Wed, Feb 26, 2025 at 04:09:30AM +0100, jiashengjiangcool@xxxxxxxxx wrote:
>Since the driver is broken in the case that src->freq_supported is not
>NULL but src->freq_supported_num is 0, add an assertion for it.
>
>Fixes: 830ead5fb0c5 ("dpll: fix pin dump crash for rebound module")
It's not a real bug in current kernel. I don't think it's worth "fixes"
line and -net tree. I think it should be just sent to -net-next.
>Cc: <stable@xxxxxxxxxxxxxxx> # v6.8+
>Signed-off-by: Jiasheng Jiang <jiashengjiangcool@xxxxxxxxx>
>---
>Changelog:
>
>v1 -> v2:
>
>1. Replace the check with an assertion.
>---
> drivers/dpll/dpll_core.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
>diff --git a/drivers/dpll/dpll_core.c b/drivers/dpll/dpll_core.c
>index 32019dc33cca..3296776c1ebb 100644
>--- a/drivers/dpll/dpll_core.c
>+++ b/drivers/dpll/dpll_core.c
>@@ -443,8 +443,9 @@ static void dpll_pin_prop_free(struct dpll_pin_properties *prop)
> static int dpll_pin_prop_dup(const struct dpll_pin_properties *src,
> struct dpll_pin_properties *dst)
> {
>+ BUG_ON(src->freq_supported && !src->freq_supported_num);
Warnon-return please.
> memcpy(dst, src, sizeof(*dst));
>- if (src->freq_supported && src->freq_supported_num) {
>+ if (src->freq_supported) {
> size_t freq_size = src->freq_supported_num *
> sizeof(*src->freq_supported);
> dst->freq_supported = kmemdup(src->freq_supported,
>--
>2.25.1
>
