-----邮件原件----- 发件人: Toke Høiland-Jørgensen [mailto:toke@xxxxxxxxxx] 发送时间: 2022年5月4日 19:20 收件人: shaozhengchao <shaozhengchao@xxxxxxxxxx>; netdev@xxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx; bpf@xxxxxxxxxxxxxxx; davem@xxxxxxxxxxxxx; edumazet@xxxxxxxxxx; kuba@xxxxxxxxxx; pabeni@xxxxxxxxxx 抄送: ast@xxxxxxxxxx; daniel@xxxxxxxxxxxxx; hawk@xxxxxxxxxx; john.fastabend@xxxxxxxxx; andrii@xxxxxxxxxx; kafai@xxxxxx; songliubraving@xxxxxx; yhs@xxxxxx; kpsingh@xxxxxxxxxx; bigeasy@xxxxxxxxxxxxx; imagedong@xxxxxxxxxxx; petrm@xxxxxxxxxx; memxor@xxxxxxxxx; arnd@xxxxxxxx; weiyongjun (A) <weiyongjun1@xxxxxxxxxx>; shaozhengchao <shaozhengchao@xxxxxxxxxx>; yuehaibing <yuehaibing@xxxxxxxxxx> 主题: Re: [PATCH bpf-next] bpf/xdp: Can't detach BPF XDP prog if not exist Zhengchao Shao <shaozhengchao@xxxxxxxxxx> writes: > if user sets nonexistent xdp_flags to detach xdp prog, kernel should > return err and tell user that detach failed with detail info. > > Signed-off-by: Zhengchao Shao <shaozhengchao@xxxxxxxxxx> I kinda see your point, but this will change user-visible behaviour that applications might be relying on, so I don't think we can make this change at this stage. Why can't your application just query the link for whether a program is attached? -Toke Thank you for your reply. I wiil change sample application firstly. But if kernel does nothing and return 0, maybe user will think setup is OK, actually It failed. Is this acceptable?