On 2/9/24 19:13, John Seth Thielemann wrote:
> Building with a gcc nolibc / musl / linux setup found a null pointer
dereference in the pred_and handling.
>
> Take care,
> J. Seth Thielemann
Thanks for the patch, but ...
> diff -Naur /musl/opt/findutils-4.9.0/build/findutils-4.9.0.pristine/find/pred.c
/musl/opt/findutils-4.9.0/build>
/findutils-4.9.0/find/pred.c
> --- /musl/opt/findutils-4.9.0/build/findutils-4.9.0.pristine/find/pred.c
2022-01-02 23:34:22.000000000 +0000
> +++ /musl/opt/findutils-4.9.0/build/findutils-4.9.0/find/pred.c 2024-02-09
01:32:12.395992857 +0000
> @@ -150,7 +150,14 @@
> bool
> pred_and (const char *pathname, struct stat *stat_buf, struct predicate
*pred_ptr)
> {
> - if (pred_ptr->pred_left == NULL
> + if (pred_ptr->pred_left == NULL) {
> + if (pred_ptr->pred_right)
> + return apply_predicate(pathname, stat_buf,
pred_ptr->pred_right);
> + else
> + return false;
> + }
> +
> + if (pred_ptr->pred_left->pred_func
> || apply_predicate (pathname, stat_buf, pred_ptr->pred_left))
> {
> return apply_predicate (pathname, stat_buf, pred_ptr->pred_right);
... this breaks most of the tests.
Have a nice day,
Berny
