Resolving one assembly resolves its whole reference closure, and every
reference in it asked the same framework directories the same questions.
The worst of it was the scan for the closest version folder of a shared
framework: a directory listing plus a recursive file search, repeated per
reference and per runtime pack - 42 scans for two distinct answers when
decompiling ICSharpCode.ILSpyX.dll.
The scan result is only safe to keep for a bounded time: a runtime can be
installed or removed while ILSpy runs, and reloading an assembly list has
to see that. So it is kept for the length of an explicitly opened scope,
which the type system opens around the closure it resolves and closes
again afterwards; outside a scope the file system is read as before. The
scope owns what was read, so two of them on one resolver do not stack -
the first to end takes it, and the other reads the file system again.
BeginSnapshot is on IAssemblyResolver rather than an interface of its
own: it is core functionality of a resolver, and one implementation is
not an abstraction. This breaks the interface for implementors outside
this repository, who opt out by returning null - which is what the three
resolvers here that hold nothing do.
The remaining probes cost nothing to fix: the preferred runtime pack was
listed among the defaults it already belongs to, so its directory was
scanned twice for every reference that is not in it, and one package
folder was probed once per assembly the package contains.
Measured over 27 references with a fresh resolver each time: 3.3 ms per
assembly before, 3.1 ms without a scope, 1.1 ms with one.
Assisted-by: Claude:claude-opus-5:Claude Code
A chain of type forwarders is followed by assembly name, and every name resolves relative
to the assembly being decompiled - so a chain that leaves for another framework can be
pulled straight back into the directory it started in. A .NET Standard 2.0 assembly
sitting among .NET Framework 4.6.1 facades lost System.Linq.Enumerable that way: the
chain went netstandard -> System.Core (from the shared framework) -> System.Linq (back to
the input directory) -> netstandard, arriving at an assembly it had already passed
through. Nothing in the closure defines the type, so it stayed unknown and every LINQ
call decompiled as a static call with a delegate cast.
Once the closure is loaded, chains that return to an assembly they already visited are
walked a second time, resolving each hop next to the assembly that forwards it. The
assembly ending the repaired chain is loaded only once it is confirmed to declare the
type; it then wins the deduplication against the assembly of the same name it displaces,
which version order says nothing about. An assembly that neither forwards nor declares
the type ends the walk with nothing loaded, so a failed repair cannot displace anything.
Only chains that are already broken are walked twice. Preferring the forwarder's own
directory as a resolution policy was tried first and rejected: measured against a corpus,
it moved a .NET 8 facade's System.Runtime reference out of a net4x compilation, splitting
type identities so that overrides printed as virtual. The two cases cannot be told apart
where references are resolved, because that layer sees assembly names, not the type whose
chain is or is not terminating.
AssemblyReference now knows the module that declares it, which is what lets a hop be
resolved next to its forwarder, and its metadata reader is that module's.
A chain that cannot be repaired is reported in the reference load log the UI already
shows, once per reference: a facade forwards hundreds of types and they all fail together.
Assisted-by: Claude:claude-opus-5:Claude Code
TypeSystemInitStop now reports the size of the final reference set passed
to Init() instead of the raw resolve count, which included same-name
duplicates that the version dedup later drops. The schema smoke tests
asserted exact global event counts, but the providers are process-wide
and the decompiler fixtures run in parallel, so unrelated decompilations
could inflate the counts; every string payload now carries a unique
marker (and the string-less snapshot events a sentinel count) that the
assertions filter on.
Assisted-by: Claude:claude-fable-5:Claude Code
The events introduced in #2519 timed only the five per-entity DoDecompile
overloads, allocated a Stopwatch and the member's FullName even when no
trace session was attached, and reported whole milliseconds, which rounds
almost every member to zero. Flat one-shot events also gave PerfView and
dotnet-trace no way to show durations or nesting, and the actually
expensive stages (type system initialization, assembly resolution probing,
the IL/AST transform pipelines, whole-project decompilation) were not
instrumented at all.
Start/Stop event pairs let trace viewers derive duration and nesting from
event timestamps, keywords let a session enable only the areas of
interest, and every call site is gated on IsEnabled() so tracing costs a
branch when disabled. Per-transform events are Verbose because of their
volume; unlike the STEP/Stepper mechanism they work in Release builds.
EventSource is in-box for netstandard2.0 and flows over both ETW and
EventPipe, so this stays cross-platform with no new dependency.
Assisted-by: Claude:claude-fable-5:Claude Code
Runtime async is a compiler feature that emits ordinary async/await (a
C# 5 construct), so reconstructing it should not require selecting C# 15.
The dedicated RuntimeAsync setting was also redundant: AsyncAwaitDecompiler
already runs the runtime-async transforms only when AsyncAwait is enabled.
Fold the behavior into the AsyncAwait setting and drop the separate toggle.
Assisted-by: Claude:claude-opus-4-8:Claude Code
Detect MethodImplOptions.Async (0x2000) in ILReader and unpack Task/Task<T>
return types so the IL Leave value and function.AsyncReturnType match the
source signature. Add CSharp15_0 (Preview also bumped to 1500) and a
RuntimeAsync setting (default on, gated to >=CSharp15_0), expose it in the
Languages dropdown, mask the synthetic MethodImplAsync bit out of the
decompiled [MethodImpl], and add a .runtimeasync test suffix.
DecompilerTypeSystem uses this to resolve/load multiple assemblies in parallel.
Unfortunately this doesn't gain us any performance yet in ILSpy because there we have a global assembly-loader-lock :(
While support for multi-module assemblies isn't fully working yet; it is clear at this point that we want
to treat each module in a multi-module assembly separately for the purposes of the type system.
This is a work-in-progress; ILSpy is not functional with this commit.
The old code path still exists but is broken because some classes were modified for the new system.
The new system is still highly incomplete (types only have fields, but no methods).
Siegfried Pammer
Christoph Wille
Peter Crabtree
ds5678
Daniel Grunwald
dotnet format
Andrey Shchekin