On 8/10/23 2:19 PM, Jakub Kicinski wrote:
On Wed, 9 Aug 2023 22:51:23 -0700 Martin KaFai Lau wrote:bpf: fix bpf_dynptr_slice() to stop return an ERR_PTR.This one looks like solid bpf material TBH, any reason it's here?
There is an earlier bpf_dynptr_check_off_len() call which should have caught that already, so ERR_PTR case should not happen.
https://lore.kernel.org/bpf/e0e8bf3b-70af-3827-2fa3-30f3d48bcf46@xxxxxxxxx/