The SimpliciTI can't be executed after increasing MAX_APP_PAYLOAD and MAX_NWK_PAYLOAD

Simon Huang1

Other Parts Discussed in Thread: SIMPLICITI, CC1200, MSP430F5438A, MSP430F5438

Hi all,

In order to transmit some bigger data packet, I tried to increase MAX_APP_PAYLOAD and MAX_NWK_PAYLOAD. I understand transmitting big data packet might need some modification of SimpliciTI to support that. With some modification, It can transmit 280-byte packet now although there are still some problems in stress test.

I want to test bigger size such as 1000 bytes. After I change the MAX_APP_PAYLOAD to 900, MAX_NWK_PAYLOAD to 1000, the image can be built without any error. But it can't be executed. In IAR debug mode, I processed 'Download and Debug', the image downloaded but the cursor didn't stop at main function. It can't 'Go'.

The following is my environment:

CC1200 TrxEB

MSP430F5438A - 256KB flash, 16KB SRAM

IAR Universal Linker V5.3.2.26

Modified SimpliciTI-IAR-1.2.1-Beta

I also attach my map files: Ap_as_Data_Hub - MAX_NWK_PAYLOAD=300.map is the workable version, Ap_as_Data_Hub - MAX_NWK_PAYLOAD=1000.map isn't.

3247.map_files.rar

by the summary of Ap_as_Data_Hub - MAX_NWK_PAYLOAD=1000.map:

****************************************
* *
* END OF CROSS REFERENCE *
* *
****************************************

15 914 bytes of CODE memory
9 510 bytes of DATA memory (+ 120 absolute )
1 831 bytes of CONST memory

I though the size of DATA memory is much less than the size of SRAM of MSP430F5438A. What's the problem? Could any body tell me. Thanks!

Best regards,

Simon

over 9 years ago

0 Leo Hendrawan over 9 years ago

TI__Guru 56650 points

Hi,

Currently SimpliciTI doesn't support CC1200. Second thing is that for now, SimpliciTI doesn't support packet fragmentation and the maximum packet size is the FIFO size of the transceiver.

0 TER over 9 years ago in reply to Leo Hendrawan

TI__Guru**** 317180 points

Simon, if I understand you correctly you have ported Simplicity to CC1200 and also included the infinite packet mode example into the code and this works but you run into issues with the space the data structures take?

0 Simon Huang1 over 9 years ago in reply to TER

Prodigy 180 points

Hi TER,

Yes, you are right. I think my problem can be simplified to be an issue that declaring too many big variables will cause system to be unworkable. The system even can't start up.

According to the summary of the map file and as I thought, the size of SRAM shall be enough. If I was right, what's the problem?

15 914 bytes of CODE memory
9 510 bytes of DATA memory (+ 120 absolute )
1 831 bytes of CONST memory

Any suggestion? Thanks!

Best regards,

Simon

0 Simon Huang1 over 9 years ago in reply to Simon Huang1

Prodigy 180 points

Can anybody give my a hand? ><

0 Simon Huang1 over 9 years ago in reply to Simon Huang1

Prodigy 180 points

upload a snapshot for the reference.

0 TIABO over 9 years ago in reply to Simon Huang1

TI__Genius 12900 points

Hi Simon,

Looking at the map files, I couldn't see anything suspicious.

You say that the code doesn't run to main. Have you tried disabling "Run to main" and stepped through the initialization code? This may give you an indication of where it breaks down. It may be that one of your large variables' (inFrameQ, outFrameQ, mrfiIncomingPacket, mrfiOutgoinggPacket) causes a problem.

You should be able to disable "Run to main" from Project > Options > Debugger.

Br,
TIABO

0 Simon Huang1 over 9 years ago in reply to TIABO

Prodigy 180 points

Dear TIABO,

Thanks for your reply.

According to your suggestion, I disabled "Run to main". By stepping the program, it stuck in __data16_memzero, as the attached file.

What can I do? Please help.

Best regards,

Simon

0 TIABO over 9 years ago in reply to Simon Huang1

TI__Genius 12900 points

Based on my assembler knowledge, the loop in __data16_memzero

clears the data (writes 0) to the address given by R12,
increments R12 and
compares it to the value in R15. If R12!=R15, it repeats the process

When R12==R15, the function should return - does it not? How does the R12 and R15 registers behave, does R12 somehow never reach the value of R15 (0x4219)?

If the function does not return as you'd expect, how does it behave if you reduce the packet buffer sizes?

Br,
TIABO

0 Simon Huang1 over 9 years ago in reply to TIABO

Prodigy 180 points

Dear TIABO,

It seems to be caused by the watch dog timer, please refer to the following discussion:

http://www.embeddedrelated.com/groups/msp430/show/38619.php

So far, it works to me. FYI.

Anyway, thank you very much.

Best regards,

Simon

0 TIABO over 9 years ago in reply to Simon Huang1

TI__Genius 12900 points

Hi Simon,

I'm glad to hear you solved it.

You can perhaps use the __low_level_init() function to disable WDT before initialization. Please refer to the IAR C/C++ Compiler reference manual, http://www.iar.com/Products/IAR-Embedded-Workbench/TI-MSP430/User-guides/.

Br,
TIABO

0 Leo Hendrawan over 9 years ago in reply to TIABO

TI__Guru 56650 points

Hi all,

please also refer here:

http://processors.wiki.ti.com/index.php/MSP430_FAQ#WDT_fires_during_C_startup_code

0 Bob Jiang over 9 years ago

Prodigy 190 points

Hi

I am using the MSP430F5438+cc1200 to implement the simple_peer_to_peer. I have debug the codes into the two device. But the two devices can not link to each other.

The SMPL_Link returns SMPL_TIMEOUT.

If possible, can you share the code related with us?

Sub-1 GHz

Sub-1 GHz forum

The SimpliciTI can't be executed after increasing MAX_APP_PAYLOAD and MAX_NWK_PAYLOAD