linux-kernel - Re: [PATCH 2/2] vfio: type1: conditionally check MSI remapping at irq domain level

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Thu, 2 Mar 2017 14:46:15 +0100
From:   Auger Eric <eric.auger@...hat.com>
To:     Mian Yousaf Kaukab <yousaf.kaukab@...e.com>,
        linux-kernel@...r.kernel.org, kvm@...r.kernel.org,
        marc.zyngier@....com, alex.williamson@...hat.com
Cc:     will.deacon@....com
Subject: Re: [PATCH 2/2] vfio: type1: conditionally check MSI remapping at irq
 domain level

Hi,

On 02/03/2017 13:38, Mian Yousaf Kaukab wrote:
> On 03/02/2017 11:24 AM, Auger Eric wrote:
>> Hi,
>>
>> On 02/03/2017 11:01, Mian Yousaf Kaukab wrote:
>>> Check only if irq domains are available.
>>>
>>> Signed-off-by: Mian Yousaf Kaukab <yousaf.kaukab@...e.com>
>>> ---
>>>   drivers/vfio/vfio_iommu_type1.c | 5 +++--
>>>   1 file changed, 3 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/drivers/vfio/vfio_iommu_type1.c
>>> b/drivers/vfio/vfio_iommu_type1.c
>>> index bd6f293c4ebd..e3ed50e40ead 100644
>>> --- a/drivers/vfio/vfio_iommu_type1.c
>>> +++ b/drivers/vfio/vfio_iommu_type1.c
>>> @@ -1287,8 +1287,9 @@ static int vfio_iommu_type1_attach_group(void
>>> *iommu_data,
>>>       INIT_LIST_HEAD(&domain->group_list);
>>>       list_add(&group->next, &domain->group_list);
>>>   -    msi_remap = resv_msi ? irq_domain_check_msi_remap() :
>>> -                iommu_capable(bus, IOMMU_CAP_INTR_REMAP);
>>> +    msi_remap = resv_msi && IS_ENABLED(CONFIG_IRQ_DOMAIN) ?
>>> +            irq_domain_check_msi_remap() :
>>> +            iommu_capable(bus, IOMMU_CAP_INTR_REMAP);
>> Is that patch actually needed after [PATCH 1/2] irqdomain: add empty
>> irq_domain_check_msi_remap. irq_domain_check_msi_remap() should be
>> defined and if you follow my suggestion, would return false. Anyway in
>> your case resv_msi should be false.
> I agree its an overkill if resv_msi is guaranteed to be false. What I am
> unsure about is that, if iommu have IOMMU_RESV_MSI regions that would
> mean that irq domains are selected in the build. If this is not
> guaranteed, then we need to add this check.

Currently only ARM SMMUs advertise IOMMU_RESV_MSI regions. If attempting
to do passthrough on an ARM platform not implementing IRQ_DOMAIN the
unsafe IRQ assignment mode would need to be chosen (if
irq_domain_check_msi_remap() returns false as discussed before). Anyway
checking the  interrupt remapping capability on IOMMU side would report
false as well since the capability is not exposed by ARM SMMU anymore.

Thanks

Eric
> 
>>
>> Thanks
>>
>> Eric
> 
> BR,
> Yousaf