On Sun, May 31, 2020 at 8:10 AM Jiri Olsa <jolsa@xxxxxxxxxx> wrote:
>
> On Fri, May 29, 2020 at 01:48:58PM -0700, Andrii Nakryiko wrote:
> > On Thu, May 28, 2020 at 10:24 AM Jiri Olsa <jolsa@xxxxxxxxxx> wrote:
> > >
> > > On Thu, May 14, 2020 at 03:46:26PM -0700, Andrii Nakryiko wrote:
> > >
> > > SNIP
> > >
> > > > > I was thinking of putting the names in __init section and generate the BTF
> > > > > ids on kernel start, but the build time generation seemed more convenient..
> > > > > let's see the linking times with 'real size' whitelist and we can reconsider
> > > > >
> > > >
> > > > Being able to record such places where to put BTF ID in code would be
> > > > really nice, as Alexei mentioned. There are many potential use cases
> > > > where it would be good to have BTF IDs just put into arbitrary
> > > > variables/arrays. This would trigger compilation error, if someone
> > > > screws up the name, or function is renamed, or if function can be
> > > > compiled out under some configuration. E.g., assuming some reasonable
> > > > implementation of the macro
> > >
> > > hi,
> > > I'm struggling with this part.. to get some reasonable reference
> > > to function/name into 32 bits? any idea? ;-)
> > >
> >
> > Well, you don't have to store actual pointer, right? E.g, emitting
> > something like this in assembly:
> >
> > .global __BTF_ID___some_function
> > .type __BTF_ID___some_function, @object
> > .size __BTF_ID___some_function, 4
> > __BTF_ID___some_function:
> > .zero 4
> >
> > Would reserve 4 bytes and emit __BTF_ID___some_function symbol. If we
> > can then post-process vmlinux image and for all symbols starting with
> > __BTF_ID___ find some_function BTF type id and put it into those 4
> > bytes, that should work, no?
> >
> > Maybe generalize it to __BTF_ID__{func,struct,typedef}__some_function,
> > whatever, not sure. Just an idea.
>
> nice, so something like below?
>
> it'd be in .S file, or perhaps in inline asm, assuming I'll be
> able to pass macro arguments to asm("")
I'd do inline asm, there are no arguments you need to pass into
asm("") itself, everything can be done through macro string
interpolation, I think. Having everything in .c file would be way more
convenient and obvious.
>
> with externs defined in some header file:
>
> extern const int bpf_skb_output_btf_ids[];
> extern const int btf_whitelist_d_path[];
>
> $ objdump -x ./kernel/bpf/whitelist.o
> ...
> 0000000000000000 l O .data 0000000000000004 __BTF_ID__func__vfs_truncate
> 0000000000000004 l O .data 0000000000000004 __BTF_ID__func__vfs_fallocate
> 0000000000000008 l O .data 0000000000000004 __BTF_ID__func__krava
> 0000000000000010 l O .data 0000000000000004 __BTF_ID__struct__sk_buff
> 0000000000000000 g .data 0000000000000000 btf_whitelist_d_path
> 0000000000000010 g .data 0000000000000000 bpf_skb_output_btf_ids
>
> also it'd be nice to get rid of BTF_ID__ symbols in the final link
>
> thanks,
> jirka
>
>
> ---
> #define BTF_ID(prefix, name) \
> .local __BTF_ID__##prefix##__##name; \
> .type __BTF_ID__##prefix##__##name, @object; \
> .size __BTF_ID__##prefix##__##name, 4; \
> __BTF_ID__##prefix##__##name: \
> .zero 4
>
> #define BTF_ID_LIST(name) \
> .global name; \
> name:
>
> #define ZERO .zero 4
>
> .data
>
> BTF_ID_LIST(btf_whitelist_d_path)
> BTF_ID(func, vfs_truncate)
> BTF_ID(func, vfs_fallocate)
> BTF_ID(func, krava)
> ZERO
>
> BTF_ID_LIST(bpf_skb_output_btf_ids)
> BTF_ID(struct, sk_buff)
>
