Commit Graph

5 Commits

Author SHA1 Message Date
Alex Crichton
7a1b7cdf92 Implement RFC 11: Redesigning Wasmtime's APIs (#2897)
Implement Wasmtime's new API as designed by RFC 11. This is quite a large commit which has had lots of discussion externally, so for more information it's best to read the RFC thread and the PR thread.
2021-06-03 09:10:53 -05:00
Alex Crichton
5b55ba8053 Use sigsetjmp instead of setjmp (#2645)
Apparently on macOS `setjmp` manipulates the process-wide signal mask
which adds a good deal of overhead. We don't actually need this
functionality so this commit switches to using the `sig` version of
setjmp/longjmp where we can explicitly ask the signal mask to not get
preserved. This came out of poking around on #2644 and on macOS locally
thi sdropped the overhead from 721ns to 55ns.
2021-02-08 12:05:11 -06:00
Alex Crichton
962f057c8a Remove no-longer-needed C shims (#1686)
The published version of `libc` now has all that's necessary to natively
read these fields!
2020-05-12 16:01:13 -05:00
Chris Fallin
bab0c79c31 ARM64 backend, part 9 / 11: wasmtime support.
This commit adds a few odds and ends required to build wasmtime on ARM64
with the new backend. In particular, it adds:

- Support for the `Arm64Call` relocation type.
- Support for fetching the trap PC when a signal is received.
- A hook for `SIGTRAP`, which is sent by the `brk` opcode (in contrast to
  x86's `SIGILL`).

With the patch sequence up to and including this patch applied,
`wasmtime` can now compile and successfully execute code on arm64. Not
all tests pass yet, but basic Wasm/WASI tests work correctly.
2020-04-11 17:52:44 -07:00
Alex Crichton
f700efeb03 Remove C++ dependency from wasmtime (#1365)
* Remove C++ dependency from `wasmtime`

This commit removes the last wads of C++ that we have in wasmtime,
meaning that building wasmtime no longer requires a C++ compiler. It
still does require a C toolchain for some minor purposes, but hopefully
we can remove that over time too!

The motivation for doing this is to consolidate all our signal-handling
code into one location in one language so you don't have to keep
crossing back and forth when understanding what's going on. This also
allows us to remove some extra cruft that wasn't necessary from the C++
original implementation. Additionally this should also make building
wasmtime a bit more portable since it's often easier to acquire a C
toolchain than it is to acquire a C++ toolchain. (e.g. if you're
cross-compiling to a musl target)

* Typos
2020-03-20 15:21:42 -05:00