Re: [PATCH net-next v3 1/9] xsk: introduce XDP_GENERIC_XMIT_BATCH setsockopt

[Simon Horman <horms@kernel.org>]

On Sat, Oct 25, 2025 at 05:08:39PM +0800, Jason Xing wrote:
> Hi Simon,
> 
> On Fri, Oct 24, 2025 at 9:30 PM Simon Horman <horms@kernel.org> wrote:
> >
> > On Tue, Oct 21, 2025 at 09:12:01PM +0800, Jason Xing wrote:
> >
> > ...
> >
> > > index 7b0c68a70888..ace91800c447 100644
> >
> > ...
> >
> > > @@ -1544,6 +1546,55 @@ static int xsk_setsockopt(struct socket *sock, int level, int optname,
> > >               WRITE_ONCE(xs->max_tx_budget, budget);
> > >               return 0;
> > >       }
> > > +     case XDP_GENERIC_XMIT_BATCH:
> > > +     {
> > > +             struct xsk_buff_pool *pool = xs->pool;
> > > +             struct xsk_batch *batch = &xs->batch;
> > > +             struct xdp_desc *descs;
> > > +             struct sk_buff **skbs;
> > > +             unsigned int size;
> > > +             int ret = 0;
> > > +
> > > +             if (optlen != sizeof(size))
> > > +                     return -EINVAL;
> > > +             if (copy_from_sockptr(&size, optval, sizeof(size)))
> > > +                     return -EFAULT;
> > > +             if (size == batch->generic_xmit_batch)
> > > +                     return 0;
> > > +             if (size > xs->max_tx_budget || !pool)
> > > +                     return -EACCES;
> > > +
> > > +             mutex_lock(&xs->mutex);
> > > +             if (!size) {
> > > +                     kfree(batch->skb_cache);
> > > +                     kvfree(batch->desc_cache);
> > > +                     batch->generic_xmit_batch = 0;
> > > +                     goto out;
> > > +             }
> > > +
> > > +             skbs = kmalloc(size * sizeof(struct sk_buff *), GFP_KERNEL);
> > > +             if (!skbs) {
> > > +                     ret = -ENOMEM;
> > > +                     goto out;
> > > +             }
> > > +             descs = kvcalloc(size, sizeof(struct xdp_desc), GFP_KERNEL);
> > > +             if (!descs) {
> > > +                     kfree(skbs);
> > > +                     ret = -ENOMEM;
> > > +                     goto out;
> > > +             }
> > > +             if (batch->skb_cache)
> > > +                     kfree(batch->skb_cache);
> > > +             if (batch->desc_cache)
> > > +                     kvfree(batch->desc_cache);
> >
> > Hi Jason,
> >
> > nit: kfree and kvfree are no-ops when passed NULL,
> >      so the conditions above seem unnecessary.
> 
> Yep, but the checkpatch complains. I thought it might be good to keep
> it because normally we need to check the validation of the pointer
> first and then free it. WDYT?

I don't feel particularly strongly about this.
But I would lean to wards removing the if() conditions
because they are unnecessary: less is more.