Re: [PATCH] bpf: Replace BUG_ON when fp_old is NULL

[Daniel Borkmann]

On 12/19/19 6:39 PM, Aditya Pakki wrote:
> On 12/16/19 5:17 AM, Daniel Borkmann wrote:
> > On 12/15/19 11:08 PM, Yonghong Song wrote:
> > > On 12/15/19 7:44 AM, Aditya Pakki wrote:
> > > > If fp_old is NULL in bpf_prog_realloc, the program does an assertion
> > > > and crashes. However, we can continue execution by returning NULL to
> > > > the upper callers. The patch fixes this issue.
> > >
> > > Could you share how to reproduce the assertion and crash? I would
> > > like to understand the problem first before making changes in the code.
> > > Thanks!
> >
> > Fully agree, Aditya, please elaborate if you have seen a crash!
>
> Thanks for your responses Alexei and Daniel. We identified this issue via static analysis
> and have not seen a crash. However, by looking at the callers of bpf_prog_realloc, I do
> agree that fp_old is never NULL.
>
> Would you recommend removing the BUG_ON assertion altogether ?

If it would ever happen, we'd already crash in fp_old->pages there, so yes,
lets remove the unneeded BUG_ON().

Thanks,
Daniel