Testing so far with recent Wasmtime has not been able to show the need
for avoiding the process-wide mmap lock in real-world use-cases. As
such, the technique of using an anonymous file and ftruncate() to extend
it seems unnecessary; instead, memfd can always use anonymous zeroed
memory for heap backing where the CoW image is not present, and
mprotect() to extend the heap limit by changing page protections.
As first suggested by Jan on the Zulip here [1], a cheap and effective
way to obtain copy-on-write semantics of a "backing image" for a Wasm
memory is to mmap a file with `MAP_PRIVATE`. The `memfd` mechanism
provided by the Linux kernel allows us to create anonymous,
in-memory-only files that we can use for this mapping, so we can
construct the image contents on-the-fly then effectively create a CoW
overlay. Furthermore, and importantly, `madvise(MADV_DONTNEED, ...)`
will discard the CoW overlay, returning the mapping to its original
state.
By itself this is almost enough for a very fast
instantiation-termination loop of the same image over and over,
without changing the address space mapping at all (which is
expensive). The only missing bit is how to implement
heap *growth*. But here memfds can help us again: if we create another
anonymous file and map it where the extended parts of the heap would
go, we can take advantage of the fact that a `mmap()` mapping can
be *larger than the file itself*, with accesses beyond the end
generating a `SIGBUS`, and the fact that we can cheaply resize the
file with `ftruncate`, even after a mapping exists. So we can map the
"heap extension" file once with the maximum memory-slot size and grow
the memfd itself as `memory.grow` operations occur.
The above CoW technique and heap-growth technique together allow us a
fastpath of `madvise()` and `ftruncate()` only when we re-instantiate
the same module over and over, as long as we can reuse the same
slot. This fastpath avoids all whole-process address-space locks in
the Linux kernel, which should mean it is highly scalable. It also
avoids the cost of copying data on read, as the `uffd` heap backend
does when servicing pagefaults; the kernel's own optimized CoW
logic (same as used by all file mmaps) is used instead.
[1] https://bytecodealliance.zulipchat.com/#narrow/stream/206238-general/topic/Copy.20on.20write.20based.20instance.20reuse/near/266657772
* Don't copy `VMBuiltinFunctionsArray` into each `VMContext`
This is another PR along the lines of "let's squeeze all possible
performance we can out of instantiation". Before this PR we would copy,
by value, the contents of `VMBuiltinFunctionsArray` into each
`VMContext` allocated. This array of function pointers is modestly-sized
but growing over time as we add various intrinsics. Additionally it's
the exact same for all `VMContext` allocations.
This PR attempts to speed up instantiation slightly by instead storing
an indirection to the function array. This means that calling a builtin
intrinsic is a tad bit slower since it requires two loads instead of one
(one to get the base pointer, another to get the actual address).
Otherwise though `VMContext` initialization is now simply setting one
pointer instead of doing a `memcpy` from one location to another.
With some macro-magic this commit also replaces the previous
implementation with one that's more `const`-friendly which also gets us
compile-time type-checks of libcalls as well as compile-time
verification that all libcalls are defined.
Overall, as with #3739, the win is very modest here. Locally I measured
a speedup from 1.9us to 1.7us taken to instantiate an empty module with
one function. While small at these scales it's still a 10% improvement!
* Review comments
This commit updates the allocation of a `VMExternRefActivationsTable`
structure to perform zero malloc memory allocations. Previously it would
allocate a page-size of `chunk` plus some space in hash sets for future
insertions. The main trick here implemented is that after the first gc
during the slow path the fast chunk allocation is allocated and
configured.
The motivation for this PR is that given our recent work to further
refine and optimize the instantiation process this allocation started to
show up in a nontrivial fashion. Most modules today never touch this
table anyway as almost none of them use reference types, so the time
spent allocation and deallocating the table per-store was largely wasted
time.
Concretely on a microbenchmark this PR speeds up instantiation of a
module with one function by 30%, decreasing the instantiation cost from
1.8us to 1.2us. Overall a pretty minor win but when the instantiation
times we're measuring start being in the single-digit microseconds this
win ends up getting magnified!
When we GC, we assert the invariant that all `externref`s we find on the stack
have a corresponding entry in the `VMExternRefActivationsTable`. However, we
also might be in code that is in the process of fixing up this invariant and
adding an entry to the table, but the table's bump chunk is full, and so we do a
GC and then add the entry into the table. This will ultimately maintain our
desired invariant, but there is a moment in time when we are doing the GC where
the invariant is relaxed which is okay because the reference will be in the
table before we return to Wasm or do anything else. This isn't a possible UAF,
in other words. To make it so that the assertion won't trip, we explicitly
insert the reference into the table *before* we GC, so that the invariant is not
relaxed across a possibly-GCing operation (even though it would be safe in this
particular case).
`ptr::cast` has the advantage of being unable to silently cast
`*const T` to `*mut T`. This turned up several places that were
performing such casts, which this PR also fixes.
This PR introduces a new way of performing cooperative timeslicing that
is intended to replace the "fuel" mechanism. The tradeoff is that this
mechanism interrupts with less precision: not at deterministic points
where fuel runs out, but rather when the Engine enters a new epoch. The
generated code instrumentation is substantially faster, however, because
it does not need to do as much work as when tracking fuel; it only loads
the global "epoch counter" and does a compare-and-branch at backedges
and function prologues.
This change has been measured as ~twice as fast as fuel-based
timeslicing for some workloads, especially control-flow-intensive
workloads such as the SpiderMonkey JS interpreter on Wasm/WASI.
The intended interface is that the embedder of the `Engine` performs an
`engine.increment_epoch()` call periodically, e.g. once per millisecond.
An async invocation of a Wasm guest on a `Store` can specify a number of
epoch-ticks that are allowed before an async yield back to the
executor's event loop. (The initial amount and automatic "refills" are
configured on the `Store`, just as for fuel.) This call does only
signal-safe work (it increments an `AtomicU64`) so could be invoked from
a periodic signal, or from a thread that wakes up once per period.
* Update to cap-std 0.22.0.
The main change relevant to Wasmtime here is that this includes the
rustix fix for compilation errors on Rust nightly with the `asm!` macro.
* Add itoa to deny.toml.
* Update the doc and fuzz builds to the latest Rust nightly.
* Update to libc 0.2.112 to pick up the `POLLRDHUP` fix.
* Update to cargo-fuzz 0.11, for compatibility with Rust nightly.
This appears to be the fix for rust-fuzz/cargo-fuzz#277.
This pulls in a fix for Android, where Android's seccomp policy on older
versions is to make `openat2` irrecoverably crash the process, so we have
to do a version check up front rather than relying on `ENOSYS` to
determine if `openat2` is supported.
And it pulls in the fix for the link errors when multiple versions of
rsix/rustix are linked in.
And it has updates for two crate renamings: rsix has been renamed to
rustix, and unsafe-io has been renamed to io-extras.
This commit adds the `pooling-allocator` feature to both the `wasmtime` and
`wasmtime-runtime` crates.
The feature controls whether or not the pooling allocator implementation is
built into the runtime and exposed as a supported instance allocation strategy
in the wasmtime API.
The feature is on by default for the `wasmtime` crate.
Closes#3513.
* Add a configuration option to force "static" memories
In poking around at some things earlier today I realized that one
configuration option for memories we haven't exposed from embeddings
like the CLI is to forcibly limit the size of memory growth and force
using a static memory style. This means that the CLI, for example, can't
limit memory growth by default and memories are only limited in size by
what the OS can give and the wasm's own memory type. This configuration
option means that the CLI can artificially limit the size of wasm linear
memories.
Additionally another motivation for this is for testing out various
codegen ramifications of static/dynamic memories. This is the only way
to force a static memory, by default, for wasm64 memories with no
maximum size listed for example.
* Review feedback
* Adjust dependency directives between crates
This commit is a preparation for the release process for Wasmtime. The
specific changes here are to delineate which crates are "public", and
all version requirements on non-public crates will now be done with
`=A.B.C` version requirements instead of today's `A.B.C` version
requirements.
The purpose for doing this is to assist with patch releases that might
happen in the future. Patch releases of wasmtime are already required to
not break the APIs of "public" crates, but no such guarantee is given
about "internal" crates. This means that a patch release runs the risk,
for example, of breaking an internal API. In doing so though we would
also need to release a new major version of the internal crate, but we
wouldn't have a great hole in the number scheme of major versions to do
so. By using `=A.B.C` requirements for internal crates it means we can
safely ignore strict semver-compatibility between releases of internal
crates for patch releases, since the only consumers of the crate will be
the corresponding patch release of the `wasmtime` crate itself (or other
public crates).
The `publish.rs` script has been updated with a check to verify that
dependencies on internal crates are all specified with an `=`
dependency, and dependnecies on all public crates are without a `=`
dependency. This will hopefully make it so we don't have to worry about
what to use where, we just let CI tell us what to do. Using this
modification all version dependency declarations have been updated.
Note that some crates were adjusted to simply remove their `version`
requirement in cases such as the crate wasn't published anyway (`publish
= false` was specified) or it's in the `dev-dependencies` section which
doesn't need version specifiers for path dependencies.
* Switch to normal sever deps for cranelift dependencies
These crates will now all be considered "public" where in patch releases
they will be guaranteed to not have breaking changes.
* Fix some nightly dead code warnings
Looks like the "struct field not used" lint has improved on nightly and
caught a few more instances of fields that were never actually read.
* Fix windows
* Add `*_unchecked` variants of `Func` APIs for the C API
This commit is what is hopefully going to be my last installment within
the saga of optimizing function calls in/out of WebAssembly modules in
the C API. This is yet another alternative approach to #3345 (sorry) but
also contains everything necessary to make the C API fast. As in #3345
the general idea is just moving checks out of the call path in the same
style of `TypedFunc`.
This new strategy takes inspiration from previously learned attempts
effectively "just" exposes how we previously passed `*mut u128` through
trampolines for arguments/results. This storage format is formalized
through a new `ValRaw` union that is exposed from the `wasmtime` crate.
By doing this it made it relatively easy to expose two new APIs:
* `Func::new_unchecked`
* `Func::call_unchecked`
These are the same as their checked equivalents except that they're
`unsafe` and they work with `*mut ValRaw` rather than safe slices of
`Val`. Working with these eschews type checks and such and requires
callers/embedders to do the right thing.
These two new functions are then exposed via the C API with new
functions, enabling C to have a fast-path of calling/defining functions.
This fast path is akin to `Func::wrap` in Rust, although that API can't
be built in C due to C not having generics in the same way that Rust
has.
For some benchmarks, the benchmarks here are:
* `nop` - Call a wasm function from the host that does nothing and
returns nothing.
* `i64` - Call a wasm function from the host, the wasm function calls a
host function, and the host function returns an `i64` all the way out to
the original caller.
* `many` - Call a wasm function from the host, the wasm calls
host function with 5 `i32` parameters, and then an `i64` result is
returned back to the original host
* `i64` host - just the overhead of the wasm calling the host, so the
wasm calls the host function in a loop.
* `many` host - same as `i64` host, but calling the `many` host function.
All numbers in this table are in nanoseconds, and this is just one
measurement as well so there's bound to be some variation in the precise
numbers here.
| Name | Rust | C (before) | C (after) |
|-----------|------|------------|-----------|
| nop | 19 | 112 | 25 |
| i64 | 22 | 207 | 32 |
| many | 27 | 189 | 34 |
| i64 host | 2 | 38 | 5 |
| many host | 7 | 75 | 8 |
The main conclusion here is that the C API is significantly faster than
before when using the `*_unchecked` variants of APIs. The Rust
implementation is still the ceiling (or floor I guess?) for performance
The main reason that C is slower than Rust is that a little bit more has
to travel through memory where on the Rust side of things we can
monomorphize and inline a bit more to get rid of that. Overall though
the costs are way way down from where they were originally and I don't
plan on doing a whole lot more myself at this time. There's various
things we theoretically could do I've considered but implementation-wise
I think they'll be much more weighty.
* Tweak `wasmtime_externref_t` API comments
* Exports symbols to be shared with external GDB/JIT debugging interface tools.
Windows O/S specific requirement.
* Moved comments into platform specific compiler directive sections.
* Use `__builtin_setjmp` instead of `sigsetjmp`.
Use [`__builtin_setjmp`] instead of `sigsetjmp`, as it is implemented in
the compiler, performed inline, and saves much less state. This speeds up
calls into wasm by about 8% on my machine.
[`__builtin_setjmp`]: https://gcc.gnu.org/onlinedocs/gcc/Nonlocal-Gotos.html
* Add a comment confirming that 5 really is the documented size.
* Add a comment about callee-saved state and __builtin_setjmp.
* On clang on aarch64, use sigsetjmp.
* Fix a stray `#endif`.
* Use rsix to make system calls in Wasmtime.
`rsix` is a system call wrapper crate that we use in `wasi-common`,
which can provide the following advantages in the rest of Wasmtime:
- It eliminates some `unsafe` blocks in Wasmtime's code. There's
still an `unsafe` block in the library, but this way, the `unsafe`
is factored out and clearly scoped.
- And, it makes error handling more consistent, factoring out code for
checking return values and `io::Error::last_os_error()`, and code that
does `errno::set_errno(0)`.
This doesn't cover *all* system calls; `rsix` doesn't implement
signal-handling APIs, and this doesn't cover calls made through `std` or
crates like `userfaultfd`, `rand`, and `region`.
* allow the ResourceLimiter to reject a memory grow before the
memory's own maximum.
* add a hook so a ResourceLimiter can detect any reason that
a memory grow fails, including if the OS denies additional memory
* add tests for this new functionality. I only took the time to
test the OS denial on Linux, it should be possible on Mac OS
as well but I don't have a test setup. I have no idea how to
do this on windows.
We _must not_ trigger a GC when moving refs from host code into
Wasm (e.g. returned from a host function or passed as arguments to a Wasm
function). After insertion into the table, this reference is no longer
rooted. If multiple references are being sent from the host into Wasm and we
allowed GCs during insertion, then the following events could happen:
* Reference A is inserted into the activations table. This does not trigger a
GC, but does fill the table to capacity.
* The caller's reference to A is removed. Now the only reference to A is from
the activations table.
* Reference B is inserted into the activations table. Because the table is at
capacity, a GC is triggered.
* A is reclaimed because the only reference keeping it alive was the activation
table's reference (it isn't inside any Wasm frames on the stack yet, so stack
scanning and stack maps don't increment its reference count).
* We transfer control to Wasm, giving it A and B. Wasm uses A. That's a use
after free.
To prevent uses after free, we cannot GC when moving refs into the
`VMExternRefActivationsTable` because we are passing them from the host to Wasm.
On the other hand, when we are *cloning* -- as opposed to moving -- refs from
the host to Wasm, then it is fine to GC while inserting into the activations
table, because the original referent that we are cloning from is still alive and
rooting the ref.
Currently wasm-calls work with `Result<T, Trap>` internally but `Trap`
is an enum defined in `wasmtime-runtime` which is actually quite large.
Since traps are supposed to be rare this commit changes these functions
to return a `Box<Trap>` which is un-boxed later up in the `wasmtime`
crate within a `#[cold]` function.
This makes Cranelift use the Rust `alloc` API its allocations,
rather than directly calling into `libc`, which makes it respect
the `#[global_allocator]` configuration.
Also, use `region::page::ceil` instead of having our own copies of
that logic.
* Add a `Module::deserialize_file` method
This commit adds a new method to the `wasmtime::Module` type,
`deserialize_file`. This is intended to be the same as the `deserialize`
method except for the serialized module is present as an on-disk file.
This enables Wasmtime to internally use `mmap` to avoid copying bytes
around and generally makes loading a module much faster.
A C API is added in this commit as well for various bindings to use this
accelerated path now as well. Another option perhaps for a Rust-based
API is to have an API taking a `File` itself to allow for a custom file
descriptor in one way or another, but for now that's left for a possible
future refactoring if we find a use case.
* Fix compat with main - handle readdonly mmap
* wip
* Try to fix Windows support
* Add differential fuzzing against V8
This commit adds a differential fuzzing target to Wasmtime along the
lines of the wasmi and spec interpreters we already have, but with V8
instead. The intention here is that wasmi is unlikely to receive updates
over time (e.g. for SIMD), and the spec interpreter is not suitable for
fuzzing against in general due to its performance characteristics. The
hope is that V8 is indeed appropriate to fuzz against because it's
naturally receiving updates and it also is expected to have good
performance.
Here the `rusty_v8` crate is used which provides bindings to V8 as well
as precompiled binaries by default. This matches exactly the use case we
need and at least for now I think the `rusty_v8` crate will be
maintained by the Deno folks as they continue to develop it. If it
becomes an issue though maintaining we can evaluate other options to
have differential fuzzing against.
For now this commit enables the SIMD and bulk-memory feature of
fuzz-target-generation which should enable them to get
differentially-fuzzed with V8 in addition to the compilation fuzzing
we're already getting.
* Use weak linkage for GDB jit helpers
This should help us deduplicate our symbol with other JIT runtimes, if
any. For now this leans on some C helpers to define the weak linkage
since Rust doesn't support that on stable yet.
* Don't use rusty_v8 on MinGW
They don't have precompiled libraries there.
* Fix msvc build
* Comment about execution
* Remove some allocations in `CodeMemory`
This commit removes the `FinishedFunctions` type as well as allocations
associated with trampolines when allocating inside of a `CodeMemory`.
The main goal of this commit is to improve the time spent in
`CodeMemory` where currently today a good portion of time is spent
simply parsing symbol names and trying to extract function indices from
them. Instead this commit implements a new strategy (different from #3236)
where compilation records offset/length information for all
functions/trampolines so this doesn't need to be re-learned from the
object file later.
A consequence of this commit is that this offset information will be
decoded/encoded through `bincode` unconditionally, but we can also
optimize that later if necessary as well.
Internally this involved quite a bit of refactoring since the previous
map for `FinishedFunctions` was relatively heavily relied upon.
* comments
