Skip to contentSkip to conversation

crypto/af_alg02: fixed read() being stuck

PATCH New Christian Amann 3 participants 3 messages May 8, 2019 – May 14, 2019

lore
Share & tools
lore
#0 lore raw

PATCH v3 1/1 crypto/af_alg02: fixed read() being stuck

On kernels < 4.14 (missing commit 2d97591ef43d) reading from the request socket does not return and the testcase does not finish.

This fix moves the logic to a child thread in order for the parent to handle the timeout and report a message to the user.

---

Notes:
Hi Li,

> We could set LTP_ATTRIBUTE_UNUSED at the behind of unused parameter to get > rid of compiling warning

Thats very useful but I couldn't find it anywhere in the documentation. IMHO it should be put in there, because I stumbled upon this problem a couple of times.

Anyway, I implemented your suggestions. I hope it's an alright patch now. Thanks for your feedback!

Regards,
Christian

 testcases/kernel/crypto/Makefile   |  2 ++
 testcases/kernel/crypto/af_alg02.c | 37 +++++++++++++++++++++++++++++++++++--
 2 files changed, 37 insertions(+), 2 deletions(-)
testcases/kernel/crypto/Makefile +2 -0
index 76f9308c2..6547e1cb6 100644
@@ -20,3 +20,5 @@ include $(top_srcdir)/include/mk/testcases.mk
 CFLAGS			+= -D_GNU_SOURCE
 
 include $(top_srcdir)/include/mk/generic_leaf_target.mk
+
+af_alg02: CFLAGS += -pthread
testcases/kernel/crypto/af_alg02.c +35 -2
index a9e820423..1c725212a 100644
@@ -7,23 +7,56 @@
  * Regression test for commit ecaaab564978 ("crypto: salsa20 - fix
  * blkcipher_walk API usage"), or CVE-2017-17805.  This test verifies that an
  * empty message can be encrypted with Salsa20 without crashing the kernel.
+ *
+ * Fix for kernels < 4.14:
+ * With kernels missing commit 2d97591ef43d ("crypto: af_alg - consolidation
+ * of duplicate code") read() does not return in this situation. The call is
+ * now moved to a child thread in order to cancel it in case read() takes an
+ * unusual long amount of time.
  */
 
 #include "tst_test.h"
 #include "tst_af_alg.h"
+#include "tst_safe_pthread.h"
+#include <pthread.h>
+#include <errno.h>
 
-static void run(void)
+void *verify_encrypt(void *arg LTP_ATTRIBUTE_UNUSED)
 {
 	char buf[16];
 	int reqfd = tst_alg_setup_reqfd("skcipher", "salsa20", NULL, 16);
 
+	TST_CHECKPOINT_WAKE(0);
+
 	/* With the bug the kernel crashed here */
 	if (read(reqfd, buf, 16) == 0)
 		tst_res(TPASS, "Successfully \"encrypted\" an empty message");
 	else
-		tst_res(TBROK, "read() didn't return 0");
+		tst_res(TFAIL, "read() didn't return 0");
+	return arg;
+}
+
+static void run(void)
+{
+	pthread_t thr;
+
+	pthread_setcancelstate(PTHREAD_CANCEL_ENABLE, NULL);
+	SAFE_PTHREAD_CREATE(&thr, NULL, verify_encrypt, NULL);
+
+	TST_CHECKPOINT_WAIT(0);
+
+	while (pthread_kill(thr, 0) != ESRCH) {
+		if (tst_timeout_remaining() <= 10) {
+			pthread_cancel(thr);
+			tst_brk(TBROK,
+				"Timed out while reading from request socket.");
+		}
+		usleep(1000);
+	}
 }
 
 static struct tst_test test = {
 	.test_all = run,
+	.timeout = 20,
+	.needs_checkpoints = 1,
 };
Signature · 1 line
2.16.4
#1 lore raw

PATCH v3 1/1 crypto/af_alg02: fixed read() being stuck

Quoted 9 lines · Christian Amann

On kernels < 4.14 (missing commit 2d97591ef43d) reading from the request socket does not return and the testcase does not finish.

This fix moves the logic to a child thread in order for the parent to handle the timeout and report a message to the user.

---

Quoted 4 lines

Notes:
Hi Li,

Quoted 8 lines

> We could set LTP_ATTRIBUTE_UNUSED at the behind of unused parameter
to get

Quoted 1 line

rid of compiling warning

Thats very useful but I couldn't find it anywhere in the documentation. IMHO it should be put in there, because I stumbled upon this problem a couple of times.

This is just a definition for variable attribute which supports by GNU compiler, I'm not sure if we should add it to LTP documents.

$ grep LTP_ATTRIBUTE_UNUSED . -r
tst_common.h:25:#define LTP_ATTRIBUTE_UNUSED __attribute__((unused))
Quoted 66 lines

Anyway, I implemented your suggestions. I hope it's an alright patch
now.
Thanks for your feedback!

Regards,
Christian

 testcases/kernel/crypto/Makefile   |  2 ++
 testcases/kernel/crypto/af_alg02.c | 37
+++++++++++++++++++++++++++++++++++--
 2 files changed, 37 insertions(+), 2 deletions(-)
testcases/kernel/crypto/Makefile +2 -0
index 76f9308c2..6547e1cb6 100644
@@ -20,3 +20,5 @@ include $(top_srcdir)/include/mk/testcases.mk
 CFLAGS                 += -D_GNU_SOURCE
 
 include $(top_srcdir)/include/mk/generic_leaf_target.mk
+
+af_alg02: CFLAGS += -pthread
testcases/kernel/crypto/af_alg02.c +0 -0
index a9e820423..1c725212a 100644
@@ -7,23 +7,56 @@
  * Regression test for commit ecaaab564978 ("crypto: salsa20 - fix
  * blkcipher_walk API usage"), or CVE-2017-17805.  This test verifies

that an
* empty message can be encrypted with Salsa20 without crashing the
kernel.
+ *
+ * Fix for kernels < 4.14:
+ * With kernels missing commit 2d97591ef43d ("crypto: af_alg - consolidation
+ * of duplicate code") read() does not return in this situation. The call is
+ * now moved to a child thread in order to cancel it in case read() takes an
+ * unusual long amount of time.
*/

#include "tst_test.h"
#include "tst_af_alg.h"

testcases/kernel/crypto/af_alg02.c +6 -1
+#include "tst_safe_pthread.h"
+#include <pthread.h>
+#include <errno.h>
 
-static void run(void)
+void *verify_encrypt(void *arg LTP_ATTRIBUTE_UNUSED)
 {
        char buf[16];
        int reqfd = tst_alg_setup_reqfd("skcipher", "salsa20", NULL, 16);
 
+       TST_CHECKPOINT_WAKE(0);
+
        /* With the bug the kernel crashed here */
        if (read(reqfd, buf, 16) == 0)
                tst_res(TPASS, "Successfully \"encrypted\" an empty

message");
else

testcases/kernel/crypto/af_alg02.c line 36

Quoted 3 lines
testcases/kernel/crypto/af_alg02.c +2 -1
-               tst_res(TBROK, "read() didn't return 0");
+               tst_res(TFAIL, "read() didn't return 0");
+       return arg;

Since arg has been marked as unused, so here better to return NULL.

Anyway, patch V3 looks good to me.

Signature · 5 lines
Regards,
Li Wang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20190509/b74f1cc3/attachment.html>
#2 lore raw

PATCH v3 1/1 crypto/af_alg02: fixed read() being stuck

In reply to Christian Amann
Quoted 5 lines
testcases/kernel/crypto/af_alg02.c +1 -1
-static void run(void)
+void *verify_encrypt(void *arg LTP_ATTRIBUTE_UNUSED)
 {
 	char buf[16];
 	int reqfd = tst_alg_setup_reqfd("skcipher", "salsa20", NULL, 16);

testcases/kernel/crypto/af_alg02.c line 36

Quoted 10 lines
testcases/kernel/crypto/af_alg02.c +4 -1
 
+	TST_CHECKPOINT_WAKE(0);
+
 	/* With the bug the kernel crashed here */
 	if (read(reqfd, buf, 16) == 0)
 		tst_res(TPASS, "Successfully \"encrypted\" an empty message");
 	else
-		tst_res(TBROK, "read() didn't return 0");
+		tst_res(TFAIL, "read() didn't return 0");
+	return arg;

Actually there is no point in adding the LTP_ATTRIBUTE_UNUSED since you do return arg; at the end of the function.

So I've removed the LTP_ATTRIBUTE_UNUSED, changed the function to be static and pushed. thanks.

Quoted 32 lines
testcases/kernel/crypto/af_alg02.c +21 -0
+}
+
+static void run(void)
+{
+	pthread_t thr;
+
+	pthread_setcancelstate(PTHREAD_CANCEL_ENABLE, NULL);
+	SAFE_PTHREAD_CREATE(&thr, NULL, verify_encrypt, NULL);
+
+	TST_CHECKPOINT_WAIT(0);
+
+	while (pthread_kill(thr, 0) != ESRCH) {
+		if (tst_timeout_remaining() <= 10) {
+			pthread_cancel(thr);
+			tst_brk(TBROK,
+				"Timed out while reading from request socket.");
+		}
+		usleep(1000);
+	}
 }
 
 static struct tst_test test = {
 	.test_all = run,
+	.timeout = 20,
+	.needs_checkpoints = 1,
 };
Signature · 5 lines
2.16.4


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp
Signature · 2 lines
Cyril Hrubis
[email protected]