diff options
| author | Stanislaw Gruszka <stf_xl@wp.pl> | 2009-01-15 13:50:44 -0800 |
|---|---|---|
| committer | Linus Torvalds <torvalds@linux-foundation.org> | 2009-01-15 16:39:35 -0800 |
| commit | 5b96f1729064453e09805a387378e0644da1c937 (patch) | |
| tree | e2ef80291f2ed0176e9d8330c0b698aae790ecd4 /drivers/spi | |
| parent | 1f14081d8b31b2780e2e6e2ae2766264e1f7d38b (diff) | |
| download | kernel-mfld-blackbay-5b96f1729064453e09805a387378e0644da1c937.tar.gz kernel-mfld-blackbay-5b96f1729064453e09805a387378e0644da1c937.tar.bz2 kernel-mfld-blackbay-5b96f1729064453e09805a387378e0644da1c937.zip | |
atmel_spi: allow transfer when max_speed_hz = 0
For some reason I have to slowdown clock to touchscreen device.
In atmel_spi_setup() there is comment that max_speed_hz == 0 means as slow
as possible and divider is set to maximum value. But in
atmel_spi_transfer() function is check against not zero max_speed_hz with
EINVAL returned.
Probably driver should setup divider for each transfer based on
transfer->speed_hz value, but I think that would be not necessary overhead
as all used devices have constant clock.
Below patch works fine for me.
Signed-off-by: Stanislaw Gruszka <stf_xl@wp.pl>
Signed-off-by: Haavard Skinnemoen <haavard.skinnemoen@atmel.com>
Cc: David Brownell <david-b@pacbell.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'drivers/spi')
| -rw-r--r-- | drivers/spi/atmel_spi.c | 3 |
1 files changed, 1 insertions, 2 deletions
diff --git a/drivers/spi/atmel_spi.c b/drivers/spi/atmel_spi.c index 5e39bac9c51..56ff3e6864e 100644 --- a/drivers/spi/atmel_spi.c +++ b/drivers/spi/atmel_spi.c @@ -670,8 +670,7 @@ static int atmel_spi_transfer(struct spi_device *spi, struct spi_message *msg) dev_dbg(controller, "new message %p submitted for %s\n", msg, spi->dev.bus_id); - if (unlikely(list_empty(&msg->transfers) - || !spi->max_speed_hz)) + if (unlikely(list_empty(&msg->transfers))) return -EINVAL; if (as->stopping) |