catchsegv

Run program with segfault catching

$ catchsegv [./program]

$ catchsegv [./program] [arg1] [arg2]

catchsegv program [args...]

catchsegv is a debugging wrapper that intercepts segmentation faults in a program and prints a symbolic backtrace to help diagnose the crash. It works by using LD_PRELOAD to inject libSegFault.so, which installs a signal handler for SIGSEGV and other fatal signals.
When the wrapped program crashes, catchsegv displays a stack trace with function names, source file locations (if debug symbols are available), and memory map information. This provides a lightweight alternative to running programs under a full debugger when you only need crash diagnostics.
The tool was part of the glibc debug utilities but has been removed from glibc 2.35 and later. Modern alternatives include AddressSanitizer and core dump analysis with GDB.

SEGFAULT_OUTPUT_NAME

Output file (default: stderr)

Signals to handle: segv, ill, bus, stkflt, abrt, fpe

Use alternate stack for handler

Compile with -g flag for function names and line numbers in backtrace:

$ gcc -g -o program program.c

- GDB with core dumps
- AddressSanitizer (-fsanitize=address)
- Custom signal handlers with backtrace()
- Valgrind

Removed from glibc 2.35+. Modern alternatives include out-of-process debuggers and post-mortem analysis tools. Uses non-async-signal-safe functions in handler.

gdb(1), addr2line(1), backtrace(3)