What are deferred versus asynchronous cancellation?

A strong answer

A strong candidate frames What are deferred versus asynchronous cancellation? around the exact memory layout, compiler semantics, and hardware execution model covered in POSIX Thread Cancellation and Cleanup Handlers. Explain deferred cancellation (cancellation points only) versus asynchronous (immediate, unsafe) and why deferred is preferred.

In multi-threaded applications, long-running tasks frequently become obsolete before completing:

• A user clicks Cancel on a 500 MB file download or a ray-tracing render.
• A search query finds its answer on worker thread 3, making results from workers 1, 2, and 4 irrelevant.
• A server initiates a graceful shutdown, needing worker threads to abort idle blocking reads.

What happens when worker_tid is canceled in the middle of read_stream(buffer)?

1. The 1 MB buffer allocated on the heap is never freed (memory leak).
2. The data_lock mutex remains locked indefinitely (abandoned lock).
3. The next thread that calls pthread_mutex_lock(&data_lock) stalls forever in a permanent deadlock.

In a production interview, always connect the specification rule to its concrete runtime consequence: state what lives on the stack versus the heap or static storage segment, how the compiler or operating system validates alignments, lifetimes, and error codes, and what happens under adversarial or boundary inputs (such as zero-length buffers, integer overflow, concurrent mutation, or partial I/O). Backing your explanation with a small trace of register/memory state transitions immediately separates senior systems engineers from candidates who only memorized syntax.

What the interviewer is listening for

• Mastery of core invariant: Request target thread termination using pthread_cancel and verify PTHREAD_CANCELED exit state.
• Mastery of core invariant: Differentiate deferred cancellation points from dangerous asynchronous cancellation.
• Mastery of core invariant: Register and execute stack cleanup routines using pthread_cleanup_push and pthread_cleanup_pop.
• Hardware and ABI awareness: Explaining how sizes, alignments, calling conventions, or kernel boundaries behave on real 64-bit targets rather than relying on implementation-defined assumptions.
• Defensive engineering: Naming the exact compiler flags (-Wall -Wextra -Werror), static analyzers, or runtime sanitizers (-fsanitize=address,undefined,thread) that verify correctness in CI.

Common mistakes

• Explaining only surface syntax without tracing the underlying memory addresses, ownership lifetime, or kernel/user-space state transitions.
• Assuming undefined behavior (such as signed overflow, uninitialized reads, strict-aliasing violations, or dangling pointers) will "just work" because it passed an unoptimized -O0 debug build.
• Ignoring return codes, partial reads/writes, alignment padding, or cleanup paths when an allocation or syscall fails midway through a function.