Re: [PATCH 1/1] spi: altera: Add empty implementation of setup_transfer callback

[Pelle Nilsson]

On 2015-04-09 17:40, Mark Brown wrote:
> Why is the callback mandatory if an empty implementation is OK?

Ask the author of spi-bitbang. :-)

In spi_bitbang_start() we have this chunk of code:

	if (!bitbang->txrx_bufs) {
		bitbang->use_dma = 0;
		bitbang->txrx_bufs = spi_bitbang_bufs;
		if (!master->setup) {
			if (!bitbang->setup_transfer)
				bitbang->setup_transfer =
					 spi_bitbang_setup_transfer;
			master->setup = spi_bitbang_setup;
			master->cleanup = spi_bitbang_cleanup;
		}
	}

As can be seen here, if setup_transfer is NULL (not set by the
specific driver), it is filled in with the default callback function
spi_bitbang_setup_transfer(), but only if txrx_bufs is also NULL,
which is not the case here.

There is a comment in spi-xilinx also stating this fact (though their
implementation isn't actually empty anymore):

/* spi_bitbang requires custom setup_transfer() to be defined if there
is a
 * custom txrx_bufs().
 */
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/