libaio: implement the Linux async I/O interface

[gburd]

What

Implements the Linux async I/O interface (libaio), which was a latent crash
hazard. io_setup() returned success, but io_submit(), io_getevents(),
io_destroy() and io_cancel() were UNIMPL() stubs that called abort(). A
program that probed libaio, saw io_setup() succeed, and then submitted any I/O
would take down the whole unikernel.

How

OSv has no user/kernel boundary, so "async" I/O is done the same way
core/io_uring.cc does it: each submitted iocb runs on a short-lived detached
worker thread that performs the blocking read/write/fsync through the VFS
(sys_read/sys_write/sys_fsync, the same primitives io_uring uses), then
queues an io_event and wakes any thread parked in io_getevents().

• io_setup(): allocate a real io_context capped at nr_events in-flight ops.
• io_submit(): dispatch each iocb to a detached worker; honor the in-flight cap
  (partial count if some were accepted, otherwise EAGAIN, matching Linux).
  Supports PREAD, PWRITE, PREADV, PWRITEV, FSYNC, FDSYNC, NOOP;
  unknown opcodes complete with res == -EINVAL rather than aborting.
• io_getevents(): block until min_nr events are available or the optional
  timeout elapses, then copy out up to nr events. A bad fd surfaces as
  res == -EBADF in the event, not a crash.
• io_destroy(): mark the context draining and wait for all in-flight ops to
  retire (detached workers self-reap) before freeing, so no worker can touch a
  freed context.
• io_cancel(): return EINVAL. Ops run to completion on a worker with no safe
  mid-flight cancellation point; EINVAL is a valid Linux response.

Best-effort eventfd notification is delivered when an iocb sets
IOCB_FLAG_RESFD.

include/api/libaio.h is filled in with the real Linux aio ABI (struct iocb,
struct io_event, IO_CMD_* opcodes, IOCB_FLAG_*), which was previously only
forward-declared.

Testing

tests/tst-libaio.cc covers the write/read round-trip (data and aio_data
echo), an 8-op batch, bad-fd (-EBADF, no abort), and the io_getevents
timeout path. It passes on OSv under KVM on both single- and multi-vCPU
configurations.

Note

This is a self-contained worker-thread-per-iocb design, which is simple and
correct for typical libaio concurrency. If a caller submits very large batches
and thread-creation cost shows up, the worker loop can be swapped for a bounded
pool like the one in core/io_uring.cc; that is left as a follow-up.

(Recreated from #1413, which GitHub auto-closed when its branch was rebased onto current master. Same change, rebased and verified on master 3aba46c.)

[Copilot]

Pull request overview

This PR implements the Linux libaio ABI/API in OSv to eliminate abort-based stubs and provide functional async I/O via detached worker threads, plus adds a basic regression test and wires it into the test build.

Changes:

• Implement io_setup/io_submit/io_getevents/io_destroy/io_cancel in core/libaio.cc with worker-thread execution and completion queuing.
• Populate include/api/libaio.h with the Linux aio ABI structures/opcodes/flags needed for compatibility.
• Add and register tests/tst-libaio.cc in the tests module build.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 8 comments.

File	Description
tests/tst-libaio.cc	Adds a regression test covering basic submit/getevents behavior and key safety paths.
modules/tests/Makefile	Registers the new libaio test module in the build.
include/api/libaio.h	Defines the libaio ABI types/opcodes/flags for consumers.
core/libaio.cc	Implements the libaio syscall-like entry points and completion machinery.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[gburd]

Thanks (and thanks to Copilot). Addressed in the current tip (41d79a2, rebased onto master and squashed into one commit):

• Missing <unistd.h> for ::write: added.
• io_setup allowing nr_events == 0 (uncapped) and null ctxp: now rejects nr_events <= 0 with EINVAL (a zero cap would mean "unbounded" and allow unbounded worker-thread creation), and a null ctxp with EFAULT.
• io_submit not validating ios / ios[i]: a null ios array returns EFAULT; a null iocb entry returns EFAULT if nothing was accepted, otherwise the partial submitted count (Linux-style).
• sched::thread::make() returning nullptr: a failed thread allocation is no longer counted as inflight (which would make io_destroy hang forever); it returns the partial count or EAGAIN.
• io_getevents null events / invalid timespec: null events returns EFAULT; a negative or tv_nsec >= 1e9 timespec returns EINVAL.
• io_destroy use-after-free vs a concurrent io_getevents: io_getevents now tracks itself in a waiters counter, and io_destroy sets destroying, wakes all, and waits for both inflight == 0 and waiters == 0 before freeing the context, so no worker or parked waiter still references ctx->mtx/cv at free time.

On the test-coverage note (RESFD notification and a non-read/write op): agreed that is worth adding; I will extend tst-libaio once I can run the image. Needs a build to validate; queued.

[gburd]

Build-validated on x86_64: image=tests builds (EXIT 0) and tst-libaio passes. The io_destroy-vs-concurrent-io_getevents waiters handshake was exercised specifically (no use-after-free, no hang, clean exit), and the new EINVAL/EFAULT input-validation cases pass without breaking the existing ones.