AnsweredAssumed Answered

Suddenly CCES v2.8.0/ICE-1000 v1.2 detect a core hang on SHARC ADSP-21479?

Question asked by jb@hme on Jun 13, 2018
Latest reply on Jun 29, 2018 by Reva@14

All of a sudden, CrossCore Embedded Studio v2.8.0 through ICE-1000 v1.2 started detecting that my SHARC ADSP-21479 core was hung almost every time I run my app, always in the same place, where it never complained about this before. A window pops-up saying:

   Core access to a SPORT/LINK Port is hung at PC:0x127c09

   CLEAR will abort the hung operation.
   ACKNOWLEDGE will allow the user to remedy the hung operation.
   RESET will perform a software reset on the target processor.

 

and another then pops-up that says:

   An internal error occurred during: "Launching mySHARC".
   [TpsdkServer] Failed to reset processor 0

 

and the CCES Console windows says: "Command control shut down."

 

This TWI (not SPORT/Link Port) code at PC:0x127c09, corresponding to this 'C' app statement:

 

   *pTWIIMASK = TWITXINT | TWIMCOM;      // enable master transfer complete and TX FIFO interrupts


has been running for several weeks without generating this error.  Only after that statement does the app do:

 

   *pTWIMCTL =  TWIMEN;

 

Why does it happen all of a sudden?  How can writing to a TWI peripheral register (or is it the SPORT ISR?), hang the SHARC core, even before enabling it in Master mode to initiate transmission?  What can I do about this? I can't debug much of anything anymore, even with the latest CCES v2.8.0.201803091800 (w1804121) and ICE-1000 v1.2.

Outcomes