This is the reproducer for the kworker hang bug.

Reproduction Steps:
  1) A user task calls io_uring_queue_exit().

  2) Suspend the task with SIGSTOP / SIGTRAP before the ring exit is
     finished (do it as soon as step (1) is done).

  3) Wait for `/proc/sys/kernel/hung_task_timeout_secs` seconds
     elapsed.

  4) We get a complaint from the khungtaskd because the kworker is
     stuck in an uninterruptible state (D).

The kworkers waiting on ring exit are not progressing as the task
cannot proceed. When the user task is continued (e.g. get SIGCONT
after SIGSTOP, or continue after SIGTRAP breakpoint), the kworkers
then can finish the ring exit.

We need a special handling for this case to avoid khungtaskd
complaint. Currently we don't have the fix for this.

The dmesg says:

  [247390.432294] INFO: task kworker/u8:2:358488 blocked for more than 10 seconds.
  [247390.432314]       Tainted: G           OE     5.15.0-stable #5
  [247390.432322] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
  [247390.432329] task:kworker/u8:2    state:D stack:    0 pid:358488 ppid:     2 flags:0x00004000
  [247390.432341] Workqueue: events_unbound io_ring_exit_work
  [247390.432354] Call Trace:
  [247390.432368]  __schedule+0x453/0x1850
  [247390.432388]  ? lock_acquire+0xc8/0x2d0
  [247390.432404]  ? usleep_range+0x90/0x90
  [247390.432412]  schedule+0x59/0xc0
  [247390.432420]  schedule_timeout+0x1aa/0x1f0
  [247390.432429]  ? mark_held_locks+0x49/0x70
  [247390.432439]  ? lockdep_hardirqs_on_prepare+0xff/0x180
  [247390.432445]  ? _raw_spin_unlock_irq+0x24/0x40
  [247390.432456]  __wait_for_common+0xc2/0x170
  [247390.432473]  io_ring_exit_work+0x1d9/0x750
  [247390.432486]  ? io_uring_del_tctx_node+0xe0/0xe0
  [247390.432502]  ? verify_cpu+0xf0/0x100
  [247390.432520]  process_one_work+0x23b/0x550
  [247390.432540]  worker_thread+0x55/0x3c0
  [247390.432546]  ? process_one_work+0x550/0x550
  [247390.432556]  kthread+0x140/0x160
  [247390.432564]  ? set_kthread_struct+0x40/0x40
  [247390.432574]  ret_from_fork+0x1f/0x30
  [247390.432605] INFO: task kworker/u8:0:359615 blocked for more than 10 seconds.
  [247390.432613]       Tainted: G           OE     5.15.0-stable #5
  [247390.432620] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
  [247390.432626] task:kworker/u8:0    state:D stack:    0 pid:359615 ppid:     2 flags:0x00004000
  [247390.432635] Workqueue: events_unbound io_ring_exit_work
  [247390.432643] Call Trace:
  [247390.432653]  __schedule+0x453/0x1850
  [247390.432676]  ? usleep_range+0x90/0x90
  [247390.432684]  schedule+0x59/0xc0
  [247390.432691]  schedule_timeout+0x1aa/0x1f0
  [247390.432700]  ? mark_held_locks+0x49/0x70
  [247390.432710]  ? lockdep_hardirqs_on_prepare+0xff/0x180
  [247390.432717]  ? _raw_spin_unlock_irq+0x24/0x40
  [247390.432728]  __wait_for_common+0xc2/0x170
  [247390.432744]  io_ring_exit_work+0x1d9/0x750
  [247390.432758]  ? io_uring_del_tctx_node+0xe0/0xe0
  [247390.432772]  ? verify_cpu+0xf0/0x100
  [247390.432788]  process_one_work+0x23b/0x550
  [247390.432807]  worker_thread+0x55/0x3c0
  [247390.432813]  ? process_one_work+0x550/0x550
  [247390.432824]  kthread+0x140/0x160
  [247390.432830]  ? set_kthread_struct+0x40/0x40
  [247390.432839]  ret_from_fork+0x1f/0x30
  [247390.432870]
                  Showing all locks held in the system:
  [247390.432877] 1 lock held by khungtaskd/40:
  [247390.432880]  #0: ffffffff82976700 (rcu_read_lock){....}-{1:2}, at: debug_show_all_locks+0x15/0x174
  [247390.432911] 1 lock held by in:imklog/922:
  [247390.432915]  #0: ffff8881041cfcf0 (&f->f_pos_lock){+.+.}-{3:3}, at: __fdget_pos+0x4a/0x60
  [247390.432977] 2 locks held by pager/318088:
  [247390.432981]  #0: ffff8881208d4898 (&tty->ldisc_sem){++++}-{0:0}, at: tty_ldisc_ref_wait+0x24/0x50
  [247390.433001]  #1: ffffc900010fd2e8 (&ldata->atomic_read_lock){+.+.}-{3:3}, at: n_tty_read+0x49e/0x660
  [247390.433024] 1 lock held by htop/341462:
  [247390.433032] 2 locks held by kworker/u8:2/358488:
  [247390.433035]  #0: ffff888100106938 ((wq_completion)events_unbound){+.+.}-{0:0}, at: process_one_work+0x1c1/0x550
  [247390.433053]  #1: ffffc90003797e70 ((work_completion)(&ctx->exit_work)){+.+.}-{0:0}, at: process_one_work+0x1c1/0x550
  [247390.433071] 2 locks held by kworker/u8:0/359615:
  [247390.433075]  #0: ffff888100106938 ((wq_completion)events_unbound){+.+.}-{0:0}, at: process_one_work+0x1c1/0x550
  [247390.433092]  #1: ffffc90003597e70 ((work_completion)(&ctx->exit_work)){+.+.}-{0:0}, at: process_one_work+0x1c1/0x550
  [247390.433110] 1 lock held by dmesg/361178:
  [247390.433113]  #0: ffff88810b5300d0 (&user->lock){+.+.}-{3:3}, at: devkmsg_read+0x4b/0x230

  [247390.433134] =============================================

Cc: Pavel Begunkov <asml.silence@gmail.com>
Link: https://github.com/axboe/liburing/issues/448


Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

-EINTR relies on io-wq, we can also find and cancel before that. Make
sure the test allows it, fixing the following case failure on newer
kernels:

1 -125
child failed 1
test_cancel_req_across_fork() failed

Signed-off-by: Jens Axboe <axboe@kernel.dk>

* 'pr1' of https://github.com/GalaxySnail/liburing:
  man/io_uring_setup.2: add missing `wq_fd` field

Add missing `wq_fd` field for `struct io_uring_params`.

Fixes: https://github.com/axboe/liburing/issues/473


Signed-off-by: Luchen Yang <ylc991@163.com>

The Makefile for test is not efficient and reads bad. We have to
specify the name for each test 2 times (filename and name without
the ext) while we could have just elided the extension from the
source filename.

Let's make it simpler and easier to manage.

Changes summary:
 - Clean up and reorder things.
 - Sort the `test_srcs` alphabetically.
 - Remove `test_objs` (it turned out unused).
 - Generate `test_targets` variable from `test_srcs` by simply
   removing the `.c` and `.cc` file extension.

Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Link: https://lore.kernel.org/r/20211030114858.320116-3-ammar.faizi@intel.com


Signed-off-by: Jens Axboe <axboe@kernel.dk>

Several things go wrong with this Makefile:
  1) Using `test_srcs` to generate `test_objs`.
  2) `test_objs` is an unused variable. So (1) is pointless.
  3) `make clean` does not remove `ucontext-cp` binary.

I assume (1) and (2) were blindly copied from the test Makefile.

For 3, the `make clean` removes $(all_targets) and $(test_objs). But
`ucontext-cp` only exists in $(all_targets) if we have
`CONFIG_HAVE_UCONTEXT`. When the target goal is `clean`, we will not
have any of `CONFIG_*` variables. Thus, `ucontext-cp` is not removed.

Clean them up!

Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Link: https://lore.kernel.org/r/20211030114858.320116-2-ammar.faizi@intel.com


Signed-off-by: Jens Axboe <axboe@kernel.dk>

* 'pr-template' of https://github.com/ammarfaizi2/liburing:
  .github: Create pull_request_template.md

GitHub throws away the relevant information like having even a valid
email address for the person asking to pull.

Another common problem on GitHub, people don't do any of the simplest
things that the kernel people expect from a commit message, like:

 - No "short one-line description in the first line".
 - No sane word-wrap of the long description. GitHub commit messages
   tend to be (if they have any description at all) one long
   unreadable line.
 - No sign-off and other tags that we require for commit message.

Let's fix this by creating a pull request template that contains pull
request guidelines for GitHub Pull Request.

Link: https://github.com/axboe/liburing/pull/464#issuecomment-950950688


Signed-off-by: Ammar Faizi <ammarfaizi2@gmail.com>

* 'fix463' of https://github.com/tchaloupka/liburing:
  Fix invalid opcode used in io_uring_unregister_iowq_aff

Just a trivial fix of used opcode IORING_REGISTER_IOWQ_AFF -> IORING_UNREGISTER_IOWQ_AFF.

Signed-off-by: Tomáš Chaloupka <chalucha@gmail.com>
Fixes: https://github.com/axboe/liburing/issues/463

* 'pr2' of https://github.com/romange/liburing:
  Add io_uring_prep_readv2 and io_uring_prep_writev2 helpers

These correspond to preadv2 and pwritev2 calls and include flags
argument missing in the original helpers.

Signed-off-by: Roman Gershman <romange@gmail.com>
Link: https://github.com/axboe/liburing/issues/461

This reverts commit af31ce2c.

The commit is buggy and causes ucontext-cp to crash. Just revert it
for now.

Fixes: https://github.com/axboe/liburing/issues/459


Signed-off-by: Jens Axboe <axboe@kernel.dk>

Fix this:
```
  In file included from timeout-overflow.c:12:
  timeout-overflow.c: In function ‘test_timeout_overflow’:
  ../src/include/liburing.h:406:9: error: ‘num’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
    406 |         io_uring_prep_rw(IORING_OP_TIMEOUT, sqe, -1, ts, 1, count);
        |         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  timeout-overflow.c:104:26: note: ‘num’ was declared here
    104 |                 unsigned num;
        |                          ^~~
```

Fixes: a4b46553

 ("Add test for overflow of timeout request's sequence")
Cc: yangerkun <yangerkun@huawei.com>
Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Link: https://lore.kernel.org/r/BULrcXMbevM-ammarfaizi2@gnuweeb.org


Signed-off-by: Jens Axboe <axboe@kernel.dk>

Fix this:
```
  timeout.c: In function ‘test_multi_timeout’:
  timeout.c:590:20: warning: ‘user_data’ may be used uninitialized in this function [-Wmaybe-uninitialized]
    590 |                 if (cqe->user_data != user_data) {
        |                    ^
  timeout.c:601:51: warning: ‘time’ may be used uninitialized in this function [-Wmaybe-uninitialized]
    601 |                 if (exp < time / 2 || exp > (time * 3) / 2) {
        |                                             ~~~~~~^~~~
```

Fixes: 37136cb4

 ("test/timeout: add multi timeout reqs test with different timeout")
Cc: zhangyi (F) <yi.zhang@huawei.com>
Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Link: https://lore.kernel.org/r/hgUsvvrR9xY-ammarfaizi2@gnuweeb.org


Signed-off-by: Jens Axboe <axboe@kernel.dk>

* 'master' of https://github.com/Tindarid/liburing:
  setup.c: rewrite `__fls()`

Rewrite handmade `__fls()`: use `__builtin_clz()`.
Also add `inline` specifier.

Signed-off-by: Nursan Valeyev <tindarid@gmail.com>

Add `__attribute__((__aligned__))` to the `user_p` to guarantee
pointer returned by the `malloc()` is properly aligned for user.

This attribute asks the compiler to align a type to the maximum
useful alignment for the target machine we are compiling for,
which is often, but by no means always, 8 or 16 bytes [1].

Link: https://gcc.gnu.org/onlinedocs/gcc-11.2.0/gcc/Common-Variable-Attributes.html#Common-Variable-Attributes [1]
Fixes: https://github.com/axboe/liburing/issues/454


Reported-by: Louvian Lyndal <louvianlyndal@gmail.com>
Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Link: https://lore.kernel.org/r/20211011064927.444704-1-ammar.faizi@students.amikom.ac.id


Signed-off-by: Jens Axboe <axboe@kernel.dk>

* 'add-ci-nolibc-x86-64' of https://github.com/ammarfaizi2/liburing:
  .github/workflows/build.yml: Add nolibc build x86-64 for CI

Fixes: https://github.com/axboe/liburing/issues/443


Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>

It's for conditonal variable and macro to enable nolibc build.
Also add `--nolibc` option to `configure` to enable it.

Link: https://github.com/axboe/liburing/issues/443


Reviewed-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Link: https://lore.kernel.org/r/20211010135338.397115-4-ammar.faizi@students.amikom.ac.id


Signed-off-by: Jens Axboe <axboe@kernel.dk>

Create `src/nolibc.c` to provide libc functions like `memset()`,
`malloc()` and `free()`. Only build this file when we build liburing
without libc.

Add `get_page_size()` function to get the page size. When we build
with libc, it uses `sysconf(_SC_PAGESIZE)`, otherwise it uses arch
dependent implementation function that we provide at `src/arch` dir.

Add conditional preprocessor in `src/syscall.h` to use arch dependent
syscalls written in Assembly when we build liburing without libc.

Extra notes for tests:
 1) Functions in `src/syscall.c` require libc to work.
 2) Tests require functions in `src/syscall.c`.

So we build `src/syscall.c` manually from the test's Makefile.

The Makefile in `src/` dir still builds `src/syscall.c` when we
build liburing with libc.

Link: https://github.com/axboe/liburing/issues/443


Reviewed-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Link: https://lore.kernel.org/r/20211010135338.397115-3-ammar.faizi@students.amikom.ac.id


Signed-off-by: Jens Axboe <axboe@kernel.dk>

Create a new directory `src/arch` to save arch dependent sources.
Add support start from x86-64, add syscalls crafted in Assembly code
and lib (currently the lib only contains get page size function).

Link: https://github.com/axboe/liburing/issues/443


Reviewed-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Link: https://lore.kernel.org/r/20211010135338.397115-2-ammar.faizi@students.amikom.ac.id


Signed-off-by: Jens Axboe <axboe@kernel.dk>

When I add support for nolibc x86-64, I find this test failed.

Long story short, we provide our own `free()` that always unmaps the VM
with `munmap()`. It makes the CQE return -EFAULT because the kernel
reads unmapped user memory from the pending `write()` SQE.

I believe this test can run properly with libc build because `free()`
from libc doesn't always unmap the memory, instead it uses free list on
the userspace and the freed heap may still be userspace addressable.

Fix this by deferring the free.

Cc: Jens Axboe <axboe@kernel.dk>
Fixes: 2edfa3f8

 ("Add test case for thread exiting with pending IO")
Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Link: https://lore.kernel.org/r/20211010063906.341014-2-ammar.faizi@students.amikom.ac.id


Signed-off-by: Jens Axboe <axboe@kernel.dk>

When we build liburing without libc, we can't check `errno` variable
with respect to liburing's functions. Don't do that it in test.

Note:
The tests themselves can still use `errno` to check error from
functions that come from the libc, but not liburing.

Link: https://github.com/axboe/liburing/issues/443
Fixes: https://github.com/axboe/liburing/issues/449


Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

`io_uring_free_probe()` should really be used to free the return value
of `io_uring_get_probe_ring()`. As we may not always allocate it with
`malloc()`. For example, to support no libc build [1].

Link: https://github.com/axboe/liburing/issues/443

 [1]
Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

struct open_how and friends reside in the openat2.h header, so include
that for the test case.

Fixes: https://github.com/axboe/liburing/issues/360


Signed-off-by: Jens Axboe <axboe@kernel.dk>

Cc: Jens Axboe <axboe@kernel.dk>
Cc: Olivier Langlois <olivier@trillion01.com>
Fixes: a060c8e5

 ("liburing: Add io_uring_submit_and_wait_timeout function in API")
Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

Before commit 0ea4ccd1 ("src/queue: don't flush SQ ring for new wait
interface"), io_uring_wait_cqes() was serving the purpose of submit sqe
and wait for cqe up to a certain timeout value.

Since the commit, a new function is needed to fill this gap.

Fixes: https://github.com/axboe/liburing/issues/440


Signed-off-by: Olivier Langlois <olivier@trillion01.com>
Link: https://lore.kernel.org/r/1bbde6755535cb7b0bdfc0846254e7c06faf04e0.1633366467.git.olivier@trillion01.com


[axboe: minor style and commit message fixup]
Signed-off-by: Jens Axboe <axboe@kernel.dk>

In commit 0c210dbae26a80ee82dbc7430828ab6fd7012548 ("Wrap all syscalls
in a kernel style return value"), we forgot to add a syscall wrapper
for `close()`. Add it.

Fixes: cccf0fa1

 ("Wrap all syscalls in a kernel style return value")
Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Link: https://lore.kernel.org/r/20211004013510.428077-1-ammar.faizi@students.amikom.ac.id


Signed-off-by: Jens Axboe <axboe@kernel.dk>

Add new syscall wrapper functions:
  1) `uring_mmap`
  2) `uring_munmap`
  3) `uring_madvise`
  4) `uring_getrlimit`
  5) `uring_setrlimit`

All of them are `static inline`.

Use them to wrap the syscalls in a kernel style return value. The
main purpose of this change is to make it possible to remove the
dependency of `errno` variable in liburing C sources, so that later
we can support build without libc.

Extra new helpers:
  1) `ERR_PTR()`
  2) `PTR_ERR()`
  3) `IS_ERR()`

These helpers are used to deal with syscalls that return a pointer.
Currently only `uring_mmap()` that depends on these.

Link: https://github.com/axboe/liburing/issues/443#issuecomment-927873932


Cc: Bedirhan KURT <windowz414@gnuweeb.org>
Suggested-by: Louvian Lyndal <louvianlyndal@gmail.com>
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Link: https://lore.kernel.org/r/20211003153428.369258-4-ammar.faizi@students.amikom.ac.id


Signed-off-by: Jens Axboe <axboe@kernel.dk>

Don't use `__sys_io_uring*` for liburing internal. These functions
are now supposed for user backward compatibility.

For now, we use `____sys_io_uring*` (4 underscores). These are
`static inline` functions that wrap the `errno` variable in a kernel
style return value (directly returns negative error code when errors).

The main purpose of this change is to make it possible to remove the
`errno` variable dependency from liburing C sources, so that later
we will be able to build liburing without libc which doesn't use
`errno` variable at all.

Cc: Bedirhan KURT <windowz414@gnuweeb.org>
Cc: Louvian Lyndal <louvianlyndal@gmail.com>
Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Link: https://lore.kernel.org/r/20211003153428.369258-3-ammar.faizi@students.amikom.ac.id


Signed-off-by: Jens Axboe <axboe@kernel.dk>

Make it possible to remove the dependency of `errno` variable (which
comes from libc).

Currently, we expose these functions to userland:
  1) `__sys_io_uring_register`
  2) `__sys_io_uring_setup`
  3) `__sys_io_uring_enter2`
  4) `__sys_io_uring_enter`

The tests in `test/io_uring_{enter,register,setup}.c` are the examples
of it. Since the userland needs to check the `errno` value to use them
properly, this means those functions always depend on libc. So we
cannot change their behavior. Don't touch them all, this ensures the
changes only affect liburing internal and no visible functionality
changes for the users.

Then we introduce new functions with the same name (with extra
underscore as prefix, 4 underscores):
  1) `____sys_io_uring_register`
  2) `____sys_io_uring_setup`
  3) `____sys_io_uring_enter2`
  4) `____sys_io_uring_enter`

These functions do not use `errno` variable *on the caller*, they use
the kernel style return value (return a negative value of error code
when errors).

These functions are defined as `static inline` in `src/syscall.h`.
They are just a wrapper to make sure liburing internal sources do not
touch `errno` variable from C files directly. We need to make C files
not to touch the `errno` variable to support build without libc.

Link: https://github.com/axboe/liburing/issues/443#issuecomment-927873932


Cc: Bedirhan KURT <windowz414@gnuweeb.org>
Suggested-by: Louvian Lyndal <louvianlyndal@gmail.com>
Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Link: https://lore.kernel.org/r/20211003153428.369258-2-ammar.faizi@students.amikom.ac.id


Signed-off-by: Jens Axboe <axboe@kernel.dk>

Happens pretty rarely, but there were cases when CQE waitinig in
test_return_before_timeout() time outs before the SQPOLL thread kicks in
and executes submitted requests, give SQPOLL a little bit more time.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/5faf8ade7e161931d76a47b809650e68a1b361ba.1633259449.git.asml.silence@gmail.com


Signed-off-by: Jens Axboe <axboe@kernel.dk>

Make sure we don't fail links on ETIME when IORING_TIMEOUT_ETIME_SUCCESS
is set.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/d1a5b6bdbcfa1ec6b5ca014248e12b3b1edb4e5d.1633259449.git.asml.silence@gmail.com


Signed-off-by: Jens Axboe <axboe@kernel.dk>

`io_uring_free_probe()` should really be used to free the return value
of `io_uring_get_probe_ring()`. As we may not always allocate it with
`malloc()`. For example, to support no libc build [1].

Link: https://github.com/axboe/liburing/issues/443

 [1]
Signed-off-by: Ammar Faizi <ammar.faizi@students.amikom.ac.id>
Link: https://lore.kernel.org/r/20211003010608.58380-1-ammar.faizi@students.amikom.ac.id


Signed-off-by: Jens Axboe <axboe@kernel.dk>

When we call `bind()` or `connect()`, the `addr` and `port` should be
in big endian value representation.

Portability notes:
  - Do not hard code the address with integer like this `0x0100007fU`.
    Instead, use `inet_addr("127.0.0.1")` to ensure portability for
    the machine with different endianess. It's also cleaner and more
    readable to use `inet_addr()` from `#include <arpa/inet.h>`.

  - Use `htons(port_number)` to make sure the port_number is properly
    choosen instead directly assign it with integer (still about
    endianess problem).

This commit fixes endianess issue in these files:
  test/232c93d07b74-test.c
  test/accept-link.c
  test/accept.c
  test/poll-link.c
  test/shutdown.c
  test/socket-rw-eagain.c
  test/socket-rw.c

Fixes: 08bd8151 ("Un-DOSify test/232c93d07b74-test.c")
Fixes: 4bce856d ("Improve reliability of poll/accept-link tests")
Fixes: 76e3b792 ("Add nonblock empty socket read test")
Fixes: 7de62535 ("test/accept: code reuse cleanup")
Fixes: 8a3a8d74 ("Add IORING_OP_SHUTDOWN test case")
Fixes: e3adbfc2

 ("Moves function calls out of assert().")
Suggested-by: Louvian Lyndal <louvianlyndal@gmail.com>
Signed-off-by: Ammar Faizi <ammarfaizi2@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

* 'master' of https://github.com/CarterLi/liburing:
  liburing.h: add io_uring_prep_close_direct

Signed-off-by: 李通洲 <carter.li@eoitek.com>

Signed-off-by: Jens Axboe <axboe@kernel.dk>