[Mono-dev] ObjectPool

[Greg Young]

We have had some moving forward in our mono stability issues. I
figured I would start up a chat here about one place that seems to be
causing many problems.


https://github.com/mono/mono/blob/master/mcs/class/corlib/System.Collections.Concurrent/ObjectPool.cs

I am not really sure the code here is actually thread safe... Maybe it
doesn't need to be for some reason?

We have added volatiles for add/remove index (yuriy pointed out that
threads were often stuck at line 86) but even with that im still not
sure its actually threadsafe. There are some weird edge conditions
that seem to be here (like if CompareExchange fails 3 times it will
just never set the remove index and continue on).

Somebody has obviously spent a lot of time thinking about this code
and optimizing it. I want to make sure we get the full "why" behind
things so we dont a) waste time b) introduce issues.

Greg

-- 
Le doute n'est pas une condition agréable, mais la certitude est absurde.
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list

Re: [Mono-dev] ObjectPool

[Greg Young]

Yuriy will reply with more details but by just making Take return
Creator() and nooping release many of our issues have gone away
(including visible issues with concurrentstack). From what I have seen
we also saw massive performance gains.

On Wed, Jul 25, 2012 at 6:54 AM, Greg Young  wrote:
> We have had some moving forward in our mono stability issues. I
> figured I would start up a chat here about one place that seems to be
> causing many problems.
>
>
> https://github.com/mono/mono/blob/master/mcs/class/corlib/System.Collections.Concurrent/ObjectPool.cs
>
> I am not really sure the code here is actually thread safe... Maybe it
> doesn't need to be for some reason?
>
> We have added volatiles for add/remove index (yuriy pointed out that
> threads were often stuck at line 86) but even with that im still not
> sure its actually threadsafe. There are some weird edge conditions
> that seem to be here (like if CompareExchange fails 3 times it will
> just never set the remove index and continue on).
>
> Somebody has obviously spent a lot of time thinking about this code
> and optimizing it. I want to make sure we get the full "why" behind
> things so we dont a) waste time b) introduce issues.
>
> Greg
>
> --
> Le doute n'est pas une condition agréable, mais la certitude est absurde.



-- 
Le doute n'est pas une condition agréable, mais la certitude est absurde.
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list

Re: [Mono-dev] ObjectPool

[Konrad Kruczyński]

On Wed, 2012-07-25 at 07:17 -0400, Greg Young wrote:
> Yuriy will reply with more details but by just making Take return
> Creator() and nooping release many of our issues have gone away
> (including visible issues with concurrentstack). From what I have seen
> we also saw massive performance gains.
> 

Just for the curiosity: are you using SGEN? I've always thought that
ObjectPool in ConcurrentStack was used to reduce the pressure on the
garbage collector (Boehm, specifically). It is probably much less
important with generational collector like SGEN.

--
Regards,
 Konrad


___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list

Re: [Mono-dev] ObjectPool

[Greg Young]

Yes we are using sgen.

Good to know the history there as well.

Cheers,

Greg

On Wed, Jul 25, 2012 at 7:22 AM, Konrad Kruczyński
 wrote:
> On Wed, 2012-07-25 at 07:17 -0400, Greg Young wrote:
>> Yuriy will reply with more details but by just making Take return
>> Creator() and nooping release many of our issues have gone away
>> (including visible issues with concurrentstack). From what I have seen
>> we also saw massive performance gains.
>>
>
> Just for the curiosity: are you using SGEN? I've always thought that
> ObjectPool in ConcurrentStack was used to reduce the pressure on the
> garbage collector (Boehm, specifically). It is probably much less
> important with generational collector like SGEN.
>
> --
> Regards,
>  Konrad
>
>



-- 
Le doute n'est pas une condition agréable, mais la certitude est absurde.
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list

Re: [Mono-dev] ObjectPool

[Konrad Kruczyński]

On Wed, 2012-07-25 at 07:28 -0400, Greg Young wrote:
> Yes we are using sgen.
> 
> Good to know the history there as well.
> 
> Cheers,
> 
> Greg
> 

Well, *actually* this is my guess ;)

--
Regards, 
 Konrad

___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list

[Mono-dev] EV

[Greg Young]

Does anyone already have/know of a socket lib on top of EV for eventing?

Cheers,

Greg

-- 
Le doute n'est pas une condition agréable, mais la certitude est absurde.
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list

Re: [Mono-dev] ObjectPool

[Jérémie Laval]

There were indeed two bugs in there, a missing barrier and the issue you
described.

I also commented out the code if you want more details.

--
Jérémie Laval
http://neteril.org



On Wed, Jul 25, 2012 at 11:54 AM, Greg Young wrote:

> We have had some moving forward in our mono stability issues. I
> figured I would start up a chat here about one place that seems to be
> causing many problems.
>
>
>
> https://github.com/mono/mono/blob/master/mcs/class/corlib/System.Collections.Concurrent/ObjectPool.cs
>
> I am not really sure the code here is actually thread safe... Maybe it
> doesn't need to be for some reason?
>
> We have added volatiles for add/remove index (yuriy pointed out that
> threads were often stuck at line 86) but even with that im still not
> sure its actually threadsafe. There are some weird edge conditions
> that seem to be here (like if CompareExchange fails 3 times it will
> just never set the remove index and continue on).
>
> Somebody has obviously spent a lot of time thinking about this code
> and optimizing it. I want to make sure we get the full "why" behind
> things so we dont a) waste time b) introduce issues.
>
> Greg
>
> --
> Le doute n'est pas une condition agréable, mais la certitude est absurde.
> ___
> Mono-devel-list mailing list
> Mono-devel-list@lists.ximian.com
> http://lists.ximian.com/mailman/listinfo/mono-devel-list
>
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list

Re: [Mono-dev] EV

[Jérémie Laval]

You can look at manos code https://github.com/jacksonh/manos

--
Jérémie Laval
http://neteril.org



2012/7/25 Greg Young 

> Does anyone already have/know of a socket lib on top of EV for eventing?
>
> Cheers,
>
> Greg
>
> --
> Le doute n'est pas une condition agréable, mais la certitude est absurde.
> ___
> Mono-devel-list mailing list
> Mono-devel-list@lists.ximian.com
> http://lists.ximian.com/mailman/listinfo/mono-devel-list
>
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list

Re: [Mono-dev] ObjectPool

[Yuriy Solodkyy]

Hi Jérémie,

thank you for so quick fix!

I'll build mono with your fix and let you know how it works.

It seems that a lot of our previously reported problems with sockets
are due to ObjectPool.

-yuriy

On Wed, Jul 25, 2012 at 2:43 PM, Jérémie Laval  wrote:
> There were indeed two bugs in there, a missing barrier and the issue you
> described.
>
> I also commented out the code if you want more details.
>
> --
> Jérémie Laval
> http://neteril.org
>
>
>
> On Wed, Jul 25, 2012 at 11:54 AM, Greg Young 
> wrote:
>>
>> We have had some moving forward in our mono stability issues. I
>> figured I would start up a chat here about one place that seems to be
>> causing many problems.
>>
>>
>>
>> https://github.com/mono/mono/blob/master/mcs/class/corlib/System.Collections.Concurrent/ObjectPool.cs
>>
>> I am not really sure the code here is actually thread safe... Maybe it
>> doesn't need to be for some reason?
>>
>> We have added volatiles for add/remove index (yuriy pointed out that
>> threads were often stuck at line 86) but even with that im still not
>> sure its actually threadsafe. There are some weird edge conditions
>> that seem to be here (like if CompareExchange fails 3 times it will
>> just never set the remove index and continue on).
>>
>> Somebody has obviously spent a lot of time thinking about this code
>> and optimizing it. I want to make sure we get the full "why" behind
>> things so we dont a) waste time b) introduce issues.
>>
>> Greg
>>
>> --
>> Le doute n'est pas une condition agréable, mais la certitude est absurde.
>> ___
>> Mono-devel-list mailing list
>> Mono-devel-list@lists.ximian.com
>> http://lists.ximian.com/mailman/listinfo/mono-devel-list
>
>
>
> ___
> Mono-devel-list mailing list
> Mono-devel-list@lists.ximian.com
> http://lists.ximian.com/mailman/listinfo/mono-devel-list
>



-- 
Yuriy Solodkyy
(y.solod...@gmail.com)
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list

Re: [Mono-dev] ObjectPool

[Greg Young]

We will be stress testing this. This seems to be the root cause of the
other issues we have been experiencing, literally hundreds of things
are broken because of this one little issue (anything with callbacks,
concurrentstack/queue, etc etc).


On Wed, Jul 25, 2012 at 7:43 AM, Jérémie Laval  wrote:
> There were indeed two bugs in there, a missing barrier and the issue you
> described.
>
> I also commented out the code if you want more details.
>
> --
> Jérémie Laval
> http://neteril.org
>
>
>
> On Wed, Jul 25, 2012 at 11:54 AM, Greg Young 
> wrote:
>>
>> We have had some moving forward in our mono stability issues. I
>> figured I would start up a chat here about one place that seems to be
>> causing many problems.
>>
>>
>>
>> https://github.com/mono/mono/blob/master/mcs/class/corlib/System.Collections.Concurrent/ObjectPool.cs
>>
>> I am not really sure the code here is actually thread safe... Maybe it
>> doesn't need to be for some reason?
>>
>> We have added volatiles for add/remove index (yuriy pointed out that
>> threads were often stuck at line 86) but even with that im still not
>> sure its actually threadsafe. There are some weird edge conditions
>> that seem to be here (like if CompareExchange fails 3 times it will
>> just never set the remove index and continue on).
>>
>> Somebody has obviously spent a lot of time thinking about this code
>> and optimizing it. I want to make sure we get the full "why" behind
>> things so we dont a) waste time b) introduce issues.
>>
>> Greg
>>
>> --
>> Le doute n'est pas une condition agréable, mais la certitude est absurde.
>> ___
>> Mono-devel-list mailing list
>> Mono-devel-list@lists.ximian.com
>> http://lists.ximian.com/mailman/listinfo/mono-devel-list
>
>



-- 
Le doute n'est pas une condition agréable, mais la certitude est absurde.
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list

Re: [Mono-dev] EV

[Greg Young]

Yeah I figured someone might have already built one and could save us
a bit of time :) Rather generic code. If we decide to build we will
OSS.

On Wed, Jul 25, 2012 at 7:45 AM, Jérémie Laval  wrote:
> You can look at manos code https://github.com/jacksonh/manos
>
> --
> Jérémie Laval
> http://neteril.org
>
>
>
> 2012/7/25 Greg Young 
>>
>> Does anyone already have/know of a socket lib on top of EV for eventing?
>>
>> Cheers,
>>
>> Greg
>>
>> --
>> Le doute n'est pas une condition agréable, mais la certitude est absurde.
>> ___
>> Mono-devel-list mailing list
>> Mono-devel-list@lists.ximian.com
>> http://lists.ximian.com/mailman/listinfo/mono-devel-list
>
>



-- 
Le doute n'est pas une condition agréable, mais la certitude est absurde.
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list

Re: [Mono-dev] EV

[Greg Young]

action jackson to the rescue as always!

If anyone is looking for similar check out
https://github.com/kersny/libuv-csharp from google summer of code
student last year

On Wed, Jul 25, 2012 at 8:01 AM, Greg Young  wrote:
> Yeah I figured someone might have already built one and could save us
> a bit of time :) Rather generic code. If we decide to build we will
> OSS.
>
> On Wed, Jul 25, 2012 at 7:45 AM, Jérémie Laval  
> wrote:
>> You can look at manos code https://github.com/jacksonh/manos
>>
>> --
>> Jérémie Laval
>> http://neteril.org
>>
>>
>>
>> 2012/7/25 Greg Young 
>>>
>>> Does anyone already have/know of a socket lib on top of EV for eventing?
>>>
>>> Cheers,
>>>
>>> Greg
>>>
>>> --
>>> Le doute n'est pas une condition agréable, mais la certitude est absurde.
>>> ___
>>> Mono-devel-list mailing list
>>> Mono-devel-list@lists.ximian.com
>>> http://lists.ximian.com/mailman/listinfo/mono-devel-list
>>
>>
>
>
>
> --
> Le doute n'est pas une condition agréable, mais la certitude est absurde.



-- 
Le doute n'est pas une condition agréable, mais la certitude est absurde.
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list

Re: [Mono-dev] EV

[Greg Young]

Thanks!

On Wed, Jul 25, 2012 at 8:19 AM, Alfred Hall  wrote:
> I think https://github.com/txdv/LibuvSharp is more up to date, Bentkus seems
> to be working on this actively..
>
> On 25 Jul 2012, at 12:17, Greg Young wrote:
>
> action jackson to the rescue as always!
>
> If anyone is looking for similar check out
> https://github.com/kersny/libuv-csharp from google summer of code
> student last year
>
> On Wed, Jul 25, 2012 at 8:01 AM, Greg Young  wrote:
>
> Yeah I figured someone might have already built one and could save us
>
> a bit of time :) Rather generic code. If we decide to build we will
>
> OSS.
>
>
> On Wed, Jul 25, 2012 at 7:45 AM, Jérémie Laval 
> wrote:
>
> You can look at manos code https://github.com/jacksonh/manos
>
>
> --
>
> Jérémie Laval
>
> http://neteril.org
>
>
>
>
> 2012/7/25 Greg Young 
>
>
> Does anyone already have/know of a socket lib on top of EV for eventing?
>
>
> Cheers,
>
>
> Greg
>
>
> --
>
> Le doute n'est pas une condition agréable, mais la certitude est absurde.
>
> ___
>
> Mono-devel-list mailing list
>
> Mono-devel-list@lists.ximian.com
>
> http://lists.ximian.com/mailman/listinfo/mono-devel-list
>
>
>
>
>
>
> --
>
> Le doute n'est pas une condition agréable, mais la certitude est absurde.
>
>
>
>
> --
> Le doute n'est pas une condition agréable, mais la certitude est absurde.
> ___
> Mono-devel-list mailing list
> Mono-devel-list@lists.ximian.com
> http://lists.ximian.com/mailman/listinfo/mono-devel-list
>
>



-- 
Le doute n'est pas une condition agréable, mais la certitude est absurde.
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list

Re: [Mono-dev] EV

[Alfred Hall]

I think https://github.com/txdv/LibuvSharp is more up to date, Bentkus seems to 
be working on this actively..

On 25 Jul 2012, at 12:17, Greg Young wrote:

> action jackson to the rescue as always!
> 
> If anyone is looking for similar check out
> https://github.com/kersny/libuv-csharp from google summer of code
> student last year
> 
> On Wed, Jul 25, 2012 at 8:01 AM, Greg Young  wrote:
>> Yeah I figured someone might have already built one and could save us
>> a bit of time :) Rather generic code. If we decide to build we will
>> OSS.
>> 
>> On Wed, Jul 25, 2012 at 7:45 AM, Jérémie Laval  
>> wrote:
>>> You can look at manos code https://github.com/jacksonh/manos
>>> 
>>> --
>>> Jérémie Laval
>>> http://neteril.org
>>> 
>>> 
>>> 
>>> 2012/7/25 Greg Young 
 
 Does anyone already have/know of a socket lib on top of EV for eventing?
 
 Cheers,
 
 Greg
 
 --
 Le doute n'est pas une condition agréable, mais la certitude est absurde.
 ___
 Mono-devel-list mailing list
 Mono-devel-list@lists.ximian.com
 http://lists.ximian.com/mailman/listinfo/mono-devel-list
>>> 
>>> 
>> 
>> 
>> 
>> --
>> Le doute n'est pas une condition agréable, mais la certitude est absurde.
> 
> 
> 
> -- 
> Le doute n'est pas une condition agréable, mais la certitude est absurde.
> ___
> Mono-devel-list mailing list
> Mono-devel-list@lists.ximian.com
> http://lists.ximian.com/mailman/listinfo/mono-devel-list

___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list

Re: [Mono-dev] Mono and Microsoft open sourced Entity Framework

[Rodrigo Kumpera]

Companies, like people, can grow up and have a better understanding of the
world around us.

On Tuesday, July 24, 2012, Brandon Perry  wrote:
> Crazy from the company that said Linux was a cancer!
>
> On Tue, Jul 24, 2012 at 3:55 PM, Stifu  wrote:
>> I somehow missed that news.
>> Great stuff!
>>
>>
>> Marek Safar-2 wrote
>>>
>>> Hi,
>>>
>>> Yes, we are working on EF integration.
>>>
>>> Marek
>>>
>>> On Tue, Jul 24, 2012 at 9:09 PM, Daniel Morgan wrote:
>>>
 Since Microsoft has open sourced the Entity Framework, is anyone
working
 on getting it to work on Mono?

 ___
 Mono-devel-list mailing list
 Mono-devel-list@.ximian
 http://lists.ximian.com/mailman/listinfo/mono-devel-list


>>>
>>> ___
>>> Mono-devel-list mailing list
>>> Mono-devel-list@.ximian
>>> http://lists.ximian.com/mailman/listinfo/mono-devel-list
>>>
>>
>>
>>
>>
>> --
>> View this message in context:
http://mono.1490590.n4.nabble.com/Mono-and-Microsoft-open-sourced-Entity-Framework-tp4650539p4650541.html
>> Sent from the Mono - Dev mailing list archive at Nabble.com.
>> ___
>> Mono-devel-list mailing list
>> Mono-devel-list@lists.ximian.com
>> http://lists.ximian.com/mailman/listinfo/mono-devel-list
>
>
>
> --
> http://volatile-minds.blogspot.com -- blog
> http://www.volatileminds.net -- website
> ___
> Mono-devel-list mailing list
> Mono-devel-list@lists.ximian.com
> http://lists.ximian.com/mailman/listinfo/mono-devel-list
>
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list

Re: [Mono-dev] Fail to run aspx web application on mod_mono-2.10 ( .Net 4 framework )

[Mike Morano]

Hi Anas,

Have you had any luck getting your application running, I believe I am
running into the same issue.


Thanks,
Mike
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list