Run MPSoCBench

81 views
Skip to first unread message

erwan....@gmail.com

unread,
Oct 21, 2015, 12:29:27 PM10/21/15
to MPSoCBench
Hi MPSoCBench Developpers,

I have a problem with the command : git submodule update --init --recursive.
When I launch   sh setup.sh ../ArchC-2.4.0/ when I am following your tutorial I get this output :
-ne --------------------------------------------------------------------------------------------
-ne SETUP: Initiating the submodules...
-ne SETUP: Copying the Environment file from ArchC to MPSoCBench...
-ne SETUP: SUCCESSFUL!
-ne --------------------------------------------------------------------------------------------
-ne The new file called 'env.sh' will set the environment variables to MPSoCBench work properly.
-ne So, whenever you are in a new terminal (bash), run:
-ne     source env.sh
-ne Or put the following command in your ~/.bashrc:
-ne      source /home/erwan/Documents/Logiciels/MPSoCBench/MPSoCBench-2.0.0/env.sh

At the beginning I had a "fatal" with the git command because I haven't git so I installed it and I did the command "git init" at the directory /home/erwan/Documents/Logiciels/MPSoCBench/MPSoCBench-2.0.0
And after I get the successful "sh setup.sh ../ArchC-2.4.0/" above

After that I use the command source env.sh (there is no errrors)

Finally, I enter the command ./MPSoCBench -h and it tells me to do :  git submodule update --init --recursive

May you tell me what is the problem?

Thank you in advance, best regards.
Erwan


erwan....@gmail.com

unread,
Oct 22, 2015, 5:06:35 AM10/22/15
to MPSoCBench
In order to find a solution I downloaded a clone with  git clone https://github.com/ArchC/MPSoCBench.git
It worked fine and after that I launch ./setup.sh ../ArchC-2.4.0/
and I get this error :
--------------------------------------------------------------------------------------------
SETUP: Initiating the submodules...
Clonage dans 'processors/arm'...
Warning: Permanently added the RSA host key for IP address '192.30.252.129' to the list of known hosts.
Permission denied (publickey).
fatal: Could not read from remote repository.
Please make sure you have the correct access rights
and the repository exists.
Le clonage de 'g...@github.com:ArchC/arm.git' dans le chemin de sous-module 'processors/arm' a échoué

SETUP: Copying the Environment file from ArchC to MPSoCBench...
SETUP: SUCCESSFUL!
--------------------------------------------------------------------------------------------

The new file called 'env.sh' will set the environment variables to MPSoCBench work properly.
So, whenever you are in a new terminal (bash), run:
    source env.sh

Or put the following command in your ~/.bashrc:
     source /home/erwan/Documents/Logiciels/MPSoCBench/MPSoCBench/env.sh

I'm sorry but I don't know how to fix that problem, may you help me please?

Thank you in advance,
Erwan




Liana Duenha

unread,
Oct 22, 2015, 6:52:36 AM10/22/15
to mpsoc...@googlegroups.com
Hello Erwan,

I hope you are pretty well!
After reading your second e-mail, it seems that the submodules (the processor modules) were properly stored in your MPSoCBench. 


The env.sh script handles the environment variables in order to run the ArchC and other tools. 

Please run the following commands and send to us the output messages, and we will help you to solve the problem.

$ cd MPSoCBench
$ source env.sh 
$ ./MPSoCBench -h 

Best regards,
Liana.


--
You received this message because you are registered to the group "MPSoCBench" Google Groups.
To post to this group, send email to mpsoc...@googlegroups.com
Visit this group http://groups.google.com/group/mpsocbench?hl=en
---
You received this message because you are subscribed to the Google Groups "MPSoCBench" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mpsocbench+...@googlegroups.com.
To post to this group, send email to mpsoc...@googlegroups.com.
Visit this group at http://groups.google.com/group/mpsocbench.



--
-------------------------------------------------------------------------
Liana Duenha

Faculdade de Computação - FACOM/UFMS
Website Institucional: www.facom.ufms.br


Faculty of Computing - FACOM/UFMS
Faculty Website: www.facom.ufms.br

erwan....@gmail.com

unread,
Oct 22, 2015, 9:37:44 AM10/22/15
to MPSoCBench
Hello Liana,

I'm fine and I hope your are fine too! Thanks for your reply. Below is my terminal view after I executed your commands :

erwan@lo-lester-040:~/Documents/Logiciels/MPSoCBench$ ls
ArchC-2.4.0  MPSoCBench  tools
erwan@lo-lester-040:~/Documents/Logiciels/MPSoCBench$ cd ./MPSoCBench/
erwan@lo-lester-040:~/Documents/Logiciels/MPSoCBench/MPSoCBench$ ls
defines.h  ip  LICENSE         MPSoCBench  processors  setup.sh   sw
env.sh     is  Makefile.rules  platforms   README.md   setup.sh~  wrappers
erwan@lo-lester-040:~/Documents/Logiciels/MPSoCBench/MPSoCBench$ source env.sh
erwan@lo-lester-040:~/Documents/Logiciels/MPSoCBench/MPSoCBench$
erwan@lo-lester-040:~/Documents/Logiciels/MPSoCBench/MPSoCBench$ ./MPSoCBench -h
The code of the models does not exist; you probably forgot to run:

    git submodule update --init --recursive

erwan@lo-lester-040:~/Documents/Logiciels/MPSoCBench/MPSoCBench$

I hope this will helps you. I don't know if it's usefull but I'll give you the content of my env.sh file :
#!/bin/sh
prefix=/home/erwan/Documents/Logiciels/MPSoCBench/ArchC-2.4.0
exec_prefix=${prefix}
libdir=${exec_prefix}/lib
includedir=${prefix}/include

export LD_LIBRARY_PATH="/home/erwan/Documents/systemc-2.3.1/lib-linux64:$LD_LIBRARY_PATH"
export PKG_CONFIG_PATH="/home/erwan/Documents/systemc-2.3.1/lib-linux64/pkgconfig:$PKG_CONFIG_PATH"

export PATH="$exec_prefix/bin:$PATH"
export LD_LIBRARY_PATH="$libdir:$LD_LIBRARY_PATH"
export PKG_CONFIG_PATH="$libdir/pkgconfig:$PKG_CONFIG_PATH"
export ARCHC_PREFIX="$prefix"

Thank you again Liana, best regards.
Erwan

Liana Duenha

unread,
Oct 22, 2015, 2:50:14 PM10/22/15
to mpsoc...@googlegroups.com
Hello Erwan,

The first problem happened because you did not have the "Git" installed. I think that you solved this problem after installing it. 
But, anyway, I think you also did not have permission to clone them. 
So, the ArchC support updated the way to clone the git submodules, in order to allow that any user clone them. 

I recommend that you start the process again:

- delete all MPSoCBench infrastructure

$ rm -rf MPSoCBench

- unzip it
- run the script:    

$ sh setup.sh

(we hope that the submodules will be cloned now )


- run the environment script to configure your environment

$ source env.sh

- run the MPSoCBench , for example:

$ ./MPSoCBench -p=mips -i=router.lt -n=2 -s=dijkstra -b


Please let me know if you still have problems.


Liana.








<li>Download the <a target="_BLANK" href="https://github.com/ArchC/MPSoCBench/releases"> MPSoCBench 2.0</a></li>
<li>Unzip in the opt/ directory:</li>
<pre><span class="inner-pre" style="font-size: 16px">$ tar xvf mpsocbench.tar.gz </span></pre>
<li>Use the setup.sh script to download automatically the newest version of the ArchC processor models </li>
      <pre><span class="inner-pre" style="font-size: 16px">$ sh setup.sh </span></pre>
      <li>Use the env.sh script to configure your system environment automatically </li>
      <pre><span class="inner-pre" style="font-size: 16px">$ source env.sh </span></pre>
      <pre><span class="inner-pre" style="font-size: 16px">$ bash -l </span></pre>




Erwan Moréac

unread,
Oct 23, 2015, 8:09:41 AM10/23/15
to mpsoc...@googlegroups.com
Hello Liana,

I followed your instructions, retry the process. I deleted MPSoCBench files and directories and downloaded to http://www.archc.org/benchs/mpsocbench/downloads.html  

This what I get :

erwan@lo-lester-040:~/Documents/Logiciels/MPSoCBench/MPSoCBench-2.0.0$ sh setup.sh /home/erwan/Documents/Logiciels/MPSoCBench/ArchC-2.4.0
-ne --------------------------------------------------------------------------------------------
-ne SETUP: Initiating the submodules...
fatal: Not a git repository (or any of the parent directories): .git

-ne SETUP: Copying the Environment file from ArchC to MPSoCBench...
-ne SETUP: SUCCESSFUL!
-ne --------------------------------------------------------------------------------------------
-ne The new file called 'env.sh' will set the environment variables to MPSoCBench work properly.
-ne So, whenever you are in a new terminal (bash), run:
-ne     source env.sh
-ne Or put the following command in your ~/.bashrc:
-ne      source /home/erwan/Documents/Logiciels/MPSoCBench/MPSoCBench-2.0.0/env.sh
erwan@lo-lester-040:~/Documents/Logiciels/MPSoCBench/MPSoCBench-2.0.0$ ./MPSoCBench -p=mips -i=router.lt -n=2 -s=dijkstra -b

The code of the models does not exist; you probably forgot to run:
    git submodule update --init --recursive
erwan@lo-lester-040:~/Documents/Logiciels/MPSoCBench/MPSoCBench-2.0.0$


So, I deleted again the MPSoCBench folder and I did the command :  git clone https://github.com/ArchC/MPSoCBench.git
And this time setup.sh worked! :)

erwan@lo-lester-040:~/Documents/Logiciels/MPSoCBench/MPSoCBench$ sh setup.sh /home/erwan/Documents/Logiciels/MPSoCBench/ArchC-2.4.0
-ne --------------------------------------------------------------------------------------------
-ne SETUP: Initiating the submodules...
Sous-module 'processors/arm' (https://github.com/ArchC/arm.git) enregistré pour le chemin 'processors/arm'
Sous-module 'processors/mips' (https://github.com/ArchC/mips.git) enregistré pour le chemin 'processors/mips'
Sous-module 'processors/powerpc' (https://github.com/ArchC/powerpc.git) enregistré pour le chemin 'processors/powerpc'
Sous-module 'processors/sparc' (https://github.com/ArchC/sparc.git) enregistré pour le chemin 'processors/sparc'
Clonage dans 'processors/arm'...
remote: Counting objects: 161, done.
remote: Compressing objects: 100% (3/3), done.
remote: Total 161 (delta 14), reused 13 (delta 13), pack-reused 145
Réception d'objets: 100% (161/161), 84.03 KiB | 0 bytes/s, done.
Résolution des deltas: 100% (91/91), done.
Vérification de la connectivité... fait.
Chemin de sous-module 'processors/arm' : 'ec75ede4e11a1b6618c95b8b0e0da7e9b22dcef2' extrait
Clonage dans 'processors/mips'...
remote: Counting objects: 232, done.
remote: Compressing objects: 100% (3/3), done.
remote: Total 232 (delta 0), reused 0 (delta 0), pack-reused 229
Réception d'objets: 100% (232/232), 59.19 KiB | 0 bytes/s, done.
Résolution des deltas: 100% (128/128), done.
Vérification de la connectivité... fait.
Chemin de sous-module 'processors/mips' : 'b18f2408339372a8a8966707f83e2142dad95ab0' extrait
Clonage dans 'processors/powerpc'...
remote: Counting objects: 122, done.
remote: Compressing objects: 100% (3/3), done.
remote: Total 122 (delta 14), reused 14 (delta 14), pack-reused 105
Réception d'objets: 100% (122/122), 46.01 KiB | 0 bytes/s, done.
Résolution des deltas: 100% (66/66), done.
Vérification de la connectivité... fait.
Chemin de sous-module 'processors/powerpc' : '2eb3444901f11ffc4c1964b10e989d41483d65c1' extrait
Clonage dans 'processors/sparc'...
remote: Counting objects: 184, done.
remote: Total 184 (delta 1), reused 1 (delta 1), pack-reused 182
Réception d'objets: 100% (184/184), 56.90 KiB | 0 bytes/s, done.
Résolution des deltas: 100% (101/101), done.
Vérification de la connectivité... fait.
Chemin de sous-module 'processors/sparc' : '784c09145af20434d9627bc8e0be93a9e2df86b6' extrait

-ne SETUP: Copying the Environment file from ArchC to MPSoCBench...
-ne SETUP: SUCCESSFUL!
-ne --------------------------------------------------------------------------------------------
-ne The new file called 'env.sh' will set the environment variables to MPSoCBench work properly.
-ne So, whenever you are in a new terminal (bash), run:
-ne     source env.sh
-ne Or put the following command in your ~/.bashrc:
-ne      source /home/erwan/Documents/Logiciels/MPSoCBench/MPSoCBench/env.sh
erwan@lo-lester-040:~/Documents/Logiciels/MPSoCBench/MPSoCBench$

There is a progress but it's not totally solved.

erwan@lo-lester-040:~/Documents/Logiciels/MPSoCBench/MPSoCBench$ ./MPSoCBench -p=mips -i=router.lt -n=2 -s=dijkstra -b
[['mips'], ['2'], ['dijkstra'], ['router.lt']]
ERROR: Cross-compiler mips-newlib-elf-gcc is not in the PATH
erwan@lo-lester-040:~/Documents/Logiciels/MPSoCBench/MPSoCBench$

I guess MPSoCBench do not find compilers that are in
PATH=/home/erwan/Documents/Logiciels/MPSoCBench/tools/compilers/mips-newlib-elf/bin:$PATH

Thank you for your support.
Erwan



You received this message because you are subscribed to a topic in the Google Groups "MPSoCBench" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/mpsocbench/7C59AzDmQes/unsubscribe.
To unsubscribe from this group and all its topics, send an email to mpsocbench+...@googlegroups.com.

Liana Duenha

unread,
Oct 23, 2015, 10:19:08 AM10/23/15
to mpsoc...@googlegroups.com
Hello Erwan,
Yes, It was a progress! 
To solve this issue in my system, I used the following commands in the ~/.profile file:

export PATH=/opt/tools/compilers/mips-newlib-elf/bin:$PATH
export PATH=/opt/tools/compilers/arm-newlib-eabi/bin:$PATH
export PATH=/opt/tools/compilers/powerpc-newlib-elf/bin:$PATH
export PATH=/opt/tools/compilers/sparc-newlib-elf/bin:$PATH

(/opt/tools/compilers/ is the folder where I put these compilers)


Please try to use the command below in your MPSoCBench folder:

$ bash -l 
(the argument is a tiny L  :-) )

and try to run the command again:
 ./MPSoCBench -p=mips -i=router.lt -n=2 -s=dijkstra -b

While you try to follow these directions, I will verify if there are other causes for this problem. 
We keep talking.

Liana





Rodolfo Azevedo

unread,
Oct 24, 2015, 10:48:28 AM10/24/15
to mpsoc...@googlegroups.com
 parte?Oi Liana,

  o env.sh já devia setar os caminhos dos compiladores. Ele está lá para isto.

  Quanto ao git, parece que você não testou baixar o pacote geral. Pelo visto, git submodules só funciona se o principal for baixado do git também. Pode checar esta parte?

    Rodolfo
Rodolfo Jardim de Azevedo
http://www.ic.unicamp.br/~rodolfo
IC - University of Campinas - UNICAMP

Felippi Crominski

unread,
Oct 24, 2015, 9:38:41 PM10/24/15
to MPSoCBench
For solve that problem with git submodules I use:

On folder of MPSocBench,

rm processors -r
git init
git submodule add https://github.com/ArchC/mips.git processors/mips
git submodule add https://github.com/ArchC/sparc.git processors/sparc
git submodule add https://github.com/ArchC/powerpc.git processors/powerpc
git submodule add https://github.com/ArchC/arm.git processors/arm


I don't know if that solution is the best, but work for me.

Liana Duenha

unread,
Oct 25, 2015, 6:17:34 AM10/25/15
to mpsoc...@googlegroups.com

Hi Rodolfo,

The env.sh is not prepared for set the compilers path yet (it starts path only for the ArchC, SystemC and MPSoCBench). So, in the currently version we need to edit the .profile once or the PATH variable. We will improve this issue  enabling cloning the compilers as submodules.

I think that the git clone submodules problems were solved by Maxiwel. It was a permission problems when cloning submodules from outside the Unicamp.

Liana.

Liana Duenha

unread,
Oct 25, 2015, 6:18:17 AM10/25/15
to mpsoc...@googlegroups.com
Thank you, Felipe!
Liana.

--
You received this message because you are registered to the group "MPSoCBench" Google Groups.
To post to this group, send email to mpsoc...@googlegroups.com
Visit this group http://groups.google.com/group/mpsocbench?hl=en
---
You received this message because you are subscribed to the Google Groups "MPSoCBench" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mpsocbench+...@googlegroups.com.
To post to this group, send email to mpsoc...@googlegroups.com.
Visit this group at http://groups.google.com/group/mpsocbench.

Erwan Moréac

unread,
Oct 27, 2015, 6:34:11 AM10/27/15
to mpsoc...@googlegroups.com
Hi Liana!

Sorry for slowly reply. I'm currently in Hollydays! ^^
I restart the debugging the next week.
Thank you again for the help!

Best regards,
Erwan


erwan....@gmail.com

unread,
Nov 2, 2015, 12:02:34 PM11/2/15
to MPSoCBench
Hi Liana,

I hope you are fine, your last instructions were good.
It finally works! :)
Thank you very much for the support!

Best regards,
Erwan
To unsubscribe from this group and stop receiving emails from it, send an email to mpsocbench+unsubscribe@googlegroups.com.

To post to this group, send email to mpsoc...@googlegroups.com.
Visit this group at http://groups.google.com/group/mpsocbench.



--
-------------------------------------------------------------------------
Liana Duenha

Faculdade de Computação - FACOM/UFMS
Website Institucional: www.facom.ufms.br


Faculty of Computing - FACOM/UFMS
Faculty Website: www.facom.ufms.br

--
You received this message because you are registered to the group "MPSoCBench" Google Groups.
To post to this group, send email to mpsoc...@googlegroups.com
Visit this group http://groups.google.com/group/mpsocbench?hl=en
---
You received this message because you are subscribed to a topic in the Google Groups "MPSoCBench" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/mpsocbench/7C59AzDmQes/unsubscribe.
To unsubscribe from this group and all its topics, send an email to mpsocbench+unsubscribe@googlegroups.com.

To post to this group, send email to mpsoc...@googlegroups.com.
Visit this group at http://groups.google.com/group/mpsocbench.

--
You received this message because you are registered to the group "MPSoCBench" Google Groups.
To post to this group, send email to mpsoc...@googlegroups.com
Visit this group http://groups.google.com/group/mpsocbench?hl=en
---
You received this message because you are subscribed to the Google Groups "MPSoCBench" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mpsocbench+unsubscribe@googlegroups.com.

To post to this group, send email to mpsoc...@googlegroups.com.
Visit this group at http://groups.google.com/group/mpsocbench.



--
-------------------------------------------------------------------------
Liana Duenha

Faculdade de Computação - FACOM/UFMS
Website Institucional: www.facom.ufms.br


Faculty of Computing - FACOM/UFMS
Faculty Website: www.facom.ufms.br

--
You received this message because you are registered to the group "MPSoCBench" Google Groups.
To post to this group, send email to mpsoc...@googlegroups.com
Visit this group http://groups.google.com/group/mpsocbench?hl=en
---
You received this message because you are subscribed to a topic in the Google Groups "MPSoCBench" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/mpsocbench/7C59AzDmQes/unsubscribe.
To unsubscribe from this group and all its topics, send an email to mpsocbench+unsubscribe@googlegroups.com.

Liana Duenha

unread,
Nov 5, 2015, 7:14:59 AM11/5/15
to mpsoc...@googlegroups.com
Hello Erwan! 
Feel free to write whenever you need.
Liana.

You received this message because you are subscribed to a topic in the Google Groups "MPSoCBench" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/mpsocbench/7C59AzDmQes/unsubscribe.
To unsubscribe from this group and all its topics, send an email to mpsocbench+...@googlegroups.com.
You received this message because you are subscribed to a topic in the Google Groups "MPSoCBench" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/mpsocbench/7C59AzDmQes/unsubscribe.
To unsubscribe from this group and all its topics, send an email to mpsocbench+...@googlegroups.com.

Erwan Moréac

unread,
Nov 9, 2015, 12:16:54 PM11/9/15
to mpsoc...@googlegroups.com
Hi Liana,

MPSoCBench works well and I have one or two questions if you have the time to answer them. I'm currently studying the platform NoC at because I want to get data payload of the packages injected by IPs in this network.

This purpose led me to  tlm_node.cpp file and tlm_generic_payload variables.
I found functions to use "properly" payload's attributes at this link : https://www.doulos.com/knowhow/systemc/tlm2/tutorial__1/ . The problem is that I cannot say if data I can read on my terminal are good or not.

My question is : Can I know with MPSoCBench all packages at the bit-level sent by all IPs in the NoC during an application ?

Thank you in advance for your answer.

Best regards,
Erwan


Liana Duenha

unread,
Nov 10, 2015, 1:13:35 PM11/10/15
to mpsoc...@googlegroups.com
Hi Erwan,

The MPSoCBench uses TLM2.0 to manage communication among processors and IPs. The TLM generic payload encapsulates necessary data (such as: the command, address, and the data pointer) and the communication between SystemC processes is done through function call (using a reference to the payload as a parameter). 

For example, 

The initiator stores data in the payload:


    payload_global->set_address((uint64_t)address);
    payload_global->set_data_length(sizeof(uint8_t));
    payload_global->set_data_ptr(p);

and the target reads from payload:

    uint32_t addr = (uint32_t) payload.get_address();
    unsigned int len = payload.get_data_length();
    unsigned char* data_pointer = payload.get_data_ptr();



So, this communication model does not support packages evaluation at a bit-level.

Best regards,
Liana.

Erwan Moréac

unread,
Nov 12, 2015, 9:53:57 AM11/12/15
to mpsoc...@googlegroups.com
Hi Liana,

Thanks for your reply. I think there is something I didn't understand. Your TLM communication used by processors with the NoC, how processors can they realize applications without real data?

The data pointer cannot give us the data used by processors? If yes, we could know the content of each packages, isn't it?

Thank you again for your help, best regards.
Erwan

Liana Duenha

unread,
Nov 17, 2015, 6:02:11 AM11/17/15
to mpsoc...@googlegroups.com
Hello Erwan,
The processors execute applications with real data. So,  you can use the data pointer to explore the data transferred between the IPs (memory, for instance) and processors. 
Best regards!
Liana.

Erwan Moréac

unread,
Nov 18, 2015, 11:24:13 AM11/18/15
to mpsoc...@googlegroups.com
Hello Liana,

Thanks for your confirmation. In order to get data as soon as they are injected in the noc.at platform, I focused my work in tlm_node.cpp in addToBuffer where the packages are stored in the network.

Here is the function prototype: void tlm_node::addToBuffer (ac_tlm2_payload &p, tlm::tlm_phase& phase, sc_core::sc_time &t)

The payload p is one of the parameters of the function so I used functions to look inside this payload p with the followings lines:
unsigned char*   ptr = p.get_data_ptr();
unsigned int     len = p.get_data_length();
unsigned char*   byt = p.get_byte_enable_ptr();
unsigned int     raw_data = 0;

printf("\nNode %d,%d add package in buffer has %d packages, length %d, payload data : "
,getX(), getY(), getNumberOfPackagesInBuffer(),len);
                if(len > 0){
                    //data recovery with the ptr pointer
                    for(int cpt=(len-1);cpt>=0;cpt--){
                    //data are divided in several bytes that we need to regroup
                        if(cpt < (len-1)){
                            raw_data = raw_data << 8;
                        }
                        raw_data = raw_data + ptr[cpt];
                    }
                    printf("%u ",raw_data);
                    printf("  en hexa %x ",raw_data);
                }

In order to test this data extraction I use this command of MPSoCBench :
./MPSoCBench -p=mips -i=noc.at -n=4 -s=basicmath -b    then,
./MPSoCBench -p=mips -i=noc.at -n=4 -s=basicmath -r

Now this is a screenshot of a short time during the run :

Node 0,0 add package in buffer has 1 packages, length 4, payload data : 32449152   en hexa 1ef2280
Node 1,2 add package in buffer has 1 packages, length 4, payload data : 32407232   en hexa 1ee7ec0
Node 0,2 add package in buffer has 1 packages, length 4, payload data : 32407232   en hexa 1ee7ec0
Node 0,1 add package in buffer has 1 packages, length 4, payload data : 32407232   en hexa 1ee7ec0
Node 0,0 add package in buffer has 1 packages, length 4, payload data : 32407232   en hexa 1ee7ec0
Node 1,1 add package in buffer has 1 packages, length 4, payload data : 32365424   en hexa 1eddb70
Node 0,1 add package in buffer has 1 packages, length 4, payload data : 32365424   en hexa 1eddb70
Node 0,0 add package in buffer has 2 packages, length 4, payload data : 32365424   en hexa 1eddb70
Node 1,0 add package in buffer has 1 packages, length 4, payload data : 568901635   en hexa 21e8c003
Node 0,0 add package in buffer has 3 packages, length 4, payload data : 568901635   en hexa 21e8c003
Node 0,0 add package in buffer has 4 packages, length 4, payload data : 0   en hexa 0
Node 0,0 add package in buffer has 4 packages, length 4, payload data : 0   en hexa 0
Node 0,0 add package in buffer has 4 packages, length 4, payload data : 0   en hexa 0
Node 0,0 add package in buffer has 4 packages, length 4, payload data : 67157647   en hexa 400be8f
Node 1,0 add package in buffer has 1 packages, length 4, payload data : 0   en hexa 0
Node 2,0 add package in buffer has 1 packages, length 4, payload data : 0   en hexa 0

I cannot find any of these numbers read in the output file so I wonder if there are good or not. Should I look the data pointer in the main thread in tlm_node or elsewhere ?

Thank you for the time you take to help me.

Best regards,
Erwan









Liana Duenha

unread,
Nov 23, 2015, 3:18:51 PM11/23/15
to mpsoc...@googlegroups.com
Hello Erwan,
It seems that the better place to explore the payload data is in the thread_node method or in the addToBuffer method (as you have done). Both methods are in the tlm_node.cpp file. 
Recently we fixed a bug in the XY routing algorithm and provided the new code in the 2.1 MPSoCBench.
As you are interested in these interconnections resources, it would be better if you download the new version. 
Best regards,
Liana.





Reply all
Reply to author
Forward
0 new messages