[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <bef24789-819c-4a7b-bbb0-f38ffe9f67f0@gmail.com>
Date: Fri, 20 Oct 2023 10:53:52 -0700
From: Kui-Feng Lee <sinquersw@...il.com>
To: Martin KaFai Lau <martin.lau@...ux.dev>, thinker.li@...il.com
Cc: kuifeng@...a.com, netdev@...r.kernel.org, bpf@...r.kernel.org,
ast@...nel.org, song@...nel.org, kernel-team@...a.com, andrii@...nel.org,
drosen@...gle.com
Subject: Re: [PATCH bpf-next v5 6/9] bpf, net: switch to dynamic registration
On 10/20/23 08:12, Kui-Feng Lee wrote:
>
>
> On 10/18/23 18:49, Martin KaFai Lau wrote:
>> On 10/17/23 9:23 AM, thinker.li@...il.com wrote:
>>> From: Kui-Feng Lee <thinker.li@...il.com>
>>> static const struct bpf_struct_ops *
>>> bpf_struct_ops_find_value(struct btf *btf, u32 value_id)
>>> {
>>> + const struct bpf_struct_ops *st_ops = NULL;
>>> + const struct bpf_struct_ops **st_ops_list;
>>> unsigned int i;
>>> + u32 cnt = 0;
>>> if (!value_id || !btf_vmlinux)
>>
>> The "!btf_vmlinux" should have been changed to "!btf" in the earlier
>> patch (patch 2?),
>
> This is not btf. It mean to check if btf_vmlinux is initialized.
> It is not necessary anymore.
> For checking btf, the following btf_get_struct_ops() will keep cnt zero
> if btf is NULL, so it is unnecessary as well.
Forget my previous comment. I think you are right!
Powered by blists - more mailing lists