Re: 答复: 答复: [PATCH] iommu/vt-d: Don't reject nvme host due to scope mismatch

[Lu Baolu <baolu.lu@linux.intel.com>]

Hi Jim,

On 2019/12/24 11:24, Jim,Yan wrote:
>> -----邮件原件-----
>> 发件人: Lu Baolu [mailto:baolu.lu@linux.intel.com]
>> 发送时间: 2019年12月23日 21:05
>> 收件人: Jim,Yan <jimyan@baidu.com>; Jerry Snitselaar <jsnitsel@redhat.com>
>> 抄送: iommu@lists.linux-foundation.org; linux-kernel@vger.kernel.org
>> 主题: Re: 答复: [PATCH] iommu/vt-d: Don't reject nvme host due to scope
>> mismatch
>>
>> Hi,
>>
>> On 2019/12/23 15:59, Jim,Yan wrote:
>>>> -----邮件原件-----
>>>> 发件人: Jerry Snitselaar [mailto:jsnitsel@redhat.com]
>>>> 发送时间: 2019年12月20日 17:23
>>>> 收件人: Jim,Yan <jimyan@baidu.com>
>>>> 抄送: joro@8bytes.org; iommu@lists.linux-foundation.org;
>>>> linux-kernel@vger.kernel.org
>>>> 主题: Re: [PATCH] iommu/vt-d: Don't reject nvme host due to scope
>>>> mismatch
>>>>
>>>> On Fri Dec 20 19, jimyan wrote:
>>>>> On a system with an Intel PCIe port configured as a nvme host
>>>>> device, iommu initialization fails with
>>>>>
>>>>>      DMAR: Device scope type does not match for 0000:80:00.0
>>>>>
>>>>> This is because the DMAR table reports this device as having scope 2
>>>>> (ACPI_DMAR_SCOPE_TYPE_BRIDGE):
>>>>>
>>>>
>>>> Isn't that a problem to be fixed in the DMAR table then?
>>>>
>>>>> but the device has a type 0 PCI header:
>>>>> 80:00.0 Class 0600: Device 8086:2020 (rev 06)
>>>>> 00: 86 80 20 20 47 05 10 00 06 00 00 06 10 00 00 00
>>>>> 10: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>>>>> 20: 00 00 00 00 00 00 00 00 00 00 00 00 86 80 00 00
>>>>> 30: 00 00 00 00 90 00 00 00 00 00 00 00 00 01 00 00
>>>>>
>>>>> VT-d works perfectly on this system, so there's no reason to bail
>>>>> out on initialization due to this apparent scope mismatch. Add the
>>>>> class
>>>>> 0x600 ("PCI_CLASS_BRIDGE_HOST") as a heuristic for allowing DMAR
>>>>> initialization for non-bridge PCI devices listed with scope bridge.
>>>>>
>>>>> Signed-off-by: jimyan <jimyan@baidu.com>
>>>>> ---
>>>>> drivers/iommu/dmar.c | 1 +
>>>>> 1 file changed, 1 insertion(+)
>>>>>
>>>>> diff --git a/drivers/iommu/dmar.c b/drivers/iommu/dmar.c index
>>>>> eecd6a421667..9faf2f0e0237 100644
>>>>> --- a/drivers/iommu/dmar.c
>>>>> +++ b/drivers/iommu/dmar.c
>>>>> @@ -244,6 +244,7 @@ int dmar_insert_dev_scope(struct
>>>> dmar_pci_notify_info *info,
>>>>> 		     info->dev->hdr_type != PCI_HEADER_TYPE_NORMAL) ||
>>>>> 		    (scope->entry_type == ACPI_DMAR_SCOPE_TYPE_BRIDGE
>> &&
>>>>> 		     (info->dev->hdr_type == PCI_HEADER_TYPE_NORMAL &&
>>>>> +			  info->dev->class >> 8 != PCI_CLASS_BRIDGE_HOST &&
>>>>> 		      info->dev->class >> 8 != PCI_CLASS_BRIDGE_OTHER))) {
>>>>> 			pr_warn("Device scope type does not match for %s\n",
>>>>> 				pci_name(info->dev));
>>>>> --
>>>>> 2.11.0
>>>>>
>>>>> _______________________________________________
>>>>> iommu mailing list
>>>>> iommu@lists.linux-foundation.org
>>>>> https://lists.linuxfoundation.org/mailman/listinfo/iommu
>>>>>
>>> Actually this patch is similar to the commit: ffb2d1eb88c3("iommu/vt-d: Don't
>> reject NTB devices due to scope mismatch"). Besides, modifying DMAR table
>> need OEM update BIOS. It is hard to implement.
>>>
>>
>> For both cases, a quirk flag seems to be more reasonable, so that unrelated
>> devices will not be impacted.
>>
>> Best regards,
>> baolu
> 
> Hi Baolu,
> 	Thanks for your advice. And I modify the patch as follow.

I just posted a patch for both NTG and NVME cases. Can you please take a
look? Does it work for you?

Best regards,
baolu

> 
>      On a system with an Intel PCIe port configured as a nvme host device, iommu
>      initialization fails with
>      
>          DMAR: Device scope type does not match for 0000:80:00.0
>      
>      This is because the DMAR table reports this device as having scope 2
>      (ACPI_DMAR_SCOPE_TYPE_BRIDGE):
>      
>      but the device has a type 0 PCI header:
>      80:00.0 Class 0600: Device 8086:2020 (rev 06)
>      00: 86 80 20 20 47 05 10 00 06 00 00 06 10 00 00 00
>      10: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>      20: 00 00 00 00 00 00 00 00 00 00 00 00 86 80 00 00
>      30: 00 00 00 00 90 00 00 00 00 00 00 00 00 01 00 00
>      
>      VT-d works perfectly on this system, so there's no reason to bail out
>      on initialization due to this apparent scope mismatch. Add the class
>      0x06 ("PCI_BASE_CLASS_BRIDGE") as a heuristic for allowing DMAR
>      initialization for non-bridge PCI devices listed with scope bridge.
>      
>      Signed-off-by: jimyan <jimyan@baidu.com>
> 
> diff --git a/drivers/iommu/dmar.c b/drivers/iommu/dmar.c
> index eecd6a421667..50c92eb23ee4 100644
> --- a/drivers/iommu/dmar.c
> +++ b/drivers/iommu/dmar.c
> @@ -244,7 +244,7 @@ int dmar_insert_dev_scope(struct dmar_pci_notify_info *info,
>                       info->dev->hdr_type != PCI_HEADER_TYPE_NORMAL) ||
>                      (scope->entry_type == ACPI_DMAR_SCOPE_TYPE_BRIDGE &&
>                       (info->dev->hdr_type == PCI_HEADER_TYPE_NORMAL &&
> -                     info->dev->class >> 8 != PCI_CLASS_BRIDGE_OTHER))) {
> +                     info->dev->class >> 16 != PCI_BASE_CLASS_BRIDGE))) {
>                          pr_warn("Device scope type does not match for %s\n",
>                                  pci_name(info->dev));
>                          return -EINVAL;
> 
> 
> Jim
>