Hi Tom, Cornor, Jagan, I would like to follow up on this patch.
The original patch introduced a Device Tree binding for bits-per-word, but based on the feedback from Tom and Conor, I reworked the implementation to follow the Linux model instead. Specifically, the updated implementation: Adds bits_per_word to struct spi_slave. Reads slave->bits_per_word in dw_spi_xfer() before each transfer. Validates the requested value against the controller capabilities. Defaults to 8-bit transfers when bits_per_word is not set, preserving backward compatibility. The updated patch has been submitted here: https://patchwork.ozlabs.org/project/uboot/patch/ [email protected]/ For reference, the original discussion and patch are here: https://patchwork.ozlabs.org/project/uboot/patch/ [email protected]/ When you have some time, could you please take a look at the new patch? I'd appreciate any feedback or comments, and if the approach looks good, I'd be grateful if it could be considered for merging. Thank you for your time and guidance throughout this discussion. Best regards, Boon Khai
