I have build up the application and the video is displayed correctly on the PC.
My configuration is 8 bit video data (grayscale) using an 8 bit data bus at the GPIF-II interface. The coding is YUV2, which means that for every pixel sent, I transfer with the first 8-bit a video byte (0xNN) which is followed by a second transfer with the value 0x80. This works well !
Now I want to improve the performance and want to configure the GPIF-II interface to use 16 bit for the transfer. Every pixel is transferred as a 16 bit word with 0x80 in the upper byte and the video data (0xNN) in the lower byte. This should result in a data sequence according the Intel little endian mode.
I have only changed the data bus width in GPIF-II to 16 bit. My expectation was that now the DMA buffer will filled up with a sequence of data like: 0xNN, 0x80, 0xNN, 0x80, ... The amount of transferred data is the same therefore I did not change any there parameter.
The screen shows only black and there is no error message in the debug console. I have also tried to switch the byte order but this did not help.
It seems for me that something is missing in the configuration or the supplied software.
Q: What do I have to change, that the 16 bit version will work like the 8 bit version before?