Hello All,
Currently I am trying to implement Dynamic Caching of functions using Memory Overlays with tables on DM6446. This is how I have grouped the code and loaded it into the external memory.
UNION
{
GROUP
{
.CCODE3
.CCODE4
}load=DDR2, table(_table1)
GROUP
{
.CCODE5
.CCODE6
}load=DDR2, table(_table2)
} run=L2RAM
I load the function in the L2RAM using copy_in () function which is called just before the actual function. This works fine, but if I use L1P RAM as run location instead of L2RAM, the copy_in() doesn't work. I have set the L1 Cache to 0KB using BCACHE, so the L1P is programmable, I observed that if I used #pragma CODE_SECTION directive to load the code in L1P RAM it works, but "dynamically I can't copy a new code into L1P RAM". So I want to know, at runtime whether a user can change L1P RAM's content.
regards
Amit