Re: [zeromq-dev] GSoC ideas inquiry

2011-03-07 Thread Pieter Hintjens 
On Tue, Mar 8, 2011 at 3:51 AM, Steven McCoy  wrote:

> Forgot about this.  The paradigm follows atomic multicast and for unicast
> deliveries it effectively follows certified mail from the postal service...

Indeed, and this is what we're covering (though not fully) with the
Ch4 Titanic pattern, http://zguide.zeromq.org/page:all#toc78

-Pieter
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

[zeromq-dev] Android Support?

2011-03-07 Thread Wolfgang Richter 
Has there been any progress beyond what's reported on this page:

http://www.zeromq.org/distro:android

?

I was thinking about taking a fresh look at this, and trying to port to the
Android Platform properly (using standard NDK).  That might be too difficult
though, I'm not sure on their current STL support (although I know it's
increased in functionality).

--
Wolf
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] [PATCH] Update to OpenPGM 5.1.110.

2011-03-07 Thread Steven McCoy 
On 8 March 2011 11:36, Steven McCoy  wrote:

> 32-bit build (supported):
>
> http://miru.hk/archive/OpenPGM-5.1.110-win32.exe
>
> 64-bit build (unsupported, repo revision 1337):
>
> http://miru.hk/archive/OpenPGM-5.1.111-win64.exe
>
>
64-bit build using 64-bit intrinsics instead of 32-bit functions
(unsupported, repo revision 1338):

http://miru.hk/archive/OpenPGM-5.1.112-win64.exe

-- 
Steve-o
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] [PATCH] Update to OpenPGM 5.1.110.

2011-03-07 Thread Alex Forster 
> Just minor tweaks, already noted on stackoverflow that ML64 does not have the 
> same features as ML, the examples in the SDK have an invalid syntax when 
> ported to x64.
> 
> Also, CMake doesn't generate the MSVC project files including the asm, 
> probably at a later date as it is plastered with notices.

Not a huge deal to make the libs beforehand, it will still get me 95% of the 
way there. I'll give it a try sometime this week, thanks!
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] [PATCH] Update to OpenPGM 5.1.110.

2011-03-07 Thread Steven McCoy 
On 8 March 2011 11:44, Alex Forster  wrote:

> How well do the asm files fit into the build system? Is it an issue?
>
> I spent more time than I'd like to admit yesterday googling, since I was
> only ~30 lines away from a successful build. Essentially the solution you
> came up with is the only solution there is if you want to maintain MSVC
> compatibility: use an external assembler. I haven't tried to compile trunk
> yet, but I'm sure it will be fine. It looks from the commit message that you
> did exactly what I did - check out how TBB does it.
> http://people.athenacr.com/~wchow/repos/tbb/tbb-2.0r014/src/tbb/ia32-masm/atomic_support.asm


I have a hunch that moving from 16 & 32-bit atomic functions to 32 & 64-bit
atomic intrinsics would be faster though so I'm moving to that in 5.1.112,
configurable 32-bit or 64-bit ticket locks.

-- 
Steve-o
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] [PATCH] Update to OpenPGM 5.1.110.

2011-03-07 Thread Alex Forster 
> The trunk builds but skips the asm because CMake 2.8.3 is broken, I've 
> updated to 2.8.4 and CMake is too stupid to find MSVC 2010 Professional and 
> asks me to install the 7.1 SDK so I end up with I guess six compilers now.

Aha.

> The 8-bit and 16-bit atomics are a part of how ticket locks work, they 
> improve latency in low contention situations, i.e. PGM bus scenarios.

?? didn't know what a ticket lock was until yesterday, so I'll take your word 
for it.

> Out of interest in playing with new integration features I'll see how long I 
> can put up with the MASM before dumping it.

How well do the asm files fit into the build system? Is it an issue?

I spent more time than I'd like to admit yesterday googling, since I was only 
~30 lines away from a successful build. Essentially the solution you came up 
with is the only solution there is if you want to maintain MSVC compatibility: 
use an external assembler. I haven't tried to compile trunk yet, but I'm sure 
it will be fine. It looks from the commit message that you did exactly what I 
did - check out how TBB does it. 
http://people.athenacr.com/~wchow/repos/tbb/tbb-2.0r014/src/tbb/ia32-masm/atomic_support.asm
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] [PATCH] Update to OpenPGM 5.1.110.

2011-03-07 Thread Steven McCoy 
Just minor tweaks, already noted on stackoverflow that ML64 does not have
the same features as ML, the examples in the SDK have an invalid syntax when
ported to x64.

Also, CMake doesn't generate the MSVC project files including the asm,
probably at a later date as it is plastered with notices.

32-bit build (supported):

http://miru.hk/archive/OpenPGM-5.1.110-win32.exe

64-bit build (unsupported, repo revision 1337):

http://miru.hk/archive/OpenPGM-5.1.111-win64.exe

-- 
Steve-o

On 8 March 2011 10:40, Steven McCoy  wrote:

> The trunk builds but skips the asm because CMake 2.8.3 is broken, I've
> updated to 2.8.4 and CMake is too stupid to find MSVC 2010 Professional and
> asks me to install the 7.1 SDK so I end up with I guess six compilers now.
>
> I'll get there eventually.
>
> The 8-bit and 16-bit atomics are a part of how ticket locks work, they
> improve latency in low contention situations, i.e. PGM bus scenarios.
>
> However the 8-bit atomics are only used for read-write ticket locks which
> are too slow in locking, the code should be ejected sometime in the future.
>  It looks like I can easily upgrade the 16-bit atomics to 32-bit atomics if
> I also use 64-bit atomics but they're not available everywhere.  Out of
> interest in playing with new integration features I'll see how long I can
> put up with the MASM before dumping it.  I'm guessing the limitations by the
> Windows API are from IA64 though when regarding alignment requirements.
>
> Bottom line is no one is paying for Win64 support so I'm not in a hurry.
>
> --
> Steve-o
>
>
> On 8 March 2011 08:48, Alex Forster  wrote:
>
>> Ah, is that the trunk version you were saying you got to build
>> successfully on x64?
>>
>>
>>
>> On Mar 6, 2011, at 11:08 PM, Steven McCoy  wrote:
>>
>> On 7 March 2011 01:31, Alex Forster < 
>> a...@alexforster.com> wrote:
>>
>>> Steven McCoy  miru.hk> writes:
>>>
>>> > Clean builds under Win64.
>>> > ...
>>> > Improve spinlock performance with inline ticket based spinlock
>>> implementation.
>>>
>>> 
>>> http://www.google.com/search?q=__asm+keyword+not+supported+on+this+architecture
>>>
>>> ticket.h is coughing up because Visual C++ doesn't support inline
>>> assembly when
>>> targeting 64-bit.
>>>
>>>
>> Oh that's nice.
>>
>>
>>>  
>>> http://code.google.com/p/openpgm/source/diff?
>>>
>>> spec=svn1294&r=1294&format=side&path=/trunk/openpgm/pgm/include/impl/ticket.h
>>>
>>> This diff points out where the problem areas are-
>>>
>>> pgm_atomic_add8
>>> pgm_atomic_fetch_and_add8
>>> pgm_atomic_inc8
>>>
>>> pgm_atomic_add16
>>> pgm_atomic_fetch_and_add16
>>> pgm_atomic_inc16
>>>
>>>
>> It was a clean build before ticket locks :-)
>>
>> I've added some MASM64 to trunk to implement the assembler as functions
>> but I'm without 64-bit MSVC to test, I wouldn't say I"m overly confident on
>> the 64-bit assembler either.
>>
>> --
>> Steve-o
>>
>>
>
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] GSoC ideas inquiry

2011-03-07 Thread Steven McCoy 
On 7 March 2011 00:56, Martin Sustrik  wrote:

> > 2) Certified sockets I love this idea because I like extremely robust
> > designs :-) Unfortunately, I don't have much clue about it. Could you
> > give a specific example about this? Besides, is there any restriction
> > on the type of the underlying transport if we try to implement
> > certified sockets?
>
> This idea have been suggested by Steven McCoy. I don't know what is
> actually meant to be. Steven, can you please be more specific about the
> idea?
>
>
Forgot about this.  The paradigm follows atomic multicast and for unicast
deliveries it effectively follows certified mail from the postal service.

   - Any sender records interest from recipients and ensures it keeps a copy
   of the message until the recipient has confirmed delivery.
   - Any recipient records state of sender message sequences so it knows the
   last message it has confirmed.
   - When a recipient receives a message from a sender it sends a
   confirmation back.  This confirmation maybe automatic from the API, but is
   usually better if manually invoked after the application has actually
   finished processing the message.
   - An API is usually required to pre-register recipients at the sender.
   - When the sender has recorded that every registered recipient has
   received and confirmed a message sequence it can notify the sender
   application that that message is complete and purge it from the message
   store.
   - When a sender restarts it can attempt redelivery of non-completed
   messages.
   - When a recipient restarts it can request delivery of messages since the
   last recorded sequence.

-- 
Steve-o
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] [PATCH] Update to OpenPGM 5.1.110.

2011-03-07 Thread Steven McCoy 
The trunk builds but skips the asm because CMake 2.8.3 is broken, I've
updated to 2.8.4 and CMake is too stupid to find MSVC 2010 Professional and
asks me to install the 7.1 SDK so I end up with I guess six compilers now.

I'll get there eventually.

The 8-bit and 16-bit atomics are a part of how ticket locks work, they
improve latency in low contention situations, i.e. PGM bus scenarios.

However the 8-bit atomics are only used for read-write ticket locks which
are too slow in locking, the code should be ejected sometime in the future.
 It looks like I can easily upgrade the 16-bit atomics to 32-bit atomics if
I also use 64-bit atomics but they're not available everywhere.  Out of
interest in playing with new integration features I'll see how long I can
put up with the MASM before dumping it.  I'm guessing the limitations by the
Windows API are from IA64 though when regarding alignment requirements.

Bottom line is no one is paying for Win64 support so I'm not in a hurry.

-- 
Steve-o

On 8 March 2011 08:48, Alex Forster  wrote:

> Ah, is that the trunk version you were saying you got to build successfully
> on x64?
>
>
>
> On Mar 6, 2011, at 11:08 PM, Steven McCoy  wrote:
>
> On 7 March 2011 01:31, Alex Forster < 
> a...@alexforster.com> wrote:
>
>> Steven McCoy  miru.hk> writes:
>>
>> > Clean builds under Win64.
>> > ...
>> > Improve spinlock performance with inline ticket based spinlock
>> implementation.
>>
>> 
>> http://www.google.com/search?q=__asm+keyword+not+supported+on+this+architecture
>>
>> ticket.h is coughing up because Visual C++ doesn't support inline assembly
>> when
>> targeting 64-bit.
>>
>>
> Oh that's nice.
>
>
>>  
>> http://code.google.com/p/openpgm/source/diff?
>>
>> spec=svn1294&r=1294&format=side&path=/trunk/openpgm/pgm/include/impl/ticket.h
>>
>> This diff points out where the problem areas are-
>>
>> pgm_atomic_add8
>> pgm_atomic_fetch_and_add8
>> pgm_atomic_inc8
>>
>> pgm_atomic_add16
>> pgm_atomic_fetch_and_add16
>> pgm_atomic_inc16
>>
>>
> It was a clean build before ticket locks :-)
>
> I've added some MASM64 to trunk to implement the assembler as functions but
> I'm without 64-bit MSVC to test, I wouldn't say I"m overly confident on the
> 64-bit assembler either.
>
> --
> Steve-o
>
>
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] [PATCH] Update to OpenPGM 5.1.110.

2011-03-07 Thread Alex Forster 
Ah, is that the trunk version you were saying you got to build successfully on 
x64?



On Mar 6, 2011, at 11:08 PM, Steven McCoy  wrote:

> On 7 March 2011 01:31, Alex Forster  wrote:
> Steven McCoy  miru.hk> writes:
> 
> > Clean builds under Win64.
> > ...
> > Improve spinlock performance with inline ticket based spinlock 
> > implementation.
> 
> http://www.google.com/search?q=__asm+keyword+not+supported+on+this+architecture
> 
> ticket.h is coughing up because Visual C++ doesn't support inline assembly 
> when
> targeting 64-bit.
> 
> 
> Oh that's nice.
>  
> http://code.google.com/p/openpgm/source/diff?
> spec=svn1294&r=1294&format=side&path=/trunk/openpgm/pgm/include/impl/ticket.h
> 
> This diff points out where the problem areas are-
> 
> pgm_atomic_add8
> pgm_atomic_fetch_and_add8
> pgm_atomic_inc8
> 
> pgm_atomic_add16
> pgm_atomic_fetch_and_add16
> pgm_atomic_inc16
> 
> 
> It was a clean build before ticket locks :-)
> 
> I've added some MASM64 to trunk to implement the assembler as functions but 
> I'm without 64-bit MSVC to test, I wouldn't say I"m overly confident on the 
> 64-bit assembler either.
> 
> -- 
> Steve-o
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] [PATCH] Update to OpenPGM 5.1.110.

2011-03-07 Thread Alex Forster 
cmake -G "Visual Studio 10 Win64" ...

See the first link in my parent post. It's MSDN documentation explaining how 
it's impossible to use inline asm with MSVC 2010 64bit - if it compiles for you 
then it's not producing a 64bit binary.

I can't figure out why it's so important that the ticket struct uses 8/16bit 
ints that you (or whoever) would drop into assembly to write your own atomic 
ops. Couldn't the struct just use native sized ints so that all the 
platform-provided atomic APIs could be used?


On Mar 7, 2011, at 5:45 AM, Steven McCoy  wrote:

> Ok, installed MSVC x64 and it nows builds in trunk with cmake -G "NMake 
> Makefiles".  Still might be broken though :D
> 
> -- 
> Steve-o
> 
> On 7 March 2011 12:08, Steven McCoy  wrote:
> On 7 March 2011 01:31, Alex Forster  wrote:
> Steven McCoy  miru.hk> writes:
> 
> > Clean builds under Win64.
> > ...
> > Improve spinlock performance with inline ticket based spinlock 
> > implementation.
> 
> http://www.google.com/search?q=__asm+keyword+not+supported+on+this+architecture
> 
> ticket.h is coughing up because Visual C++ doesn't support inline assembly 
> when
> targeting 64-bit.
> 
> 
> Oh that's nice.
>  
> http://code.google.com/p/openpgm/source/diff?
> spec=svn1294&r=1294&format=side&path=/trunk/openpgm/pgm/include/impl/ticket.h
> 
> This diff points out where the problem areas are-
> 
> pgm_atomic_add8
> pgm_atomic_fetch_and_add8
> pgm_atomic_inc8
> 
> pgm_atomic_add16
> pgm_atomic_fetch_and_add16
> pgm_atomic_inc16
> 
> 
> It was a clean build before ticket locks :-)
> 
> I've added some MASM64 to trunk to implement the assembler as functions but 
> I'm without 64-bit MSVC to test, I wouldn't say I"m overly confident on the 
> 64-bit assembler either.
> 
> -- 
> Steve-o
> 
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] 0MQ V2.1.2 released on OpenVMS - see http://zeromqonopenvms.blogspot.com/ for further information

2011-03-07 Thread Apps, John 

Here is the URI for those who cannot read it in the subject line: 
http://zeromqonopenvms.blogspot.com/

All Brett's work despite the earthquake!

 John Apps | Hewlett-Packard | john.a...@hp.com | +49 171 869 1813 


-Original Message-
From: zeromq-dev-boun...@lists.zeromq.org 
[mailto:zeromq-dev-boun...@lists.zeromq.org] On Behalf Of Pieter Hintjens
Sent: Montag, 7. März 2011 18:59
To: ZeroMQ development list
Subject: Re: [zeromq-dev] 0MQ V2.1.2 released on OpenVMS - see 
http://zeromqonopenvms.blogspot.com/ for further information

Excellent stuff!

BTW, URIs in subjects aren't clickable in many email clients...

On Mon, Mar 7, 2011 at 6:46 PM, Apps, John  wrote:
>
>
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Making push-pull connections

2011-03-07 Thread Timothy Shead 
On 3/7/11 11:33 AM, Pieter Hintjens wrote:

OK, so this is comparable to the PUB-SUB connection issues described in 
detail in the Guide ... that gives me what I need to proceed.

Many thanks!
Tim

> There is a trick, yes. Connects take a finite time, so the first pull
> socket past the gate will grab a lot of messages before the others
> arrive. You need to have all pull sockets connected before pushing out
> messages.
>
> -Pieter
>
> On 7 Mar 2011 19:25, "Timothy Shead"  > wrote:
>  > Folks:
>  >
>  > Is there a trick when connecting multiple PULL sockets to a single PUSH
>  > socket? What I'm seeing is that sometimes my messages are load-balanced
>  > among all the connected PULL sockets, but other times go to just one
>  > PULL socket, or a subset of PULL sockets. My assumption is that some of
>  > my PULL sockets aren't getting a chance to connect, so aren't receiving
>  > messages, but I'm not clear on what to do differently.
>  >
>  > Thanks in advance,
>  > Tim
>  >
>  >
>  > --
>  > Timothy M. Shead
>  > Sandia National Laboratories
>  > 1461, Scalable Analysis and Visualization
>  >
>  > ___
>  > zeromq-dev mailing list
>  > zeromq-dev@lists.zeromq.org 
>  > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>  >


-- 
Timothy M. Shead
Sandia National Laboratories
1461, Scalable Analysis and Visualization

___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Making push-pull connections

2011-03-07 Thread Pieter Hintjens 
There is a trick, yes. Connects take a finite time, so the first pull socket
past the gate will grab a lot of messages before the others arrive. You need
to have all pull sockets connected before pushing out messages.

-Pieter
On 7 Mar 2011 19:25, "Timothy Shead"  wrote:
> Folks:
>
> Is there a trick when connecting multiple PULL sockets to a single PUSH
> socket? What I'm seeing is that sometimes my messages are load-balanced
> among all the connected PULL sockets, but other times go to just one
> PULL socket, or a subset of PULL sockets. My assumption is that some of
> my PULL sockets aren't getting a chance to connect, so aren't receiving
> messages, but I'm not clear on what to do differently.
>
> Thanks in advance,
> Tim
>
>
> --
> Timothy M. Shead
> Sandia National Laboratories
> 1461, Scalable Analysis and Visualization
>
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

[zeromq-dev] Making push-pull connections

2011-03-07 Thread Timothy Shead 
Folks:

Is there a trick when connecting multiple PULL sockets to a single PUSH 
socket?  What I'm seeing is that sometimes my messages are load-balanced 
among all the connected PULL sockets, but other times go to just one 
PULL socket, or a subset of PULL sockets.  My assumption is that some of 
my PULL sockets aren't getting a chance to connect, so aren't receiving 
messages, but I'm not clear on what to do differently.

Thanks in advance,
Tim


-- 
Timothy M. Shead
Sandia National Laboratories
1461, Scalable Analysis and Visualization

___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] 0MQ V2.1.2 released on OpenVMS - see http://zeromqonopenvms.blogspot.com/ for further information

2011-03-07 Thread Pieter Hintjens 
Excellent stuff!

BTW, URIs in subjects aren't clickable in many email clients...

On Mon, Mar 7, 2011 at 6:46 PM, Apps, John  wrote:
>
>
> ___
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

[zeromq-dev] 0MQ V2.1.2 released on OpenVMS - see http://zeromqonopenvms.blogspot.com/ for further information

2011-03-07 Thread Apps, John 

___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

[zeromq-dev] zmsg C++ Translation

2011-03-07 Thread Hoelzlwimmer Andreas 
As I'm going for some Guide Chapter 4 C++ Translations, here's the Translation 
of zmsg.h into C++.


zmsg.hpp
Description: zmsg.hpp
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] Problem building application in Suse 11.3 (uuid lib issues)

2011-03-07 Thread PJ Durai 
On Mon, Mar 7, 2011 at 12:50 AM, Dhammika Pathirana  wrote:
> Hi PJ,
>
> On Sun, Mar 6, 2011 at 10:34 PM, PJ Durai  wrote:
>> Greetings,
>>
>> I am having some issues in building my application in Suse 11.3 instalation.
>>
>> Zero MQ library itself has no problems building.
>>
>> But when I build my application, I end up with the following error
>>
>> /home/pi/dev/lib/libzmq.a(libzmq_la-uuid.o): In function `uuid_t':
>> /home/pi/tmp/zeromq-2.1.2/src/uuid.cpp:80: undefined reference to
>> `uuid_generate'
>> /home/pi/tmp/zeromq-2.1.2/src/uuid.cpp:81: undefined reference to 
>> `uuid_unparse'
>>
>> I have installed, reinstalled uuid-dev packages from repositories. To no 
>> avail.
>>
>
>
> You have to link your app with -luuid
> :-P

Indeed I tried that. But my Make fails with an error.
I do have the Header and the SO files in proper loctions. (But I can't
find  .a file)
But Make fails saying 'unable to find uuid'.  I think Suse 11.3 has
some weirdness in that area.
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] [PATCH] Update to OpenPGM 5.1.110.

2011-03-07 Thread Martin Sustrik 
On 03/07/2011 11:45 AM, Steven McCoy wrote:
> Ok, installed MSVC x64 and it nows builds in trunk with cmake -G "NMake
> Makefiles".  Still might be broken though :D

I am not sure now. Should the patch be applied to not?

Martin
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] List of ideas for GSoC

2011-03-07 Thread Martin Sustrik 
On 03/06/2011 06:49 PM, Pieter Hintjens wrote:

>> Also, it would be helpful to expand the text. Say: "Implement a delivery
>> mechanism that can withstand restarts of both the sender and receiver,"
>> doesn't make much sense for me.
>
> Seems clear enough. This particular functionality is almost done for
> Ch4 of the Guide, Titanic pattern running over Majordomo protocol.

I think it has to do with PGM. RendezVous certified messages or somesuch.

Anyway, the point is that ideas presented like this are unhelpful for 
students looking for ideas for GSoC.

Keep in mind that organisations to participate on GSoC are chosen mainly 
on the basis of their ideas page. Thus, messy ideas page can make the 
whole GSoC thing fail.

So, please, everyone, make your ideas clearly articulated and take care 
to choose such ideas that can be implemented by a student during the summer.

Thanks!
Martin
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Re: [zeromq-dev] [PATCH] Update to OpenPGM 5.1.110.

2011-03-07 Thread Steven McCoy 
Ok, installed MSVC x64 and it nows builds in trunk with cmake -G "NMake
Makefiles".  Still might be broken though :D

-- 
Steve-o

On 7 March 2011 12:08, Steven McCoy  wrote:

> On 7 March 2011 01:31, Alex Forster  wrote:
>
>> Steven McCoy  miru.hk> writes:
>>
>> > Clean builds under Win64.
>> > ...
>> > Improve spinlock performance with inline ticket based spinlock
>> implementation.
>>
>>
>> http://www.google.com/search?q=__asm+keyword+not+supported+on+this+architecture
>>
>> ticket.h is coughing up because Visual C++ doesn't support inline assembly
>> when
>> targeting 64-bit.
>>
>>
> Oh that's nice.
>
>
>> http://code.google.com/p/openpgm/source/diff?
>>
>> spec=svn1294&r=1294&format=side&path=/trunk/openpgm/pgm/include/impl/ticket.h
>>
>> This diff points out where the problem areas are-
>>
>> pgm_atomic_add8
>> pgm_atomic_fetch_and_add8
>> pgm_atomic_inc8
>>
>> pgm_atomic_add16
>> pgm_atomic_fetch_and_add16
>> pgm_atomic_inc16
>>
>>
> It was a clean build before ticket locks :-)
>
> I've added some MASM64 to trunk to implement the assembler as functions but
> I'm without 64-bit MSVC to test, I wouldn't say I"m overly confident on the
> 64-bit assembler either.
>
> --
> Steve-o
>
___
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev