)) {
		root->osc_support_set = support_set;
		root->osc_control_qry = result;
		root->osc_queried = 1;
	}

The comment says "do _OSC query for all possible controls", but it 
doesn't look like that's what the code is actually doing. The first time 
this gets called, assuming osc_support_set is 0, it looks like 
support_set passed into _OSC will also be 0. Then osc_queried would be 
set to 1, _OSC would never get called again and the cached query result 
is based on 0 support flags being passed into _OSC, which is wrong. Or 
am I missing something?

     Name (PEHP, One)
     Name (SHPC, Zero)
     Name (PEPM, One)
     Name (PEER, One)
     Name (PECS, One)

...

         Method (_OSC, 4, NotSerialized)
         {
             Name (SUPP, Zero)
             Name (CTRL, Zero)
             CreateDWordField (Arg3, Zero, CDW1)
             CreateDWordField (Arg3, 0x04, CDW2)
             CreateDWordField (Arg3, 0x08, CDW3)
             If (LEqual (Arg0, Buffer (0x10)
                     {
                         /* 0000 */    0x5B, 0x4D, 0xDB, 0x33, 0xF7, 
0x1F, 0x1C, 0x40,
                         /* 0008 */    0x96, 0x57, 0x74, 0x41, 0xC0, 
0x3D, 0xD7, 0x66
                     }))
             {
                 Store (CDW2, SUPP)
                 Store (CDW3, CTRL)
                 If (LNotEqual (And (SUPP, 0x16), 0x16))
                 {
                     And (CTRL, 0x1E, CTRL)
                 }

                 If (LNot (PEHP))
                 {
                     And (CTRL, 0x1E, CTRL)
                 }

                 If (LNot (SHPC))
                 {
                     And (CTRL, 0x1D, CTRL)
                 }

                 If (LNot (PEPM))
                 {
                     And (CTRL, 0x1B, CTRL)
                 }

                 If (LNot (PEER))
                 {
                     And (CTRL, 0x15, CTRL)
                 }

                 If (LNot (PECS))
                 {
                     And (CTRL, 0x0F, CTRL)
                 }

                 If (Not (And (CDW1, One)))
                 {
                     If (And (CTRL, One)) {}
                     If (And (CTRL, 0x04)) {}
                     If (And (CTRL, 0x10)) {}
                 }

                 If (LNotEqual (Arg1, One))
                 {
                     Or (CDW1, 0x08, CDW1)
                 }

                 If (LNotEqual (CDW3, CTRL))
                 {
                     Or (CDW1, 0x10, CDW1)
                 }

                 And (CTRL, 0xEF, CTRL)
                 Store (CTRL, CDW3)
                 Return (Arg3)
             }
             Else
             {
                 Or (CDW1, 0x04, CDW1)
                 Return (Arg3)
             }
         }


^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: AER/hotplug _OSC wierdness?
  2009-12-07  3:55 AER/hotplug _OSC wierdness? Robert Hancock
@ 2009-12-07  9:26 ` Kenji Kaneshige
  2009-12-08  2:37   ` Robert Hancock
  2009-12-07 10:55 ` Kenji Kaneshige
  1 sibling, 1 reply; 5+ messages in thread
From: Kenji Kaneshige @ 2009-12-07  9:26 UTC (permalink / raw)
  To: Robert Hancock; +Cc: linux-acpi

Robert Hancock wrote:
> I've been looking into why I get these messages on bootup on an Asus 
> P7P55D PRO motherboard:
> 
> Firmware did not grant requested _OSC control
> aer 0000:00:03.0:pcie02: AER service couldn't init device: no _OSC support
> pci_hotplug: PCI Hot Plug PCI Core version: 0.5
> Firmware did not grant requested _OSC control
> Firmware did not grant requested _OSC control
> Firmware did not grant requested _OSC control
> Firmware did not grant requested _OSC control
> 
> The AML code for _OSC in the DSDT (below) is almost identical to the 
> example code in the ACPI spec. From looking at the _OSC method, I don't 
> see why it should reject any requests for PCI Express hotplug or AER 
> control.
> 
> I can't quite make sense of the ACPI code that is supposed to handle 
> this. When the AER driver calls acpi_pci_osc_control_set, it first does 
> this:
> 
>     /* Need to query controls first before requesting them */
>     if (!root->osc_queried) {
>         status = acpi_pci_query_osc(root, root->osc_support_set);
>         if (ACPI_FAILURE(status))
>             goto out;
>     }


Just in case, does removing this 'if' statement fix the problem?


>     if ((root->osc_control_qry & control_req) != control_req) {
>         printk(KERN_DEBUG
>                "Firmware did not grant requested _OSC control\n");
>         status = AE_SUPPORT;
>         goto out;
>     }
> 
> Inside acpi_pci_query_osc:
> 
>     /* do _OSC query for all possible controls */
>     support_set = root->osc_support_set | (flags & OSC_SUPPORT_MASKS);
>     capbuf[OSC_QUERY_TYPE] = OSC_QUERY_ENABLE;
>     capbuf[OSC_SUPPORT_TYPE] = support_set;
>     capbuf[OSC_CONTROL_TYPE] = OSC_CONTROL_MASKS;
> 
>     status = acpi_pci_run_osc(root->device->handle, capbuf, &result);
>     if (ACPI_SUCCESS(status)) {
>         root->osc_support_set = support_set;
>         root->osc_control_qry = result;
>         root->osc_queried = 1;
>     }
> 
> The comment says "do _OSC query for all possible controls", but it 
> doesn't look like that's what the code is actually doing. The first time 
> this gets called, assuming osc_support_set is 0, it looks like 
> support_set passed into _OSC will also be 0. Then osc_queried would be 
> set to 1, _OSC would never get called again and the cached query result 
> is based on 0 support flags being passed into _OSC, which is wrong. Or 
> am I missing something?
> 

Though I might be missing something, I think _OSC is evaluated
and osc_control_qry is updated, whenever acpi_pci_osc_support()
gets called.

Thanks,
Kenji Kaneshige


^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: AER/hotplug _OSC wierdness?
  2009-12-07  3:55 AER/hotplug _OSC wierdness? Robert Hancock
  2009-12-07  9:26 ` Kenji Kaneshige
@ 2009-12-07 10:55 ` Kenji Kaneshige
  2009-12-08  2:57   ` Robert Hancock
  1 sibling, 1 reply; 5+ messages in thread
From: Kenji Kaneshige @ 2009-12-07 10:55 UTC (permalink / raw)
  To: Robert Hancock; +Cc: linux-acpi

Robert Hancock wrote:
> I've been looking into why I get these messages on bootup on an Asus 
> P7P55D PRO motherboard:
> 
> Firmware did not grant requested _OSC control
> aer 0000:00:03.0:pcie02: AER service couldn't init device: no _OSC support
> pci_hotplug: PCI Hot Plug PCI Core version: 0.5
> Firmware did not grant requested _OSC control
> Firmware did not grant requested _OSC control
> Firmware did not grant requested _OSC control
> Firmware did not grant requested _OSC control
> 
> The AML code for _OSC in the DSDT (below) is almost identical to the 
> example code in the ACPI spec. From looking at the _OSC method, I don't 
> see why it should reject any requests for PCI Express hotplug or AER 
> control.
> 
> I can't quite make sense of the ACPI code that is supposed to handle 
> this. When the AER driver calls acpi_pci_osc_control_set, it first does 
> this:
> 
>     /* Need to query controls first before requesting them */
>     if (!root->osc_queried) {
>         status = acpi_pci_query_osc(root, root->osc_support_set);
>         if (ACPI_FAILURE(status))
>             goto out;
>     }
>     if ((root->osc_control_qry & control_req) != control_req) {
>         printk(KERN_DEBUG
>                "Firmware did not grant requested _OSC control\n");
>         status = AE_SUPPORT;
>         goto out;
>     }
> 
> Inside acpi_pci_query_osc:
> 
>     /* do _OSC query for all possible controls */
>     support_set = root->osc_support_set | (flags & OSC_SUPPORT_MASKS);
>     capbuf[OSC_QUERY_TYPE] = OSC_QUERY_ENABLE;
>     capbuf[OSC_SUPPORT_TYPE] = support_set;
>     capbuf[OSC_CONTROL_TYPE] = OSC_CONTROL_MASKS;
> 
>     status = acpi_pci_run_osc(root->device->handle, capbuf, &result);
>     if (ACPI_SUCCESS(status)) {
>         root->osc_support_set = support_set;
>         root->osc_control_qry = result;
>         root->osc_queried = 1;
>     }
> 
> The comment says "do _OSC query for all possible controls", but it 
> doesn't look like that's what the code is actually doing. The first time 
> this gets called, assuming osc_support_set is 0, it looks like 
> support_set passed into _OSC will also be 0. Then osc_queried would be 
> set to 1, _OSC would never get called again and the cached query result 
> is based on 0 support flags being passed into _OSC, which is wrong. Or 
> am I missing something?
> 
>     Name (PEHP, One)
>     Name (SHPC, Zero)
>     Name (PEPM, One)
>     Name (PEER, One)
>     Name (PECS, One)
> 
> ...
> 
>         Method (_OSC, 4, NotSerialized)
>         {
>             Name (SUPP, Zero)
>             Name (CTRL, Zero)
>             CreateDWordField (Arg3, Zero, CDW1)
>             CreateDWordField (Arg3, 0x04, CDW2)
>             CreateDWordField (Arg3, 0x08, CDW3)
>             If (LEqual (Arg0, Buffer (0x10)
>                     {
>                         /* 0000 */    0x5B, 0x4D, 0xDB, 0x33, 0xF7, 
> 0x1F, 0x1C, 0x40,
>                         /* 0008 */    0x96, 0x57, 0x74, 0x41, 0xC0, 
> 0x3D, 0xD7, 0x66
>                     }))
>             {
>                 Store (CDW2, SUPP)
>                 Store (CDW3, CTRL)
>                 If (LNotEqual (And (SUPP, 0x16), 0x16))
>                 {
>                     And (CTRL, 0x1E, CTRL)
>                 }
> 
>                 If (LNot (PEHP))
>                 {
>                     And (CTRL, 0x1E, CTRL)
>                 }
> 
>                 If (LNot (SHPC))
>                 {
>                     And (CTRL, 0x1D, CTRL)
>                 }
> 
>                 If (LNot (PEPM))
>                 {
>                     And (CTRL, 0x1B, CTRL)
>                 }
> 
>                 If (LNot (PEER))
>                 {
>                     And (CTRL, 0x15, CTRL)
>                 }
> 
>                 If (LNot (PECS))
>                 {
>                     And (CTRL, 0x0F, CTRL)
>                 }
> 
>                 If (Not (And (CDW1, One)))
>                 {
>                     If (And (CTRL, One)) {}
>                     If (And (CTRL, 0x04)) {}
>                     If (And (CTRL, 0x10)) {}
>                 }
> 
>                 If (LNotEqual (Arg1, One))
>                 {
>                     Or (CDW1, 0x08, CDW1)
>                 }
> 
>                 If (LNotEqual (CDW3, CTRL))
>                 {
>                     Or (CDW1, 0x10, CDW1)
>                 }
> 
>                 And (CTRL, 0xEF, CTRL)
>                 Store (CTRL, CDW3)

Your ACPI DSDT seems to clear "PCI Express Capability Structure
control" bit in returned value. Both AER and hotplug driver
requests this control. I think this is the reason why you
encountered "Firmware did not grant requested _OSC control".

Please double check.

Thanks,
Kenji Kaneshige



^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: AER/hotplug _OSC wierdness?
  2009-12-07  9:26 ` Kenji Kaneshige
@ 2009-12-08  2:37   ` Robert Hancock
  0 siblings, 0 replies; 5+ messages in thread
From: Robert Hancock @ 2009-12-08  2:37 UTC (permalink / raw)
  To: Kenji Kaneshige; +Cc: linux-acpi

On Mon, Dec 7, 2009 at 3:26 AM, Kenji Kaneshige
<kaneshige.kenji@jp.fujitsu.com> wrote:
> Robert Hancock wrote:
>>
>> I've been looking into why I get these messages on bootup on an Asus
>> P7P55D PRO motherboard:
>>
>> Firmware did not grant requested _OSC control
>> aer 0000:00:03.0:pcie02: AER service couldn't init device: no _OSC support
>> pci_hotplug: PCI Hot Plug PCI Core version: 0.5
>> Firmware did not grant requested _OSC control
>> Firmware did not grant requested _OSC control
>> Firmware did not grant requested _OSC control
>> Firmware did not grant requested _OSC control
>>
>> The AML code for _OSC in the DSDT (below) is almost identical to the
>> example code in the ACPI spec. From looking at the _OSC method, I don't see
>> why it should reject any requests for PCI Express hotplug or AER control.
>>
>> I can't quite make sense of the ACPI code that is supposed to handle this.
>> When the AER driver calls acpi_pci_osc_control_set, it first does this:
>>
>>    /* Need to query controls first before requesting them */
>>    if (!root->osc_queried) {
>>        status = acpi_pci_query_osc(root, root->osc_support_set);
>>        if (ACPI_FAILURE(status))
>>            goto out;
>>    }
>
>
> Just in case, does removing this 'if' statement fix the problem?
>
>
>>    if ((root->osc_control_qry & control_req) != control_req) {
>>        printk(KERN_DEBUG
>>               "Firmware did not grant requested _OSC control\n");
>>        status = AE_SUPPORT;
>>        goto out;
>>    }
>>
>> Inside acpi_pci_query_osc:
>>
>>    /* do _OSC query for all possible controls */
>>    support_set = root->osc_support_set | (flags & OSC_SUPPORT_MASKS);
>>    capbuf[OSC_QUERY_TYPE] = OSC_QUERY_ENABLE;
>>    capbuf[OSC_SUPPORT_TYPE] = support_set;
>>    capbuf[OSC_CONTROL_TYPE] = OSC_CONTROL_MASKS;
>>
>>    status = acpi_pci_run_osc(root->device->handle, capbuf, &result);
>>    if (ACPI_SUCCESS(status)) {
>>        root->osc_support_set = support_set;
>>        root->osc_control_qry = result;
>>        root->osc_queried = 1;
>>    }
>>
>> The comment says "do _OSC query for all possible controls", but it doesn't
>> look like that's what the code is actually doing. The first time this gets
>> called, assuming osc_support_set is 0, it looks like support_set passed into
>> _OSC will also be 0. Then osc_queried would be set to 1, _OSC would never
>> get called again and the cached query result is based on 0 support flags
>> being passed into _OSC, which is wrong. Or am I missing something?
>>
>
> Though I might be missing something, I think _OSC is evaluated
> and osc_control_qry is updated, whenever acpi_pci_osc_support()
> gets called.

Ahh, I missed that acpi_pci_query_osc is also called from acpi_pci_osc_support.
--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: AER/hotplug _OSC wierdness?
  2009-12-07 10:55 ` Kenji Kaneshige
@ 2009-12-08  2:57   ` Robert Hancock
  0 siblings, 0 replies; 5+ messages in thread
From: Robert Hancock @ 2009-12-08  2:57 UTC (permalink / raw)
  To: Kenji Kaneshige; +Cc: linux-acpi

On Mon, Dec 7, 2009 at 4:55 AM, Kenji Kaneshige
<kaneshige.kenji@jp.fujitsu.com> wrote:
> Robert Hancock wrote:
>>
>> I've been looking into why I get these messages on bootup on an Asus
>> P7P55D PRO motherboard:
>>
>> Firmware did not grant requested _OSC control
>> aer 0000:00:03.0:pcie02: AER service couldn't init device: no _OSC support
>> pci_hotplug: PCI Hot Plug PCI Core version: 0.5
>> Firmware did not grant requested _OSC control
>> Firmware did not grant requested _OSC control
>> Firmware did not grant requested _OSC control
>> Firmware did not grant requested _OSC control
>>
>> The AML code for _OSC in the DSDT (below) is almost identical to the
>> example code in the ACPI spec. From looking at the _OSC method, I don't see
>> why it should reject any requests for PCI Express hotplug or AER control.
>>
>> I can't quite make sense of the ACPI code that is supposed to handle this.
>> When the AER driver calls acpi_pci_osc_control_set, it first does this:
>>
>>    /* Need to query controls first before requesting them */
>>    if (!root->osc_queried) {
>>        status = acpi_pci_query_osc(root, root->osc_support_set);
>>        if (ACPI_FAILURE(status))
>>            goto out;
>>    }
>>    if ((root->osc_control_qry & control_req) != control_req) {
>>        printk(KERN_DEBUG
>>               "Firmware did not grant requested _OSC control\n");
>>        status = AE_SUPPORT;
>>        goto out;
>>    }
>>
>> Inside acpi_pci_query_osc:
>>
>>    /* do _OSC query for all possible controls */
>>    support_set = root->osc_support_set | (flags & OSC_SUPPORT_MASKS);
>>    capbuf[OSC_QUERY_TYPE] = OSC_QUERY_ENABLE;
>>    capbuf[OSC_SUPPORT_TYPE] = support_set;
>>    capbuf[OSC_CONTROL_TYPE] = OSC_CONTROL_MASKS;
>>
>>    status = acpi_pci_run_osc(root->device->handle, capbuf, &result);
>>    if (ACPI_SUCCESS(status)) {
>>        root->osc_support_set = support_set;
>>        root->osc_control_qry = result;
>>        root->osc_queried = 1;
>>    }
>>
>> The comment says "do _OSC query for all possible controls", but it doesn't
>> look like that's what the code is actually doing. The first time this gets
>> called, assuming osc_support_set is 0, it looks like support_set passed into
>> _OSC will also be 0. Then osc_queried would be set to 1, _OSC would never
>> get called again and the cached query result is based on 0 support flags
>> being passed into _OSC, which is wrong. Or am I missing something?
>>
>>    Name (PEHP, One)
>>    Name (SHPC, Zero)
>>    Name (PEPM, One)
>>    Name (PEER, One)
>>    Name (PECS, One)
>>
>> ...
>>
>>        Method (_OSC, 4, NotSerialized)
>>        {
>>            Name (SUPP, Zero)
>>            Name (CTRL, Zero)
>>            CreateDWordField (Arg3, Zero, CDW1)
>>            CreateDWordField (Arg3, 0x04, CDW2)
>>            CreateDWordField (Arg3, 0x08, CDW3)
>>            If (LEqual (Arg0, Buffer (0x10)
>>                    {
>>                        /* 0000 */    0x5B, 0x4D, 0xDB, 0x33, 0xF7, 0x1F,
>> 0x1C, 0x40,
>>                        /* 0008 */    0x96, 0x57, 0x74, 0x41, 0xC0, 0x3D,
>> 0xD7, 0x66
>>                    }))
>>            {
>>                Store (CDW2, SUPP)
>>                Store (CDW3, CTRL)
>>                If (LNotEqual (And (SUPP, 0x16), 0x16))
>>                {
>>                    And (CTRL, 0x1E, CTRL)
>>                }
>>
>>                If (LNot (PEHP))
>>                {
>>                    And (CTRL, 0x1E, CTRL)
>>                }
>>
>>                If (LNot (SHPC))
>>                {
>>                    And (CTRL, 0x1D, CTRL)
>>                }
>>
>>                If (LNot (PEPM))
>>                {
>>                    And (CTRL, 0x1B, CTRL)
>>                }
>>
>>                If (LNot (PEER))
>>                {
>>                    And (CTRL, 0x15, CTRL)
>>                }
>>
>>                If (LNot (PECS))
>>                {
>>                    And (CTRL, 0x0F, CTRL)
>>                }
>>
>>                If (Not (And (CDW1, One)))
>>                {
>>                    If (And (CTRL, One)) {}
>>                    If (And (CTRL, 0x04)) {}
>>                    If (And (CTRL, 0x10)) {}
>>                }
>>
>>                If (LNotEqual (Arg1, One))
>>                {
>>                    Or (CDW1, 0x08, CDW1)
>>                }
>>
>>                If (LNotEqual (CDW3, CTRL))
>>                {
>>                    Or (CDW1, 0x10, CDW1)
>>                }
>>
>>                And (CTRL, 0xEF, CTRL)
>>                Store (CTRL, CDW3)
>
> Your ACPI DSDT seems to clear "PCI Express Capability Structure
> control" bit in returned value. Both AER and hotplug driver
> requests this control. I think this is the reason why you
> encountered "Firmware did not grant requested _OSC control".
>
> Please double check.

You're right here as well, I missed which bit that And statement was clearing.
--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2009-12-08  2:57 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2009-12-07  3:55 AER/hotplug _OSC wierdness? Robert Hancock
2009-12-07  9:26 ` Kenji Kaneshige
2009-12-08  2:37   ` Robert Hancock
2009-12-07 10:55 ` Kenji Kaneshige
2009-12-08  2:57   ` Robert Hancock

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox