Viewing profile — andyayers

Microsoft | Compiler Engineer, .NET JIT | C++, C# | Redmond WA / Other locations possible We are a small tight-knit team, happy to both teach and learn new ways of making code run …

.NET JIT developer here. What else would we call it? It is what it is. I believe there are some differences between what .NET does and what mainstream Java does. For instance, obje…

Sometimes, yes. Linq contains a goodly number of hand-crafted special-case enumerators for common collections, or collections with certain interfaces, or span projections that are …

Here is one that has some historical comparison, though it does not show perf on Framework, and no .NET 10 yet. https://endjin.com/blog/2024/11/how-dotnet-9-boosted-ais-dot...

HPUX compilers were doing this back in 1993.

Unfortunately, those improvements don't work for Linq. Some notes on why this is so here: https://github.com/dotnet/runtime/blob/main/docs/design/core...

There indeed is a regression if the method is only called a few times. But not if it is called frequently. With BenchmarkDotNet it may not be obvious which scenario you intend to m…

Long-running methods (like the one here) transition mid-execution to more optimized versions, via on-stack replacement (OSR), after roughly 50K iterations. So you end up running op…

For the .NET JIT, at least, speculation on types seems beneficial even if we're only right maybe 30% of the time. See eg https://github.com/dotnet/runtime/blob/main/docs/design/cor…

Or https://learn.microsoft.com/en-us/defender-endpoint/microsof... (DevDrive + Defender's "performance mode")

In .NET, even in optimized methods, there can be "untracked" lifetimes where a stack slot is reported live to GC throughout the extent of a method, so presumably these can lead to …

Not these days, but yes, years ago.

There are a few interesting places where Galois Theory touches on compilation/programming. Abstract interpretation models a potentially infinite set of program behaviors onto a sim…

Something closer to a "pure codegen/runtime" example perhaps: I have data showing Roslyn (the C# compiler, itself written in C#) speeds up between ~2x and ~3x running on .NET 8 vs …

Feel free to open an issue for this over at github.com/dotnet/runtime, if you have code you can share.

IIRC with .NET 7 we will be able to improve this as we can now dump the jit-generated assembly from official builds. .NET 7 will be officially released in a month or so. Currently …

Yes, we need to provide more guidance. For now, you might find more info by searching for COMPlus_TieredPGO. The DOTNET_ prefix is a recent addition.

Please let us know how well (or poorly) it works for you.

FWIW [peachpie]( https://www.peachpie.io/2020/09/peachpie-1-0-preview1.html ) is a fairly complete PHP stack running on top of .NET core, leveraging the .NET runtime and jit. It ca…

FWIW, MSVC has had this ability for a few years now. It helps with both normal compilation unit compiles and with link-time code generation. See for instance Bruce Dawson's notes: …