2018-02-28 06:48:35

by Zhang, Ning A

Subject: maybe a bug in SELinux: security_context_to_sid_core

Hi,

Before SELinux is initialized, get scontext by secid by using:

security_secctx_to_secid() may return wrong numbe

eg:
security_secctx_to_secid("devnull", strlen("devnull"), &sid);

sid here will be 1

because:

in security_context_to_sid_core:

...
if (!ss_initialized) {
int i;

for (i = 1; i < SECINITSID_NUM; i++) {
if (!strcmp(initial_sid_to_string[i],
scontext)) {
*sid = i;
return 0;
}
}
*sid = SECINITSID_KERNEL;
return 0;
}
...

and SECINITSID_DEVNULL equals to SECINITSID_NUM, and it will never get
right secid for "devnull".

is this by design or bug?

BR.
Ning.

2018-03-09 02:10:57

by Zhang, Ning A

Subject: Re: maybe a bug in SELinux: security_context_to_sid_core

Hi, Bill

For below SELinux behavior, do you know why.

BR.
Ning.


在 2018-02-28三的 14:47 +0800，Zhang Ning写道：
> Hi,
>
> Before SELinux is initialized, get scontext by secid by using:
>
> security_secctx_to_secid() may return wrong numbe
>
> eg:
> security_secctx_to_secid("devnull", strlen("devnull"), &sid);
>
> sid here will be 1
>
> because:
>
> in security_context_to_sid_core:
>
> ...
> if (!ss_initialized) {
> int i;
>
> for (i = 1; i < SECINITSID_NUM; i++) {
> if (!strcmp(initial_sid_to_string[i],
> scontext)) {
> *sid = i;
> return 0;
> }
> }
> *sid = SECINITSID_KERNEL;
> return 0;
> }
> ...
>
> and SECINITSID_DEVNULL equals to SECINITSID_NUM, and it will never get
> right secid for "devnull".
>
> is this by design or bug?
>
> BR.
> Ning.