From TRM Table 3-9. Global Shared RAM, we got the concept about that CPU1 fetch the program code at this time, CPU2 can’t fetch it.
But, from 220.127.116.11 Access Arbitration and Figure 3-17. Arbitration Scheme on Global Shared Memories, we saw the access arbitration, CPU1 and CPU1 can ran the program fetch by round robin.
If customer hope to get more resource when program fetch, could F2838xD run the program fetch on both cores no matter CPU1 or CPU2? (so that, we can just copy one program code in GSx rather than 2 copies, one for CPU1, another for CPU2).
If CPU1 and CPU2 can share the same GSx to program fetch (not only read or write), how could we set up it?
If you have any suggestions, please feel free to let us know.
Thanks a lot.