C Systems Programming

C Systems Programming

Systems programming in C provides direct access to operating system resources through system calls, enabling control over files, processes, signals, and inter-process communication. This skill covers essential systems programming patterns for building robust low-level applications.

File I/O Operations

C provides both standard library I/O (buffered) and system-level I/O (unbuffered) operations. Understanding when to use each is crucial for performance and correctness.

Standard I/O Functions

Standard I/O provides buffering and convenience functions for common operations.

#include <stdio.h>
#include <stdlib.h>

// Reading and writing with standard I/O
int file_copy_stdio(const char *src, const char *dst) {
    FILE *source = fopen(src, "rb");
    if (!source) {
        perror("Failed to open source");
        return -1;
    }

    FILE *dest = fopen(dst, "wb");
    if (!dest) {
        perror("Failed to open destination");
        fclose(source);
        return -1;
    }

    char buffer[4096];
    size_t bytes;
    while ((bytes = fread(buffer, 1, sizeof(buffer), source)) > 0) {
        if (fwrite(buffer, 1, bytes, dest) != bytes) {
            perror("Write failed");
            fclose(source);
            fclose(dest);
            return -1;
        }
    }

    if (ferror(source)) {
        perror("Read failed");
        fclose(source);
        fclose(dest);
        return -1;
    }

    fclose(source);
    fclose(dest);
    return 0;
}

System-Level I/O

System calls provide direct access to kernel I/O operations without buffering.

#include <fcntl.h>
#include <unistd.h>
#include <errno.h>
#include <string.h>

// Reading and writing with system calls
int file_copy_syscall(const char *src, const char *dst) {
    int source_fd = open(src, O_RDONLY);
    if (source_fd == -1) {
        perror("Failed to open source");
        return -1;
    }

    int dest_fd = open(dst, O_WRONLY | O_CREAT | O_TRUNC, 0644);
    if (dest_fd == -1) {
        perror("Failed to open destination");
        close(source_fd);
        return -1;
    }

    char buffer[4096];
    ssize_t bytes_read, bytes_written;

    while ((bytes_read = read(source_fd, buffer, sizeof(buffer))) > 0) {
        bytes_written = write(dest_fd, buffer, bytes_read);
        if (bytes_written != bytes_read) {
            perror("Write failed");
            close(source_fd);
            close(dest_fd);
            return -1;
        }
    }

    if (bytes_read == -1) {
        perror("Read failed");
        close(source_fd);
        close(dest_fd);
        return -1;
    }

    close(source_fd);
    close(dest_fd);
    return 0;
}

Process Management

Creating and managing processes is fundamental to systems programming. The fork() and exec() family of functions enable process creation and execution.

#include <sys/types.h>
#include <sys/wait.h>
#include <unistd.h>
#include <stdio.h>
#include <stdlib.h>

// Create child process and execute command
int execute_command(const char *program, char *const argv[]) {
    pid_t pid = fork();

    if (pid == -1) {
        perror("fork failed");
        return -1;
    }

    if (pid == 0) {
        // Child process
        execvp(program, argv);
        // execvp only returns on error
        perror("execvp failed");
        exit(EXIT_FAILURE);
    }

    // Parent process
    int status;
    pid_t waited = waitpid(pid, &status, 0);

    if (waited == -1) {
        perror("waitpid failed");
        return -1;
    }

    if (WIFEXITED(status)) {
        return WEXITSTATUS(status);
    } else if (WIFSIGNALED(status)) {
        fprintf(stderr, "Child terminated by signal %d\n",
                WTERMSIG(status));
        return -1;
    }

    return -1;
}

Signal Handling

Signals provide asynchronous event notification. Proper signal handling is essential for graceful shutdown and error recovery.

#include <signal.h>
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <stdbool.h>

// Global flag for signal handling
static volatile sig_atomic_t keep_running = 1;

// Signal handler for graceful shutdown
void signal_handler(int signum) {
    if (signum == SIGINT || signum == SIGTERM) {
        keep_running = 0;
    }
}

// Setup signal handlers
int setup_signals(void) {
    struct sigaction sa;
    sa.sa_handler = signal_handler;
    sigemptyset(&sa.sa_mask);
    sa.sa_flags = 0;

    if (sigaction(SIGINT, &sa, NULL) == -1) {
        perror("sigaction SIGINT");
        return -1;
    }

    if (sigaction(SIGTERM, &sa, NULL) == -1) {
        perror("sigaction SIGTERM");
        return -1;
    }

    return 0;
}

// Main loop with signal handling
void run_with_signals(void) {
    if (setup_signals() == -1) {
        return;
    }

    while (keep_running) {
        // Do work
        sleep(1);
        printf("Working...\n");
    }

    printf("Shutting down gracefully\n");
}

Inter-Process Communication

Pipes enable communication between related processes, commonly used for command pipelines and parent-child communication.

#include <unistd.h>
#include <stdio.h>
#include <string.h>
#include <sys/wait.h>

// Create a pipeline: ls | wc -l
int pipeline_example(void) {
    int pipefd[2];

    if (pipe(pipefd) == -1) {
        perror("pipe");
        return -1;
    }

    pid_t pid1 = fork();
    if (pid1 == -1) {
        perror("fork");
        return -1;
    }

    if (pid1 == 0) {
        // First child: ls
        close(pipefd[0]);  // Close read end
        dup2(pipefd[1], STDOUT_FILENO);
        close(pipefd[1]);
        execlp("ls", "ls", NULL);
        perror("execlp ls");
        exit(EXIT_FAILURE);
    }

    pid_t pid2 = fork();
    if (pid2 == -1) {
        perror("fork");
        return -1;
    }

    if (pid2 == 0) {
        // Second child: wc -l
        close(pipefd[1]);  // Close write end
        dup2(pipefd[0], STDIN_FILENO);
        close(pipefd[0]);
        execlp("wc", "wc", "-l", NULL);
        perror("execlp wc");
        exit(EXIT_FAILURE);
    }

    // Parent
    close(pipefd[0]);
    close(pipefd[1]);

    waitpid(pid1, NULL, 0);
    waitpid(pid2, NULL, 0);

    return 0;
}

File Locking

File locking prevents concurrent access issues in multi-process environments.

#include <fcntl.h>
#include <unistd.h>
#include <stdio.h>
#include <errno.h>

// Advisory file locking
int lock_file(int fd, int lock_type) {
    struct flock fl;
    fl.l_type = lock_type;    // F_RDLCK, F_WRLCK, F_UNLCK
    fl.l_whence = SEEK_SET;
    fl.l_start = 0;
    fl.l_len = 0;             // Lock entire file
    fl.l_pid = getpid();

    if (fcntl(fd, F_SETLKW, &fl) == -1) {
        perror("fcntl");
        return -1;
    }

    return 0;
}

// Write to file with exclusive lock
int write_locked(const char *filename, const char *data) {
    int fd = open(filename, O_WRONLY | O_CREAT | O_APPEND, 0644);
    if (fd == -1) {
        perror("open");
        return -1;
    }

    // Acquire exclusive lock
    if (lock_file(fd, F_WRLCK) == -1) {
        close(fd);
        return -1;
    }

    // Write data
    ssize_t written = write(fd, data, strlen(data));
    if (written == -1) {
        perror("write");
        lock_file(fd, F_UNLCK);
        close(fd);
        return -1;
    }

    // Release lock
    lock_file(fd, F_UNLCK);
    close(fd);

    return 0;
}

Directory Operations

Working with directories requires understanding directory streams and entry manipulation.

#include <dirent.h>
#include <sys/stat.h>
#include <stdio.h>
#include <string.h>

// List all files in directory recursively
void list_directory(const char *path, int indent) {
    DIR *dir = opendir(path);
    if (!dir) {
        perror("opendir");
        return;
    }

    struct dirent *entry;
    while ((entry = readdir(dir)) != NULL) {
        // Skip . and ..
        if (strcmp(entry->d_name, ".") == 0 ||
            strcmp(entry->d_name, "..") == 0) {
            continue;
        }

        // Print with indentation
        for (int i = 0; i < indent; i++) {
            printf("  ");
        }
        printf("%s", entry->d_name);

        // Check if directory
        char fullpath[1024];
        snprintf(fullpath, sizeof(fullpath), "%s/%s", path,
                 entry->d_name);

        struct stat statbuf;
        if (stat(fullpath, &statbuf) == 0) {
            if (S_ISDIR(statbuf.st_mode)) {
                printf("/\n");
                list_directory(fullpath, indent + 1);
            } else {
                printf(" (%ld bytes)\n", statbuf.st_size);
            }
        } else {
            printf("\n");
        }
    }

    closedir(dir);
}

Error Handling in System Calls

Proper error handling is critical in systems programming. System calls indicate errors through return values and errno.

#include <stdio.h>
#include <errno.h>
#include <string.h>
#include <fcntl.h>
#include <unistd.h>

// Robust file operation with error handling
int safe_file_operation(const char *filename) {
    int fd = open(filename, O_RDONLY);
    if (fd == -1) {
        switch (errno) {
            case ENOENT:
                fprintf(stderr, "File not found: %s\n", filename);
                break;
            case EACCES:
                fprintf(stderr, "Permission denied: %s\n", filename);
                break;
            default:
                fprintf(stderr, "Error opening %s: %s\n",
                        filename, strerror(errno));
        }
        return -1;
    }

    char buffer[1024];
    ssize_t bytes_read;

    while ((bytes_read = read(fd, buffer, sizeof(buffer))) > 0) {
        // Process data
        write(STDOUT_FILENO, buffer, bytes_read);
    }

    if (bytes_read == -1) {
        fprintf(stderr, "Error reading: %s\n", strerror(errno));
        close(fd);
        return -1;
    }

    if (close(fd) == -1) {
        fprintf(stderr, "Error closing file: %s\n", strerror(errno));
        return -1;
    }

    return 0;
}

Best Practices

1. Always check return values from system calls and handle errors appropriately
2. Use errno and strerror() for detailed error reporting
3. Close file descriptors and free resources in all code paths including errors
4. Use sigaction() instead of deprecated signal() for signal handling
5. Avoid blocking operations in signal handlers; use sig_atomic_t flags
6. Prefer standard I/O for buffered operations; use syscalls for direct control
7. Use advisory locks consistently across all processes accessing shared files
8. Set appropriate file permissions using umask or explicit mode bits
9. Handle EINTR errors by retrying interrupted system calls when appropriate
10. Use waitpid() to prevent zombie processes and handle child termination

Common Pitfalls

1. Forgetting to check return values from system calls leading to silent errors
2. Using signal() instead of sigaction(), missing important control flags
3. Performing complex operations in signal handlers causing race conditions
4. Not closing file descriptors in error paths, causing resource leaks
5. Mixing buffered and unbuffered I/O on same file, causing data corruption
6. Forgetting to wait for child processes, creating zombie processes
7. Using unsafe functions in signal handlers (printf, malloc, etc.)
8. Not handling EINTR errors, causing premature operation termination
9. Ignoring race conditions in file operations without proper locking
10. Using kill() without checking if process exists, sending signals to wrong processes

When to Use C Systems Programming

Use C systems programming when you need:

• Direct access to operating system resources and kernel interfaces
• Maximum control over process execution and resource management
• Low-level I/O operations without buffering overhead
• Building system utilities, daemons, or services
• Inter-process communication using pipes, signals, or shared memory
• Fine-grained control over file operations and permissions
• Signal handling for graceful shutdown and error recovery
• High-performance applications requiring minimal overhead
• Interfacing with legacy systems or porting Unix utilities
• Understanding how higher-level abstractions work under the hood

Resources

• Advanced Programming in the UNIX Environment
• The Linux Programming Interface
• POSIX Standards Documentation
• Linux System Programming