netdev - Re: [RFC bpf-next v1 3/7] bpf: Support pulling non-linear xdp data

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAMB2axP59H5h5pLw4x-uu-CyJkds=BvzD6pa51PtAD2QRNhSfQ@mail.gmail.com>
Date: Tue, 26 Aug 2025 06:44:12 -0700
From: Amery Hung <ameryhung@...il.com>
To: Jakub Kicinski <kuba@...nel.org>
Cc: bpf@...r.kernel.org, netdev@...r.kernel.org, alexei.starovoitov@...il.com, 
	andrii@...nel.org, daniel@...earbox.net, martin.lau@...nel.org, 
	mohsin.bashr@...il.com, saeedm@...dia.com, tariqt@...dia.com, 
	mbloch@...dia.com, maciej.fijalkowski@...el.com, kernel-team@...a.com
Subject: Re: [RFC bpf-next v1 3/7] bpf: Support pulling non-linear xdp data

On Tue, Aug 26, 2025 at 6:20 AM Jakub Kicinski <kuba@...nel.org> wrote:
>
> On Mon, 25 Aug 2025 22:12:21 -0700 Amery Hung wrote:
> > > > +     data_end = xdp->data + len;
> > > > +     delta = data_end - xdp->data_end;
> > > > +
> > > > +     if (delta <= 0)
> > > > +             return 0;
> > > > +
> > > > +     if (unlikely(data_end > data_hard_end))
> > > > +             return -EINVAL;
> > >
> > > Is this safe against pointers wrapping on 32b systems?
> > >
> >
> > You are right. This may be a problem.
> >
> > > Maybe it's better to do:
> > >
> > >          if (unlikely(data_hard_end - xdp->data_end < delta))
> > >
> > > ?
> >
> > But delta may be negative if the pointer wraps around and then the
> > function will still continue. How about adding data_end < xdp->data
> > check and reorganizing the checks like this?
>
> You already checked that delta is positive in the previous if (),
> so I think it's safe. Admittedly having 3 separate conditions is
> more readable but it's not strictly necessary. Up to you.

Got it. I will change to the new set of checks. The original kfunc
would return 0 when pointer wrapping happens and delta <= 0.