Re: [PATCH bpf-next v7] bpf: add new helper get_file_path for mapping a file descriptor to a pathname

[Alexei Starovoitov <alexei.starovoitov@gmail.com>]

On Sun, Nov 03, 2019 at 02:54:17AM -0500, Wenbo Zhang wrote:
> When people want to identify which file system files are being opened,
> read, and written to, they can use this helper with file descriptor as
> input to achieve this goal. Other pseudo filesystems are also supported.
> 
> This requirement is mainly discussed here:
> 
>   https://github.com/iovisor/bcc/issues/237
> 
> v6->v7:
> - fix missing signed-off-by line
> 
> v5->v6: addressed Andrii's feedback
> - avoid unnecessary goto end by having two explicit returns
> 
> v4->v5: addressed Andrii and Daniel's feedback
> - rename bpf_fd2path to bpf_get_file_path to be consistent with other
> helper's names
> - when fdget_raw fails, set ret to -EBADF instead of -EINVAL
> - remove fdput from fdget_raw's error path
> - use IS_ERR instead of IS_ERR_OR_NULL as d_path ether returns a pointer
> into the buffer or an error code if the path was too long
> - modify the normal path's return value to return copied string length
> including NUL
> - update this helper description's Return bits.
> 
> v3->v4: addressed Daniel's feedback
> - fix missing fdput()
> - move fd2path from kernel/bpf/trace.c to kernel/trace/bpf_trace.c
> - move fd2path's test code to another patch
> - add comment to explain why use fdget_raw instead of fdget
> 
> v2->v3: addressed Yonghong's feedback
> - remove unnecessary LOCKDOWN_BPF_READ
> - refactor error handling section for enhanced readability
> - provide a test case in tools/testing/selftests/bpf
> 
> v1->v2: addressed Daniel's feedback
> - fix backward compatibility
> - add this helper description
> - fix signed-off name
> 
> Signed-off-by: Wenbo Zhang <ethercflow@gmail.com>
> ---
>  include/uapi/linux/bpf.h       | 15 ++++++++++-
>  kernel/trace/bpf_trace.c       | 48 ++++++++++++++++++++++++++++++++++
>  tools/include/uapi/linux/bpf.h | 15 ++++++++++-
>  3 files changed, 76 insertions(+), 2 deletions(-)
> 
> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index a6bf19dabaab..d618a914c6fe 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> @@ -2777,6 +2777,18 @@ union bpf_attr {
>   * 		restricted to raw_tracepoint bpf programs.
>   * 	Return
>   * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_get_file_path(char *path, u32 size, int fd)
> + *	Description
> + *		Get **file** atrribute from the current task by *fd*, then call
> + *		**d_path** to get it's absolute path and copy it as string into
> + *		*path* of *size*. The **path** also support pseudo filesystems
> + *		(whether or not it can be mounted). The *size* must be strictly
> + *		positive. On success, the helper makes sure that the *path* is
> + *		NUL-terminated. On failure, it is filled with zeroes.
> + *	Return
> + *		On success, returns the length of the copied string INCLUDING
> + *		the trailing NUL, or a negative error in case of failure.
>   */
>  #define __BPF_FUNC_MAPPER(FN)		\
>  	FN(unspec),			\
> @@ -2890,7 +2902,8 @@ union bpf_attr {
>  	FN(sk_storage_delete),		\
>  	FN(send_signal),		\
>  	FN(tcp_gen_syncookie),		\
> -	FN(skb_output),
> +	FN(skb_output),			\
> +	FN(get_file_path),
>  
>  /* integer value in 'imm' field of BPF_CALL instruction selects which helper
>   * function eBPF program intends to call
> diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c
> index f50bf19f7a05..41be1c5989af 100644
> --- a/kernel/trace/bpf_trace.c
> +++ b/kernel/trace/bpf_trace.c
> @@ -683,6 +683,52 @@ static const struct bpf_func_proto bpf_send_signal_proto = {
>  	.arg1_type	= ARG_ANYTHING,
>  };
>  
> +BPF_CALL_3(bpf_get_file_path, char *, dst, u32, size, int, fd)
> +{
> +	struct fd f;
> +	char *p;
> +	int ret = -EBADF;
> +
> +	/* Use fdget_raw instead of fdget to support O_PATH, and
> +	 * fdget_raw doesn't have any sleepable code, so it's ok
> +	 * to be here.
> +	 */
> +	f = fdget_raw(fd);
> +	if (!f.file)
> +		goto error;
> +
> +	/* d_path doesn't have any sleepable code, so it's ok to
> +	 * be here. But it uses the current macro to get fs_struct
> +	 * (current->fs). So this helper shouldn't be called in
> +	 * interrupt context.
> +	 */
> +	p = d_path(&f.file->f_path, dst, size);
> +	if (IS_ERR(p)) {
> +		ret = PTR_ERR(p);
> +		fdput(f);
> +		goto error;
> +	}

This is definitely very useful helper that bpf tracing community has
been asking for long time, but I have few concerns with implementation:
- fdget_raw is only used inside fs/, so it doesn't look right to skip the layers.
- accessing current->fs is not always correct, so the code should somehow
  check that it's ok to do so, but I'm not sure if (in_irq()) would be enough.
- some implementations of d_dname do sleep.  For example: dmabuffs_dname.
  Though it seems to me that it's a bug in that particular FS. But I'd like
  to hear clear yes from VFS experts that fdget_raw() + d_path() is ok
  from preempt_disabled section.

The other alternative is to wait for sleepable and preemptible BPF programs to
appear. Which is probably a month or so away. Then all these issues will
disappear.