• State Resolved
  • Locked Locked
  • Replies 1 reply
  • Subscribers 82 subscribers
  • Views 97 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Original question:

TMS320F28375D: Why Was Only the Lower 16 Bytes Referenced After Defining Argument 1 as 32 Bytes in a Function?

TMS320F28379S: Regarding ALIGN Settings and Their Impact on Read and Program/ECC Operations

K.Z
Genius 3215 points

Part Number: TMS320F28379S

Hi All,

Please tell me about the ALIGN setting.

When the linker file is configured with ALIGN(8), I suspect that differences may occur between reading and programming.
As a result, I believe it may be more appropriate to use ALIGN(4) instead.

  • Read operation
    → Since the hardware bus width is 128 bits, using ALIGN(8) (128‑bit boundary) should not cause any issues.

  • Program / ECC operation
    → Because ECC is generated in 64‑bit units, enforcing ALIGN(8) may create “gaps” that do not align with the ECC calculation unit.

Previously, we observed a phenomenon where abnormal behavior was corrected simply by placing a NOP instruction.
I suspect the root cause may have been that the ALIGN(8) setting changed the padding size, causing the programmed addresses to shift unexpectedly.

Best Regards,

Ito

  • +1
    TI__Prodigy 470 points

    instead of using NOPS, `palign` could have been used. It does not only make sure that the starting address is aligned to the given number, but also the size of that section is aligned. But I think ALIGN(8) should be the only correct setting.