wait_for_completion_timeout returns unsigned long not int so the check for
<= 0 should be == 0 here, and the type unsigned long. The function return
is set to -ETIMEDOUT to reflect the actual problem.
Signed-off-by: Nicholas Mc Guire <hof...@osadl.org>
---
the only call site of pic32_sqi_one_messag() is in
drivers/spi/spi.c:__spi_pump_messages() (via struct spi_master
master->transfer_one_message) which is checking for non-zero
return only so technically -EIO is fine, for consistency this
should probably be -ETIMEDOUT though.
Compile tested with: pic32mzda_defconfig + CONFIG_SPI=y,
CONFIG_SPI_PIC32_SQI=m
Patch is against 4.7.0-rc7 (localversion-next -next-20160719)
drivers/spi/spi-pic32-sqi.c | 7 ++++---
1 file changed, 4 insertions(+), 3 deletions(-)
diff --git a/drivers/spi/spi-pic32-sqi.c b/drivers/spi/spi-pic32-sqi.c
index ca3c8d9..c41abdd 100644
--- a/drivers/spi/spi-pic32-sqi.c
+++ b/drivers/spi/spi-pic32-sqi.c
@@ -354,6 +354,7 @@ static int pic32_sqi_one_message(struct spi_master *master,
struct spi_transfer *xfer;
struct pic32_sqi *sqi;
int ret = 0, mode;
+ unsigned long timeout;
u32 val;
sqi = spi_master_get_devdata(master);
@@ -419,10 +420,10 @@ static int pic32_sqi_one_message(struct spi_master
*master,
writel(val, sqi->regs + PESQI_BD_CTRL_REG);
/* wait for xfer completion */
- ret = wait_for_completion_timeout(&sqi->xfer_done, 5 * HZ);
- if (ret <= 0) {
+ timeout = wait_for_completion_timeout(&sqi->xfer_done, 5 * HZ);
+ if (timeout == 0) {
dev_err(&sqi->master->dev, "wait timedout/interrupted\n");
- ret = -EIO;
+ ret = -ETIMEDOUT;
msg->status = ret;
} else {
/* success */
--
2.1.4
