On Tue, Jun 16, 2020 at 3:07 AM Jiri Olsa <[email protected]> wrote:
>
> Adding d_path helper function that returns full path
> for give 'struct path' object, which needs to be the
> kernel BTF 'path' object.
>
> The helper calls directly d_path function.
>
> Updating also bpf.h tools uapi header and adding
> 'path' to bpf_helpers_doc.py script.
>
> Signed-off-by: Jiri Olsa <[email protected]>
> ---
> include/linux/bpf.h | 4 ++++
> include/uapi/linux/bpf.h | 14 ++++++++++++-
> kernel/bpf/btf_ids.c | 11 ++++++++++
> kernel/trace/bpf_trace.c | 38 ++++++++++++++++++++++++++++++++++
> scripts/bpf_helpers_doc.py | 2 ++
> tools/include/uapi/linux/bpf.h | 14 ++++++++++++-
> 6 files changed, 81 insertions(+), 2 deletions(-)
>
> diff --git a/include/linux/bpf.h b/include/linux/bpf.h
> index a94e85c2ec50..d35265b6c574 100644
> --- a/include/linux/bpf.h
> +++ b/include/linux/bpf.h
> @@ -1752,5 +1752,9 @@ extern int bpf_skb_output_btf_ids[];
> extern int bpf_seq_printf_btf_ids[];
> extern int bpf_seq_write_btf_ids[];
> extern int bpf_xdp_output_btf_ids[];
> +extern int bpf_d_path_btf_ids[];
> +
> +extern int btf_whitelist_d_path[];
> +extern int btf_whitelist_d_path_cnt;
So with suggestion from previous patch, this would be declared as:
extern const struct btf_id_set btf_whitelist_d_path;
>
> #endif /* _LINUX_BPF_H */
> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index c65b374a5090..e308746b9344 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> @@ -3252,6 +3252,17 @@ union bpf_attr {
> * case of **BPF_CSUM_LEVEL_QUERY**, the current skb->csum_level
> * is returned or the error code -EACCES in case the skb is not
> * subject to CHECKSUM_UNNECESSARY.
> + *
> + * int bpf_d_path(struct path *path, char *buf, u32 sz)
> + * Description
> + * Return full path for given 'struct path' object, which
> + * needs to be the kernel BTF 'path' object. The path is
> + * returned in buffer provided 'buf' of size 'sz'.
> + *
> + * Return
> + * length of returned string on success, or a negative
> + * error in case of failure
> + *
> */
> #define __BPF_FUNC_MAPPER(FN) \
> FN(unspec), \
> @@ -3389,7 +3400,8 @@ union bpf_attr {
> FN(ringbuf_submit), \
> FN(ringbuf_discard), \
> FN(ringbuf_query), \
> - FN(csum_level),
> + FN(csum_level), \
> + FN(d_path),
>
> /* integer value in 'imm' field of BPF_CALL instruction selects which helper
> * function eBPF program intends to call
> diff --git a/kernel/bpf/btf_ids.c b/kernel/bpf/btf_ids.c
> index d8d0df162f04..853c8fd59b06 100644
> --- a/kernel/bpf/btf_ids.c
> +++ b/kernel/bpf/btf_ids.c
> @@ -13,3 +13,14 @@ BTF_ID(struct, seq_file)
>
> BTF_ID_LIST(bpf_xdp_output_btf_ids)
> BTF_ID(struct, xdp_buff)
> +
> +BTF_ID_LIST(bpf_d_path_btf_ids)
> +BTF_ID(struct, path)
> +
> +BTF_WHITELIST_ENTRY(btf_whitelist_d_path)
> +BTF_ID(func, vfs_truncate)
> +BTF_ID(func, vfs_fallocate)
> +BTF_ID(func, dentry_open)
> +BTF_ID(func, vfs_getattr)
> +BTF_ID(func, filp_close)
> +BTF_WHITELIST_END(btf_whitelist_d_path)
Oh, so that's why you added btf_ids.c. Do you think centralizing all
those BTF ID lists in one file is going to be more convenient? I lean
towards keeping them closer to where they are used, as it was with all
those helper BTF IDS. But I wonder what others think...
> diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c
> index c1866d76041f..0ff5d8434d40 100644
> --- a/kernel/trace/bpf_trace.c
> +++ b/kernel/trace/bpf_trace.c
> @@ -1016,6 +1016,42 @@ static const struct bpf_func_proto
> bpf_send_signal_thread_proto = {
> .arg1_type = ARG_ANYTHING,
> };
>
[...]
