kernel/net/unix, branch linux-5.1.y

Merge tag 'io_uring-2019-03-06' of git://git.kernel.dk/linux-block

2019-03-08T22:48:40Z

Pull io_uring IO interface from Jens Axboe: "Second attempt at adding the io_uring interface. Since the first one, we've added basic unit testing of the three system calls, that resides in liburing like the other unit tests that we have so far. It'll take a while to get full coverage of it, but we're working towards it. I've also added two basic test programs to tools/io_uring. One uses the raw interface and has support for all the various features that io_uring supports outside of standard IO, like fixed files, fixed IO buffers, and polled IO. The other uses the liburing API, and is a simplified version of cp(1). This adds support for a new IO interface, io_uring. io_uring allows an application to communicate with the kernel through two rings, the submission queue (SQ) and completion queue (CQ) ring. This allows for very efficient handling of IOs, see the v5 posting for some basic numbers: https://lore.kernel.org/linux-block/20190116175003.17880-1-axboe@kernel.dk/ Outside of just efficiency, the interface is also flexible and extendable, and allows for future use cases like the upcoming NVMe key-value store API, networked IO, and so on. It also supports async buffered IO, something that we've always failed to support in the kernel. Outside of basic IO features, it supports async polled IO as well. This particular feature has already been tested at Facebook months ago for flash storage boxes, with 25-33% improvements. It makes polled IO actually useful for real world use cases, where even basic flash sees a nice win in terms of efficiency, latency, and performance. These boxes were IOPS bound before, now they are not. This series adds three new system calls. One for setting up an io_uring instance (io_uring_setup(2)), one for submitting/completing IO (io_uring_enter(2)), and one for aux functions like registrating file sets, buffers, etc (io_uring_register(2)). Through the help of Arnd, I've coordinated the syscall numbers so merge on that front should be painless. Jon did a writeup of the interface a while back, which (except for minor details that have been tweaked) is still accurate. Find that here: https://lwn.net/Articles/776703/ Huge thanks to Al Viro for helping getting the reference cycle code correct, and to Jann Horn for his extensive reviews focused on both security and bugs in general. There's a userspace library that provides basic functionality for applications that don't need or want to care about how to fiddle with the rings directly. It has helpers to allow applications to easily set up an io_uring instance, and submit/complete IO through it without knowing about the intricacies of the rings. It also includes man pages (thanks to Jeff Moyer), and will continue to grow support helper functions and features as time progresses. Find it here: git://git.kernel.dk/liburing Fio has full support for the raw interface, both in the form of an IO engine (io_uring), but also with a small test application (t/io_uring) that can exercise and benchmark the interface" * tag 'io_uring-2019-03-06' of git://git.kernel.dk/linux-block: io_uring: add a few test tools io_uring: allow workqueue item to handle multiple buffered requests io_uring: add support for IORING_OP_POLL io_uring: add io_kiocb ref count io_uring: add submission polling io_uring: add file set registration net: split out functions related to registering inflight socket files io_uring: add support for pre-mapped user IO buffers block: implement bio helper to add iter bvec pages to bio io_uring: batch io_kiocb allocation io_uring: use fget/fput_many() for file references fs: add fget_many() and fput_many() io_uring: support for IO polling io_uring: add fsync support Add io_uring IO interface

net: split out functions related to registering inflight socket files

2019-02-28T15:24:23Z

We need this functionality for the io_uring file registration, but we cannot rely on it since CONFIG_UNIX can be modular. Move the helpers to a separate file, that's always builtin to the kernel if CONFIG_UNIX is m/y. No functional changes in this patch, just moving code around. Reviewed-by: Hannes Reinecke Acked-by: David S. Miller Signed-off-by: Jens Axboe

Add io_uring IO interface

2019-02-28T15:24:23Z

The submission queue (SQ) and completion queue (CQ) rings are shared between the application and the kernel. This eliminates the need to copy data back and forth to submit and complete IO. IO submissions use the io_uring_sqe data structure, and completions are generated in the form of io_uring_cqe data structures. The SQ ring is an index into the io_uring_sqe array, which makes it possible to submit a batch of IOs without them being contiguous in the ring. The CQ ring is always contiguous, as completion events are inherently unordered, and hence any io_uring_cqe entry can point back to an arbitrary submission. Two new system calls are added for this: io_uring_setup(entries, params) Sets up an io_uring instance for doing async IO. On success, returns a file descriptor that the application can mmap to gain access to the SQ ring, CQ ring, and io_uring_sqes. io_uring_enter(fd, to_submit, min_complete, flags, sigset, sigsetsize) Initiates IO against the rings mapped to this fd, or waits for them to complete, or both. The behavior is controlled by the parameters passed in. If 'to_submit' is non-zero, then we'll try and submit new IO. If IORING_ENTER_GETEVENTS is set, the kernel will wait for 'min_complete' events, if they aren't already available. It's valid to set IORING_ENTER_GETEVENTS and 'min_complete' == 0 at the same time, this allows the kernel to return already completed events without waiting for them. This is useful only for polling, as for IRQ driven IO, the application can just check the CQ ring without entering the kernel. With this setup, it's possible to do async IO with a single system call. Future developments will enable polled IO with this interface, and polled submission as well. The latter will enable an application to do IO without doing ANY system calls at all. For IRQ driven IO, an application only needs to enter the kernel for completions if it wants to wait for them to occur. Each io_uring is backed by a workqueue, to support buffered async IO as well. We will only punt to an async context if the command would need to wait for IO on the device side. Any data that can be accessed directly in the page cache is done inline. This avoids the slowness issue of usual threadpools, since cached data is accessed as quickly as a sync interface. Sample application: http://git.kernel.dk/cgit/fio/plain/t/io_uring.c Reviewed-by: Hannes Reinecke Signed-off-by: Jens Axboe

missing barriers in some of unix_sock ->addr and ->path accesses

2019-02-21T04:06:28Z

Several u->addr and u->path users are not holding any locks in common with unix_bind(). unix_state_lock() is useless for those purposes. u->addr is assign-once and *(u->addr) is fully set up by the time we set u->addr (all under unix_table_lock). u->path is also set in the same critical area, also before setting u->addr, and any unix_sock with ->path filled will have non-NULL ->addr. So setting ->addr with smp_store_release() is all we need for those "lockless" users - just have them fetch ->addr with smp_load_acquire() and don't even bother looking at ->path if they see NULL ->addr. Users of ->addr and ->path fall into several classes now: 1) ones that do smp_load_acquire(u->addr) and access *(u->addr) and u->path only if smp_load_acquire() has returned non-NULL. 2) places holding unix_table_lock. These are guaranteed that *(u->addr) is seen fully initialized. If unix_sock is in one of the "bound" chains, so's ->path. 3) unix_sock_destructor() using ->addr is safe. All places that set u->addr are guaranteed to have seen all stores *(u->addr) while holding a reference to u and unix_sock_destructor() is called when (atomic) refcount hits zero. 4) unix_release_sock() using ->path is safe. unix_bind() is serialized wrt unix_release() (normally - by struct file refcount), and for the instances that had ->path set by unix_bind() unix_release_sock() comes from unix_release(), so they are fine. Instances that had it set in unix_stream_connect() either end up attached to a socket (in unix_accept()), in which case the call chain to unix_release_sock() and serialization are the same as in the previous case, or they never get accept'ed and unix_release_sock() is called when the listener is shut down and its queue gets purged. In that case the listener's queue lock provides the barriers needed - unix_stream_connect() shoves our unix_sock into listener's queue under that lock right after having set ->path and eventual unix_release_sock() caller picks them from that queue under the same lock right before calling unix_release_sock(). 5) unix_find_other() use of ->path is pointless, but safe - it happens with successful lookup by (abstract) name, so ->path.dentry is guaranteed to be NULL there. earlier-variant-reviewed-by: "Paul E. McKenney" Signed-off-by: Al Viro Signed-off-by: David S. Miller

Revert "net: simplify sock_poll_wait"

2018-10-23T17:57:06Z

This reverts commit dd979b4df817e9976f18fb6f9d134d6bc4a3c317. This broke tcp_poll for SMC fallback: An AF_SMC socket establishes an internal TCP socket for the initial handshake with the remote peer. Whenever the SMC connection can not be established this TCP socket is used as a fallback. All socket operations on the SMC socket are then forwarded to the TCP socket. In case of poll, the file->private_data pointer references the SMC socket because the TCP socket has no file assigned. This causes tcp_poll to wait on the wrong socket. Signed-off-by: Karsten Graul Signed-off-by: David S. Miller

net: fix warning in af_unix

2018-10-18T04:57:28Z

This fixes the "'hash' may be used uninitialized in this function" net/unix/af_unix.c:1041:20: warning: 'hash' may be used uninitialized in this function [-Wmaybe-uninitialized] addr->hash = hash ^ sk->sk_type; Signed-off-by: Kyeongdon Kim Signed-off-by: David S. Miller

af_unix: ensure POLLOUT on remote close() for connected dgram socket

2018-08-03T23:44:19Z

Applications use -ECONNREFUSED as returned from write() in order to determine that a socket should be closed. However, when using connected dgram unix sockets in a poll/write loop, a final POLLOUT event can be missed when the remote end closes. Thus, the poll is stuck forever: thread 1 (client) thread 2 (server) connect() to server write() returns -EAGAIN unix_dgram_poll() -> unix_recvq_full() is true close() ->unix_release_sock() ->wake_up_interruptible_all() unix_dgram_poll() (due to the wake_up_interruptible_all) -> unix_recvq_full() still is true ->free all skbs Now thread 1 is stuck and will not receive anymore wakeups. In this case, when thread 1 gets the -EAGAIN, it has not queued any skbs otherwise the 'free all skbs' step would in fact cause a wakeup and a POLLOUT return. So the race here is probably fairly rare because it means there are no skbs that thread 1 queued and that thread 1 schedules before the 'free all skbs' step. This issue was reported as a hang when /dev/log is closed. The fix is to signal POLLOUT if the socket is marked as SOCK_DEAD, which means a subsequent write() will get -ECONNREFUSED. Reported-by: Ian Lance Taylor Cc: David Rientjes Cc: Rainer Weikusat Cc: Eric Dumazet Signed-off-by: Jason Baron Signed-off-by: David S. Miller

net: simplify sock_poll_wait

2018-07-30T16:10:25Z

The wait_address argument is always directly derived from the filp argument, so remove it. Signed-off-by: Christoph Hellwig Signed-off-by: David S. Miller

Revert changes to convert to ->poll_mask() and aio IOCB_CMD_POLL

2018-06-28T17:40:47Z

The poll() changes were not well thought out, and completely unexplained. They also caused a huge performance regression, because "->poll()" was no longer a trivial file operation that just called down to the underlying file operations, but instead did at least two indirect calls. Indirect calls are sadly slow now with the Spectre mitigation, but the performance problem could at least be largely mitigated by changing the "->get_poll_head()" operation to just have a per-file-descriptor pointer to the poll head instead. That gets rid of one of the new indirections. But that doesn't fix the new complexity that is completely unwarranted for the regular case. The (undocumented) reason for the poll() changes was some alleged AIO poll race fixing, but we don't make the common case slower and more complex for some uncommon special case, so this all really needs way more explanations and most likely a fundamental redesign. [ This revert is a revert of about 30 different commits, not reverted individually because that would just be unnecessarily messy - Linus ] Cc: Al Viro Cc: Christoph Hellwig Signed-off-by: Linus Torvalds

Merge branch 'work.aio-1' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs

2018-06-04T20:57:43Z

Pull aio updates from Al Viro: "Majority of AIO stuff this cycle. aio-fsync and aio-poll, mostly. The only thing I'm holding back for a day or so is Adam's aio ioprio - his last-minute fixup is trivial (missing stub in !CONFIG_BLOCK case), but let it sit in -next for decency sake..." * 'work.aio-1' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (46 commits) aio: sanitize the limit checking in io_submit(2) aio: fold do_io_submit() into callers aio: shift copyin of iocb into io_submit_one() aio_read_events_ring(): make a bit more readable aio: all callers of aio_{read,write,fsync,poll} treat 0 and -EIOCBQUEUED the same way aio: take list removal to (some) callers of aio_complete() aio: add missing break for the IOCB_CMD_FDSYNC case random: convert to ->poll_mask timerfd: convert to ->poll_mask eventfd: switch to ->poll_mask pipe: convert to ->poll_mask crypto: af_alg: convert to ->poll_mask net/rxrpc: convert to ->poll_mask net/iucv: convert to ->poll_mask net/phonet: convert to ->poll_mask net/nfc: convert to ->poll_mask net/caif: convert to ->poll_mask net/bluetooth: convert to ->poll_mask net/sctp: convert to ->poll_mask net/tipc: convert to ->poll_mask ...