Re: [PATCH v6 bpf-next 6/7] bpf: introduce bpf_prog_pack allocator

From: Song Liu
Date: Tue Jan 25 2022 - 20:29:10 EST


On Tue, Jan 25, 2022 at 5:20 PM Alexei Starovoitov
<alexei.starovoitov@xxxxxxxxx> wrote:
>
> On Tue, Jan 25, 2022 at 4:50 PM Song Liu <song@xxxxxxxxxx> wrote:
> >
> > On Tue, Jan 25, 2022 at 4:38 PM Alexei Starovoitov
> > <alexei.starovoitov@xxxxxxxxx> wrote:
> > >
[...]
> > > >
> > > > In bpf_jit_binary_hdr(), we calculate header as image & PAGE_MASK.
> > > > If we want s/PAGE_MASK/63 for x86_64, we will have different versions
> > > > of bpf_jit_binary_hdr(). It is not on any hot path, so we can use __weak for
> > > > it. Other than this, I think the solution works fine.
> > >
> > > I think it can stay generic.
> > >
> > > The existing bpf_jit_binary_hdr() will do & PAGE_MASK
> > > while bpf_jit_binary_hdr_pack() will do & 63.
> >
> > The problem with this approach is that we need bpf_prog_ksym_set_addr
> > to be smart to pick bpf_jit_binary_hdr() or bpf_jit_binary_hdr_pack().
>
> We can probably add a true JIT image size to bpf_prog_aux.
> bpf_prog_ksym_set_addr() is approximating the end:
> prog->aux->ksym.end = addr + hdr->pages * PAGE_SIZE
> which doesn't have to include all the 'int 3' padding after the end.
>
> Or add a flag to bpf_prog_aux.
> Ideally bpf_jit_free() would stay generic too.

Both ideas sound promising. Let me try to implement them and see
which is better (or maybe we get both).

Thanks for the suggestions!
Song