The link to tokio::main was relative to tokio_macros crate in the source
directory. This is why it worked in local build of documentation and not
in doc.rs.
Refs: #1473
This enables `block_in_place` to be used in more contexts. Specifically,
it allows you to block whenever you are off the tokio runtime (like if
you are not using tokio, are in a `spawn_blocking` closure, etc.), and
in the threaded scheduler's `block_on`. Blocking in `LocalSet` and the
basic scheduler's` block_on` is still disallowed.
Fixes#2327.
Fixes#2393.
Included changes
- all simple references like `<type>.<name>.html` for these types
- enum
- fn
- struct
- trait
- type
- simple references for methods, like struct.DelayQueue.html#method.poll
Refs: #1473
Some reasons to prefer logical count as the default:
- Chips reporting many logical CPUs vs physical, such as via
hyperthreading, probably know better than us about the workload the CPUs
can handle.
- The logical count (`num_cpus::get()`) takes into consideration
schedular affinity, and cgroups CPU quota, in case the user wants to
limit the amount of CPUs a process can use.
Closes#2269
Fixes a couple bugs in the work-stealing queue introduced as
part of #2315. First, the cursor needs to be able to represent more
values than the size of the buffer. This is to be able to track if
`tail` is ahead of `head` or if they are identical. This bug resulted in
the "overflow" path being taken before the buffer was full.
The second bug can happen when a queue is being stolen from concurrently
with stealing into. In this case, it is possible for buffer slots to be
overwritten before they are released by the stealer. This is harder to
happen in practice due to the first bug preventing the queue from
filling up 100%, but could still happen. It triggered an assertion in
`steal_into`. This bug slipped through due to a bug in loom not
correctly catching the case. The loom bug is fixed as part of
tokio-rs/loom#119.
Fixes: #2382
Previously, the `Mutex::lock`, `RwLock::{read, write}`, and
`Semaphore::acquire` futures in `tokio::sync` implemented `Send + Sync`
automatically. This was by virtue of being implemented using a `poll_fn`
that only closed over `Send + Sync` types. However, this broke in
PR #2325, which rewrote those types using the new `batch_semaphore`.
Now, they await an `Acquire` future, which contains a `Waiter`, which
internally contains an `UnsafeCell`, and thus does not implement `Sync`.
Since removing previously implemented traits breaks existing code, this
inadvertantly caused a breaking change. There were tests ensuring that
the `Mutex`, `RwLock`, and `Semaphore` types themselves were `Send +
Sync`, but no tests that the _futures they return_ implemented those
traits.
I've fixed this by adding an explicit impl of `Sync` for the
`batch_semaphore::Acquire` future. Since the `Waiter` type held by this
struct is only accessed when borrowed mutably, it is safe for it to
implement `Sync`.
Additionally, I've added to the bounds checks for the effected
`tokio::sync` types to ensure that returned futures continue to
implement `Send + Sync` in the future.
The new queue uses `u8` to track offsets. Cursors are expected to wrap.
An operation was performed with `+` instead of `wrapping_add`. This was
not _obviously_ issue before as it is difficult to wrap a `usize` on
64bit platforms, but wrapping a `u8` is trivial.
The fix is to use `wrapping_add` instead of `+`. A new test is added
that catches the issue.
Fixes#2361
The work-stealing scheduler includes an optimization where each worker
includes a single slot to store the **last** scheduled task. Tasks in
scheduler's LIFO slot are executed next. This speeds up and reduces
latency with message passing patterns.
Previously, this optimization was susceptible to starving other tasks in
certain cases. If two tasks ping-ping between each other without ever
yielding, the worker would never execute other tasks.
An early PR (#2160) introduced a form of pre-emption. Each task is
allocated a per-poll operation budget. Tokio resources will return ready
until the budget is depleted, at which point, Tokio resources will
always return `Pending`.
This patch leverages the operation budget to limit the LIFO scheduler
optimization. When executing tasks from the LIFO slot, the budget is
**not** reset. Once the budget goes to zero, the task in the LIFO slot
is pushed to the back of the queue.
## Motivation
When cancelling futures which are waiting to acquire semaphore permits,
there is a possible dangling pointer if notified futures are dropped
after the notified wakers have been split into a separate list. Because
these futures' wait queue nodes are no longer in the main list guarded
by the lock, their `Drop` impls will complete immediately, and they may
be dropped while still in the list of tasks to notify.
## Solution
This branch fixes this by popping from the wait list inside the lock.
The wakers of popped nodes are temporarily stored in a stack array,
so that they can be notified after the lock is released. Since the
size of the stack array is fixed, we may in some cases have to loop
multiple times, acquiring and releasing the lock, until all permits
have been released. This may also have the possible side advantage of
preventing a thread releasing a very large number of permits from
starving other threads that need to enqueue waiters.
I've also added a loom test that can reliably reproduce a segfault
on master, but passes on this branch (after a lot of iterations).
Signed-off-by: Eliza Weisman <[email protected]>
This reverts commit 826fc21abf.
The code was intentional. Holding the lock while notifying is
unnecessary. Also change the code to use `drop` so clippy doesn't
confuse people against their will.
Fixes a test from a PR that was written before the recent loom upgrade.
A change in the details how loom executes models resulted in the test to
start failing. The fix is to reduce the number of iterations performed
by the test.
This commit improves the memory ordering in the implementation of
Inner's increment function. The former code did a sequentially
consistent load of self.num, then entered a loop with a sequentially
consistent compare and swap on the same, bailing out with and Err only
if the loaded value was MAX_TIMEOUTS. The use of SeqCst means that all
threads must observe all relevant memory operations in the same order,
implying synchronization between all CPUs.
This commit adjusts the implementation in two key ways. First, the
initial load of self.num is now down with Relaxed ordering. If two
threads entered this code simultaneously, formerly, tokio required
that one proceed before the other, negating their parallelism. Now,
either thread may proceed without coordination. Second, the SeqCst
compare_and_swap is changed to a Release, Relaxed
compare_exchange_weak. The first memory ordering referrs to success:
if the value is swapped the load of that value for comparison will be
Relaxed and the store will be Release. The second memory ordering
referrs to failure: if the value is not swapped the load is
Relaxed. The _weak variant may spuriously fail but will generate
better code.
These changes mean that it is possible for more loops to be taken per
call than strictly necessary but with greater parallelism available on
this operation, improved energy consumption as CPUs don't have to
coordinate as much.
When the queue was polled and yielded an index from the wheel, the delay
until the next item was never updated. As a result, when one item was
yielded from `poll_idx` the following insert erronously updated the
delay to the instant of the inserted item.
Fixes: #1700
Loom is having a big refresh to improve performance and tighten up the
concurrency model. This diff tracks those changes.
Included in the changes is the removal of `CausalCell` deferred checks.
This is due to it technically being undefined behavior in the C++11
memory model. To address this, the work-stealing queue is updated to
avoid needing this behavior. This is done by limiting the queue to have
one concurrent stealer.
Geoffry Song