On 25/02/20 08:24PM, Pu Lehui wrote: > On 2025/2/20 13:50, Ethan Carter Edwards wrote: > > In most code paths variable move_kern_type_id remains uninitialized upon > > return. By moving it to the goto, it is initialized in these code paths. > > As well as others. Caught by Coverity. > > > > Closes: https://scan5.scan.coverity.com/#/project-view/63874/10063?selectedIssue=1595567 > > Fixes: e2b3c4ff5d183d ("bpf: add __arg_trusted global func arg tag") > > Signed-off-by: Ethan Carter Edwards <ethan@xxxxxxxxxxxxxxxxx> > > --- > > kernel/bpf/btf.c | 2 +- > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c > > index 9de6acddd479b4f5e32a5e6ba43cf369de4cee29..8c82ced7da299ad1ad769024fe097898c269013b 100644 > > --- a/kernel/bpf/btf.c > > +++ b/kernel/bpf/btf.c > > @@ -7496,9 +7496,9 @@ static int btf_get_ptr_to_btf_id(struct bpf_verifier_log *log, int arg_idx, > > err = -EOPNOTSUPP; > > goto cand_cache_unlock; > > } > > - kern_type_id = cc->cands[0].id; > > cand_cache_unlock: > > + kern_type_id = cc->cands[0].id; > > Hi, for goto's branches, it will always `return err`, no need to make this > move. You are right. My apologies. I should probably do less coding at 2AM. Thanks, Ethan