Re: [PATCHv3 bpf-next 3/9] bpf: Use file object build id in stackmap

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Thu, Mar 16, 2023 at 10:02 AM Jiri Olsa <jolsa@xxxxxxxxxx> wrote:
>
> Use build id from file object in stackmap if it's available.
>
> The file's f_build_id is available (for CONFIG_FILE_BUILD_ID option)
> when the file is mmap-ed, so it will be available (if present) when
> used by stackmap.
>
> Signed-off-by: Jiri Olsa <jolsa@xxxxxxxxxx>
> ---

LGTM.

Acked-by: Andrii Nakryiko <andrii@xxxxxxxxxx>

>  kernel/bpf/stackmap.c | 24 +++++++++++++++++++++++-
>  1 file changed, 23 insertions(+), 1 deletion(-)
>
> diff --git a/kernel/bpf/stackmap.c b/kernel/bpf/stackmap.c
> index 0f1d8dced933..14d27bd83081 100644
> --- a/kernel/bpf/stackmap.c
> +++ b/kernel/bpf/stackmap.c
> @@ -124,6 +124,28 @@ static struct bpf_map *stack_map_alloc(union bpf_attr *attr)
>         return ERR_PTR(err);
>  }
>
> +#ifdef CONFIG_FILE_BUILD_ID
> +static int vma_get_build_id(struct vm_area_struct *vma, unsigned char *build_id)
> +{
> +       struct build_id *bid;
> +
> +       if (!vma->vm_file)
> +               return -EINVAL;
> +       bid = vma->vm_file->f_build_id;
> +       if (IS_ERR_OR_NULL(bid))
> +               return bid ? PTR_ERR(bid) : -ENOENT;
> +       if (bid->sz > BUILD_ID_SIZE_MAX)
> +               return -EINVAL;
> +       memcpy(build_id, bid->data, bid->sz);
> +       return 0;
> +}
> +#else
> +static int vma_get_build_id(struct vm_area_struct *vma, unsigned char *build_id)
> +{
> +       return build_id_parse(vma, build_id, NULL);
> +}
> +#endif
> +
>  static void stack_map_get_build_id_offset(struct bpf_stack_build_id *id_offs,
>                                           u64 *ips, u32 trace_nr, bool user)
>  {
> @@ -156,7 +178,7 @@ static void stack_map_get_build_id_offset(struct bpf_stack_build_id *id_offs,
>                         goto build_id_valid;
>                 }
>                 vma = find_vma(current->mm, ips[i]);
> -               if (!vma || build_id_parse(vma, id_offs[i].build_id, NULL)) {
> +               if (!vma || vma_get_build_id(vma, id_offs[i].build_id)) {
>                         /* per entry fall back to ips */
>                         id_offs[i].status = BPF_STACK_BUILD_ID_IP;
>                         id_offs[i].ip = ips[i];
> --
> 2.39.2
>




[Index of Archives]     [Linux Ext4 Filesystem]     [Union Filesystem]     [Filesystem Testing]     [Ceph Users]     [Ecryptfs]     [NTFS 3]     [AutoFS]     [Kernel Newbies]     [Share Photos]     [Security]     [Netfilter]     [Bugtraq]     [Yosemite News]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux Cachefs]     [Reiser Filesystem]     [Linux RAID]     [NTFS 3]     [Samba]     [Device Mapper]     [CEPH Development]

  Powered by Linux