TMS320F28379D: Instructions of accelerator

1. Regarding to the use of FPU, if I use fpu32 for float_support of Compiler options, can FPU be implemented automatically?

---Yes, that will lead the compiler to generate FPU32 native instructions.

2. For FPU configuration, float support = 64bit, I can find ABI be set to EABI in TMS320C28x Optimizing C/C++ Compiler v20.2.0.LTS」.
is it possible to perform 64 bit floating point arithmetic in 28379D 32-bit FPU.

---28379D does not contain an FPU64 hardware module. 28388D does. So on 28379D you will not generate native 64-bit FPU HW instructions.

3. I also have the same queries on TMU, VCU.
4. Regarding to the use of TMU,
If I set the compiler option to "-- TMU support = tmu0", will TMU automatically run into implementation status?
Or have to write with TMU in existing C source code?

---You will need to set TMU to tmu0 and also, under optimization, set --fp_mode to "relaxed"

5.Regarding to the use of VCU, Compiler options for VCU support, what is the difference between "0" and "2"?
6.Regarding to the use of VCU, if I set the compiler option to "-- VCU support = vcu0", can VCU be implemented automatically?

Or have to write with VCU in existing C source code?

---Correct, this will be sufficient.

7. If I set 'on' for all support of FPU, VCU, TMU, How to calculate an existing C source code will be assigned to each accelerators?

---To support each of the cases, C2000Ware has specific libraries containing assembly optimized C-callable functions, and C examples illustrating how they can be used. So you can use them in your project. Alternatively, you can use intrinsics directly in your code.

Thanks,

Sira