This commit performs a small cleanup in the AArch64 backend - after
the MAdd and MSub variants have been extracted, the ALUOp enum is
used purely for binary integer operations.
Also, Inst::Mov has been renamed to Inst::Mov64 for consistency.
Copyright (c) 2020, Arm Limited.
Baldrdash's API requires that there is at most one result in a register,
across all the possible register classes: in particular, it's not
possible to return an i64 value in a register while returning an v128
value in another register.
This patch adds a notion of "remaining register values", so this is
properly taking into account when choosing whether a return value may be
put into a register or not.
The logic for generation of shifts-by-immediate was not quite right. The result was that even
shifts by an amount known at compile time were being done by moving the shift immediate into %cl
and then doing a variable shift by %cl. The effect is worse than it sounds, because all of
those shift constants are small and often used in multiple places, so they were GVN'd up and
often ended up at the entry block of the function. Hence these were connected to the use points
by long live ranges which got spilled. So all in all, most of the win here comes from avoiding
spilling.
The problem was caused by this line, in the `Opcode::Ishl | Opcode::Ushr ..` case:
```
let (count, rhs) = if let Some(cst) = ctx.get_constant(inputs[1].insn) {
```
`inputs[]` appears to refer to this CLIF instruction's inputs, and bizarrely `inputs[].insn` all
refer to the instruction (the shift) itself. Hence `ctx.get_constant(inputs[1].insn)` asks
"does this shift instruction produce a constant" to which the answer is always "no", so the
shift-by-unknown amount code is always generated. The fix here is to change that expression to
```
let (count, rhs) = if let Some(cst) = ctx.get_input(insn, 1).constant {
```
`get_input`'s result conveniently includes a `constant` field of type `Option<u64>`, so we just
use that instead.
The pinned register hack didn't work because the GetPinnedReg is marked
as having side-effects, so that GVN wouldn't try to common it out.
This commit tweaks the function used during lowering to vcode, so that
the GetPinnedReg opcode is specially handled. It's a bit lame, but it
makes the hack work again.
Also, use_input needs to be a no-op for real registers.
It does this by providing an implementation of the CLIF instructions `AtomicRmw`, `AtomicCas`,
`AtomicLoad`, `AtomicStore` and `Fence`.
The translation is straightforward. `AtomicCas` is translated into x64 `cmpxchg`, `AtomicLoad`
becomes a normal load because x64-TSO provides adequate sequencing, `AtomicStore` becomes a
normal store followed by `mfence`, and `Fence` becomes `mfence`. `AtomicRmw` is the only
complex case: it becomes a normal load, followed by a loop which computes an updated value,
tries to `cmpxchg` it back to memory, and repeats if necessary.
This is a minimum-effort initial implementation. `AtomicRmw` could be implemented more
efficiently using LOCK-prefixed integer read-modify-write instructions in the case where the old
value in memory is not required. Subsequent work could add that, if required.
The x64 emitter has been updated to emit the new instructions, obviously. The `LegacyPrefix`
mechanism has been revised to handle multiple prefix bytes, not just one, since it is now
sometimes necessary to emit both 0x66 (Operand Size Override) and F0 (Lock).
In the aarch64 implementation of atomics, there has been some minor renaming for the sake of
clarity, and for consistency with this x64 implementation.
This change primarily adds the ability to lower packed `[move|load|store]` instructions (the vector types were previously unimplemented), but with the addition of the utility `Inst::[move|load|store]` functions it became possible to remove duplicated code (e.g. `stack_load` and `stack_store`) and use these utility functions elsewhere (though not exhaustively).
This upgrade pulls in one memory-allocation reduction improvement
(bytecodealliance/regalloc.rs#95). There should be no change in behavior
as a result of this.
When storing an argument to a stack location for consumption by a
callee, or storing a return value to an on-stack return slot for
consumption by the caller, the ABI implementation was properly extending
the value but was then performing a store with only the original width.
This fixes the issue by always performing a 64-bit store of the extended
value.
Issue reported by @uweigand (thanks!).
We have observed that the ABI implementations for AArch64 and x64 are
very similar; in fact, x64's implementation started as a modified copy
of AArch64's implementation. This is an artifact of both a similar ABI
(both machines pass args and return values in registers first, then the
stack, and both machines give considerable freedom with stack-frame
layout) and a too-low-level ABI abstraction in the existing design. For
machines that fit the mainstream or most common ABI-design idioms, we
should be able to do much better.
This commit factors AArch64 into machine-specific and
machine-independent parts, but does not yet modify x64; that will come
next.
This should be completely neutral with respect to compile time and
generated code performance.
This change is a pure refactoring--no change to functionality. It removes `use crate::ir::types::*` imports and uses instead `types::I32`, e.g., throughout the x64 code. Though it increases code verbosity, this change makes it more clear where the type identifiers come from (they are generated by `cranelif-codegen-meta` so without a prefix it is difficult to find their origin), avoids IDE confusion (e.g. CLion flags the un-prefixed identifiers as errors), and avoids importing unwanted identifiers into the namespace.
This change is a pure refactoring--no change to functionality. It removes newlines between the `use ...` statements in the x64 backend so that rustfmt can format them according to its convention. I noticed some files had followed a manual convention but subsequent additions did not seem to fit; this change fixes that and lightly coalesces some of the occurrences of `use a::b; use a::c;` into `use::{b, c}`.
The implementation is pretty straightforward. Wasm atomic instructions fall
into 5 groups
* atomic read-modify-write
* atomic compare-and-swap
* atomic loads
* atomic stores
* fences
and the implementation mirrors that structure, at both the CLIF and AArch64
levels.
At the CLIF level, there are five new instructions, one for each group. Some
comments about these:
* for those that take addresses (all except fences), the address is contained
entirely in a single `Value`; there is no offset field as there is with
normal loads and stores. Wasm atomics require alignment checks, and
removing the offset makes implementation of those checks a bit simpler.
* atomic loads and stores get their own instructions, rather than reusing the
existing load and store instructions, for two reasons:
- per above comment, makes alignment checking simpler
- reuse of existing loads and stores would require extension of `MemFlags`
to indicate atomicity, which sounds semantically unclean. For example,
then *any* instruction carrying `MemFlags` could be marked as atomic, even
in cases where it is meaningless or ambiguous.
* I tried to specify, in comments, the behaviour of these instructions as
tightly as I could. Unfortunately there is no way (per my limited CLIF
knowledge) to enforce the constraint that they may only be used on I8, I16,
I32 and I64 types, and in particular not on floating point or vector types.
The translation from Wasm to CLIF, in `code_translator.rs` is unremarkable.
At the AArch64 level, there are also five new instructions, one for each
group. All of them except `::Fence` contain multiple real machine
instructions. Atomic r-m-w and atomic c-a-s are emitted as the usual
load-linked store-conditional loops, guarded at both ends by memory fences.
Atomic loads and stores are emitted as a load preceded by a fence, and a store
followed by a fence, respectively. The amount of fencing may be overkill, but
it reflects exactly what the SM Wasm baseline compiler for AArch64 does.
One reason to implement r-m-w and c-a-s as a single insn which is expanded
only at emission time is that we must be very careful what instructions we
allow in between the load-linked and store-conditional. In particular, we
cannot allow *any* extra memory transactions in there, since -- particularly
on low-end hardware -- that might cause the transaction to fail, hence
deadlocking the generated code. That implies that we can't present the LL/SC
loop to the register allocator as its constituent instructions, since it might
insert spills anywhere. Hence we must present it as a single indivisible
unit, as we do here. It also has the benefit of reducing the total amount of
work the RA has to do.
The only other notable feature of the r-m-w and c-a-s translations into
AArch64 code, is that they both need a scratch register internally. Rather
than faking one up by claiming, in `get_regs` that it modifies an extra
scratch register, and having to have a dummy initialisation of it, these new
instructions (`::LLSC` and `::CAS`) simply use fixed registers in the range
x24-x28. We rely on the RA's ability to coalesce V<-->R copies to make the
cost of the resulting extra copies zero or almost zero. x24-x28 are chosen so
as to be call-clobbered, hence their use is less likely to interfere with long
live ranges that span calls.
One subtlety regarding the use of completely fixed input and output registers
is that we must be careful how the surrounding copy from/to of the arg/result
registers is done. In particular, it is not safe to simply emit copies in
some arbitrary order if one of the arg registers is a real reg. For that
reason, the arguments are first moved into virtual regs if they are not
already there, using a new method `<LowerCtx for Lower>::ensure_in_vreg`.
Again, we rely on coalescing to turn them into no-ops in the common case.
There is also a ridealong fix for the AArch64 lowering case for
`Opcode::Trapif | Opcode::Trapff`, which removes a bug in which two trap insns
in a row were generated.
In the patch as submitted there are 6 "FIXME JRS" comments, which mark things
which I believe to be correct, but for which I would appreciate a second
opinion. Unless otherwise directed, I will remove them for the final commit
but leave the associated code/comments unchanged.
* Refactor where results of compilation are stored
This commit refactors the internals of compilation in Wasmtime to change
where results of individual function compilation are stored. Previously
compilation resulted in many maps being returned, and compilation
results generally held all these maps together. This commit instead
switches this to have all metadata stored in a `CompiledFunction`
instead of having a separate map for each item that can be stored.
The motivation for this is primarily to help out with future
module-linking-related PRs. What exactly "module level" is depends on
how we interpret modules and how many modules are in play, so it's a bit
easier for operations in wasmtime to work at the function level where
possible. This means that we don't have to pass around multiple
different maps and a function index, but instead just one map or just
one entry representing a compiled function.
Additionally this change updates where the parallelism of compilation
happens, pushing it into `wasmtime-jit` instead of `wasmtime-environ`.
This is another goal where `wasmtime-jit` will have more knowledge about
module-level pieces with module linking in play. User-facing-wise this
should be the same in terms of parallel compilation, though.
The ultimate goal of this refactoring is to make it easier for the
results of compilation to actually be a set of wasm modules. This means
we won't be able to have a map-per-metadata where the primary key is the
function index, because there will be many modules within one "object
file".
* Don't clear out fields, just don't store them
Persist a smaller set of fields in `CompilationArtifacts` instead of
trying to clear fields out and dynamically not accessing them.
When one branch target label in a MachBuffer is redirected to another,
we eventually fix up branches targetting the first to refer to the
redirected target instead. Separately, we have a branch-folding
optimization that, when an unconditional branch occurs as the only
instruction in a block (right at a label) and the previous instruction
is also an unconditional branch (hence no fallthrough), we can elide
that block entirely and redirect the label. Finally, we prevented
infinite loops when resolving label aliases by chasing only one alias
deep.
Unfortunately, these three facts interacted poorly, and this is a result
of our correctness arguments assuming a fully-general "redirect" that
was not limited to one indirection level. In particular, we could have
some label A that redirected to B, then remove the block at B because it
is just a single branch to C, redirecting B to C. A would still redirect
to B, though, without chasing to C, and hence a branch to B would fall
through to the unrelated block that came after block B.
Thanks to @bnjbvr for finding this bug while debugging the x64 backend
and reducing a failure to the function in issue #2082. (This is a very
subtle bug and it seems to have been quite difficult to chase; my
apologies!)
The fix is to (i) chase redirects arbitrarily deep, but also (ii) ensure
that we do not form a cycle of redirects. The latter is done by very
carefully checking the existing fully-resolved target of the label we
are about to redirect *to*; if it resolves back to the branch that
is causing this redirect, then we avoid making the alias. The comments
in this patch make a slightly more detailed argument why this should be
correct.
Unfortunately we cannot directly test the CLIF that @bnjbvr reduced
because we don't have a way to assert anything about the machine-code
that comes after the branch folding and emission. However, the dedicated
unit tests in this patch replicate an equivalent folding case, and also
test that we handle branch cycles properly (as argued above).
Fixes#2082.
