On Sat, Jun 8, 2024 at 2:01 AM Benjamin Tissoires <bentiss@xxxxxxxxxx> wrote:
>
> The idea is to provide a list of stucts and their editable fields.
>
> Currently no functional changes are introduced here, we will add some
> more writeable fields in the next patch.
>
> Signed-off-by: Benjamin Tissoires <bentiss@xxxxxxxxxx>
>
> ---
>
> changes in v3:
> - rewrote WRITE_RANGE macro to not deal with offset from the caller side
>
> new in v2
> ---
> drivers/hid/bpf/hid_bpf_struct_ops.c | 91 +++++++++++++++++++++++++++---------
> 1 file changed, 69 insertions(+), 22 deletions(-)
>
> diff --git a/drivers/hid/bpf/hid_bpf_struct_ops.c b/drivers/hid/bpf/hid_bpf_struct_ops.c
> index 056d05d96962..b14eccb121e0 100644
> --- a/drivers/hid/bpf/hid_bpf_struct_ops.c
> +++ b/drivers/hid/bpf/hid_bpf_struct_ops.c
> @@ -16,6 +16,7 @@
> #include <linux/hid_bpf.h>
> #include <linux/init.h>
> #include <linux/module.h>
> +#include <linux/stddef.h>
> #include <linux/workqueue.h>
> #include "hid_bpf_dispatch.h"
>
> @@ -52,40 +53,86 @@ static int hid_bpf_ops_check_member(const struct btf_type *t,
> return 0;
> }
>
> +struct hid_bpf_offset_write_range {
> + const char *struct_name;
> + u32 struct_length;
> + u32 start;
> + u32 end;
> +};
> +
> static int hid_bpf_ops_btf_struct_access(struct bpf_verifier_log *log,
> const struct bpf_reg_state *reg,
> int off, int size)
> {
> - const struct btf_type *state;
> - const struct btf_type *t;
> - s32 type_id;
> +#define WRITE_RANGE(_name, _field, _is_string) \
> + { \
> + .struct_name = #_name, \
> + .struct_length = sizeof(struct _name), \
> + .start = offsetof(struct _name, _field), \
> + .end = offsetofend(struct _name, _field) - !!(_is_string),
so it works because char name[128]; had last byte as zero
before prog writes into it (in addition to potentially having
earlier 0 bytes), so the string is guaranteed
to be null-terminated regardless of what prog writes into it.
Right?
Overall:
Acked-by: Alexei Starovoitov <ast@xxxxxxxxxx>
