Hi all:
I develop on dm8168 use dvrrdk 4.0.
I have some questions about algorithm optimization on c674x of dm8168.
1. how to find out algorithm bottlenecks quickly, it is useful to help us to find which parts of our algorithm should be optimized.
2. we use some assemble from dm642 assemble imglib and some assemble fun we write before on dm642 platform.
and now our platform is dm8168,with dsp c674x, do the dm642 assemble code can be run effectively?
3.Can VLIB use to process 1080P video on realtime?
4.How to init Cache mapps to the external mem, as internal mem in c674x is only 256k, only 64k used for cache in configure_2g.bld.
5.in dvrrdk DMA_Fast2D2D has the same efficiency with DMA_Fast2D1D？
6.How to enable assemble code debug step by step in CCS？
best regards
xavier
