Any single-threaded task benchmark will spend a good chunk of time in `kqueue()` on osx and `epoll()` on linux, and the reason for this is that each time a task is terminated it will hit the syscall. When a task terminates, it context switches back to the scheduler thread, and the scheduler thread falls out of `run_sched_once` whenever it figures out that it did some work.
If we know that `epoll()` will return nothing, then we can continue to do work locally (only while there's work to be done). We must fall back to `epoll()` whenever there's active I/O in order to check whether it's ready or not, but without that (which is largely the case in benchmarks), we can prevent the costly syscall and can get a nice speedup.
I've separated the commits into preparation for this change and then the change itself, the last commit message has more details.
The green scheduler can optimize its runtime based on this by deciding to not go
to sleep in epoll() if there is no active I/O and there is a task to be stolen.
This is implemented for librustuv by keeping a count of the number of tasks
which are currently homed. If a task is homed, and then performs a blocking I/O
operation, the count will be nonzero while the task is blocked. The homing count
is intentionally 0 when there are I/O handles, but no handles currently blocked.
The reason for this is that epoll() would only be used to wake up the scheduler
anyway.
The crux of this change was to have a `HomingMissile` contain a mutable borrowed
reference back to the `HomeHandle`. The rest of the change was just dealing with
this fallout. This reference is used to decrement the homed handle count in a
HomingMissile's destructor.
Also note that the count maintained is not atomic because all of its
increments/decrements/reads are all on the same I/O thread.
This, the Nth rewrite of channels, is not a rewrite of the core logic behind
channels, but rather their API usage. In the past, we had the distinction
between oneshot, stream, and shared channels, but the most recent rewrite
dropped oneshots in favor of streams and shared channels.
This distinction of stream vs shared has shown that it's not quite what we'd
like either, and this moves the `std::comm` module in the direction of "one
channel to rule them all". There now remains only one Chan and one Port.
This new channel is actually a hybrid oneshot/stream/shared channel under the
hood in order to optimize for the use cases in question. Additionally, this also
reduces the cognitive burden of having to choose between a Chan or a SharedChan
in an API.
My simple benchmarks show no reduction in efficiency over the existing channels
today, and a 3x improvement in the oneshot case. I sadly don't have a
pre-last-rewrite compiler to test out the old old oneshots, but I would imagine
that the performance is comparable, but slightly slower (due to atomic reference
counting).
This commit also brings the bonus bugfix to channels that the pending queue of
messages are all dropped when a Port disappears rather then when both the Port
and the Chan disappear.
Beforehand, using a concurrent queue always mandated that the "shared state" be
stored internally to the queues in order to provide a safe interface. This isn't
quite as flexible as one would want in some circumstances, so instead this
commit moves the queues to not containing the shared state.
The queues no longer have a "default useful safe" interface, but rather a
"default safe" interface (minus the useful part). The queues have to be shared
manually through an Arc or some other means. This allows them to be a little
more flexible at the cost of a usability hindrance.
I plan on using this new flexibility to upgrade a channel to a shared channel
seamlessly.
This is part of the overall strategy I would like to take when approaching
issue #11165. The only two I/O objects that reasonably want to be "split" are
the network stream objects. Everything else can be "split" by just creating
another version.
The initial idea I had was the literally split the object into a reader and a
writer half, but that would just introduce lots of clutter with extra interfaces
that were a little unnnecssary, or it would return a ~Reader and a ~Writer which
means you couldn't access things like the remote peer name or local socket name.
The solution I found to be nicer was to just clone the stream itself. The clone
is just a clone of the handle, nothing fancy going on at the kernel level.
Conceptually I found this very easy to wrap my head around (everything else
supports clone()), and it solved the "split" problem at the same time.
The cloning support is pretty specific per platform/lib combination:
* native/win32 - uses some specific WSA apis to clone the SOCKET handle
* native/unix - uses dup() to get another file descriptor
* green/all - This is where things get interesting. When we support full clones
of a handle, this implies that we're allowing simultaneous writes
and reads to happen. It turns out that libuv doesn't support two
simultaneous reads or writes of the same object. It does support
*one* read and *one* write at the same time, however. Some extra
infrastructure was added to just block concurrent writers/readers
until the previous read/write operation was completed.
I've added tests to the tcp/unix modules to make sure that this functionality is
supported everywhere.
The general consensus is that we want to move away from conditions for I/O, and I propose a two-step plan for doing so:
1. Warn about unused `Result` types. When all of I/O returns `Result`, it will require you inspect the return value for an error *only if* you have a result you want to look at. By default, for things like `write` returning `Result<(), Error>`, these will all go silently ignored. This lint will prevent blind ignorance of these return values, letting you know that there's something you should do about them.
2. Implement a `try!` macro:
```
macro_rules! try( ($e:expr) => (match $e { Ok(e) => e, Err(e) => return Err(e) }) )
```
With these two tools combined, I feel that we get almost all the benefits of conditions. The first step (the lint) is a sanity check that you're not ignoring return values at callsites. The second step is to provide a convenience method of returning early out of a sequence of computations. After thinking about this for awhile, I don't think that we need the so-called "do-notation" in the compiler itself because I think it's just *too* specialized. Additionally, the `try!` macro is super lightweight, easy to understand, and works almost everywhere. As soon as you want to do something more fancy, my answer is "use match".
Basically, with these two tools in action, I would be comfortable removing conditions. What do others think about this strategy?
----
This PR specifically implements the `unused_result` lint. I actually added two lints, `unused_result` and `unused_must_use`, and the first commit has the rationale for why `unused_result` is turned off by default.
The `malloc` family of functions may return a null pointer for a
zero-size allocation, which should not be interpreted as an
out-of-memory error.
If the implementation does not return a null pointer, then handling
this will result in memory savings for zero-size types.
This also switches some code to `malloc_raw` in order to maintain a
centralized point for handling out-of-memory in `rt::global_heap`.
Closes#11634
Right now on linux, an empty executable with LTO still depends on librt becaues
of the clock_gettime function in rust_builtin.o, but this commit moves this
dependency into a rust function which is subject to elimination via LTO.
At the same time, this also drops libstd's dependency on librt on unices that
are not OSX because the library is only used by extra::time (and now the
dependency is listed in that module instead).
* vec::raw::to_ptr is gone
* Pausible => Pausable
* Removing @
* Calling the main task "<main>"
* Removing unused imports
* Removing unused mut
* Bringing some libextra tests up to date
* Allowing compiletest to work at stage0
* Fixing the bootstrap-from-c rmake tests
* assert => rtassert in a few cases
* printing to stderr instead of stdout in fail!()
These functions are all unnecessary now, and they only have meaning in the M:N
context. Removing these functions uncovered a bug in the librustuv timer
bindings, but it was fairly easy to cover (and the test is already committed).
These cannot be completely removed just yet due to their usage in the WaitQueue
of extra::sync, and until the mutex in libextra is rewritten it will not be
possible to remove the deferred sends for channels.
The uv loop was being destroyed before the async handle was being destroyed, so
closing the async handle was causing a use-after-free in the uv loop. This was
fixed by moving destruction of the queue's async handle to an earlier location
and then actually freeing it once the loop has been dropped.
The queue has an async handle which must be destroyed before the loop is
destroyed, so use a little bit of an option dance to get around this
requirement.
Note that this removes a number of run-pass tests which are exercising behavior
of the old runtime. This functionality no longer exists and is thoroughly tested
inside of libgreen and libnative. There isn't really the notion of "starting the
runtime" any more. The major notion now is "bootstrapping the initial task".
This allows creation of different sched pools with different io factories.
Namely, this will be used to test the basic I/O loop in the green crate. This
can also be used to override the global default.
Use the previous commit's new scheduler pool abstraction in libgreen to write
some homing tests which force an I/O handle to be homed from one event loop to
another.
All tests except for the homing tests are now working again with the
librustuv/libgreen refactoring. The homing-related tests are currently commented
out and now placed in the rustuv::homing module.
I plan on refactoring scheduler pool spawning in order to enable more homing
tests in a future commit.
This reimplements librustuv without using the interfaces provided by the
scheduler in libstd. This solely uses the new Runtime trait in order to
interface with the local task and perform the necessary scheduling operations.
The largest snag in this refactoring is reimplementing homing. The new runtime
trait exposes no concept of "homing" a task or forcibly sending a task to a
remote scheduler (there is no concept of a scheduler). In order to reimplement
homing, the transferrence of tasks is now done at the librustuv level instead of
the scheduler level. This means that all I/O loops now have a concurrent queue
which receives homing messages and requests.
This allows the entire implementation of librustuv to be only dependent on the
runtime trait, severing all dependence of librustuv on the scheduler and related
green-thread functions.
This is all in preparation of the introduction of libgreen and libnative.
At the same time, I also took the liberty of removing all glob imports from
librustuv.
Understand 'pkgid' in stage0. As a bonus, the snapshot now contains now metadata
(now that those changes have landed), and the snapshot download is half as large
as it used to be!
This replaces the link meta attributes with a pkgid attribute and uses a hash
of this as the crate hash. This makes the crate hash computable by things
other than the Rust compiler. It also switches the hash function ot SHA1 since
that is much more likely to be available in shell, Python, etc than SipHash.
Fixes#10188, #8523.
This reverts commit c54427ddfb.
Leave the #[ignores] in that were added to rustpkg tests.
Conflicts:
src/librustc/driver/driver.rs
src/librustc/metadata/creader.rs
This function had type &[u8] -> ~str, i.e. it allocates a string
internally, even though the non-allocating version that take &[u8] ->
&str and ~[u8] -> ~str are all that is necessary in most circumstances.
This commit implements the support necessary for generating both intermediate
and result static rust libraries. This is an implementation of my thoughts in
https://mail.mozilla.org/pipermail/rust-dev/2013-November/006686.html.
When compiling a library, we still retain the "lib" option, although now there
are "rlib", "staticlib", and "dylib" as options for crate_type (and these are
stackable). The idea of "lib" is to generate the "compiler default" instead of
having too choose (although all are interchangeable). For now I have left the
"complier default" to be a dynamic library for size reasons.
Of the rust libraries, lib{std,extra,rustuv} will bootstrap with an
rlib/dylib pair, but lib{rustc,syntax,rustdoc,rustpkg} will only be built as a
dynamic object. I chose this for size reasons, but also because you're probably
not going to be embedding the rustc compiler anywhere any time soon.
Other than the options outlined above, there are a few defaults/preferences that
are now opinionated in the compiler:
* If both a .dylib and .rlib are found for a rust library, the compiler will
prefer the .rlib variant. This is overridable via the -Z prefer-dynamic option
* If generating a "lib", the compiler will generate a dynamic library. This is
overridable by explicitly saying what flavor you'd like (rlib, staticlib,
dylib).
* If no options are passed to the command line, and no crate_type is found in
the destination crate, then an executable is generated
With this change, you can successfully build a rust program with 0 dynamic
dependencies on rust libraries. There is still a dynamic dependency on
librustrt, but I plan on removing that in a subsequent commit.
This change includes no tests just yet. Our current testing
infrastructure/harnesses aren't very amenable to doing flavorful things with
linking, so I'm planning on adding a new mode of testing which I believe belongs
as a separate commit.
Closes#552
This adds an implementation of the Chase-Lev work-stealing deque to libstd
under std::rt::deque. I've been unable to break the implementation of the deque
itself, and it's not super highly optimized just yet (everything uses a SeqCst
memory ordering).
The major snag in implementing the chase-lev deque is that the buffers used to
store data internally cannot get deallocated back to the OS. In the meantime, a
shared buffer pool (synchronized by a normal mutex) is used to
deallocate/allocate buffers from. This is done in hope of not overcommitting too
much memory. It is in theory possible to eventually free the buffers, but one
must be very careful in doing so.
I was unable to get some good numbers from src/test/bench tests (I don't think
many of them are slamming the work queue that much), but I was able to get some
good numbers from one of my own tests. In a recent rewrite of select::select(),
I found that my implementation was incredibly slow due to contention on the
shared work queue. Upon switching to the parallel deque, I saw the contention
drop to 0 and the runtime go from 1.6s to 0.9s with the most amount of time
spent in libuv awakening the schedulers (plus allocations).
Closes#4877
It turns out that libuv was returning ENOSPC to us in our usage of the
uv_ipX_name functions. It also turns out that there may be an off-by-one in
libuv. For now just add one to the buffer size and handle the return value
correctly.
Closes#10663
This is a behavioral difference in libuv between different platforms in
different situations. It turns out that libuv on windows will immediately
allocate a buffer instead of waiting for data to be ready. What this implies is
that we must have our custom data set on the handle before we call
uv_read_start.
I wish I knew of a way to test this, but this relies to being on the windows
platform *and* reading from a true TTY handle which only happens when this is
actually attached to a terminal. I have manually verified this works.
Closes#10645
This is a behavioral difference in libuv between different platforms in
different situations. It turns out that libuv on windows will immediately
allocate a buffer instead of waiting for data to be ready. What this implies is
that we must have our custom data set on the handle before we call
uv_read_start.
I wish I knew of a way to test this, but this relies to being on the windows
platform *and* reading from a true TTY handle which only happens when this is
actually attached to a terminal. I have manually verified this works.
Closes#10645
The reasons for doing this are:
* The model on which linked failure is based is inherently complex
* The implementation is also very complex, and there are few remaining who
fully understand the implementation
* There are existing race conditions in the core context switching function of
the scheduler, and possibly others.
* It's unclear whether this model of linked failure maps well to a 1:1 threading
model
Linked failure is often a desired aspect of tasks, but we would like to take a
much more conservative approach in re-implementing linked failure if at all.
Closes#8674Closes#8318Closes#8863
The reasons for doing this are:
* The model on which linked failure is based is inherently complex
* The implementation is also very complex, and there are few remaining who
fully understand the implementation
* There are existing race conditions in the core context switching function of
the scheduler, and possibly others.
* It's unclear whether this model of linked failure maps well to a 1:1 threading
model
Linked failure is often a desired aspect of tasks, but we would like to take a
much more conservative approach in re-implementing linked failure if at all.
Closes#8674Closes#8318Closes#8863
There are issues with reading stdin when it is actually attached to a pipe, but
I have run into no problems in writing to stdout/stderr when they are attached
to pipes.
There are issues with reading stdin when it is actually attached to a pipe, but
I have run into no problems in writing to stdout/stderr when they are attached
to pipes.
This commit re-organizes the io::native module slightly in order to have a
working implementation of rtio::IoFactory which uses native implementations. The
goal is to seamlessly multiplex among libuv/native implementations wherever
necessary.
Right now most of the native I/O is unimplemented, but we have existing bindings
for file descriptors and processes which have been hooked up. What this means is
that you can now invoke println!() from libstd with no local task, no local
scheduler, and even without libuv.
There's still plenty of work to do on the native I/O factory, but this is the
first steps into making it an official portion of the standard library. I don't
expect anyone to reach into io::native directly, but rather only std::io
primitives will be used. Each std::io interface seamlessly falls back onto the
native I/O implementation if the local scheduler doesn't have a libuv one
(hurray trait ojects!)
These two attributes are no longer useful now that Rust has decided to leave
segmented stacks behind. It is assumed that the rust task's stack is always
large enough to make an FFI call (due to the stack being very large).
There's always the case of stack overflow, however, to consider. This does not
change the behavior of stack overflow in Rust. This is still normally triggered
by the __morestack function and aborts the whole process.
C stack overflow will continue to corrupt the stack, however (as it did before
this commit as well). The future improvement of a guard page at the end of every
rust stack is still unimplemented and is intended to be the mechanism through
which we attempt to detect C stack overflow.
Closes#8822Closes#10155
When a channel is destroyed, it may attempt scheduler operations which could
move a task off of it's I/O scheduler. This is obviously a bad interaction, and
some finesse is required to make it work (making destructors run at the right
time).
Closes#10375
It appears that uv's support for interacting with a stdio stream as a tty when
it's actually a pipe is pretty problematic. To get around this, promote a check
to see if the stream is a tty to the top of the tty constructor, and bail out
quickly if it's not identified as a tty.
Closes#10237
In the ideal world, uv I/O could be canceled safely at any time. In reality,
however, we are unable to do this. Right now linked failure is fairly flaky as
implemented in the runtime, making it very difficult to test whether the linked
failure mechanisms inside of the uv bindings are ready for this kind of
interaction.
Right now, all constructors will execute in a task::unkillable block, and all
homing I/O operations will prevent linked failure in the duration of the homing
operation. What this means is that tasks which perform I/O are still susceptible
to linked failure, but the I/O operations themselves will never get interrupted.
Instead, the linked failure will be received at the edge of the I/O operation.
It turns out that the uv implementation would cause use-after-free if the idle
callback was used after the call to `close`, and additionally nothing would ever
really work that well if `start()` were called twice. To change this, the
`start` and `close` methods were removed in favor of specifying the callback at
creation, and allowing destruction to take care of closing the watcher.
This renames the `file` module to `fs` because that more accurately describes
its current purpose (manipulating the filesystem, not just files).
Additionally, this adds an UnstableFileStat structure as a nested structure of
FileStat to signify that the fields should not be depended on. The structure is
currently flagged with #[unstable], but it's unlikely that it has much meaning.
Closes#10241
This adds bindings to the remaining functions provided by libuv, all of which
are useful operations on files which need to get exposed somehow.
Some highlights:
* Dropped `FileReader` and `FileWriter` and `FileStream` for one `File` type
* Moved all file-related methods to be static methods under `File`
* All directory related methods are still top-level functions
* Created `io::FilePermission` types (backed by u32) that are what you'd expect
* Created `io::FileType` and refactored `FileStat` to use FileType and
FilePermission
* Removed the expanding matrix of `FileMode` operations. The mode of reading a
file will not have the O_CREAT flag, but a write mode will always have the
O_CREAT flag.
Closes#10130Closes#10131Closes#10121
This commit moves all thread-blocking I/O functions from the std::os module.
Their replacements can be found in either std::rt::io::file or in a hidden
"old_os" module inside of native::file. I didn't want to outright delete these
functions because they have a lot of special casing learned over time for each
OS/platform, and I imagine that these will someday get integrated into a
blocking implementation of IoFactory. For now, they're moved to a private module
to prevent bitrot and still have tests to ensure that they work.
I've also expanded the extensions to a few more methods defined on Path, most of
which were previously defined in std::os but now have non-thread-blocking
implementations as part of using the current IoFactory.
The api of io::file is in flux, but I plan on changing it in the next commit as
well.
Closes#10057
The invocation for making a directory should be able to specify a mode to make
the directory with (instead of defaulting to one particular mode). Additionally,
libuv and various OSes implement efficient versions of renaming files, so this
operation is exposed as an IoFactory call.
New standards have arisen in recent months, mostly for the use of
rustpkg, but the main Rust codebase has not been altered to match these
new specifications. This changeset rectifies most of these issues.
- Renamed the crate source files `src/libX/X.rs` to `lib.rs`, for
consistency with current styles; this affects extra, rustc, rustdoc,
rustpkg, rustuv, std, syntax.
- Renamed `X/X.rs` to `X/mod.rs,` as is now recommended style, for
`std::num` and `std::terminfo`.
- Shifted `src/libstd/str/ascii.rs` out of the otherwise unused `str`
directory, to be consistent with its import path of `std::ascii`;
libstd is flat at present so it's more appropriate thus.
While this removes some `#[path = "..."]` directives, it does not remove
all of them, and leaves certain other inconsistencies, such as `std::u8`
et al. which are actually stored in `src/libstd/num/` (one subdirectory
down). No quorum has been reached on this issue, so I felt it best to
leave them all alone at present. #9208 deals with the possibility of
making libstd more hierarchical (such as changing the crate to match the
current filesystem structure, which would make the module path
`std::num::u8`).
There is one thing remaining in which this repository is not
rustpkg-compliant: rustpkg would have `src/std/` et al. rather than
`src/libstd/` et al. I have not endeavoured to change that at this point
as it would guarantee prompt bitrot and confusion. A change of that
magnitude needs to be discussed first.
Similarly to the previous commit, libuv is only used by this library, so there's
no need for it to be linked into librustrt and available to all crates by
default.
Cleaned up the source in a few places
Renamed `map_move` to `map`, removed other `map` methods
Added `as_ref` and `as_mut` adapters to `Result`
Added `fmt::Default` impl
There are a few reasons that this is a desirable move to take:
1. Proof of concept that a third party event loop is possible
2. Clear separation of responsibility between rt::io and the uv-backend
3. Enforce in the future that the event loop is "pluggable" and replacable
Here's a quick summary of the points of this pull request which make this
possible:
* Two new lang items were introduced: event_loop, and event_loop_factory.
The idea of a "factory" is to define a function which can be called with no
arguments and will return the new event loop as a trait object. This factory
is emitted to the crate map when building an executable. The factory doesn't
have to exist, and when it doesn't then an empty slot is in the crate map and
a basic event loop with no I/O support is provided to the runtime.
* When building an executable, then the rustuv crate will be linked by default
(providing a default implementation of the event loop) via a similar method to
injecting a dependency on libstd. This is currently the only location where
the rustuv crate is ever linked.
* There is a new #[no_uv] attribute (implied by #[no_std]) which denies
implicitly linking to rustuv by default
Closes#5019
