On Sat, Dec 18, 2021 at 8:33 PM Kumar Kartikeya Dwivedi
[off-list ref] wrote:

It is, but into parent_ref_obj_id, to match during release_reference.

quoted

Shouldn't r2 get a different ref_obj_id after r2 = r1->next ?

It's ref_obj_id is still 0.

Thinking about this more, we actually only need 1 extra bit of information in
reg_state, not even a new member. We can simply copy ref_obj_id and set this
bit, then we can reject this register during release but consider it during
release_reference.

It seems to me that this patch created the problem and it's trying
to fix it at the same time.

mark_btf_ld_reg() shouldn't be copying ref_obj_id.
If it keeps it as zero the problem will not happen, no?