Tag
A blog post reveals a memory over-allocation bug in Zig's std.Io.Writer.Allocating due to the `drain` function incorrectly reserving space for the splat parameter on every data slice, causing unexpected memory growth.
This blog post details three recent optimizations to LLVM's BumpPtrAllocator, reducing fast-path overhead by removing redundant alignment, null pointer checks, and per-allocation accounting, resulting in improved performance for Clang, lld, and other LLVM components.
This video uses disassembly and the WinDbg debugging tool to reveal that the eight mystery bytes that appear when calloc allocates memory in an x86 environment are actually the entry header of the Windows heap manager, containing information such as the current block size and the previous block size.
Stumpalo is a new high-performance bump allocator for Rust that benchmarks significantly faster than existing alternatives like blink and bumpalo across a wide range of allocation operations. It also features scoped stack support and is available as a Rust crate.
This blog post explains the current double free detection mechanism in glibc malloc's per-thread tcache, identifies a flaw that allows cross-thread double frees, and proposes a potential fix using randomized keys per thread.
The article introduces Raven, a new State Space Model (SSM) with selective memory allocation that achieves state-of-the-art performance on recall tasks and demonstrates superior length generalization compared to existing models like SWA.