p0nce

Adoptée à la va-vite par une sorte de **cabinet noir** de la fusion, on trouve sur ce site bien connu de tous (http://bitoduc.fr/) le terme _« métascript »_ comme traduction...

Mandataire côté client ?

1

On pourrait peut-être créer un mandataire côté client. Imaginons une page spéciale : `http://bitoduc.fr/mandataire.html?adresse=http://serveur.exemple.com` En page statique, il s'agit juste d'une page avec une bannière `bitoduc.fr`. Mais en Javascript, langage...

LFENCE not implementable with MFENCE

1

Currently LFENCE is modelled by a full memory barrier on LLVM with `llvm_memory_fence` which generates MFENCE. It was assumed MFENCE is more restrictive than LFENCE, which is a performance issue...

~~165 TODO left~~ ~~132 TODO left~~ ~~123 TODO left~~ ~~105 TODO left~~ 91 TODO left

DMD: Support D_SIMD

14

Enabling core.simd: - [x] We can enable core.simd usage with DMD today, without even using D_SIMD, which brings the performance gap of LDC vs DMD from 20x to 4x. DMD...

Deprecate loadUnaligned and storeUnaligned from public API

2

Concrete context always has a type and size information, so there isn't such a need for those in the public API. You can always use `_mm_loadu_ps` and friends.

#BONUS intrinsics that might be useful

2

Add one here every time you wish for one: - [ ] `_mm_cvtpd_epi64` that would convert 2x double using MXCSR would speed up things for arm and non-avx x86 =>...

Could _mm_pow_s/ps support 0.0f input?

1

It can make time-consuming bugs downstream, as people expect 0.0f ^ something to yield 0, but instead it yield a very large value.

Optimization opportunities probably. Also: disabled with GDC

Safety-washing

4

Function that takes a pointer which is then accessed too greedily should be `@system`. This is breaking unfortunately. Find and fix all such functions that either cast a pointer to...