On Mon, Mar 03, 2025 at 10:47:40AM +0000, Conor Dooley wrote:
When the size of a transfer exceeds the size of the FIFO (32 bytes), RX overflows will be generated and receive data will be corrupted and warnings will be produced. For example, here's an error generated by a transfer of 36 bytes:
spi_master spi0: mchp_corespi_interrupt: RX OVERFLOW: rxlen: 4, txlen: 0
The driver is currently split between handling receiving in the interrupt handler, and sending outside of it. Move all handling out of the interrupt handling, and explicitly link the number of bytes read of of the RX FIFO to the number written into the TX one. This both resolves the overflow problems as well as simplifying the flow of the driver.
...
+static inline void mchp_corespi_read_fifo(struct mchp_corespi *spi, int fifo_max) {
- for (int i = 0; i < fifo_max; i++) {
u32 data;
while (mchp_corespi_read(spi, REG_STATUS) & STATUS_RXFIFO_EMPTY);
What may go wrong with this code? :-) I think limiting the infinite loop by a timeout is a good thing to have.
data = mchp_corespi_read(spi, REG_RX_DATA);spi->rx_len -= spi->n_bytes;
mchp_corespi_write(spi, REG_FRAMESUP, len); }