SLIP using SPI ?

Using UART, both ends are some kind of masters in transmit direction, and slaves in receive direction. With SPI is a different story. One end (master) initiates both transmit and receive, the other end (slave) sends and receives data as instructed by the master.

SLIP at low end sends/receives a byte asynchronously. UART can do that by its nature, but SPI not. So you have to emulate this mode in software.

The question is however, what is a benefit of doing this, since a CPU has much more work to do for each byte it transmits or receives. (byte stuffing, checksum calculation and validation, etc.) Ethernet MAC controller performs all those actions in hardware. So the speed gain using SPI with SLIP might not be so obvious.

Using a modified SPI bus, a continuous SPI also known as PCM bus, would be a better idea.

Using UART, both ends are some kind of masters in transmit direction, and slaves in receive direction. With SPI is a different story. One end (master) initiates both transmit and receive, the other end (slave) sends and receives data as instructed by the master.

SLIP at low end sends/receives a byte asynchronously. UART can do that by its nature, but SPI not. So you have to emulate this mode in software.

The question is however, what is a benefit of doing this, since a CPU has much more work to do for each byte it transmits or receives. (byte stuffing, checksum calculation and validation, etc.) Ethernet MAC controller performs all those actions in hardware. So the speed gain using SPI with SLIP might not be so obvious.

Using a modified SPI bus, a continuous SPI also known as PCM bus, would be a better idea.