tree 48b3f41e2182984b3d18000286b020d1732c142c
parent 164c05f03ffabe36564cb8959c1fad7c56f294dd
author Angelo Dureghello <angelo.dureghello@timesys.com> 1590782276 +0200
committer Mark Brown <broonie@kernel.org> 1590792195 +0100
gpgsig -----BEGIN PGP SIGNATURE-----
 
 iQFHBAABCgAxFiEEreZoqmdXGLWf4p/qJNaLcl1Uh9AFAl7RkAMTHGJyb29uaWVA
 a2VybmVsLm9yZwAKCRAk1otyXVSH0O2fB/4zfppPEvLEPH8EfbIzxMJQHap6V2vJ
 ExP2QFKrimjqGET8ZKQGcHJWFUoVJsf0UjS1/uddfyw3LiinsXTDsZFeh4nWXC2F
 gO826eikXGxgmh/HT8W64RBu7ZYTTaBkCXgrwjjNxv458ghsSArLsfpoccImdGUJ
 Xfe8M/D0EmtzzuVReXNtPVJrXptVpI50tgDMvjCOEL6ARP6zm1L+zAsWtwoNQBoK
 +4owW0WVpm5D153ThcBTAHG5tbbfGdLrIqS+pV4i6eKlAd8VfBUZ760fxvNmkY85
 o6ho9m/2uu3QJvFWxsP2PgBxdRlRvUUg91DUbqGVrO2s0g4RCLOt61Bt
 =gsJl
 -----END PGP SIGNATURE-----

spi: spi-fsl-dspi: fix native data copy

ColdFire is a big-endian cpu with a big-endian dspi hw module,
so, it uses native access, but memcpy breaks the endianness.

So, if i understand properly, by native copy we would mean
be(cpu)->be(dspi) or le(cpu)->le(dspi) accesses, so my fix
shouldn't break anything, but i couldn't test it on LS family,
so every test is really appreciated.

Fixes: 53fadb4d90c7 ("spi: spi-fsl-dspi: Simplify bytes_per_word gymnastics")
Signed-off-by: Angelo Dureghello <angelo.dureghello@timesys.com>
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://lore.kernel.org/r/20200529195756.184677-1-angelo.dureghello@timesys.com
Signed-off-by: Mark Brown <broonie@kernel.org>