[LTP] [PATCH v2 0/3] io_uring READ(V), WRITE(v) operation tests

[LTP] [PATCH v2 0/3] io_uring READ(V), WRITE(v) operation tests

[Sachin Sant]

This patch series adds a set of test case to validate
IOURING READ & WRITE (io_uring03), READV & WRITEV (io_uring04)
operations. The patch also adds a common header file to
avoid code duplication.

This patch series also refactors existing test io_uring01
to use the common structures and function defined in header.

Signed-off-by: Sachin Sant <sachinp@linux.ibm.com>
---
V2 changes:
- Use guarded buffer(including iovec) allocation
- Move setup/cleanup code from run to appropriate path
- Add a function to map the OP to the enum name
- Define and use generic cleanup functions
- Add a buffer size 0 in the middle of the iovec
- Simplify setup_io_uring_test to use common code.
- Link to V1: https://lore.kernel.org/ltp/20260320124742.75946-1-sachinp@linux.ibm.com/T/#t

Changes after RFC patch series:
- Addressed review comments
- Refactored io_uring01 test to use common code
- Removed git tags
- Link to RFC: https://lore.kernel.org/ltp/20260318110328.52031-1-sachinp@linux.ibm.com/T/#t

Test run:

$ ./testcases/bin/io_uring03
tst_buffers.c:57: TINFO: Test is using guarded buffers
tst_tmpdir.c:308: TINFO: Using /tmp/LTP_io_o1kf2Z as tmpdir (tmpfs filesystem)
....
tst_kconfig.c:88: TINFO: Parsing kernel config '/proc/config.gz'
tst_kconfig.c:676: TINFO: CONFIG_FAULT_INJECTION kernel option detected which might slow the execution
tst_test.c:1887: TINFO: Overall timeout per run is 0h 02m 00s
io_uring03.c:62: TINFO: Testing IORING_OP_WRITE
io_uring_common.h:310: TPASS: OP=IORING_OP_WRITE (17) fd=4 buf=0x7fff83b0f000 len=4096 offset=0
io_uring03.c:68: TINFO: Testing IORING_OP_READ
io_uring_common.h:310: TPASS: OP=IORING_OP_READ (16) fd=4 buf=0x7fff83aef000 len=4096 offset=0
io_uring03.c:40: TPASS: Basic I/O data integrity verified
io_uring03.c:83: TINFO: Testing partial I/O operations
io_uring_common.h:310: TPASS: OP=IORING_OP_WRITE (17) fd=4 buf=0x7fff83b0f000 len=2048 offset=0
io_uring_common.h:310: TPASS: OP=IORING_OP_WRITE (17) fd=4 buf=0x7fff83b0f800 len=2048 offset=2048
io_uring_common.h:310: TPASS: OP=IORING_OP_READ (16) fd=4 buf=0x7fff83aef000 len=4096 offset=0
io_uring03.c:40: TPASS: Partial I/O data integrity verified

.....
ltcden8-lp8:/opt/ltp # ./testcases/bin/io_uring04
tst_buffers.c:57: TINFO: Test is using guarded buffers
tst_tmpdir.c:308: TINFO: Using /tmp/LTP_io_reAIwJ as tmpdir (tmpfs filesystem)
......
tst_kconfig.c:88: TINFO: Parsing kernel config '/proc/config.gz'
tst_kconfig.c:676: TINFO: CONFIG_FAULT_INJECTION kernel option detected which might slow the execution
tst_test.c:1887: TINFO: Overall timeout per run is 0h 02m 00s
io_uring04.c:102: TINFO: Testing IORING_OP_WRITEV and IORING_OP_READV
io_uring04.c:106: TINFO: Writing 4096 bytes using 4 vectors
io_uring_common.h:326: TPASS: OP=IORING_OP_WRITEV (02) fd=4 iovs=0x7fffb832ffc0 nvecs=4 offset=0 expected_size=4096
io_uring04.c:112: TINFO: Reading 4096 bytes using 4 vectors
io_uring_common.h:326: TPASS: OP=IORING_OP_READV (01) fd=4 iovs=0x7fffb7bcffc0 nvecs=4 offset=0 expected_size=4096
io_uring04.c:92: TPASS: Basic vectored I/O: data integrity verified across 4 vectors
io_uring04.c:126: TINFO: Testing partial vector operations
io_uring_common.h:326: TPASS: OP=IORING_OP_WRITEV (02) fd=4 iovs=0x7fffb832ffc0 nvecs=2 offset=0 expected_size=2048
io_uring_common.h:326: TPASS: OP=IORING_OP_WRITEV (02) fd=4 iovs=0x7fffb832ffe0 nvecs=2 offset=2048 expected_size=2048
io_uring_common.h:326: TPASS: OP=IORING_OP_READV (01) fd=4 iovs=0x7fffb7bcffc0 nvecs=4 offset=0 expected_size=4096
io_uring04.c:92: TPASS: Partial vector I/O: data integrity verified across 4 vectors
io_uring04.c:154: TINFO: Testing vectors with varying sizes including zero-length buffer
io_uring_common.h:326: TPASS: OP=IORING_OP_WRITEV (02) fd=4 iovs=0x7fffb7b2ffc0 nvecs=4 offset=0 expected_size=1792
io_uring_common.h:326: TPASS: OP=IORING_OP_READV (01) fd=4 iovs=0x7fffb7aaffc0 nvecs=4 offset=0 expected_size=1792
io_uring04.c:92: TPASS: Varying size vector I/O with zero-length buffer: data integrity verified across 4 vectors

.....
$

---

Sachin Sant (3):
  io_uring: Test IORING READ and WRITE operations
  io_uring: Test READV and WRITEV operations
  io_uring: Refactor io_uring01 to use common code

 runtest/syscalls                              |   2 +
 testcases/kernel/syscalls/io_uring/.gitignore |   2 +
 .../kernel/syscalls/io_uring/io_uring01.c     | 145 ++-------
 .../kernel/syscalls/io_uring/io_uring03.c     | 140 +++++++++
 .../kernel/syscalls/io_uring/io_uring04.c     | 216 ++++++++++++++
 .../syscalls/io_uring/io_uring_common.h       | 278 ++++++++++++++++++
 6 files changed, 660 insertions(+), 123 deletions(-)
 create mode 100644 testcases/kernel/syscalls/io_uring/io_uring03.c
 create mode 100644 testcases/kernel/syscalls/io_uring/io_uring04.c
 create mode 100644 testcases/kernel/syscalls/io_uring/io_uring_common.h

-- 
2.39.1


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

[LTP] [PATCH v2 1/3] io_uring: Test IORING READ and WRITE operations

[Sachin Sant]

This test validates basic read and write operations using io_uring.
It tests:
1. IORING_OP_WRITE - Writing data to a file
2. IORING_OP_READ - Reading data from a file
3. Data integrity verification

This patch also introduces a header file for common functions.

Signed-off-by: Sachin Sant <sachinp@linux.ibm.com>
---
 runtest/syscalls                              |   1 +
 testcases/kernel/syscalls/io_uring/.gitignore |   1 +
 .../kernel/syscalls/io_uring/io_uring03.c     | 140 +++++++++
 .../syscalls/io_uring/io_uring_common.h       | 278 ++++++++++++++++++
 4 files changed, 420 insertions(+)
 create mode 100644 testcases/kernel/syscalls/io_uring/io_uring03.c
 create mode 100644 testcases/kernel/syscalls/io_uring/io_uring_common.h

diff --git a/runtest/syscalls b/runtest/syscalls
index 2179e007c..7dc80fe29 100644
--- a/runtest/syscalls
+++ b/runtest/syscalls
@@ -1898,6 +1898,7 @@ membarrier01 membarrier01
 
 io_uring01 io_uring01
 io_uring02 io_uring02
+io_uring03 io_uring03
 
 # Tests below may cause kernel memory leak
 perf_event_open03 perf_event_open03
diff --git a/testcases/kernel/syscalls/io_uring/.gitignore b/testcases/kernel/syscalls/io_uring/.gitignore
index 749db17db..9382ae413 100644
--- a/testcases/kernel/syscalls/io_uring/.gitignore
+++ b/testcases/kernel/syscalls/io_uring/.gitignore
@@ -1,2 +1,3 @@
 /io_uring01
 /io_uring02
+/io_uring03
diff --git a/testcases/kernel/syscalls/io_uring/io_uring03.c b/testcases/kernel/syscalls/io_uring/io_uring03.c
new file mode 100644
index 000000000..645c96b02
--- /dev/null
+++ b/testcases/kernel/syscalls/io_uring/io_uring03.c
@@ -0,0 +1,140 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (C) 2026 IBM
+ * Author: Sachin Sant <sachinp@linux.ibm.com>
+ */
+/*
+ * Test IORING_OP_READ and IORING_OP_WRITE operations.
+ *
+ * This test validates basic read and write operations using io_uring.
+ * It tests:
+ * 1. IORING_OP_WRITE - Writing data to a file
+ * 2. IORING_OP_READ - Reading data from a file
+ * 3. Data integrity verification
+ */
+
+#include "io_uring_common.h"
+
+#define TEST_FILE "io_uring_test_file"
+#define QUEUE_DEPTH 2
+#define BLOCK_SZ 4096
+
+static char *write_buf;
+static char *read_buf;
+static struct io_uring_submit s;
+static sigset_t sig;
+
+static void init_buffer(char start_char)
+{
+	size_t i;
+
+	for (i = 0; i < BLOCK_SZ; i++)
+		write_buf[i] = start_char + (i % 26);
+}
+
+static void verify_data_integrity(const char *test_name)
+{
+	size_t i;
+
+	if (memcmp(write_buf, read_buf, BLOCK_SZ) == 0) {
+		tst_res(TPASS, "%s data integrity verified", test_name);
+	} else {
+		tst_res(TFAIL, "%s data mismatch", test_name);
+		for (i = 0; i < BLOCK_SZ && i < 64; i++) {
+			if (write_buf[i] != read_buf[i]) {
+				tst_res(TINFO, "First mismatch at offset %zu: "
+					"wrote 0x%02x, read 0x%02x",
+					i, write_buf[i], read_buf[i]);
+				break;
+			}
+		}
+	}
+}
+
+static void test_write_read(void)
+{
+	int fd;
+
+	init_buffer('A');
+
+	fd = SAFE_OPEN(TEST_FILE, O_RDWR | O_CREAT | O_TRUNC, 0644);
+
+	tst_res(TINFO, "Testing IORING_OP_WRITE");
+	io_uring_do_io_op(&s, fd, IORING_OP_WRITE, write_buf, BLOCK_SZ, 0,
+			  &sig);
+
+	SAFE_FSYNC(fd);
+
+	tst_res(TINFO, "Testing IORING_OP_READ");
+	memset(read_buf, 0, BLOCK_SZ);
+	io_uring_do_io_op(&s, fd, IORING_OP_READ, read_buf, BLOCK_SZ, 0,
+			  &sig);
+
+	verify_data_integrity("Basic I/O");
+
+	SAFE_CLOSE(fd);
+}
+
+static void test_partial_io(void)
+{
+	int fd;
+	size_t half = BLOCK_SZ / 2;
+
+	tst_res(TINFO, "Testing partial I/O operations");
+
+	init_buffer('a');
+
+	fd = SAFE_OPEN(TEST_FILE, O_RDWR | O_CREAT | O_TRUNC, 0644);
+
+	io_uring_do_io_op(&s, fd, IORING_OP_WRITE, write_buf, half, 0,
+			  &sig);
+
+	io_uring_do_io_op(&s, fd, IORING_OP_WRITE, write_buf + half, half,
+			  half, &sig);
+
+	SAFE_FSYNC(fd);
+
+	memset(read_buf, 0, BLOCK_SZ);
+	io_uring_do_io_op(&s, fd, IORING_OP_READ, read_buf, BLOCK_SZ, 0,
+			  &sig);
+
+	verify_data_integrity("Partial I/O");
+
+	SAFE_CLOSE(fd);
+}
+
+static void run(void)
+{
+	test_write_read();
+	test_partial_io();
+}
+
+static void setup(void)
+{
+	io_uring_setup_supported_by_kernel();
+	sigemptyset(&sig);
+	memset(&s, 0, sizeof(s));
+	io_uring_setup_queue(&s, QUEUE_DEPTH, 0);
+}
+
+static void cleanup(void)
+{
+	io_uring_cleanup_queue(&s, QUEUE_DEPTH);
+}
+
+static struct tst_test test = {
+	.test_all = run,
+	.setup = setup,
+	.cleanup = cleanup,
+	.needs_tmpdir = 1,
+	.bufs = (struct tst_buffers []) {
+		{&write_buf, .size = BLOCK_SZ},
+		{&read_buf, .size = BLOCK_SZ},
+		{}
+	},
+	.save_restore = (const struct tst_path_val[]) {
+		{"/proc/sys/kernel/io_uring_disabled", "0",
+			TST_SR_SKIP_MISSING | TST_SR_TCONF_RO},
+		{}
+	}
+};
diff --git a/testcases/kernel/syscalls/io_uring/io_uring_common.h b/testcases/kernel/syscalls/io_uring/io_uring_common.h
new file mode 100644
index 000000000..aa31339fb
--- /dev/null
+++ b/testcases/kernel/syscalls/io_uring/io_uring_common.h
@@ -0,0 +1,278 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (C) 2026 IBM
+ * Author: Sachin Sant <sachinp@linux.ibm.com>
+ *
+ * Common definitions and helper functions for io_uring tests
+ */
+
+#ifndef IO_URING_COMMON_H
+#define IO_URING_COMMON_H
+
+#include <stdlib.h>
+#include <string.h>
+#include <fcntl.h>
+#include "config.h"
+#include "tst_test.h"
+#include "lapi/io_uring.h"
+
+/* Common structures for io_uring ring management */
+struct io_sq_ring {
+	unsigned int *head;
+	unsigned int *tail;
+	unsigned int *ring_mask;
+	unsigned int *ring_entries;
+	unsigned int *flags;
+	unsigned int *array;
+};
+
+struct io_cq_ring {
+	unsigned int *head;
+	unsigned int *tail;
+	unsigned int *ring_mask;
+	unsigned int *ring_entries;
+	struct io_uring_cqe *cqes;
+};
+
+struct io_uring_submit {
+	int ring_fd;
+	struct io_sq_ring sq_ring;
+	struct io_uring_sqe *sqes;
+	struct io_cq_ring cq_ring;
+	void *sq_ptr;
+	size_t sq_ptr_size;
+	void *cq_ptr;
+	size_t cq_ptr_size;
+};
+
+/*
+ * Setup io_uring instance with specified queue depth and optional flags
+ * Returns 0 on success, -1 on failure
+ */
+static inline int io_uring_setup_queue(struct io_uring_submit *s,
+				       unsigned int queue_depth,
+				       unsigned int flags)
+{
+	struct io_sq_ring *sring = &s->sq_ring;
+	struct io_cq_ring *cring = &s->cq_ring;
+	struct io_uring_params p;
+
+	memset(&p, 0, sizeof(p));
+	p.flags = flags;
+	s->ring_fd = io_uring_setup(queue_depth, &p);
+	if (s->ring_fd < 0) {
+		tst_brk(TBROK | TERRNO, "io_uring_setup() failed");
+		return -1;
+	}
+
+	s->sq_ptr_size = p.sq_off.array + p.sq_entries * sizeof(unsigned int);
+
+	s->sq_ptr = SAFE_MMAP(0, s->sq_ptr_size, PROT_READ | PROT_WRITE,
+			      MAP_SHARED | MAP_POPULATE, s->ring_fd,
+			      IORING_OFF_SQ_RING);
+
+	/* Save submission queue pointers */
+	sring->head = s->sq_ptr + p.sq_off.head;
+	sring->tail = s->sq_ptr + p.sq_off.tail;
+	sring->ring_mask = s->sq_ptr + p.sq_off.ring_mask;
+	sring->ring_entries = s->sq_ptr + p.sq_off.ring_entries;
+	sring->flags = s->sq_ptr + p.sq_off.flags;
+	sring->array = s->sq_ptr + p.sq_off.array;
+
+	s->sqes = SAFE_MMAP(0, p.sq_entries * sizeof(struct io_uring_sqe),
+			    PROT_READ | PROT_WRITE, MAP_SHARED | MAP_POPULATE,
+			    s->ring_fd, IORING_OFF_SQES);
+
+	s->cq_ptr_size = p.cq_off.cqes +
+			 p.cq_entries * sizeof(struct io_uring_cqe);
+
+	s->cq_ptr = SAFE_MMAP(0, s->cq_ptr_size, PROT_READ | PROT_WRITE,
+			      MAP_SHARED | MAP_POPULATE, s->ring_fd,
+			      IORING_OFF_CQ_RING);
+
+	/* Save completion queue pointers */
+	cring->head = s->cq_ptr + p.cq_off.head;
+	cring->tail = s->cq_ptr + p.cq_off.tail;
+	cring->ring_mask = s->cq_ptr + p.cq_off.ring_mask;
+	cring->ring_entries = s->cq_ptr + p.cq_off.ring_entries;
+	cring->cqes = s->cq_ptr + p.cq_off.cqes;
+
+	return 0;
+}
+
+/*
+ * Cleanup io_uring instance and unmap all memory regions
+ */
+static inline void io_uring_cleanup_queue(struct io_uring_submit *s,
+					  unsigned int queue_depth)
+{
+	if (s->sqes)
+		SAFE_MUNMAP(s->sqes, queue_depth * sizeof(struct io_uring_sqe));
+	if (s->cq_ptr)
+		SAFE_MUNMAP(s->cq_ptr, s->cq_ptr_size);
+	if (s->sq_ptr)
+		SAFE_MUNMAP(s->sq_ptr, s->sq_ptr_size);
+	if (s->ring_fd > 0)
+		SAFE_CLOSE(s->ring_fd);
+}
+
+/*
+ * Internal helper to submit a single SQE to the submission queue
+ * Used by both vectored and non-vectored I/O operations
+ */
+static inline void io_uring_submit_sqe_internal(struct io_uring_submit *s,
+						int fd, int opcode,
+						unsigned long addr,
+						unsigned int len,
+						off_t offset,
+						unsigned long long user_data)
+{
+	struct io_sq_ring *sring = &s->sq_ring;
+	unsigned int tail, index;
+	struct io_uring_sqe *sqe;
+
+	tail = *sring->tail;
+	index = tail & *sring->ring_mask;
+	sqe = &s->sqes[index];
+
+	memset(sqe, 0, sizeof(*sqe));
+	sqe->opcode = opcode;
+	sqe->fd = fd;
+	sqe->addr = addr;
+	sqe->len = len;
+	sqe->off = offset;
+	sqe->user_data = user_data;
+
+	sring->array[index] = index;
+	tail++;
+
+	*sring->tail = tail;
+}
+
+/*
+ * Submit a single SQE to the submission queue
+ * For basic read/write operations (non-vectored)
+ */
+static inline void io_uring_submit_sqe(struct io_uring_submit *s, int fd,
+				       int opcode, void *buf, size_t len,
+				       off_t offset)
+{
+	io_uring_submit_sqe_internal(s, fd, opcode, (unsigned long)buf,
+				     len, offset, opcode);
+}
+
+/*
+ * Submit a vectored SQE to the submission queue
+ * For readv/writev operations
+ */
+static inline void io_uring_submit_sqe_vec(struct io_uring_submit *s, int fd,
+					   int opcode, struct iovec *iovs,
+					   int nr_vecs, off_t offset)
+{
+	io_uring_submit_sqe_internal(s, fd, opcode, (unsigned long)iovs,
+				     nr_vecs, offset, opcode);
+}
+
+/*
+ * Map io_uring operation code to human-readable name
+ */
+static inline const char *ioring_op_name(int op)
+{
+	switch (op) {
+	case IORING_OP_READV:
+		return "IORING_OP_READV";
+	case IORING_OP_WRITEV:
+		return "IORING_OP_WRITEV";
+	case IORING_OP_READ:
+		return "IORING_OP_READ";
+	case IORING_OP_WRITE:
+		return "IORING_OP_WRITE";
+	default:
+		return "UNKNOWN";
+	}
+}
+
+/*
+ * Wait for and validate a completion queue entry
+ * Aborts test on failure using tst_brk()
+ */
+static inline void io_uring_wait_cqe(struct io_uring_submit *s,
+				     int expected_res, int expected_opcode,
+				     sigset_t *sig)
+{
+	struct io_cq_ring *cring = &s->cq_ring;
+	struct io_uring_cqe *cqe;
+	unsigned int head;
+	int ret;
+
+	ret = io_uring_enter(s->ring_fd, 1, 1, IORING_ENTER_GETEVENTS, sig);
+	if (ret < 0)
+		tst_brk(TBROK | TERRNO, "io_uring_enter() failed");
+
+	head = *cring->head;
+	if (head == *cring->tail)
+		tst_brk(TBROK, "No completion event received");
+
+	cqe = &cring->cqes[head & *cring->ring_mask];
+
+	if (cqe->user_data != (uint64_t)expected_opcode) {
+		*cring->head = head + 1;
+		tst_brk(TBROK, "Unexpected user_data: got %llu, expected %d",
+			(unsigned long long)cqe->user_data, expected_opcode);
+	}
+
+	if (cqe->res != expected_res) {
+		*cring->head = head + 1;
+		tst_brk(TBROK, "Operation failed: res=%d, expected=%d",
+			cqe->res, expected_res);
+	}
+
+	*cring->head = head + 1;
+}
+
+/*
+ * Initialize buffer with a repeating character pattern
+ * Useful for creating test data with predictable patterns
+ */
+static inline void io_uring_init_buffer_pattern(char *buf, size_t size,
+						char pattern)
+{
+	size_t i;
+
+	for (i = 0; i < size; i++)
+		buf[i] = pattern;
+}
+
+/*
+ * Submit and wait for a non-vectored I/O operation
+ * Combines io_uring_submit_sqe() and io_uring_wait_cqe() with result reporting
+ */
+static inline void io_uring_do_io_op(struct io_uring_submit *s, int fd,
+				     int op, void *buf, size_t len,
+				     off_t offset, sigset_t *sig)
+{
+	io_uring_submit_sqe(s, fd, op, buf, len, offset);
+	io_uring_wait_cqe(s, len, op, sig);
+	tst_res(TPASS, "OP=%s (%02x) fd=%i buf=%p len=%zu offset=%jd",
+		ioring_op_name(op), op, fd, buf, len, (intmax_t)offset);
+}
+
+/*
+ * Submit and wait for a vectored I/O operation
+ * Combines io_uring_submit_sqe_vec() and io_uring_wait_cqe() with
+ * result reporting
+ */
+static inline void io_uring_do_vec_io_op(struct io_uring_submit *s, int fd,
+					 int op, struct iovec *iovs,
+					 int nvecs, off_t offset,
+					 int expected_size, sigset_t *sig)
+{
+	io_uring_submit_sqe_vec(s, fd, op, iovs, nvecs, offset);
+	io_uring_wait_cqe(s, expected_size, op, sig);
+	tst_res(TPASS, "OP=%s (%02x) fd=%i iovs=%p nvecs=%i offset=%jd "
+		"expected_size=%i",
+		ioring_op_name(op), op, fd, iovs, nvecs, (intmax_t)offset,
+		expected_size);
+}
+
+#endif /* IO_URING_COMMON_H */
-- 
2.39.1


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

[LTP] [PATCH v2 2/3] io_uring: Test READV and WRITEV operations

[Sachin Sant]

This test validates vectored read and write operations using io_uring.

1. IORING_OP_WRITEV - Writing data using multiple buffers (scatter)
2. IORING_OP_READV - Reading data into multiple buffers (gather)
3. Data integrity verification across multiple iovecs
4. Edge cases with different iovec configurations including zero
   buffer length

Signed-off-by: Sachin Sant <sachinp@linux.ibm.com>
---
 runtest/syscalls                              |   1 +
 testcases/kernel/syscalls/io_uring/.gitignore |   1 +
 .../kernel/syscalls/io_uring/io_uring04.c     | 216 ++++++++++++++++++
 3 files changed, 218 insertions(+)
 create mode 100644 testcases/kernel/syscalls/io_uring/io_uring04.c

diff --git a/runtest/syscalls b/runtest/syscalls
index 7dc80fe29..eacf946c5 100644
--- a/runtest/syscalls
+++ b/runtest/syscalls
@@ -1899,6 +1899,7 @@ membarrier01 membarrier01
 io_uring01 io_uring01
 io_uring02 io_uring02
 io_uring03 io_uring03
+io_uring04 io_uring04
 
 # Tests below may cause kernel memory leak
 perf_event_open03 perf_event_open03
diff --git a/testcases/kernel/syscalls/io_uring/.gitignore b/testcases/kernel/syscalls/io_uring/.gitignore
index 9382ae413..36cd24662 100644
--- a/testcases/kernel/syscalls/io_uring/.gitignore
+++ b/testcases/kernel/syscalls/io_uring/.gitignore
@@ -1,3 +1,4 @@
 /io_uring01
 /io_uring02
 /io_uring03
+/io_uring04
diff --git a/testcases/kernel/syscalls/io_uring/io_uring04.c b/testcases/kernel/syscalls/io_uring/io_uring04.c
new file mode 100644
index 000000000..48479df1f
--- /dev/null
+++ b/testcases/kernel/syscalls/io_uring/io_uring04.c
@@ -0,0 +1,216 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (C) 2026 IBM
+ * Author: Sachin Sant <sachinp@linux.ibm.com>
+ */
+/*
+ * Test IORING_OP_READV and IORING_OP_WRITEV operations.
+ *
+ * This test validates vectored read and write operations using io_uring.
+ * It tests:
+ * 1. IORING_OP_WRITEV - Writing data using multiple buffers (scatter)
+ * 2. IORING_OP_READV - Reading data into multiple buffers (gather)
+ * 3. Data integrity verification across multiple iovecs
+ * 4. Edge cases with different iovec configurations
+ */
+
+#include "io_uring_common.h"
+
+#define TEST_FILE "io_uring_test_file"
+#define QUEUE_DEPTH 2
+#define NUM_VECS 4
+#define VEC_SIZE 1024
+#define VAR_BUF1_SIZE 512
+#define VAR_BUF2_SIZE 1024
+#define VAR_BUF3_SIZE 256
+
+static struct iovec *write_iovs, *read_iovs;
+static struct iovec *var_write_iovs, *var_read_iovs;
+static struct io_uring_submit s;
+static sigset_t sig;
+
+static void prepare_write_buffers(void)
+{
+	size_t i, j;
+	char *buf;
+
+	for (i = 0; i < NUM_VECS; i++) {
+		buf = (char *)write_iovs[i].iov_base;
+		for (j = 0; j < write_iovs[i].iov_len; j++) {
+			/* Each vector has a different pattern */
+			buf[j] = 'A' + i + (j % 26);
+		}
+	}
+}
+
+static void prepare_read_buffers(void)
+{
+	size_t i;
+
+	for (i = 0; i < NUM_VECS; i++)
+		memset(read_iovs[i].iov_base, 0, read_iovs[i].iov_len);
+}
+
+static void clear_iovec_buffers(struct iovec *iovs, int nvecs)
+{
+	int i;
+
+	for (i = 0; i < nvecs; i++)
+		memset(iovs[i].iov_base, 0, iovs[i].iov_len);
+}
+
+static void verify_iovec_data(struct iovec *write_iovs, struct iovec *read_iovs,
+			      int nvecs, const char *test_name)
+{
+	int i;
+	size_t j;
+
+	for (i = 0; i < nvecs; i++) {
+		if (write_iovs[i].iov_len != read_iovs[i].iov_len) {
+			tst_res(TFAIL, "%s: iovec %d length mismatch: write=%zu read=%zu",
+				test_name, i, write_iovs[i].iov_len, read_iovs[i].iov_len);
+			return;
+		}
+
+		if (memcmp(write_iovs[i].iov_base, read_iovs[i].iov_base,
+			   write_iovs[i].iov_len) != 0) {
+			tst_res(TFAIL, "%s: data mismatch in vector %d", test_name, i);
+			for (j = 0; j < write_iovs[i].iov_len && j < 64; j++) {
+				char *wbuf = (char *)write_iovs[i].iov_base;
+				char *rbuf = (char *)read_iovs[i].iov_base;
+				if (wbuf[j] != rbuf[j]) {
+					tst_res(TINFO, "Vector %d: first mismatch at "
+						"offset %zu: wrote 0x%02x, read 0x%02x",
+						i, j, wbuf[j], rbuf[j]);
+					break;
+				}
+			}
+			return;
+		}
+	}
+
+	tst_res(TPASS, "%s: data integrity verified across %d vectors",
+		test_name, nvecs);
+}
+
+
+static void test_writev_readv(void)
+{
+	int fd;
+	int total_size = NUM_VECS * VEC_SIZE;
+
+	tst_res(TINFO, "Testing IORING_OP_WRITEV and IORING_OP_READV");
+
+	fd = SAFE_OPEN(TEST_FILE, O_RDWR | O_CREAT | O_TRUNC, 0644);
+
+	tst_res(TINFO, "Writing %d bytes using %d vectors", total_size, NUM_VECS);
+	io_uring_do_vec_io_op(&s, fd, IORING_OP_WRITEV, write_iovs, NUM_VECS,
+			      0, total_size, &sig);
+
+	SAFE_FSYNC(fd);
+
+	tst_res(TINFO, "Reading %d bytes using %d vectors", total_size, NUM_VECS);
+	io_uring_do_vec_io_op(&s, fd, IORING_OP_READV, read_iovs, NUM_VECS,
+			      0, total_size, &sig);
+
+	verify_iovec_data(write_iovs, read_iovs, NUM_VECS, "Basic vectored I/O");
+
+	SAFE_CLOSE(fd);
+}
+
+static void test_partial_vectors(void)
+{
+	int fd;
+	int half_size = 2 * VEC_SIZE;
+
+	tst_res(TINFO, "Testing partial vector operations");
+
+	fd = SAFE_OPEN(TEST_FILE, O_RDWR | O_CREAT | O_TRUNC, 0644);
+
+	/* Write first half using first 2 vectors at offset 0 */
+	io_uring_do_vec_io_op(&s, fd, IORING_OP_WRITEV, write_iovs, 2, 0,
+			      half_size, &sig);
+
+	/* Write second half using next 2 vectors at offset half_size */
+	io_uring_do_vec_io_op(&s, fd, IORING_OP_WRITEV, &write_iovs[2], 2,
+			      half_size, half_size, &sig);
+
+	SAFE_FSYNC(fd);
+
+	/* Read back entire file using all 4 vectors */
+	io_uring_do_vec_io_op(&s, fd, IORING_OP_READV, read_iovs, NUM_VECS, 0,
+			      NUM_VECS * VEC_SIZE, &sig);
+
+	verify_iovec_data(write_iovs, read_iovs, NUM_VECS, "Partial vector I/O");
+
+	SAFE_CLOSE(fd);
+}
+
+static void test_varying_sizes(void)
+{
+	int fd;
+	int expected_size = VAR_BUF1_SIZE + VAR_BUF2_SIZE + VAR_BUF3_SIZE;
+
+	tst_res(TINFO, "Testing vectors with varying sizes including zero-length buffer");
+
+	io_uring_init_buffer_pattern(var_write_iovs[0].iov_base, VAR_BUF1_SIZE, 'X');
+	io_uring_init_buffer_pattern(var_write_iovs[2].iov_base, VAR_BUF2_SIZE, 'Y');
+	io_uring_init_buffer_pattern(var_write_iovs[3].iov_base, VAR_BUF3_SIZE, 'Z');
+
+	clear_iovec_buffers(var_read_iovs, 4);
+
+	fd = SAFE_OPEN(TEST_FILE, O_RDWR | O_CREAT | O_TRUNC, 0644);
+
+	io_uring_do_vec_io_op(&s, fd, IORING_OP_WRITEV, var_write_iovs, 4, 0,
+			      expected_size, &sig);
+
+	SAFE_FSYNC(fd);
+
+	io_uring_do_vec_io_op(&s, fd, IORING_OP_READV, var_read_iovs, 4, 0,
+			      expected_size, &sig);
+
+	verify_iovec_data(var_write_iovs, var_read_iovs, 4, "Varying size vector I/O with zero-length buffer");
+
+	SAFE_CLOSE(fd);
+}
+
+static void run(void)
+{
+	test_writev_readv();
+	test_partial_vectors();
+	test_varying_sizes();
+}
+
+static void setup(void)
+{
+	io_uring_setup_supported_by_kernel();
+	sigemptyset(&sig);
+	memset(&s, 0, sizeof(s));
+	io_uring_setup_queue(&s, QUEUE_DEPTH, 0);
+	prepare_write_buffers();
+	prepare_read_buffers();
+}
+
+static void cleanup(void)
+{
+	io_uring_cleanup_queue(&s, QUEUE_DEPTH);
+}
+
+static struct tst_test test = {
+	.test_all = run,
+	.setup = setup,
+	.cleanup = cleanup,
+	.needs_tmpdir = 1,
+	.bufs = (struct tst_buffers []) {
+		{&write_iovs, .iov_sizes = (int[]){VEC_SIZE, VEC_SIZE, VEC_SIZE, VEC_SIZE, -1}},
+		{&read_iovs, .iov_sizes = (int[]){VEC_SIZE, VEC_SIZE, VEC_SIZE, VEC_SIZE, -1}},
+		{&var_write_iovs, .iov_sizes = (int[]){VAR_BUF1_SIZE, 0, VAR_BUF2_SIZE, VAR_BUF3_SIZE, -1}},
+		{&var_read_iovs, .iov_sizes = (int[]){VAR_BUF1_SIZE, 0, VAR_BUF2_SIZE, VAR_BUF3_SIZE, -1}},
+		{}
+	},
+	.save_restore = (const struct tst_path_val[]) {
+		{"/proc/sys/kernel/io_uring_disabled", "0",
+			TST_SR_SKIP_MISSING | TST_SR_TCONF_RO},
+		{}
+	}
+};
-- 
2.39.1


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

[LTP] [PATCH v2 3/3] io_uring: Refactor io_uring01 to use common code

[Sachin Sant]

No functional change.

Refactor io_uring01 test case to use
- common definitions from io_uring_common.h
- remove duplicate structure definitions
- Replace manual munmap/close calls with io_uring_cleanup_queue()
  helper function
- Submit SQE using common helper

Signed-off-by: Sachin Sant <sachinp@linux.ibm.com>
---
 .../kernel/syscalls/io_uring/io_uring01.c     | 145 +++---------------
 1 file changed, 22 insertions(+), 123 deletions(-)

diff --git a/testcases/kernel/syscalls/io_uring/io_uring01.c b/testcases/kernel/syscalls/io_uring/io_uring01.c
index ab1ec00d6..4c64e562b 100644
--- a/testcases/kernel/syscalls/io_uring/io_uring01.c
+++ b/testcases/kernel/syscalls/io_uring/io_uring01.c
@@ -11,13 +11,7 @@
  * registered in the kernel for long term operation using io_uring_register().
  * This tests initiates I/O operations with the help of io_uring_enter().
  */
-#include <stdlib.h>
-#include <errno.h>
-#include <string.h>
-#include <fcntl.h>
-#include "config.h"
-#include "tst_test.h"
-#include "lapi/io_uring.h"
+#include "io_uring_common.h"
 
 #define TEST_FILE "test_file"
 
@@ -32,96 +26,19 @@ static struct tcase {
 	{0, IORING_REGISTER_BUFFERS, IORING_OP_READ_FIXED},
 };
 
-struct io_sq_ring {
-	unsigned int *head;
-	unsigned int *tail;
-	unsigned int *ring_mask;
-	unsigned int *ring_entries;
-	unsigned int *flags;
-	unsigned int *array;
-};
-
-struct io_cq_ring {
-	unsigned int *head;
-	unsigned int *tail;
-	unsigned int *ring_mask;
-	unsigned int *ring_entries;
-	struct io_uring_cqe *cqes;
-};
-
-struct submitter {
-	int ring_fd;
-	struct io_sq_ring sq_ring;
-	struct io_uring_sqe *sqes;
-	struct io_cq_ring cq_ring;
-};
-
-static struct submitter sub_ring;
-static struct submitter *s = &sub_ring;
+static struct io_uring_submit s;
 static sigset_t sig;
 static struct iovec *iov;
 
-
-static void *sptr;
-static size_t sptr_size;
-static void *cptr;
-static size_t cptr_size;
-
-static int setup_io_uring_test(struct submitter *s, struct tcase *tc)
+static int setup_io_uring_test(struct io_uring_submit *s, struct tcase *tc)
 {
-	struct io_sq_ring *sring = &s->sq_ring;
-	struct io_cq_ring *cring = &s->cq_ring;
-	struct io_uring_params p;
+	int ret;
 
-	memset(&p, 0, sizeof(p));
-	p.flags |= tc->setup_flags;
-	s->ring_fd = io_uring_setup(QUEUE_DEPTH, &p);
-	if (s->ring_fd != -1) {
+	ret = io_uring_setup_queue(s, QUEUE_DEPTH, tc->setup_flags);
+	if (ret == 0)
 		tst_res(TPASS, "io_uring_setup() passed");
-	} else {
-		tst_res(TFAIL | TERRNO, "io_uring_setup() failed");
-		return 1;
-	}
-
-	sptr_size = p.sq_off.array + p.sq_entries * sizeof(unsigned int);
-
-	/* Submission queue ring buffer mapping */
-	sptr = SAFE_MMAP(0, sptr_size,
-			PROT_READ | PROT_WRITE,
-			MAP_SHARED | MAP_POPULATE,
-			s->ring_fd, IORING_OFF_SQ_RING);
-
-	/* Save global submission queue struct info */
-	sring->head = sptr + p.sq_off.head;
-	sring->tail = sptr + p.sq_off.tail;
-	sring->ring_mask = sptr + p.sq_off.ring_mask;
-	sring->ring_entries = sptr + p.sq_off.ring_entries;
-	sring->flags = sptr + p.sq_off.flags;
-	sring->array = sptr + p.sq_off.array;
-
-	/* Submission queue entries ring buffer mapping */
-	s->sqes = SAFE_MMAP(0, p.sq_entries *
-			sizeof(struct io_uring_sqe),
-			PROT_READ | PROT_WRITE,
-			MAP_SHARED | MAP_POPULATE,
-			s->ring_fd, IORING_OFF_SQES);
-
-	cptr_size = p.cq_off.cqes + p.cq_entries * sizeof(struct io_uring_cqe);
-
-	/* Completion queue ring buffer mapping */
-	cptr = SAFE_MMAP(0, cptr_size,
-			PROT_READ | PROT_WRITE,
-			MAP_SHARED | MAP_POPULATE,
-			s->ring_fd, IORING_OFF_CQ_RING);
-
-	/* Save global completion queue struct info */
-	cring->head = cptr + p.cq_off.head;
-	cring->tail = cptr + p.cq_off.tail;
-	cring->ring_mask = cptr + p.cq_off.ring_mask;
-	cring->ring_entries = cptr + p.cq_off.ring_entries;
-	cring->cqes = cptr + p.cq_off.cqes;
 
-	return 0;
+	return ret;
 }
 
 static void check_buffer(char *buffer, size_t len)
@@ -139,7 +56,7 @@ static void check_buffer(char *buffer, size_t len)
 		tst_res(TPASS, "Buffer filled in correctly");
 }
 
-static void drain_uring_cq(struct submitter *s, unsigned int exp_events)
+static void drain_uring_cq(struct io_uring_submit *s, unsigned int exp_events)
 {
 	struct io_cq_ring *cring = &s->cq_ring;
 	unsigned int head = *cring->head;
@@ -175,12 +92,10 @@ static void drain_uring_cq(struct submitter *s, unsigned int exp_events)
 	        events, exp_events);
 }
 
-static int submit_to_uring_sq(struct submitter *s, struct tcase *tc)
+static int submit_to_uring_sq(struct io_uring_submit *s, struct tcase *tc)
 {
-	unsigned int index = 0, tail = 0, next_tail = 0;
-	struct io_sq_ring *sring = &s->sq_ring;
-	struct io_uring_sqe *sqe;
 	int ret;
+	int fd;
 
 	memset(iov->iov_base, 0, iov->iov_len);
 
@@ -193,26 +108,13 @@ static int submit_to_uring_sq(struct submitter *s, struct tcase *tc)
 		return 1;
 	}
 
-	int fd = SAFE_OPEN(TEST_FILE, O_RDONLY);
-
-	/* Submission queue entry addition to SQE ring buffer tail */
-	tail = *sring->tail;
-	next_tail = tail + 1;
-	index = tail & *s->sq_ring.ring_mask;
-	sqe = &s->sqes[index];
-	sqe->flags = 0;
-	sqe->fd = fd;
-	sqe->opcode = tc->enter_flags;
-	sqe->addr = (unsigned long)iov->iov_base;
-	sqe->len = BLOCK_SZ;
-	sqe->off = 0;
-	sqe->user_data = (unsigned long long)iov;
-	sring->array[index] = index;
-	tail = next_tail;
-
-	/* Kernel to notice the tail update */
-	if (*sring->tail != tail)
-		*sring->tail = tail;
+	fd = SAFE_OPEN(TEST_FILE, O_RDONLY);
+
+	/* Submit SQE using common helper */
+	io_uring_submit_sqe_internal(s, fd, tc->enter_flags,
+				     (unsigned long)iov->iov_base,
+				     BLOCK_SZ, 0,
+				     (unsigned long long)iov);
 
 	ret = io_uring_enter(s->ring_fd, 1, 1, IORING_ENTER_GETEVENTS, &sig);
 	if (ret == 1) {
@@ -229,23 +131,20 @@ static int submit_to_uring_sq(struct submitter *s, struct tcase *tc)
 
 static void cleanup_io_uring_test(void)
 {
-	io_uring_register(s->ring_fd, IORING_UNREGISTER_BUFFERS,
+	io_uring_register(s.ring_fd, IORING_UNREGISTER_BUFFERS,
 			  NULL, QUEUE_DEPTH);
-	SAFE_MUNMAP(s->sqes, sizeof(struct io_uring_sqe));
-	SAFE_MUNMAP(cptr, cptr_size);
-	SAFE_MUNMAP(sptr, sptr_size);
-	SAFE_CLOSE(s->ring_fd);
+	io_uring_cleanup_queue(&s, QUEUE_DEPTH);
 }
 
 static void run(unsigned int n)
 {
 	struct tcase *tc = &tcases[n];
 
-	if (setup_io_uring_test(s, tc))
+	if (setup_io_uring_test(&s, tc))
 		return;
 
-	if (!submit_to_uring_sq(s, tc))
-		drain_uring_cq(s, 1);
+	if (!submit_to_uring_sq(&s, tc))
+		drain_uring_cq(&s, 1);
 
 	cleanup_io_uring_test();
 }
-- 
2.39.1


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH v2 0/3] io_uring READ(V), WRITE(v) operation tests

[Cyril Hrubis]

Hi!
I've merged 1/3 and 3/3 since these two patches were ready, thanks. I
will add a few minor comments for 2/3.

-- 
Cyril Hrubis
chrubis@suse.cz

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH v2 2/3] io_uring: Test READV and WRITEV operations

[Cyril Hrubis]

Hi!
> +#include "io_uring_common.h"
> +
> +#define TEST_FILE "io_uring_test_file"
> +#define QUEUE_DEPTH 2
> +#define NUM_VECS 4
> +#define VEC_SIZE 1024
> +#define VAR_BUF1_SIZE 512
> +#define VAR_BUF2_SIZE 1024
> +#define VAR_BUF3_SIZE 256
> +
> +static struct iovec *write_iovs, *read_iovs;
> +static struct iovec *var_write_iovs, *var_read_iovs;
> +static struct io_uring_submit s;
> +static sigset_t sig;
> +
> +static void prepare_write_buffers(void)
> +{
> +	size_t i, j;
> +	char *buf;
> +
> +	for (i = 0; i < NUM_VECS; i++) {
> +		buf = (char *)write_iovs[i].iov_base;
> +		for (j = 0; j < write_iovs[i].iov_len; j++) {
> +			/* Each vector has a different pattern */
> +			buf[j] = 'A' + i + (j % 26);
> +		}
> +	}
> +}
> +
> +static void prepare_read_buffers(void)
> +{
> +	size_t i;
> +
> +	for (i = 0; i < NUM_VECS; i++)
> +		memset(read_iovs[i].iov_base, 0, read_iovs[i].iov_len);
> +}
> +
> +static void clear_iovec_buffers(struct iovec *iovs, int nvecs)
> +{
> +	int i;
> +
> +	for (i = 0; i < nvecs; i++)
> +		memset(iovs[i].iov_base, 0, iovs[i].iov_len);
> +}

These two functions are nearly identical, we really need only one
function to clear the buffers.

> +static void verify_iovec_data(struct iovec *write_iovs, struct iovec *read_iovs,
> +			      int nvecs, const char *test_name)
> +{
> +	int i;
> +	size_t j;
> +
> +	for (i = 0; i < nvecs; i++) {
> +		if (write_iovs[i].iov_len != read_iovs[i].iov_len) {
> +			tst_res(TFAIL, "%s: iovec %d length mismatch: write=%zu read=%zu",
> +				test_name, i, write_iovs[i].iov_len, read_iovs[i].iov_len);
> +			return;
> +		}
> +
> +		if (memcmp(write_iovs[i].iov_base, read_iovs[i].iov_base,
> +			   write_iovs[i].iov_len) != 0) {
> +			tst_res(TFAIL, "%s: data mismatch in vector %d", test_name, i);
> +			for (j = 0; j < write_iovs[i].iov_len && j < 64; j++) {
> +				char *wbuf = (char *)write_iovs[i].iov_base;
> +				char *rbuf = (char *)read_iovs[i].iov_base;
> +				if (wbuf[j] != rbuf[j]) {
> +					tst_res(TINFO, "Vector %d: first mismatch at "
> +						"offset %zu: wrote 0x%02x, read 0x%02x",
> +						i, j, wbuf[j], rbuf[j]);
> +					break;
> +				}
> +			}
> +			return;
> +		}
> +	}
> +
> +	tst_res(TPASS, "%s: data integrity verified across %d vectors",
> +		test_name, nvecs);
> +}
> +
> +
> +static void test_writev_readv(void)
> +{
> +	int fd;
> +	int total_size = NUM_VECS * VEC_SIZE;
> +
> +	tst_res(TINFO, "Testing IORING_OP_WRITEV and IORING_OP_READV");
> +
> +	fd = SAFE_OPEN(TEST_FILE, O_RDWR | O_CREAT | O_TRUNC, 0644);
> +
> +	tst_res(TINFO, "Writing %d bytes using %d vectors", total_size, NUM_VECS);
> +	io_uring_do_vec_io_op(&s, fd, IORING_OP_WRITEV, write_iovs, NUM_VECS,
> +			      0, total_size, &sig);
> +
> +	SAFE_FSYNC(fd);
> +
> +	tst_res(TINFO, "Reading %d bytes using %d vectors", total_size, NUM_VECS);
> +	io_uring_do_vec_io_op(&s, fd, IORING_OP_READV, read_iovs, NUM_VECS,
> +			      0, total_size, &sig);
> +
> +	verify_iovec_data(write_iovs, read_iovs, NUM_VECS, "Basic vectored I/O");
> +
> +	SAFE_CLOSE(fd);
> +}
> +
> +static void test_partial_vectors(void)
> +{
> +	int fd;
> +	int half_size = 2 * VEC_SIZE;
> +
> +	tst_res(TINFO, "Testing partial vector operations");
> +
> +	fd = SAFE_OPEN(TEST_FILE, O_RDWR | O_CREAT | O_TRUNC, 0644);
> +
> +	/* Write first half using first 2 vectors at offset 0 */
> +	io_uring_do_vec_io_op(&s, fd, IORING_OP_WRITEV, write_iovs, 2, 0,
> +			      half_size, &sig);
> +
> +	/* Write second half using next 2 vectors at offset half_size */
> +	io_uring_do_vec_io_op(&s, fd, IORING_OP_WRITEV, &write_iovs[2], 2,
> +			      half_size, half_size, &sig);
> +
> +	SAFE_FSYNC(fd);
> +
> +	/* Read back entire file using all 4 vectors */
> +	io_uring_do_vec_io_op(&s, fd, IORING_OP_READV, read_iovs, NUM_VECS, 0,
> +			      NUM_VECS * VEC_SIZE, &sig);
> +
> +	verify_iovec_data(write_iovs, read_iovs, NUM_VECS, "Partial vector I/O");
> +
> +	SAFE_CLOSE(fd);
> +}
> +
> +static void test_varying_sizes(void)
> +{
> +	int fd;
> +	int expected_size = VAR_BUF1_SIZE + VAR_BUF2_SIZE + VAR_BUF3_SIZE;
> +
> +	tst_res(TINFO, "Testing vectors with varying sizes including zero-length buffer");
> +
> +	io_uring_init_buffer_pattern(var_write_iovs[0].iov_base, VAR_BUF1_SIZE, 'X');
> +	io_uring_init_buffer_pattern(var_write_iovs[2].iov_base, VAR_BUF2_SIZE, 'Y');
> +	io_uring_init_buffer_pattern(var_write_iovs[3].iov_base, VAR_BUF3_SIZE, 'Z');

Can we handle this with the prepare_write_buffers() as well?

> +	clear_iovec_buffers(var_read_iovs, 4);
> +
> +	fd = SAFE_OPEN(TEST_FILE, O_RDWR | O_CREAT | O_TRUNC, 0644);
> +
> +	io_uring_do_vec_io_op(&s, fd, IORING_OP_WRITEV, var_write_iovs, 4, 0,
> +			      expected_size, &sig);
> +
> +	SAFE_FSYNC(fd);
> +
> +	io_uring_do_vec_io_op(&s, fd, IORING_OP_READV, var_read_iovs, 4, 0,
> +			      expected_size, &sig);
> +
> +	verify_iovec_data(var_write_iovs, var_read_iovs, 4, "Varying size vector I/O with zero-length buffer");
> +
> +	SAFE_CLOSE(fd);
> +}
> +
> +static void run(void)
> +{
> +	test_writev_readv();
> +	test_partial_vectors();
> +	test_varying_sizes();
> +}
> +
> +static void setup(void)
> +{
> +	io_uring_setup_supported_by_kernel();
> +	sigemptyset(&sig);
> +	memset(&s, 0, sizeof(s));
> +	io_uring_setup_queue(&s, QUEUE_DEPTH, 0);
> +	prepare_write_buffers();
> +	prepare_read_buffers();
> +}
> +
> +static void cleanup(void)
> +{
> +	io_uring_cleanup_queue(&s, QUEUE_DEPTH);
> +}
> +
> +static struct tst_test test = {
> +	.test_all = run,
> +	.setup = setup,
> +	.cleanup = cleanup,
> +	.needs_tmpdir = 1,
> +	.bufs = (struct tst_buffers []) {
> +		{&write_iovs, .iov_sizes = (int[]){VEC_SIZE, VEC_SIZE, VEC_SIZE, VEC_SIZE, -1}},
> +		{&read_iovs, .iov_sizes = (int[]){VEC_SIZE, VEC_SIZE, VEC_SIZE, VEC_SIZE, -1}},
> +		{&var_write_iovs, .iov_sizes = (int[]){VAR_BUF1_SIZE, 0, VAR_BUF2_SIZE, VAR_BUF3_SIZE, -1}},
> +		{&var_read_iovs, .iov_sizes = (int[]){VAR_BUF1_SIZE, 0, VAR_BUF2_SIZE, VAR_BUF3_SIZE, -1}},
> +		{}
> +	},
> +	.save_restore = (const struct tst_path_val[]) {
> +		{"/proc/sys/kernel/io_uring_disabled", "0",
> +			TST_SR_SKIP_MISSING | TST_SR_TCONF_RO},
> +		{}
> +	}
> +};
> -- 
> 2.39.1
> 

-- 
Cyril Hrubis
chrubis@suse.cz

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH v2 2/3] io_uring: Test READV and WRITEV operations

[Sachin Sant]

>> +static void prepare_read_buffers(void)
>> +{
>> +	size_t i;
>> +
>> +	for (i = 0; i < NUM_VECS; i++)
>> +		memset(read_iovs[i].iov_base, 0, read_iovs[i].iov_len);
>> +}
>> +
>> +static void clear_iovec_buffers(struct iovec *iovs, int nvecs)
>> +{
>> +	int i;
>> +
>> +	for (i = 0; i < nvecs; i++)
>> +		memset(iovs[i].iov_base, 0, iovs[i].iov_len);
>> +}
> These two functions are nearly identical, we really need only one
> function to clear the buffers.
>
Indeed. Good catch. Will merge them.

>> +static void test_varying_sizes(void)
>> +{
>> +	int fd;
>> +	int expected_size = VAR_BUF1_SIZE + VAR_BUF2_SIZE + VAR_BUF3_SIZE;
>> +
>> +	tst_res(TINFO, "Testing vectors with varying sizes including zero-length buffer");
>> +
>> +	io_uring_init_buffer_pattern(var_write_iovs[0].iov_base, VAR_BUF1_SIZE, 'X');
>> +	io_uring_init_buffer_pattern(var_write_iovs[2].iov_base, VAR_BUF2_SIZE, 'Y');
>> +	io_uring_init_buffer_pattern(var_write_iovs[3].iov_base, VAR_BUF3_SIZE, 'Z');
> Can we handle this with the prepare_write_buffers() as well?
These two functions implement different logic. prepare_write_buffers() 
create a rotating
pattern unique to each vector while io_uring_init_buffer_pattern() fills 
a buffer with
a single repeating character.

But I see your point. I will merge these two as well.

Thank You for the detailed review.

-- 
Thanks
- Sachin



-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH v2 0/3] io_uring READ(V), WRITE(v) operation tests

[Sachin Sant]

On 15/04/26 3:05 pm, Cyril Hrubis wrote:
> Hi!
> I've merged 1/3 and 3/3 since these two patches were ready, thanks. I
> will add a few minor comments for 2/3.
Thank You Cyril.
  I will send an updated patch for 2/3.

-- 
Thanks
- Sachin



-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH v2 1/3] io_uring: Test IORING READ and WRITE operations

[Martin Doucha]

On 3/24/26 06:22, Sachin Sant wrote:
> +/*
> + * Map io_uring operation code to human-readable name
> + */
> +static inline const char *ioring_op_name(int op)
> +{
> +	switch (op) {
> +	case IORING_OP_READV:
> +		return "IORING_OP_READV";
> +	case IORING_OP_WRITEV:
> +		return "IORING_OP_WRITEV";
> +	case IORING_OP_READ:
> +		return "IORING_OP_READ";
> +	case IORING_OP_WRITE:
> +		return "IORING_OP_WRITE";

Hi,
IORING_OP_READ and IORING_OP_WRITE have been added in kernel v5.6. Some 
older distros currently fail to build LTP due to their missing 
declarations so you'll need to add those constants to lapi headers.

-- 
Martin Doucha   mdoucha@suse.cz
SW Quality Engineer
SUSE LINUX, s.r.o.
CORSO IIa
Krizikova 148/34
186 00 Prague 8
Czech Republic

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH v2 1/3] io_uring: Test IORING READ and WRITE operations

[Sachin Sant]

On 16/04/26 2:25 pm, Martin Doucha wrote:
> On 3/24/26 06:22, Sachin Sant wrote:
>> +/*
>> + * Map io_uring operation code to human-readable name
>> + */
>> +static inline const char *ioring_op_name(int op)
>> +{
>> +    switch (op) {
>> +    case IORING_OP_READV:
>> +        return "IORING_OP_READV";
>> +    case IORING_OP_WRITEV:
>> +        return "IORING_OP_WRITEV";
>> +    case IORING_OP_READ:
>> +        return "IORING_OP_READ";
>> +    case IORING_OP_WRITE:
>> +        return "IORING_OP_WRITE";
>
> Hi,
> IORING_OP_READ and IORING_OP_WRITE have been added in kernel v5.6. 
> Some older distros currently fail to build LTP due to their missing 
> declarations so you'll need to add those constants to lapi headers.
>

Thank you for the report. I will send a fix to address this.

-- 
Thanks
- Sachin



-- 
Mailing list info: https://lists.linux.it/listinfo/ltp