I understand what is going on. I cannot tell you how to fix it.
Use armofd to get some information about the sections in the file ...
% armofd --obj_display=none,sections PRJ_TMS570LS0432_ECC.out | find " Y " 1 .intvecs 0x00020000 0x00020000 0x20 4 Y 2 .text 0x00020020 0x00020020 0x41d8 4 Y 3 .const 0x000241f8 0x000241f8 0x624 4 Y 4 .cinit 0x00024820 0x00024820 0x28 8 Y 7 .data 0x08001500 0x08001500 0x19 4 Y 21 .ecc0 0xf0404000 0xf0404000 0x909 0 Y
That --obj_display option disables the display of everything except sections. Piping it through that find command limits the output to sections which have Y in the last column. Those sections section contain initialized code or data.
The important thing to notice is the memory range of these sections. It goes from 0x20000 to 0xf0404909. This binary file format has no mechanism for representing holes in memory. All of the memory holes are filled with 0. In this case, there are two large holes to consider. One is between .cinit and .data. The other is between .data and .ecc0. Total binary file size comes to 0xf03e4909 bytes. That is about 3.75 gigabytes of mostly 0. I didn't bother to let it finish either. But, if you waited it out, I think it would eventually finish.
Clearly, this is impractical. But I don't know enough about the intended use of nowECC to tell you what to do next. I'll try to get further attention to this thread.
Thanks and regards,
-George