[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] Unstable fail to load dom0


  • To: "Cui, Dexuan" <dexuan.cui@xxxxxxxxx>
  • From: Andrew Lyon <andrew.lyon@xxxxxxxxx>
  • Date: Fri, 17 Apr 2009 07:41:40 +0100
  • Cc: Xen-devel <xen-devel@xxxxxxxxxxxxxxxxxxx>, Ross Philipson <Ross.Philipson@xxxxxxxxxx>
  • Delivery-date: Thu, 16 Apr 2009 23:42:11 -0700
  • Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type:content-transfer-encoding; b=Za0LPdEALK6jzOHII52XWrbI1WJBtFAhtwiuAy9LWY9at+oTe/J4NEVUOjM1aJ8CJ1 Gs4mh0O0h/ZmFx73vLXs2iyG1vzrBAOdGTkDjL9yUbVWX2s535sI/Bmo8XWNaWM7ngpq Bc8noa3ttbCfuZUrs+iKaHQQI3NueXXZsSy3k=
  • List-id: Xen developer discussion <xen-devel.lists.xensource.com>

2009/4/17 Cui, Dexuan <dexuan.cui@xxxxxxxxx>:
> Hi Andy,
> Could you please check if your BIOS vendor (e.g., the Suppermicro website) 
> has released a new BIOS update?
> I guess there may be a new BIOS update available.
>

I've already updated the bios to the latest version (1.2) but it did
not fix the problem.


Andy

> BTW, for the xen parameter iommu_inclusive_mapping, you could see 
> http://xenbits.xensource.com/xen-unstable.hg?rev/5dee49f3a86e.
>
> Thanks,
> -- Dexuan
>
> Cui, Dexuan wrote:
>> Your SuperMicro BIOS fails to report RMRR (please see Intel VT-d spec
>> for details) correctly.
>>
>> -- Dexuan
>>
>> -----Original Message-----
>> From: xen-devel-bounces@xxxxxxxxxxxxxxxxxxx
>> [mailto:xen-devel-bounces@xxxxxxxxxxxxxxxxxxx] On Behalf Of Andrew
>> Lyon Sent: 2009年4月16日 23:26
>> To: Xen-devel
>> Subject: Re: [Xen-devel] Unstable fail to load dom0
>>
>> Hi,
>>
>> Can anybody tell me what bug it is that makes it necessary to set
>> iommu_inclusive_mapping=1 in order to run Xen 3.4 on my Supermicro
>> X7DWA-N ?
>>
>> Without it the system hangs immediately after printing "*** LOADING
>> DOMAIN 0 ***", with the setting in place it seems to work ok but I
>> would like to report this bug to Supermicro so I need to give them a
>> good description about what is wrong.
>>
>> Perhaps it is related to this error I get:
>>
>> (XEN) [VT-D]iommu.c:722: iommu_page_fault: iommu->reg =
>> ffff828bfff53000 (XEN) [VT-D]iommu.c:694: iommu_fault_status: Primary
>> Pending Fault (XEN) [VT-D]iommu.c:676: iommu_fault:DMA Read: 0:1e.0
>> addr bfc54000
>> REASON 6 iommu->reg = ffff828bfff53000
>> (XEN) print_vtd_entries: iommu = ffff83043fde1db0 bdf = 0:1e:0 gmfn =
>> bfc54 (XEN)     root_entry = ffff83043fd7a000
>> (XEN)     root_entry[0] = 43214c001
>> (XEN)     context = ffff83043214c000
>> (XEN)     context[f0] = 102_43fd5f001
>> (XEN)     l4 = ffff83043fd5f000
>> (XEN)     l4_index = 0
>> (XEN)     l4[0] = 43fd5e003
>> (XEN)     l3 = ffff83043fd5e000
>> (XEN)     l3_index = 2
>> (XEN)     l3[2] = 433f5b003
>> (XEN)     l2 = ffff830433f5b000
>> (XEN)     l2_index = 1fe
>> (XEN)     l2[1fe] = 433d5c003
>> (XEN)     l1 = ffff830433d5c000
>> (XEN)     l1_index = 54
>> (XEN)     l1[54] = 0
>> (XEN)     l1[54] not present
>>
>>
>> Andy
>>
>> On Fri, Mar 27, 2009 at 5:09 PM, Cui, Dexuan <dexuan.cui@xxxxxxxxx>
>> wrote:
>>> Hi Andy,
>>> Can you try to change the xen parameter 'iommu' to 0?
>>> If with iommu=0 you can boot Xen fine, that may be due to a known
>>> BIOS bug and you can use xen with c/s >=19270 and use "iommu=1
>>> iommu_inclusive_mapping=1".
>>>
>>> Thanks,
>>> -- Dexuan
>>>
>>> -----Original Message-----
>>> From: xen-devel-bounces@xxxxxxxxxxxxxxxxxxx
>>> [mailto:xen-devel-bounces@xxxxxxxxxxxxxxxxxxx] On Behalf Of Andrew
>>> Lyon
>>> Sent: 2009年3月27日 22:52
>>> To: Ian Campbell
>>> Cc: Xen-devel
>>> Subject: Re: [Xen-devel] Unstable fail to load dom0
>>>
>>> On Fri, Mar 27, 2009 at 2:26 PM, Andrew Lyon <andrew.lyon@xxxxxxxxx>
>>> wrote:
>>>> On Fri, Mar 27, 2009 at 2:21 PM, Ian Campbell
>>>> <Ian.Campbell@xxxxxxxxxx> wrote:
>>>>> On Fri, 2009-03-27 at 10:16 -0400, Andrew Lyon wrote:
>>>>>> Hi,
>>>>>>
>>>>>> I just tried to upgrade one of my systems to Xen unstable, Xen
>>>>>> itself boots ok but it hangs after "Loading dom0", I think the
>>>>>> cause of the problem is that the necessary support for loading
>>>>>> compressed kernel (vmlinuz/bzImage) was not compiled in properly,
>>>>>> I seem to recall that recently the compression code was updated
>>>>>> and now supports many different compression schemes, what libs
>>>>>> etc are required to build with support for bzImage and vmlinuz?
>>>>>
>>>>> For domain0 the decompression code is built into the hypervisor
>>>>> itself and only supports gzip style compression (historically the
>>>>> only choice).
>>>>>
>>>>> Therefore your kernel config must have:
>>>>>        CONFIG_KERNEL_GZIP=y
>>>>>        # CONFIG_KERNEL_BZIP2 is not set
>>>>>        # CONFIG_KERNEL_LZMA is not set
>>>>>
>>>>> Ian.
>>>>>
>>>>>
>>>>>
>>>>>
>>>>
>>>> Hmm, not that then, any ideas why it would hang immeediatly after
>>>> "loading dom0" ?
>>>>
>>>> The same grub entry works perfectly if Xen 3.3.1 is installed, it is
>>>> loading xen.gz which is symlinked to the installed version.
>>>>
>>>> Andy
>>>>
>>>
>>> This is not a build issue, the same xen-3.4-unstable.gz + grub stanza
>>> works ok on my Dell Optiplex 755, but on my Supermicro X7DWA-N it
>>> locks up at (XEN) *** LOADING DOMAIN 0 ***
>>>
>>>
>>> title Xen 3.4 / Linux 2.6.29-suse
>>> root (hd0,0)
>>> kernel /xen.gz iommu=1 com1=115200,8n1 console=vga,com1 noirqbalance
>>> watchdog dom0_vcpus_pin
>>> module /vmlinuz-2.6.29-xen root=/dev/sda2 swiotlb=256 console=xvc0
>>> console=tty1
>>>
>>> Andy
>>>
>>> _______________________________________________
>
>
>
>

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.