On 23/07/2025 18:03, Quentin Schulz wrote:
Hi Andrew,
On 7/23/25 5:17 PM, Andrew Goodbody wrote:
The function dev_read_u32_default does not return an error and the
variable 'val' is unsigned so testing for >= 0 will always be true so
remove this test.
This issue was found by Smatch.
Signed-off-by: Andrew Goodbody <andrew.goodb...@linaro.org>
---
drivers/clk/clk-cdce9xx.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/drivers/clk/clk-cdce9xx.c b/drivers/clk/clk-cdce9xx.c
index e5f74e714d5..492a196604a 100644
--- a/drivers/clk/clk-cdce9xx.c
+++ b/drivers/clk/clk-cdce9xx.c
@@ -117,8 +117,7 @@ static int cdce9xx_clk_probe(struct udevice *dev)
data->xtal_rate = clk_get_rate(&clk);
val = dev_read_u32_default(dev, "xtal-load-pf", -1);
- if (val >= 0)
- cdce9xx_reg_write(dev, CDCE9XX_REG_XCSEL, val << 3);
+ cdce9xx_reg_write(dev, CDCE9XX_REG_XCSEL, val << 3);
I think the original code was broken?
I assume the intended behavior was to return -1 if the property isn't
present. Obviously, dev_read_u32_default having a -1 as default won't
return a -1...
So I think we should migrate to dev_read_u32 instead and check the
return code and do the write based on that?
Something like:
if (!dev_read_u32(dev, "xtal-load-pf", &val))
cdce9xx_reg_write(dev, CDCE9XX_REG_XCSEL, val << 3);
maybe? (Not tested nor compiled)
Cheers,
Quentin
Fine by me. V2 coming.
Andrew
