public inbox for io-uring@vger.kernel.org
 help / color / mirror / Atom feed
* [PATCH 1/2] src/queue: don't swallow -ETIME when SQEs were submitted
@ 2026-07-12 22:10 Prateek
  2026-07-12 22:12 ` [PATCH 2/2] test/timeout-swallow: verify -ETIME is not swallowed Prateek
  0 siblings, 1 reply; 2+ messages in thread
From: Prateek @ 2026-07-12 22:10 UTC (permalink / raw)
  To: io-uring; +Cc: axboe, Prateek

If _io_uring_get_cqe() submits SQEs and then times out waiting for
completions, it returns the submit count instead of -ETIME:

  1. The first enter submits the SQEs; because submit > 0 the kernel
     returns the submit count, not -ETIME, and it is stored in err.
  2. On the next iteration the has_ts shortcut wants to report -ETIME,
     but the 'if (!err)' guard sees the non-zero submit count and keeps
     it, so -ETIME is dropped.

That contradicts io_uring_submit_and_wait_timeout(3) and
io_uring_wait_cqes(3), which document -ETIME on timeout.

At these two sites (lines 113 and 118) err is only ever 0 or a positive
submit count. A negative error from __io_uring_peek_cqe() or a prior
enter breaks out of the loop before reaching here. So the change is
functionally equivalent to dropping the err condition entirely; we change
'!err' to 'err >= 0' so -ETIME is successfully synthesized whenever no
CQE was seen.

The guards were added in 2f61e849 ("src/queue: don't wait twice if
looping in _io_uring_get_cqe()") to carry the submit count across
iterations for the partial-completion case (got some CQEs, no error);
that case still returns the count because both sites remain guarded by
!cqe.

Signed-off-by: Prateek <kprateek283@gmail.com>
---
 src/queue.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/queue.c b/src/queue.c
index fcd3c702..e2e5a061 100644
--- a/src/queue.c
+++ b/src/queue.c
@@ -110,12 +110,12 @@ static int _io_uring_get_cqe(struct io_uring *ring,
 			 * timeout, so treat any timeout the same as -ETIME here.
 			 */
 			if (data->get_flags & IORING_ENTER_EXT_ARG_REG) {
-				if (!cqe && !err)
+				if (!cqe && err >= 0)
 					err = -ETIME;
 			} else {
 				struct io_uring_getevents_arg *arg = data->arg;
 
-				if (!cqe && arg->ts && !err)
+				if (!cqe && arg->ts && err >= 0)
 					err = -ETIME;
 			}
 			break;
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 2+ messages in thread

* [PATCH 2/2] test/timeout-swallow: verify -ETIME is not swallowed
  2026-07-12 22:10 [PATCH 1/2] src/queue: don't swallow -ETIME when SQEs were submitted Prateek
@ 2026-07-12 22:12 ` Prateek
  0 siblings, 0 replies; 2+ messages in thread
From: Prateek @ 2026-07-12 22:12 UTC (permalink / raw)
  To: io-uring; +Cc: axboe, Prateek

Regression test for the previous commit. Submits an SQE and waits with a
zero timeout for more completions than can arrive; the result must be
-ETIME, not the positive submit count. Covers the normal EXT_ARG wait
path and the registered-wait path, each skipped gracefully where
unsupported.

Signed-off-by: Prateek <kprateek283@gmail.com>
---
 test/Makefile          |   1 +
 test/timeout-swallow.c | 117 +++++++++++++++++++++++++++++++++++++++++
 2 files changed, 118 insertions(+)
 create mode 100644 test/timeout-swallow.c

diff --git a/test/Makefile b/test/Makefile
index d6358a93..ae23ef6d 100644
--- a/test/Makefile
+++ b/test/Makefile
@@ -293,6 +293,7 @@ test_srcs := \
 	timerfd-short-read.c \
 	timeout.c \
 	timeout-new.c \
+	timeout-swallow.c \
 	timestamp.c \
 	timestamp-bug.c \
 	truncate.c \
diff --git a/test/timeout-swallow.c b/test/timeout-swallow.c
new file mode 100644
index 00000000..9fb4ab01
--- /dev/null
+++ b/test/timeout-swallow.c
@@ -0,0 +1,117 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Description: tests that io_uring_wait_cqes() and variants do not swallow 
+ *              -ETIME when loop-fetching CQEs if some SQEs were submitted.
+ */
+#include <stdio.h>
+#include <unistd.h>
+#include <errno.h>
+#include <sys/time.h>
+#include "liburing.h"
+#include "helpers.h"
+
+/*
+ * Test the normal -ETIME swallow path. (line 118 in queue.c)
+ */
+static int test_timeout(struct io_uring *ring)
+{
+    struct io_uring_sqe *sqe;
+    struct io_uring_cqe *cqe;
+    struct __kernel_timespec long_ts = { .tv_sec = 10, .tv_nsec = 0 };
+    struct __kernel_timespec zero_ts = { .tv_sec = 0, .tv_nsec = 0 };
+    int ret;
+
+    if (!(ring->features & IORING_FEAT_EXT_ARG))
+        return T_EXIT_SKIP;
+
+    sqe = io_uring_get_sqe(ring);
+    if (!sqe) {
+        fprintf(stderr, "get_sqe failed\n");
+        return T_EXIT_FAIL;
+    }
+    /* Long timeout, won't complete immediately */
+    io_uring_prep_timeout(sqe, &long_ts, 1, 0);
+    
+    /* Zero timeout forces immediate expiry inside the syscall on first/second pass */
+    ret = io_uring_submit_and_wait_timeout(ring, &cqe, 2, &zero_ts, NULL);
+    
+    if (ret == -ETIME) {
+        return T_EXIT_PASS;
+    }
+
+    fprintf(stderr, "test_timeout failed: expected -ETIME, got %d\n", ret);
+    return T_EXIT_FAIL;
+}
+
+/*
+ * Test the registered-wait -ETIME swallow path. (line 113 in queue.c)
+ */
+static int test_timeout_reg(struct io_uring *ring)
+{
+    struct io_uring_sqe *sqe;
+    struct io_uring_cqe *cqe;
+    struct __kernel_timespec long_ts = { .tv_sec = 10, .tv_nsec = 0 };
+    struct io_uring_reg_wait reg = { .ts = { .tv_sec = 0, .tv_nsec = 0 } };
+    int ret;
+
+    if (!(ring->features & IORING_FEAT_EXT_ARG))
+        return T_EXIT_SKIP;
+
+    ret = io_uring_register_wait_reg(ring, &reg, 1);
+    if (ret) {
+        /* Not supported on this kernel */
+        return T_EXIT_SKIP;
+    }
+
+    sqe = io_uring_get_sqe(ring);
+    if (!sqe) {
+        fprintf(stderr, "get_sqe failed\n");
+        return T_EXIT_FAIL;
+    }
+    io_uring_prep_timeout(sqe, &long_ts, 1, 0);
+    
+    ret = io_uring_submit_and_wait_reg(ring, &cqe, 2, 0);
+    if (ret == -ETIME) {
+        return T_EXIT_PASS;
+    }
+
+    fprintf(stderr, "test_timeout_reg failed: expected -ETIME, got %d\n", ret);
+    return T_EXIT_FAIL;
+}
+
+int main(int argc, char *argv[])
+{
+    struct io_uring ring;
+    int ret, ret2, final_ret = T_EXIT_PASS;
+
+    if (argc > 1)
+        return T_EXIT_SKIP;
+
+    ret = io_uring_queue_init(8, &ring, 0);
+    if (ret) {
+        if (ret == -ENOSYS)
+            return T_EXIT_SKIP;
+        fprintf(stderr, "queue_init failed: %d\n", ret);
+        return T_EXIT_FAIL;
+    }
+
+    ret = test_timeout(&ring);
+    if (ret == T_EXIT_FAIL) {
+        fprintf(stderr, "test_timeout failed\n");
+        final_ret = T_EXIT_FAIL;
+    }
+
+    ret2 = test_timeout_reg(&ring);
+    if (ret2 == T_EXIT_FAIL) {
+        fprintf(stderr, "test_timeout_reg failed\n");
+        final_ret = T_EXIT_FAIL;
+    }
+
+    io_uring_queue_exit(&ring);
+
+    if (final_ret == T_EXIT_FAIL)
+        return T_EXIT_FAIL;
+    if (ret == T_EXIT_SKIP && ret2 == T_EXIT_SKIP)
+        return T_EXIT_SKIP;
+    return T_EXIT_PASS;
+}
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2026-07-12 22:12 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-07-12 22:10 [PATCH 1/2] src/queue: don't swallow -ETIME when SQEs were submitted Prateek
2026-07-12 22:12 ` [PATCH 2/2] test/timeout-swallow: verify -ETIME is not swallowed Prateek

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox