[Mono-dev] Regressions on 3.13.0-46-generic -> 3.13.0-48-generic

[Greg Young]

So we have just seen an issue that looks oddly familiar to this in kernel 4.4.5

On Mon, Sep 21, 2015 at 10:03 PM, Greg Young <gregoryyoung1 at gmail.com> wrote:
> I am surprised to see no one else post on this.
>
> I would think a massive regression that prevents mono from running
> multi-threaded code on a multi-cpu system (including in the future on
> any v4 kernel) might be an issue that people would be concerned
> about?!
>
>
>
> On Thu, Sep 17, 2015 at 4:03 PM, Greg Young <gregoryyoung1 at gmail.com> wrote:
>> So to follow up on this.
>>
>> The fix that was put in for this has since been reverted as mentioned
>> in the bug above.
>>
>> https://github.com/torvalds/linux/commit/73459e2a1ada09a68c02cc5b73f3116fc8194b3d
>>
>> So all 4 kernels will have this issue. I'm not really sure what needs
>> to be done on the mono side here to avoid this issue but it would seem
>> to affect any multi-threaded code running on multiple cpu linux
>> systems (which I would imagine would be a vast majority of mono
>> installations).
>>
>> Greg
>>
>> On Thu, Sep 17, 2015 at 1:05 PM, Greg Young <gregoryyoung1 at gmail.com> wrote:
>>> I was just actually writing that this seems to be discussed here:
>>> https://bugzilla.xamarin.com/show_bug.cgi?id=29692 :) Thanks for the
>>> additional information.
>>>
>>>
>>>
>>> On Thu, Sep 17, 2015 at 1:01 PM, Taloth Saldono <talothsaldono at gmail.com> wrote:
>>>> This Ubuntu bug discusses the issue for 3.13.
>>>> https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1450584
>>>>
>>>> Please note that later kernel 4.0 versions exhibit similar issues, search
>>>> this mailinglist for specifics on that new regression. I gave up on
>>>> investigating further due to other obligations and the complexity of the
>>>> subject matter.
>>>>
>>>> Op 17 sep. 2015 11:49 schreef "Greg Young" <gregoryyoung1 at gmail.com>:
>>>>>
>>>>> 3.13.0-63-generic does not have the issue as well.
>>>>>
>>>>> I guess our mitigation will just be to know about the issue and
>>>>> suggest downgrade/upgrade. This may however be useful for others in
>>>>> the future. Mono 3.12.1 on 3.13.0-46-generic #77-Ubuntu SMP appears to
>>>>> not be stable. We will also test mono4 now.
>>>>>
>>>>> Greg
>>>>>
>>>>> On Thu, Sep 17, 2015 at 11:38 AM, Greg Young <gregoryyoung1 at gmail.com>
>>>>> wrote:
>>>>> > We are running into some interesting regressions with mono. We are
>>>>> > seeing many failures such as:
>>>>> >
>>>>> > Mono v3.12.1 (we have considered moving up to 4.1 or 4.2 but the last
>>>>> > time we tested it didn't seem safe to move).
>>>>> >
>>>>> > Native stacktrace:
>>>>> >
>>>>> > /home/ubuntu/EventStore-OSS-Ubuntu-v3.2.1/eventstored() [0x609a4f]
>>>>> > /home/ubuntu/EventStore-OSS-Ubuntu-v3.2.1/eventstored() [0x5bdecf]
>>>>> > /home/ubuntu/EventStore-OSS-Ubuntu-v3.2.1/eventstored() [0x453b19]
>>>>> > /lib/x86_64-linux-gnu/libpthread.so.0(+0x10340) [0x7fd9709b7340]
>>>>> > /home/ubuntu/EventStore-OSS-Ubuntu-v3.2.1/eventstored() [0x4e6145]
>>>>> > [0x40e21ceb]
>>>>> >
>>>>> >
>>>>> > [03041,21,13:43:11.379] Global Unhandled Exception occurred.
>>>>> > Object reference not set to an instance of an object
>>>>> > [ERROR] FATAL UNHANDLED EXCEPTION: System.NullReferenceException:
>>>>> > Object reference not set to an instance of an object
>>>>> >   at EventStore.Core.Bus.QueuedHandlerAutoReset.ReadFromQueue
>>>>> > (System.Object o) [0x00000] in <filename unknown>:0
>>>>> >   at System.Threading.Thread.StartInternal () [0x00000] in <filename
>>>>> > unknown>:0
>>>>> >
>>>>> > note that this isn't a real nre
>>>>> >
>>>>> > We do not see these regressions in Linux ip-172-31-37-60
>>>>> > 3.13.0-46-generic #77-Ubuntu SMP
>>>>> >
>>>>> > We do see them in Linux ip-172-31-40-115 3.13.0-48-generic #80-Ubuntu
>>>>> > SMP
>>>>> >
>>>>> > Anyone have any ideas how to debug this further? These are not
>>>>> > deterministic issues.
>>>>> >
>>>>> > Cheers,
>>>>> >
>>>>> > Greg
>>>>> >
>>>>> > --
>>>>> > Studying for the Turing test
>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> Studying for the Turing test
>>>>> _______________________________________________
>>>>> Mono-devel-list mailing list
>>>>> Mono-devel-list at lists.ximian.com
>>>>> http://lists.ximian.com/mailman/listinfo/mono-devel-list
>>>
>>>
>>>
>>> --
>>> Studying for the Turing test
>>
>>
>>
>> --
>> Studying for the Turing test
>
>
>
> --
> Studying for the Turing test



-- 
Studying for the Turing test