Create Files Using open(), write(), close() System Calls

Topic Overview

Files via Kernel Descriptors

Creating a file is open(path, O_CREAT|O_WRONLY, mode) followed by write and close. The kernel allocates an fd table slot, creates the inode on disk or tmpfs, and returns a small integer handle.

Descriptor Lifecycle

Sequence:

• open returns fd 3 (after stdin/out/err)
• write(3, buf, len) copies to page cache
• close(3) drops reference count

For example, writing 5 bytes to /tmp/test.txt creates the dentry if missing when O_CREAT is set with mode 0644.

Why This Matters in Production

open(), write(), and close() are the exact three syscalls every C library's fopen()/fwrite()/fclose() eventually wraps, and the O_CREAT|O_TRUNC flag combination here is what tools like logrotate rely on to atomically replace files. Skipping the return-value check on open() is one of the most common real-world causes of silent data loss in production logging pipelines.

Debugging Checklist

Before you call the implementation done, walk failure modes on purpose. Test empty structures, single-element edge cases, maximum concurrency, and errno paths that must not crash the program. OS code usually fails in production when happy-path tests pass but invariants break under contention or memory pressure.

Implementation Notes

Keep structures small and name fields after kernel counterparts when possible. That lets you read man pages and kernel source side by side while you work. Print observable events during development; remove noisy logs once tests pass reliably.

For This Exercise

You will create files with raw syscalls or libc wrappers and verify contents with read. The task asks you to handle EEXIST when O_EXCL is requested on an existing path.

Your Task

Create and modify files with the raw POSIX system calls open(), write(), read() and close(), using the flag combinations real programs use: O_CREAT | O_TRUNC to replace, O_APPEND to add, and O_CREAT | O_EXCL to create only if absent (how lock files and mkstemp avoid races). Check every return value and report failures by their errno name.

Sandbox

At start-up, create a private working directory with mkdtemp("/tmp/fsXXXXXX") (falling back to "fsXXXXXX" in the current directory if /tmp isn't writable), chdir into it and call umask(022). Every name in the input is a plain file name inside that directory. Delete everything you created before exiting. Never print the directory's path.

Commands

TEXT is the rest of the line after one space and may be empty. It is written without a newline.

Errors

Print COMMAND NAME: SYSCALL failed: ENAME, where SYSCALL is the call that failed (open, write, read, stat, chmod, unlink) and ENAME is the symbolic errno (ENOENT, EEXIST, EBADF, EISDIR, EACCES, …). Close any descriptor you opened even when a later step fails.

Example

Input:

cCopy
Loading…

Output:

cCopy
Loading…

Requirements

1. Use open/read/write/close from <fcntl.h> and <unistd.h>, not fopen.
2. Map errno to its name yourself (a small table covering the codes above), and print on stdout.

Hidden tests cover truncation by a second create, excl creating a new file with mode 0600, chmod, writing through a read-only descriptor (EBADF), rm of a missing file, and cat of an empty file.