s1ko/llamacpp-gguf-zero-dim-sigfpe
llama.cpp/ggml GGUF loader divides INT64_MAX by a file-supplied tensor dimension that may be zero: 53-byte file crashes x86-64 builds (SIGFPE). Fixed in b10398.
🎯 Vulnerability: Web Security Vulnerability
💡 Defense Fix: Implement strict input validation, contextual output encoding, and enforce least-privilege authorization controls.