c-language

A LWN comment discusses the relatively recent implementation of tail-call optimization in C compilers, citing historical limitations and noting that modern GCC and Clang now support it, with potential benefits for interpreter implementations.

A blog post examines an ambiguity in the C89 standard about implicit function declarations, where GCC and Clang disagree on interpretation, leading to different compile results for certain code.

An article introducing tinylisp, a Lisp interpreter written in 99 lines of C, with a companion PDF explaining how to write one yourself, covering NaN boxing and extensions.

This article explores the challenges of parsing the C `sizeof` operator, which can take either an expression or a parenthesized type, and the extra complications introduced by compound literals and postfix operators. It discusses parsing strategies and notes that C2y's newly introduced `_Countof` operator faces similar issues.

The author releases version 2.22 of ft2-clone, a portable clone of the classic Fasttracker II music tracker written in C using SDL 2, with binaries for Windows, macOS, and Linux.

QuixiAI releases embeddinggemma.c, a fast cross-platform embedding inference engine written in C, supporting multiple backends (CPU, Metal, CUDA, ROCm, SYCL) and Matryoshka embeddings with a standard HTTP API.

Describes Fil-C's InvisiCaps capability model for ensuring memory safety in C and C++ by tracking pointer permissions, while aiming for fanatical compatibility and reasonable performance.

The article explores obscure details about pre-ANSI C (K&R C) including the absence of void, different floating point types, and simplified type specifier rules. It explains how the language's context-sensitive grammar was justified by single-pass compiler constraints.

A detailed technical post exploring how to replicate Go's concurrency model in C using POSIX threads, mutexes, condition variables, and a worker pool, as part of the Solod transpiler project.

This article presents a branchless Quicksort implementation using sorting networks and discusses how modern compilers, especially Clang, optimize loops with branch-free instructions when written in the right style.

colibri is a pure C inference tool that runs the GLM-5.2 744B MoE model on ~25 GB RAM by streaming experts from disk, eliminating the need for expensive GPUs.

Colibrì is a pure C inference engine that runs the 744B GLM-5.2 MoE model on consumer hardware with ~25GB RAM by streaming experts from disk, achieving ~2.2-2.8 tokens/second with speculative decoding.

An excerpt from the tiny-c Reference Manual, describing the origins, design philosophy, and features of the tiny-c programming language, which combines elements of BASIC, LOGO, and C in a simple, integrated environment.

The article announces the results of the 2015 Underhanded C Contest, a programming contest focused on writing deceptively malicious C code, with this year's challenge involving nuclear verification and highlighting NaN poisoning attacks.

CParseC is a single-header C99 library for parser combinators inspired by Haskell's Parsec, offering zero-copy parsing, no hidden allocations, and SIMD-optimized combinators. It aims to provide a flexible, performant alternative to handwritten parsers and lex/yacc tools.

Compares Qwen3.6 27B running locally against Opus 4.8, and highlights a voxel engine built in raw C with zero frameworks.

A minimal CPU-only inference engine for Qwen 3 models implemented from scratch in pure C.

This article delves into the differences between C in the Linux kernel and ordinary userspace C, covering core techniques such as resource management, error handling, concurrency, logging, static analysis, and extensively using GNU C extensions and kernel-specific patterns.

The Linux kernel has finally removed the dangerous strncpy function after six years and over 360 patches, replacing it with safer alternatives like strscpy that guarantee NULL-termination.

A CLI tool called ffs that searches files by reading the disk directly, bypassing the OS kernel's VFS layer, offering potential speed advantages over tools like ripgrep for large, uncached directories. Supports ext4, btrfs, and APFS file systems.