Prepare global Environment singleton for PMacc #166

c-schumann-zih · 2014-01-30T13:25:31Z

Pull request to discuss how to implement a single global Environment singleton in libPMacc, related to this issue.

It seems so far that a splitting into hpp and tpp files is necessary for classes that use __eventsystem_macros which moved into the new Environment.hpp.
This has been tested for TaskKernel and Buffer which need to be included in EventSystem.hpp now (whichis not great). However, more problems remain:

/home/schuma/picongpu_src/src/libPMacc/include/cuSTL/algorithm/kernel/run-time/Foreach.hpp(103):
error: too few arguments for class template "PMacc::Environment"

Any ideas on how to solve this better would be greatly appreciated :)

psychocoderHPC · 2014-01-30T13:40:42Z

src/libPMacc/include/Environment.hpp

+
+#include "eventSystem/tasks/Factory.hpp"
+#include "eventSystem/Manager.hpp"
+#include "eventSystem/transactions/TransactionManager.hpp"


Please only include EventSystem.hpp and no other classes in this folder. There are a lot of include loops inside thus you run very fast in compile problems which are very hard to solve.

very good note, this helped a lot!

ax3l · 2014-01-30T14:01:59Z

Hi Conrad,
thank you for your first pull request! ✨

I am sorry the test-environment created some weird output, but you spotted the problem yourself already. Just as a hint: manual tests can be done as described in this video - this produces much faster results than waiting for the suite.

I assigned @psychocoderHPC to the pull request since he probably knows best about the issues you reported (and @f-schmitt-zih is reviewing it anyway I guess).

Actually the EventSystem makros should be refactored (-> reads "removed") as a preparation for #133 . We might remove this makros already in a separate pull request and replace them by some nicer objects? Would you like to do that?

ax3l · 2014-01-30T14:07:08Z

src/libPMacc/include/Environment.hpp

+
+
+namespace PMacc {
+    //using namespace gol;


please remove me :)

psychocoderHPC · 2014-01-30T15:02:00Z

Thanks for your pull request!

I think there is no need to create .def files, because before picongpu and game of life compile fine without those definitions. I will look tomorrow to your branch.
I will test to extract only your environment class and fix the compile problems without adding .def files.

ax3l · 2014-01-30T15:23:33Z

The .def file is only intended for the new Environment class. The purpose is as I described:

if you change for example the number of templates of this class at some point in the future. It reduces also the Copy&Paste portion of the code.

@psychocoderHPC That will not solve the compile problems, but cleans up the implementation in secondary singletons.

-modefied Environment.hpp -removed most of the forward declarations for class Environment -removed some .tpp files

c-schumann-zih · 2014-01-30T16:01:01Z

I modefied the Files with your comments and removed most of the forward declarations for class Environment and the code compiled well.

[100%] Built target gameOfLife ;D

ax3l · 2014-01-30T16:47:08Z

yeah, cool 👍

[100%] Built target gameOfLife

I really have to fix the error message - I am quite interested in it :D

Edit: CompileSuite Message: Ah, your branch is also outdated/away from the dev of the mainline. I guess @f-schmitt-zih will love ^ to show you how to rebase against the current dev afterwards ;) ^sarcAnote1.0

psychocoderHPC · 2014-01-30T17:43:25Z

+1 for getting gol to compile

ax3l · 2014-02-21T13:55:13Z

Hi @c-schumann-zih, are you still working on that branch?
Do you like to rebase that branch or would you rather like to open a new pull request and close this one for now?

c-schumann-zih · 2014-02-24T06:40:51Z

Hi, i'm just back from a two-weeks trainingcouse. I think i will ask f-schmitt-zih how to move on.

ax3l · 2014-02-24T09:09:00Z

Nice, welcome back!

c-schumann-zih · 2014-02-25T11:44:54Z

Hi @psychocoderHPC,
I added some more Singletons to the Environment.hpp. They compile well until I try to integrate the ParticleFactory. After adding ParticleFactory to the Environment.hpp i get errors like 'identifier "__getTransactionEvent" is undefined" ' and I can't figure out the reason. Maybe you can give me a hint.

psychocoderHPC · 2014-02-25T12:03:55Z

OK I will check it.

- make PArticleFactory getInstance private - use Environment<> to get instance of ParticleFactory - sove all dependencies - fix all example compile time errors

psychocoderHPC · 2014-02-28T16:30:04Z

I pushed an update to your pull request branch (here).

Updates:

add ParticleFactory to Environment
solve all dependencies
fix all compile time erros

…dParticleFactory add ParticleFactory to Environment

psychocoderHPC · 2014-03-03T09:18:03Z

Mh why is this pull red after @c-schumann-zih merged my pull request. I have tested all. If somene can see the error please feel free to fix them :-)

ax3l · 2014-03-03T09:27:44Z

Probably due to a merge conflict...
He might have still to rebase against dev?

c-schumann-zih · 2014-03-03T11:54:41Z

Pull request is replaced by #246

123cf3c952 Merge commit 'a1601e35c146c09d339239f076e477680982c2f7' into spec-cupla-alpaka0.5.0dev a1601e35c1 Squashed 'alpaka/' changes from c2d14d09ea..b52eefa232 047884171f Merge pull request ComputationalRadiationPhysics#167 from psychocoderHPC/fix-ambiguousNamespace a63eeb52e1 Merge pull request ComputationalRadiationPhysics#166 from psychocoderHPC/topic-moveMathDir 8564d2dc4f fix ambiguous namespace `device` a0302ff66c move `cupla/math` to `cupla/device/math` d6310a037e Merge pull request ComputationalRadiationPhysics#165 from psychocoderHPC/topic-updateGithubLinks2 e7f9261bf7 update cupla github links f4dba001d4 Merge pull request ComputationalRadiationPhysics#162 from psychocoderHPC/topic-cuplaAccMathFunctions bbbc0b0266 `BlackScholes` use cupla math functions efa8d61037 on device math functions dbc0f649b2 Merge pull request ComputationalRadiationPhysics#160 from psychocoderHPC/topic-updateAlpakaTo0.5.0dev 13b737ba6c Squashed 'alpaka/' changes from ab0b8a460..c2d14d09e b31fc4ea33 Merge commit '13b737ba6cd38400636d6a6a52dee81b21a7412c' into topic-updateAlpakaTo0.5.0dev 09d5daf1ff update travis tests 582f8ca10e update CMake requirements 59fba492fa include changes needed to use alpaka 0.5.0 git-subtree-dir: thirdParty/cupla git-subtree-split: 123cf3c952f4b5c4d2200fb9f2bb09bb84019614

b72833cbf0 Merge pull request ComputationalRadiationPhysics#169 from psychocoderHPC/topic-moreMathFunctions b9944dc139 math functions: `erf` and `pow` 047884171f Merge pull request ComputationalRadiationPhysics#167 from psychocoderHPC/fix-ambiguousNamespace a63eeb52e1 Merge pull request ComputationalRadiationPhysics#166 from psychocoderHPC/topic-moveMathDir 8564d2dc4f fix ambiguous namespace `device` a0302ff66c move `cupla/math` to `cupla/device/math` d6310a037e Merge pull request ComputationalRadiationPhysics#165 from psychocoderHPC/topic-updateGithubLinks2 e7f9261bf7 update cupla github links f4dba001d4 Merge pull request ComputationalRadiationPhysics#162 from psychocoderHPC/topic-cuplaAccMathFunctions bbbc0b0266 `BlackScholes` use cupla math functions efa8d61037 on device math functions dbc0f649b2 Merge pull request ComputationalRadiationPhysics#160 from psychocoderHPC/topic-updateAlpakaTo0.5.0dev b31fc4ea33 Merge commit '13b737ba6cd38400636d6a6a52dee81b21a7412c' into topic-updateAlpakaTo0.5.0dev 13b737ba6c Squashed 'alpaka/' changes from ab0b8a460..c2d14d09e 09d5daf1ff update travis tests 582f8ca10e update CMake requirements 59fba492fa include changes needed to use alpaka 0.5.0 git-subtree-dir: thirdParty/cupla git-subtree-split: b72833cbf0844664b26bbf553289356cfa03c18a

Prepare global Environment singleton for PMacc

6d4e925

psychocoderHPC reviewed Jan 30, 2014
View reviewed changes

ghost assigned psychocoderHPC Jan 30, 2014

ax3l reviewed Jan 30, 2014
View reviewed changes

src/libPMacc/include/Environment.hpp

namespace PMacc {

//using namespace gol;

Copy link

Member

ax3l Jan 30, 2014

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

please remove me :)

Edited after review-comments

5bd866b

-modefied Environment.hpp -removed most of the forward declarations for class Environment -removed some .tpp files

c-schumann-zih added 2 commits February 24, 2014 14:39

Made picongpu compile with the singletons from GameofLife

6db5d6d

Integrated more Singletons in the Environment.hpp

4ff8b7f

ax3l mentioned this pull request Feb 28, 2014

Fix gpu init bug #239

Merged

add ParticleFactory to Environment

b162322

- make PArticleFactory getInstance private - use Environment<> to get instance of ParticleFactory - sove all dependencies - fix all example compile time errors

Merge pull request #1 from psychocoderHPC/conrad-topic-Environment-ad…

3a97623

…dParticleFactory add ParticleFactory to Environment

c-schumann-zih mentioned this pull request Mar 3, 2014

Topic environment singleton #246

Closed

c-schumann-zih closed this Mar 3, 2014

f-schmitt mentioned this pull request Mar 31, 2014

Integrate all Singletons in master Environment #91

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Prepare global Environment singleton for PMacc #166

Prepare global Environment singleton for PMacc #166

c-schumann-zih commented Jan 30, 2014

psychocoderHPC Jan 30, 2014

c-schumann-zih Jan 30, 2014

ax3l commented Jan 30, 2014

ax3l Jan 30, 2014

psychocoderHPC commented Jan 30, 2014

ax3l commented Jan 30, 2014

c-schumann-zih commented Jan 30, 2014

ax3l commented Jan 30, 2014

psychocoderHPC commented Jan 30, 2014

ax3l commented Feb 21, 2014

c-schumann-zih commented Feb 24, 2014

ax3l commented Feb 24, 2014

c-schumann-zih commented Feb 25, 2014

psychocoderHPC commented Feb 25, 2014

psychocoderHPC commented Feb 28, 2014

psychocoderHPC commented Mar 3, 2014

ax3l commented Mar 3, 2014

c-schumann-zih commented Mar 3, 2014

Prepare global Environment singleton for PMacc #166

Prepare global Environment singleton for PMacc #166

Conversation

c-schumann-zih commented Jan 30, 2014

psychocoderHPC Jan 30, 2014

Choose a reason for hiding this comment

c-schumann-zih Jan 30, 2014

Choose a reason for hiding this comment

ax3l commented Jan 30, 2014

ax3l Jan 30, 2014

Choose a reason for hiding this comment

psychocoderHPC commented Jan 30, 2014

ax3l commented Jan 30, 2014

c-schumann-zih commented Jan 30, 2014

ax3l commented Jan 30, 2014

psychocoderHPC commented Jan 30, 2014

ax3l commented Feb 21, 2014

c-schumann-zih commented Feb 24, 2014

ax3l commented Feb 24, 2014

c-schumann-zih commented Feb 25, 2014

psychocoderHPC commented Feb 25, 2014

psychocoderHPC commented Feb 28, 2014

psychocoderHPC commented Mar 3, 2014

ax3l commented Mar 3, 2014

c-schumann-zih commented Mar 3, 2014