michaelh/zephyr
1
0
Fork 0
Code Releases Activity

mgmt: ec_host_cmd: stm32_spi: fix rx reload for STM32H7 chips

Browse source 
Disable and enable SPI module before every transaction for STM32H7
chips. It is a recommended way in the STM32 RM. In another case, a first
byte of a transaction is always 0x00.

Signed-off-by: Dawid Niedzwiecki <dawidn@google.com>
This commit is contained in:
Dawid Niedzwiecki 2024-11-22 14:00:00 +01:00 • committed by Benjamin Cabé
commit d7df7cdba6
1 changed files with 15 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

15
subsys/mgmt/ec_host_cmd/backends/ec_host_cmd_backend_spi_stm32.c
View file

@ -536,6 +536,20 @@ static int prepare_rx(struct ec_host_cmd_spi_ctx *hc_spi)
int ret;
hc_spi->prepare_rx_later = 0;
#ifdef EC_HOST_CMD_ST_STM32H7
/* As described in RM0433 "To restart the internal state machine
* properly, SPI is strongly suggested to be disabled and re-enabled
* before next transaction starts despite its setting is not changed.",
* disable and re-enable the SPI module. Without that, the SPI module
* receives the first byte on a next transaction incorrently - it is
* always 0x00.
* It also clears RX FIFO, so there is no needed to read the remianing
* bytes manually.
*/
LL_SPI_Disable(spi);
LL_SPI_Enable(spi);
#else /* EC_HOST_CMD_ST_STM32H7 */
/* Flush RX buffer. It clears the RXNE(RX not empty) flag not to trigger
* the DMA transfer at the beginning of a new SPI transfer. The flag is
* set while sending response to host. The number of bytes to read can
@ -543,6 +557,7 @@ static int prepare_rx(struct ec_host_cmd_spi_ctx *hc_spi)
* threshold.
*/
LL_SPI_ReceiveData8(spi);
#endif /* EC_HOST_CMD_ST_STM32H7 */
ret = reload_dma_rx(hc_spi);
if (!ret) {