[PATCH] Fix regression on spi-fsl-spi.c following modification on spi.c

[Christophe Leroy]

spi-fsl-spi.c made the assumption that t->bits_per_word and t->speed_hz were
set only when wanting to change the setting.
Since commit e6811d1d7a6a38ee637fe219c3b67dbfe17e8b3f, t->bits_per_word has
been set in all transfers, therefore the assumption made by the driver leads to
failure. This patch fixes the issue by comparing the t->new bits_per_word and
t->speed_hz with the previous one in order to determine if one of them changed.

Signed-off-by: Christophe Leroy <christophe.leroy@xxxxxx>

--- linux-3.8.13/drivers/spi/spi-fsl-spi.c	2013-05-11 22:57:46.000000000 +0200
+++ linux/drivers/spi/spi-fsl-spi.c	2013-06-09 00:13:30.000000000 +0200
@@ -476,11 +476,14 @@
 	unsigned int cs_change;
 	const int nsecs = 50;
 	int status;
+	u32 prev_speed_hz = 0;
+	u8 prev_bits_per_word = 0;
 
 	cs_change = 1;
 	status = 0;
 	list_for_each_entry(t, &m->transfers, transfer_list) {
-		if (t->bits_per_word || t->speed_hz) {
+		if (prev_bits_per_word != t->bits_per_word
+				|| prev_speed_hz != t->speed_hz) {
 			/* Don't allow changes if CS is active */
 			status = -EINVAL;
 
@@ -488,6 +491,8 @@
 				status = fsl_spi_setup_transfer(spi, t);
 			if (status < 0)
 				break;
+			prev_bits_per_word = t->bits_per_word;
+			prev_speed_hz = t->speed_hz;
 		}
 
 		if (cs_change) {
--
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/