diff mbox series

[v2] kill13, CVE-2018-10124: Reproduce INT_MIN negation

Message ID 20210707105954.18182-1-rpalethorpe@suse.com
State Accepted
Headers show
Series [v2] kill13, CVE-2018-10124: Reproduce INT_MIN negation | expand

Commit Message

Richard Palethorpe July 7, 2021, 10:59 a.m. UTC 
Signed-off-by: Richard Palethorpe <rpalethorpe@suse.com>
Acked-by: Joerg Vehlow <joerg.vehlow@aox-tech.de>
---

V2:
* Add more explanation
* Set signal to 0
* Require signed overflow detection in the kernel

 runtest/cve                               |  1 +
 runtest/syscalls                          |  1 +
 testcases/kernel/syscalls/kill/.gitignore |  1 +
 testcases/kernel/syscalls/kill/kill13.c   | 45 +++++++++++++++++++++++
 4 files changed, 48 insertions(+)
 create mode 100644 testcases/kernel/syscalls/kill/kill13.c

Comments

Cyril Hrubis July 8, 2021, 8:49 a.m. UTC | #1 
Hi!
Both pushed, thanks.
Cyril Hrubis July 8, 2021, 9:02 a.m. UTC | #2 
Hi!
> > Both pushed, thanks.
> >
> It seems wait403 doesn't use correct doc format.
> 
> It used the following style.
> /*
>   * [Description]

Good catch, I've missed that, will fix it ASAP.
Yang Xu \(Fujitsu\) July 8, 2021, 9:22 a.m. UTC | #3 
Hi  Cyril, Richard
> Hi!
> Both pushed, thanks.
>
It seems wait403 doesn't use correct doc format.

It used the following style.
/*
  * [Description]

Best Regards
Yang Xu
diff mbox series

Patch

diff --git a/runtest/cve b/runtest/cve
index 5a6ef966d..226b5ea44 100644
--- a/runtest/cve
+++ b/runtest/cve
@@ -49,6 +49,7 @@  cve-2018-5803 sctp_big_chunk
 cve-2018-7566 snd_seq01
 cve-2018-8897 ptrace09
 cve-2018-9568 connect02
+cve-2018-10124 kill13
 cve-2018-1000001 realpath01
 cve-2018-1000199 ptrace08
 cve-2018-1000204 ioctl_sg01
diff --git a/runtest/syscalls b/runtest/syscalls
index 98fe3c02e..0c1e16f9e 100644
--- a/runtest/syscalls
+++ b/runtest/syscalls
@@ -644,6 +644,7 @@  kill09 kill09
 kill10 kill10
 kill11 kill11
 kill12 kill12
+kill13 kill13
 
 lchown01 lchown01
 lchown01_16 lchown01_16
diff --git a/testcases/kernel/syscalls/kill/.gitignore b/testcases/kernel/syscalls/kill/.gitignore
index 75fdaa561..810ed0200 100644
--- a/testcases/kernel/syscalls/kill/.gitignore
+++ b/testcases/kernel/syscalls/kill/.gitignore
@@ -8,3 +8,4 @@ 
 /kill10
 /kill11
 /kill12
+/kill13
diff --git a/testcases/kernel/syscalls/kill/kill13.c b/testcases/kernel/syscalls/kill/kill13.c
new file mode 100644
index 000000000..66ae37bc0
--- /dev/null
+++ b/testcases/kernel/syscalls/kill/kill13.c
@@ -0,0 +1,45 @@ 
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (C) 2021 SUSE LLC <rpalethorpe@suse.com>
+ */
+
+/*\
+ * [Description]
+ *
+ * Reproducer of CVE-2018-10124; INT_MIN negation.
+ *
+ * On most two's complement CPUs negation of INT_MIN will result in
+ * INT_MIN because ~((unsigned)INT_MIN) + 1 overflows to INT_MIN
+ * (unless trapped). On one's complement ~((unsigned)INT_MIN) = INT_MAX.
+ *
+ * Without UBSAN kill will always return ESRCH. Regardless of if the
+ * bug is present as INT_MIN/INT_MAX are invalid PIDs. It checks the
+ * PID before the signal number so we can not cause EINVAL. A trivial
+ * test of kill is performed elsewhere. So we don't run the test
+ * without UBSAN to avoid giving the impression we have actually
+ * tested for the bug.
+ */
+
+#include <limits.h>
+#include <signal.h>
+#include "tst_test.h"
+
+static void run(void)
+{
+	TST_EXP_FAIL2(kill(INT_MIN, 0), ESRCH,
+		      "kill(INT_MIN, ...) fails with ESRCH");
+}
+
+static struct tst_test test = {
+	.test_all = run,
+	.taint_check = TST_TAINT_W | TST_TAINT_D,
+	.needs_kconfigs = (const char *[]) {
+		"CONFIG_UBSAN_SIGNED_OVERFLOW",
+		NULL
+	},
+	.tags = (const struct tst_tag[]) {
+		{"linux-git", "4ea77014af0d"},
+		{"CVE", "CVE-2018-10124"},
+		{}
+	}
+};