Boot hangs with "AMD-Vi Command Buffer Timeout"

Boot hangs with "AMD-Vi Command Buffer Timeout"

[ojab //]

Oh hai!

I've made a bugzilla issue [1] about it, but I'm not sure if anyone
reading bugzilla reports, so duplicating here.
I have raven ridge APU (2400g) + ASUS A320M-K motherboard and boot
(4.16.4 & 4.17-rc2) hangs printing many `[    0.001000] AMD-Vi Command
Buffer Timeout` messages. With IOMMU disabled (either via EFI/BIOS or
using amd_iommu=off) it boots fine.

What info is needed to debug/fix it?

[1] https://bugzilla.kernel.org/show_bug.cgi?id=199513

//wbr ojab

Re: Boot hangs with "AMD-Vi Command Buffer Timeout"

[Joerg Roedel]

I've added some AMD people to Cc. Suravee, Tom, any ideas here? One of
the screen-shots behind the link below show IOTLB_INV_TIMEOUT errors, so
it is likely ATS related. Maybe you guys can reproduce it on your side?


Thanks,

	Joerg


On Thu, Apr 26, 2018 at 10:28:47AM +0000, ojab // wrote:
> I've made a bugzilla issue [1] about it, but I'm not sure if anyone
> reading bugzilla reports, so duplicating here.
> I have raven ridge APU (2400g) + ASUS A320M-K motherboard and boot
> (4.16.4 & 4.17-rc2) hangs printing many `[    0.001000] AMD-Vi Command
> Buffer Timeout` messages. With IOMMU disabled (either via EFI/BIOS or
> using amd_iommu=off) it boots fine.
> 
> What info is needed to debug/fix it?
> 
> [1] https://bugzilla.kernel.org/show_bug.cgi?id=199513
> 
> //wbr ojab

Re: Boot hangs with "AMD-Vi Command Buffer Timeout"

[Gary R Hook]

On 04/26/2018 10:22 AM, Joerg Roedel wrote:
> I've added some AMD people to Cc. Suravee, Tom, any ideas here? One of
> the screen-shots behind the link below show IOTLB_INV_TIMEOUT errors, so
> it is likely ATS related. Maybe you guys can reproduce it on your side?
> 
> 
> Thanks,
> 
> 	Joerg
> 
> 
> On Thu, Apr 26, 2018 at 10:28:47AM +0000, ojab // wrote:
>> I've made a bugzilla issue [1] about it, but I'm not sure if anyone
>> reading bugzilla reports, so duplicating here.
>> I have raven ridge APU (2400g) + ASUS A320M-K motherboard and boot
>> (4.16.4 & 4.17-rc2) hangs printing many `[    0.001000] AMD-Vi Command
>> Buffer Timeout` messages. With IOMMU disabled (either via EFI/BIOS or
>> using amd_iommu=off) it boots fine.
>>
>> What info is needed to debug/fix it?

I realize this is an Asus system, but what BIOS version is on it? Did 
you look to see if an update is available (which I suspect is not, given 
that RR is quite new).

Have you made Asus aware of the problem?

Gary

Re: Boot hangs with "AMD-Vi Command Buffer Timeout"

[ojab //]

On Thu, Apr 26, 2018 at 4:25 PM, Gary R Hook <gary.hook-5C7GfCeVMHo@public.gmane.org> wrote:
> On 04/26/2018 10:22 AM, Joerg Roedel wrote:
>>
>> I've added some AMD people to Cc. Suravee, Tom, any ideas here? One of
>> the screen-shots behind the link below show IOTLB_INV_TIMEOUT errors, so
>> it is likely ATS related. Maybe you guys can reproduce it on your side?
>>
>>
>> Thanks,
>>
>>         Joerg
>>
>>
>> On Thu, Apr 26, 2018 at 10:28:47AM +0000, ojab // wrote:
>>>
>>> I've made a bugzilla issue [1] about it, but I'm not sure if anyone
>>> reading bugzilla reports, so duplicating here.
>>> I have raven ridge APU (2400g) + ASUS A320M-K motherboard and boot
>>> (4.16.4 & 4.17-rc2) hangs printing many `[    0.001000] AMD-Vi Command
>>> Buffer Timeout` messages. With IOMMU disabled (either via EFI/BIOS or
>>> using amd_iommu=off) it boots fine.
>>>
>>> What info is needed to debug/fix it?
>
>
> I realize this is an Asus system, but what BIOS version is on it? Did you
> look to see if an update is available (which I suspect is not, given that RR
> is quite new).
Latest BIOS (4008, release notes says `Update AGESA 1002a`). RR support was
added a while ago, but I've bought MB/CPU yesterday, so I'm not sure if it works
on previous kernels/bioses.
>
> Have you made Asus aware of the problem?
Just sent request to their tech support via website form, but I doubt
that I'll hear
something useful from them.
>
> Gary

//wbr ojab

Re: Boot hangs with "AMD-Vi Command Buffer Timeout"

[ojab //]

On Thu, Apr 26, 2018 at 4:51 PM, ojab // <ojab-uv7J75MkRmE@public.gmane.org> wrote:
> On Thu, Apr 26, 2018 at 4:25 PM, Gary R Hook <gary.hook-5C7GfCeVMHo@public.gmane.org> wrote:
>> On 04/26/2018 10:22 AM, Joerg Roedel wrote:
>>>
>>> I've added some AMD people to Cc. Suravee, Tom, any ideas here? One of
>>> the screen-shots behind the link below show IOTLB_INV_TIMEOUT errors, so
>>> it is likely ATS related. Maybe you guys can reproduce it on your side?
>>>
>>>
>>> Thanks,
>>>
>>>         Joerg
>>>
>>>
>>> On Thu, Apr 26, 2018 at 10:28:47AM +0000, ojab // wrote:
>>>>
>>>> I've made a bugzilla issue [1] about it, but I'm not sure if anyone
>>>> reading bugzilla reports, so duplicating here.
>>>> I have raven ridge APU (2400g) + ASUS A320M-K motherboard and boot
>>>> (4.16.4 & 4.17-rc2) hangs printing many `[    0.001000] AMD-Vi Command
>>>> Buffer Timeout` messages. With IOMMU disabled (either via EFI/BIOS or
>>>> using amd_iommu=off) it boots fine.
>>>>
>>>> What info is needed to debug/fix it?
>>
>>
>> I realize this is an Asus system, but what BIOS version is on it? Did you
>> look to see if an update is available (which I suspect is not, given that RR
>> is quite new).
> Latest BIOS (4008, release notes says `Update AGESA 1002a`). RR support was
> added a while ago, but I've bought MB/CPU yesterday, so I'm not sure if it works
> on previous kernels/bioses.
>>
>> Have you made Asus aware of the problem?
> Just sent request to their tech support via website form, but I doubt
> that I'll hear
> something useful from them.


and BTW screenshots w/ kernel panic and IOTLB_INV_TIMEOUT in bugzilla are not
mine, it's another user with quite not the same issue. I have
continuous messages
>[    0.001000] AMD-Vi Command Buffer Timeout

(i. e. time is always 0.001000) with no kernel panics or any other feedback.

//wbr ojab