This adds a tier-up mechanism at loop edges, making it so we can
seamlessly (ish) transition between interpreted and native code so we
can start running wasm code immediately after validation while
compilation happens in the background, and switching to native code
eventually once we hit a big enough function that would benefit from
being compiled to begin with.
Wasm opcodes only need one byte plus a 24-bit selector for
prefixed instructions. Store them in u32 instead of u64 and reject
selectors that would not fit. Update the Rust opcode generator to
accept the new integer suffixes used by Opcode.h.
These are part of the typed function references proposal, which is now
widely used by toolchains like wasm-bindgen. This makes sites like
wordsalad.online load in Ladybird.
The opcode entry declared i16x8_replace_lane with pushes = -1, but
replace_lane pops 2 (vector, lane value) and pushes 1 result vector.
Set pushes to 1 to match the other replace_lane opcodes.
This, along with moving the sources and destination out of the config
object, makes it so we don't have to double-deref to get to them on each
instruction, leading to a ~15% perf improvement on dispatch.
This still passes the values on the stack, but registers are now allowed
to cross a call boundary.
This is a very significant (>50%) improvement on the small call
microbenchmarks on my machine.
Largely combinations of i32.const and local.get.
This shaves off at most single-digit% number of instructions from
dispatch, which translates to at most ~10% reduced dispatch time.
Across most benchmarks, this gains around ~5% perf increase.
This commit adds a register allocator, with 8 available "register"
slots.
In testing with various random blobs, this moves anywhere from 30% to
74% of value accesses into predefined slots, and is about a ~20% perf
increase end-to-end.
To actually make this usable, a few structural changes were also made:
- we no longer do one instruction per interpret call
- trapping is an (unlikely) exit condition
- the label and frame stacks are replaced with linked lists with a huge
node cache size, as we only need to touch the last element and
push/pop is very frequent.
