superdimwit

Thanks!

Clang recently got a new calling convention that makes these tail calls much cheaper (avoids the need for the caller to preserve some registers). I can never remember the name - it’s either preserve_all or preserve_none (whose perspective is the preservation from?).

On 1: this small amount of overhead matters because the amount of work you do on each opcode can be tiny. The extra jump could be 20% of your runtime! On 2: yes, this helps the indirect branch target predictor. In a real program you’ll often get repeated sequences of opcodes (increment, then compare) that can be predicted. These branch predictors can be pattern based.

Try dear-imgui + implot. There are Python bindings

There’s a Hello World on the page

30 years?

Agreed, autovec is in my experience where clang really shines.

Disappointed this wasn't an article about bath houses

One nice property is their stable addressing. A linked list element won't move once allocated.

Amsterdam is good for options market making, most of the techniques in this book come up a lot in trading

reinterpret_cast comes up a lot when handling binary protocols - e.g. reading a message struct off a buffer / packet. How would this be handled in Rust or otherwise without copying the data?

You can just define a generic fallback method like `eff(p, d1, d2) = eff(p, d2, d1)`

Could a difference in alignment of the hot loop also have an effect here?