[PATCH v7 4/9] seccomp: move no_new_privs into seccomp
From: oleg@redhat.com (Oleg Nesterov)
Date: 2014-06-24 19:20:09
Also in:
linux-api, linux-arch, linux-mips, lkml
On 06/23, Kees Cook wrote:
quoted hunk ↗ jump to hunk
--- a/include/linux/seccomp.h +++ b/include/linux/seccomp.h@@ -3,6 +3,8 @@ #include <uapi/linux/seccomp.h> +#define SECCOMP_FLAG_NO_NEW_PRIVS 0 /* task may not gain privs */ + #ifdef CONFIG_SECCOMP #include <linux/thread_info.h>@@ -16,6 +18,7 @@ struct seccomp_filter; * system calls available to a process. * @filter: must always point to a valid seccomp-filter or NULL as it is * accessed without locking during system call entry. + * @flags: flags under task->sighand->siglock lock * * @filter must only be accessed from the context of current as there * is no read locking.@@ -23,6 +26,7 @@ struct seccomp_filter; struct seccomp { int mode; struct seccomp_filter *filter; + unsigned long flags; }; extern int __secure_computing(int);@@ -51,7 +55,9 @@ static inline int seccomp_mode(struct seccomp *s) #include <linux/errno.h> -struct seccomp { }; +struct seccomp { + unsigned long flags; +};
A bit messy ;)
I am wondering if we can simply do
static inline bool current_no_new_privs(void)
{
if (current->no_new_privs)
return true;
#ifdef CONFIG_SECCOMP
if (test_thread_flag(TIF_SECCOMP))
return true;
#endif
return false;
return test_bit(SECCOMP_FLAG_NO_NEW_PRIVS, &p->seccomp.flags);
}
instead ?
Oleg.