Re: [PATCH v3 3/3] bpf: make bpf_dump_raw_ok() based on CONFIG_KALLSYMS

["Leizhen (ThunderTown)" <thunder.leizhen@huawei.com>]

On 2023/6/5 12:07, Maninder Singh wrote:
> bpf_dump_raw_ok() depends on kallsyms_show_value() and we already
> have a false definition for the !CONFIG_KALLSYMS case. But we have
> expanded kallsyms_show_value() to work for !CONFIG_KALLSYMS case also
> in previous patch.
> 
> And so to make the code easier to follow just provide a direct
> !CONFIG_KALLSYMS definition for bpf_dump_raw_ok() as well.
> 
> As it is heavily dependent on KALLSYMS and checking based on
> kallsyms_show_value() will not work now.

This patch needs to be swapped with 2/3. To avoid unnecessary trouble
for people using "git bisect", assume that they just fall back to patch 2/3.

> 
> Co-developed-by: Onkarnath <onkarnath.1@samsung.com>
> Signed-off-by: Onkarnath <onkarnath.1@samsung.com>
> Signed-off-by: Maninder Singh <maninder1.s@samsung.com>
> Reviewed-by: Luis Chamberlain <mcgrof@kernel.org>
> ---
>  include/linux/filter.h | 14 +++++++++++---
>  1 file changed, 11 insertions(+), 3 deletions(-)
> 
> diff --git a/include/linux/filter.h b/include/linux/filter.h
> index bbce89937fde..1f237a3bb11a 100644
> --- a/include/linux/filter.h
> +++ b/include/linux/filter.h
> @@ -923,13 +923,21 @@ bool bpf_jit_supports_kfunc_call(void);
>  bool bpf_jit_supports_far_kfunc_call(void);
>  bool bpf_helper_changes_pkt_data(void *func);
>  
> +/*
> + * Reconstruction of call-sites is dependent on kallsyms,
> + * thus make dump the same restriction.
> + */
> +#ifdef CONFIG_KALLSYMS
>  static inline bool bpf_dump_raw_ok(const struct cred *cred)
>  {
> -	/* Reconstruction of call-sites is dependent on kallsyms,
> -	 * thus make dump the same restriction.
> -	 */
>  	return kallsyms_show_value(cred);
>  }
> +#else
> +static inline bool bpf_dump_raw_ok(const struct cred *cred)
> +{
> +	return false;
> +}
> +#endif
>  
>  struct bpf_prog *bpf_patch_insn_single(struct bpf_prog *prog, u32 off,
>  				       const struct bpf_insn *patch, u32 len);
> 

-- 
Regards,
  Zhen Lei