Re: [PATCH bpf v2 1/2] bpf, sockmap: Account for receive queue in FIONREAD without a verdict program

["Emil Tsalapatis" <emil@etsalapatis.com>]

On Wed Jul 8, 2026 at 9:56 AM EDT, sashiko-bot wrote:
> Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
> - [Low] The function sk_psock_msg_inq() is now dead code and should be removed.
> --
>
> --- Patch [1]: [PATCH bpf v2 1/2] bpf, sockmap: Account for receive queue in FIONREAD without a verdict program ---
> commit c5a219566b06af42fd341fd941447a9a448fa4f0
> Author: Mattia Meleleo <mattia.meleleo@coralogix.com>
>
> bpf, sockmap: Account for receive queue in FIONREAD without a verdict program
>
> This commit fixes SIOCINQ reporting for tcp_bpf_ioctl() when no verdict
> program is attached. It accounts for bytes remaining in the sk_receive_queue
> by adding tcp_inq() to the reported value.
>
>> diff --git a/net/ipv4/tcp_bpf.c b/net/ipv4/tcp_bpf.c
>> index cc0bd73f36b6d..8e905b50deadb 100644
>> --- a/net/ipv4/tcp_bpf.c
>> +++ b/net/ipv4/tcp_bpf.c
>> @@ -344,7 +345,21 @@ static int tcp_bpf_ioctl(struct sock *sk, int cmd, int *karg)
>>  		return -EINVAL;
>>  
>>  	slow = lock_sock_fast(sk);
>> -	*karg = sk_psock_msg_inq(sk);
>
> [Severity: Low]
> Does replacing this call to sk_psock_msg_inq() leave the function as
> dead code?
>
> Looking at include/linux/skmsg.h, it appears this patch removes the only
> caller of sk_psock_msg_inq() in the tree. Should the function definition 
> be removed alongside this change to clean up the unused code?

This is a nice suggestion, if we can remove the function let's do so.

>
>> +	psock = sk_psock_get(sk);
>> +	if (unlikely(!psock)) {
> [ ... ]