From: Alexander Aring <aahringo@redhat.com>
To: ltp@lists.linux.it
Subject: [LTP] [PATCH 3/5] fcntl42: test for F_SETLKW interruption case
Date: Tue, 30 May 2023 16:37:05 -0400 [thread overview]
Message-ID: <20230530203707.2965684-4-aahringo@redhat.com> (raw)
In-Reply-To: <20230530203707.2965684-1-aahringo@redhat.com>
This patch adds fcntl42 testcase to test on side effects, e.g. unlock
all acquired locks, when a lock request was interrupted by a signal.
Signed-off-by: Alexander Aring <aahringo@redhat.com>
---
testcases/kernel/syscalls/fcntl/.gitignore | 2 +
testcases/kernel/syscalls/fcntl/fcntl42.c | 153 +++++++++++++++++++++
2 files changed, 155 insertions(+)
create mode 100644 testcases/kernel/syscalls/fcntl/fcntl42.c
diff --git a/testcases/kernel/syscalls/fcntl/.gitignore b/testcases/kernel/syscalls/fcntl/.gitignore
index 4bdae1a22..abffa2967 100644
--- a/testcases/kernel/syscalls/fcntl/.gitignore
+++ b/testcases/kernel/syscalls/fcntl/.gitignore
@@ -78,3 +78,5 @@
/fcntl40_64
/fcntl41
/fcntl41_64
+/fcntl42
+/fcntl42_64
diff --git a/testcases/kernel/syscalls/fcntl/fcntl42.c b/testcases/kernel/syscalls/fcntl/fcntl42.c
new file mode 100644
index 000000000..4d66568fd
--- /dev/null
+++ b/testcases/kernel/syscalls/fcntl/fcntl42.c
@@ -0,0 +1,153 @@
+/*
+ * [Description]
+ * This test confirms that DLM posix locking has problems when a posix lock
+ * got interrupted every lock gets unlocked.
+ *
+ * man fcntl:
+ *
+ * F_SETLKW (struct flock *)
+ * As for F_SETLK, but if a conflicting lock is held on the file,
+ * then wait for that lock to be released. If a signal is caught
+ * while waiting, then the call is interrupted and (after the signal
+ * handler has returned) returns immediately (with return value -1 and
+ * errno set to EINTR; see signal(7)).
+ *
+ * The above quote of the man page of fcntl() is what this testcase tests.
+ * particulary if it has side-effects of previously acquired locks.
+ *
+ * [How to use it]
+ * Call it with TMPDIR=/mnt ./fcntl42 where TMPDIR is a gfs2 mountpoint.
+ * Try it on other filesystems to compare results.
+ *
+ * [What's it doing]
+ *
+ * What this test does is (using dlm_controld interval range interpretation):
+ *
+ * parent:
+ *
+ * 1. lock[0-0]
+ *
+ * child:
+ *
+ * 2. lock[1-1]
+ * 3. lockw[0-0] - should block (see 1. parent), but we get interrupted by SIGALRM
+ *
+ * parent:
+ *
+ * 4. trylock[1-1] - should return -1 and errno -EAGAIN because the child
+ * should still have lock[1-1] acuired and this is what
+ * the child thinks to have. If it's successful the child
+ * wrongly assumes it has the lock[1-1] still acquired and
+ * the child process is still alive.
+ *
+ */
+
+#include <sys/wait.h>
+
+#include "tst_test.h"
+
+static int fd;
+
+static void catch_alarm(int num)
+{
+ (void)num;
+
+ tst_res(TINFO, "catch alarm");
+}
+
+void do_child1(void)
+{
+ struct flock fl = {
+ .l_type = F_WRLCK,
+ .l_whence = SEEK_SET,
+ .l_start = 1L,
+ .l_len = 1L,
+ };
+ struct sigaction act;
+ int rv;
+
+ SAFE_FCNTL(fd, F_SETLK, &fl);
+ tst_res(TINFO, "child locked lock 1-1");
+
+ fl.l_type = F_WRLCK;
+ fl.l_start = 0;
+ fl.l_len = 1;
+
+ memset(&act, 0, sizeof(act));
+ act.sa_handler = catch_alarm;
+ sigemptyset(&act.sa_mask);
+ sigaddset(&act.sa_mask, SIGALRM);
+ sigaction(SIGALRM, &act, NULL);
+
+ /* interrupt SETLKW by signal in 3 secs */
+ alarm(3);
+ rv = fcntl(fd, F_SETLKW, &fl);
+ if (rv == -1 && errno == EINTR)
+ tst_res(TPASS, "Child1 interrupted 0-0");
+
+ TST_CHECKPOINT_WAKE(1);
+ /* keep child alive until parent check region */
+ TST_CHECKPOINT_WAIT(2);
+}
+
+static void fcntl40_test(void)
+{
+ struct flock fl = {
+ .l_type = F_WRLCK,
+ .l_whence = SEEK_SET,
+ .l_start = 0L,
+ .l_len = 1L,
+ };
+ pid_t pid;
+ int rv;
+
+ SAFE_FCNTL(fd, F_SETLK, &fl);
+ tst_res(TINFO, "parent lock 0-0");
+
+ pid = SAFE_FORK();
+ if (pid == 0) {
+ do_child1();
+ tst_res(TINFO, "childs exits");
+ return;
+ }
+
+ TST_CHECKPOINT_WAIT(1);
+
+ fl.l_type = F_WRLCK;
+ fl.l_start = 1;
+ fl.l_len = 1;
+ rv = fcntl(fd, F_SETLK, &fl);
+ /* parent testing childs region, the child will think
+ * it has region 1-1 locked because it was interrupted
+ * by region 0-0. Due bugs the interruption also unlocked
+ * region 1-1.
+ */
+ if (rv == -1 && errno == EAGAIN)
+ tst_res(TPASS, "region 1-1 locked");
+ else
+ tst_res(TFAIL, "region 1-1 unlocked");
+
+ TST_CHECKPOINT_WAKE(2);
+
+ wait(NULL);
+ tst_res(TINFO, "parent exits");
+}
+
+static void setup(void)
+{
+ fd = SAFE_OPEN("filename", O_RDWR | O_CREAT, 0700);
+}
+
+static void cleanup(void)
+{
+ if (fd > -1)
+ SAFE_CLOSE(fd);
+}
+
+static struct tst_test test = {
+ .forks_child = 1,
+ .needs_checkpoints = 1,
+ .test_all = fcntl40_test,
+ .setup = setup,
+ .cleanup = cleanup,
+};
--
2.31.1
--
Mailing list info: https://lists.linux.it/listinfo/ltp
next prev parent reply other threads:[~2023-05-30 20:37 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-05-30 20:37 [LTP] [PATCH 0/5] fcntl: add more testcases Alexander Aring
2023-05-30 20:37 ` [LTP] [PATCH 1/5] fcntl40: test for owner values on classic posix lock Alexander Aring
2023-06-21 9:03 ` Petr Vorel
2023-06-30 19:59 ` Alexander Aring
2023-07-02 19:18 ` Petr Vorel
2023-07-05 13:23 ` Alexander Aring
2023-07-07 8:14 ` Petr Vorel
2023-07-07 12:50 ` Alexander Aring
2023-07-07 13:17 ` Petr Vorel
2023-07-02 19:19 ` Petr Vorel
2023-05-30 20:37 ` [LTP] [PATCH 2/5] fcntl41: test for owner values on OFD posix locks Alexander Aring
2023-06-21 9:38 ` Petr Vorel
2023-06-30 20:00 ` Alexander Aring
2023-05-30 20:37 ` Alexander Aring [this message]
2023-05-30 20:37 ` [LTP] [PATCH 4/5] fcntl43: test for identical F_SETLKW lock requests Alexander Aring
2023-05-30 20:37 ` [LTP] [PATCH 5/5] fcntl44: test for kill child while others waiting Alexander Aring
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20230530203707.2965684-4-aahringo@redhat.com \
--to=aahringo@redhat.com \
--cc=ltp@lists.linux.it \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox