[PATCH] security/loadpin: Replace "kernel_read_file_str[j]" with function "kernel_read_file_id_str(j)".
Kees Cook
keescook at chromium.org
Thu Mar 4 23:50:29 UTC 2021
On Thu, Mar 04, 2021 at 08:36:38AM +0000, zhaojiele wrote:
> Actually Linux kernel already provide function "kernel_read_file_id_str()"
> for secure access in "kernel_read_file.h".And, in "parse_exclude()"
> function, there is no need for
> "BUILD_BUG_ON(ARRAY_SIZE(kernel_read_file_str) <
> ARRAY_SIZE(ignore_read_file_id))"
> when access array by functon "kernel_read_file_id_str(j)".
>
> Signed-off-by: zhaojiele <unclexiaole at gmail.com>
> ---
> security/loadpin/loadpin.c | 6 ++----
> 1 file changed, 2 insertions(+), 4 deletions(-)
>
> diff --git a/security/loadpin/loadpin.c b/security/loadpin/loadpin.c
> index b12f7d986b1e..3e8bdcd06600 100644
> --- a/security/loadpin/loadpin.c
> +++ b/security/loadpin/loadpin.c
> @@ -210,8 +210,6 @@ static void __init parse_exclude(void)
> */
> BUILD_BUG_ON(ARRAY_SIZE(exclude_read_files) !=
> ARRAY_SIZE(ignore_read_file_id));
> - BUILD_BUG_ON(ARRAY_SIZE(kernel_read_file_str) <
> - ARRAY_SIZE(ignore_read_file_id));
This should stay to make sure kernel_read_file_str doesn't diverge from
the other two. However, maybe it should be tightened to:
BUILD_BUG_ON(ARRAY_SIZE(kernel_read_file_str) - 1 ==
ARRAY_SIZE(ignore_read_file_id));
>
> for (i = 0; i < ARRAY_SIZE(exclude_read_files); i++) {
> cur = exclude_read_files[i];
> @@ -221,9 +219,9 @@ static void __init parse_exclude(void)
> continue;
>
> for (j = 0; j < ARRAY_SIZE(ignore_read_file_id); j++) {
> - if (strcmp(cur, kernel_read_file_str[j]) == 0) {
> + if (strcmp(cur, kernel_read_file_id_str(j)) == 0) {
> pr_info("excluding: %s\n",
> - kernel_read_file_str[j]);
> + kernel_read_file_id_str(j));
> ignore_read_file_id[j] = 1;
> /*
> * Can not break, because one read_file_str
I feel funny about making these into function calls when we've already
validated the index, but yeah, that would be fine. Can you send a v2
with the earlier suggestion addressed?
Thanks!
-Kees
--
Kees Cook
More information about the Linux-security-module-archive
mailing list