msm: wigig_sensing: use 32 bit transactions for SPI block read

32 bit SPI transactions are more efficient. Additionally, the data is
received in the correct endianness, as expected by the host, therefore an
endianness flip can be avoided.

Change-Id: I35c0be291c971d5f07a00e79923c2050914d9642
Signed-off-by: Gidon Studinski <gidons@codeaurora.org>
Signed-off-by: Alexei Avshalom Lazar <ailizaro@codeaurora.org>
This commit is contained in:
Gidon Studinski 2019-11-06 15:11:52 +02:00 • committed by Alexei Avshalom Lazar
commit 965ece3b02

View file

@ -244,13 +244,12 @@ static int spis_block_read_mem(struct spi_device *spi,
.tx_buf = ctx->tx_buf,
.rx_buf = ctx->rx_buf,
.len = sz,
.bits_per_word = 8,
/*.bits_per_word = 32,*/
.bits_per_word = 32,
};
u32 frame = 0xB << 24;
frame |= address & 0xFFFFFF;
frame = cpu_to_be32(frame);
frame = cpu_to_le32(frame);
/* Read length must be in 32 bit units */
if (length & 3) {