[PATCH 1/1] capabilities: oom_kill: don't set PF_SUPERPRIV for oom check

[Serge Hallyn]

With 64-bit capabilities came an inadvertent change such
that the check for a privileged process, to make it less
likely to be killed when out of memory, sets PF_SUPERPRIV,
which it did not do before.

This patch restores the original behavior of not setting the
PF_SUPERPRIV bit when checking for privileged processes.

Signed-off-by: Serge Hallyn <serue@xxxxxxxxxx>
Acked-by: Andrew G. Morgan <morgan@xxxxxxxxxx>
---
 include/linux/capability.h |    2 ++
 mm/oom_kill.c              |    3 ++-
 2 files changed, 4 insertions(+), 1 deletions(-)

diff --git a/include/linux/capability.h b/include/linux/capability.h
index 7d50ff6..0acdd65 100644
--- a/include/linux/capability.h
+++ b/include/linux/capability.h
@@ -490,6 +490,8 @@ extern const kernel_cap_t __cap_init_eff_set;
 int capable(int cap);
 int __capable(struct task_struct *t, int cap);
 
+#define CAPABLE_PROBE_ONLY(a,b)		(!security_capable(a,b))
+
 extern long cap_prctl_drop(unsigned long cap);
 
 #endif /* __KERNEL__ */
diff --git a/mm/oom_kill.c b/mm/oom_kill.c
index 9fd8d5d..cd515f8 100644
--- a/mm/oom_kill.c
+++ b/mm/oom_kill.c
@@ -128,7 +128,8 @@ unsigned long badness(struct task_struct *p, unsigned long uptime,
 	 * Superuser processes are usually more important, so we make it
 	 * less likely that we kill those.
 	 */
-	if (__capable(p, CAP_SYS_ADMIN) || __capable(p, CAP_SYS_RESOURCE))
+	if (CAPABLE_PROBE_ONLY(p, CAP_SYS_ADMIN) ||
+		CAPABLE_PROBE_ONLY(p, CAP_SYS_RESOURCE))
 		points /= 4;
 
 	/*
-- 
1.5.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/