On 2024/1/4 04:02, Andi Shyti wrote:
Hi Hans,
[...]
--- a/drivers/i2c/busses/i2c-viai2c-common.c
+++ b/drivers/i2c/busses/i2c-viai2c-common.c
@@ -5,7 +5,7 @@
#define VIAI2C_TIMEOUT (msecs_to_jiffies(1000))
#define VIAI2C_STRETCHING_TIMEOUT 200
-static int viai2c_wait_bus_ready(struct viai2c *i2c)
+int viai2c_wait_bus_ready(struct viai2c *i2c)
no, please, choose the rigth prototype at the beginning, don't
double change things in the same series.
ok, I will review my patch more carefully.
{
unsigned long timeout;
void __iomem *base = i2c->base;
@@ -23,7 +23,7 @@ static int viai2c_wait_bus_ready(struct viai2c *i2c)
return 0;
}
-static int viai2c_wait_status(struct viai2c *i2c)
+int viai2c_wait_status(struct viai2c *i2c)
{
int ret = 0;
unsigned long time_left;
@@ -71,6 +71,9 @@ static irqreturn_t viai2c_isr(int irq, void *data)
/* save the status and write-clear it */
i2c->cmd_status = readw(i2c->base + VIAI2C_REG_ISR);
+ if (!i2c->cmd_status)
+ return IRQ_NONE;
+
is this change related to this patch? can you put this at the
beginning?
Andi
The original wmt driver does not have this, maybe I should use
'if (i2c->platform == VIAI2C_PLAT_ZHAOXIN &&! i2c->cmd_status)' instead?
Hans
