utils.h

/*
 This file contains utilities shared between the different training scripts.
 In particular, we define a series of macros xxxCheck that call the corresponding
 C standard library function and check its return code. If an error was reported,
 the program prints some debug information and exits.
*/
#ifndef UTILS_H
#define UTILS_H

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

// ----------------------------------------------------------------------------
// fread convenience utils, with nice handling of error checking using macros
// simple replace fopen, fread, fclose, fseek
// with fopenCheck, freadCheck, fcloseCheck, fseekCheck

FILE *fopen_check(const char *path, const char *mode, const char *file, int line) {
    FILE *fp = fopen(path, mode);
    if (fp == NULL) {
        fprintf(stderr, "Error: Failed to open file '%s' at %s:%d\n", path, file, line);
        fprintf(stderr, "Error details:\n");
        fprintf(stderr, "  File: %s\n", file);
        fprintf(stderr, "  Line: %d\n", line);
        fprintf(stderr, "  Path: %s\n", path);
        fprintf(stderr, "  Mode: %s\n", mode);
        fprintf(stderr, "---> HINT 1: dataset files/code have moved to dev/data recently (May 20, 2024). You may have to mv them from the legacy data/ dir to dev/data/(dataset), or re-run the data preprocessing script. Refer back to the main README\n");
        fprintf(stderr, "---> HINT 2: possibly try to re-run `python train_gpt2.py`\n");
        exit(EXIT_FAILURE);
    }
    return fp;
}

#define fopenCheck(path, mode) fopen_check(path, mode, __FILE__, __LINE__)

void fread_check(void *ptr, size_t size, size_t nmemb, FILE *stream, const char *file, int line) {
    size_t result = fread(ptr, size, nmemb, stream);
    if (result != nmemb) {
        if (feof(stream)) {
            fprintf(stderr, "Error: Unexpected end of file at %s:%d\n", file, line);
        } else if (ferror(stream)) {
            fprintf(stderr, "Error: File read error at %s:%d\n", file, line);
        } else {
            fprintf(stderr, "Error: Partial read at %s:%d. Expected %zu elements, read %zu\n",
                    file, line, nmemb, result);
        }
        fprintf(stderr, "Error details:\n");
        fprintf(stderr, "  File: %s\n", file);
        fprintf(stderr, "  Line: %d\n", line);
        fprintf(stderr, "  Expected elements: %zu\n", nmemb);
        fprintf(stderr, "  Read elements: %zu\n", result);
        exit(EXIT_FAILURE);
    }
}

#define freadCheck(ptr, size, nmemb, stream) fread_check(ptr, size, nmemb, stream, __FILE__, __LINE__)

void fclose_check(FILE *fp, const char *file, int line) {
    if (fclose(fp) != 0) {
        fprintf(stderr, "Error: Failed to close file at %s:%d\n", file, line);
        fprintf(stderr, "Error details:\n");
        fprintf(stderr, "  File: %s\n", file);
        fprintf(stderr, "  Line: %d\n", line);
        exit(EXIT_FAILURE);
    }
}

#define fcloseCheck(fp) fclose_check(fp, __FILE__, __LINE__)

void fseek_check(FILE *fp, long off, int whence, const char *file, int line) {
    if (fseek(fp, off, whence) != 0) {
        fprintf(stderr, "Error: Failed to seek in file at %s:%d\n", file, line);
        fprintf(stderr, "Error details:\n");
        fprintf(stderr, "  Offset: %ld\n", off);
        fprintf(stderr, "  Whence: %d\n", whence);
        fprintf(stderr, "  File:   %s\n", file);
        fprintf(stderr, "  Line:   %d\n", line);
        exit(EXIT_FAILURE);
    }
}

#define fseekCheck(fp, off, whence) fseek_check(fp, off, whence, __FILE__, __LINE__)

// ----------------------------------------------------------------------------
// malloc error-handling wrapper util

void *malloc_check(size_t size, const char *file, int line) {
    void *ptr = malloc(size);
    if (ptr == NULL) {
        fprintf(stderr, "Error: Memory allocation failed at %s:%d\n", file, line);
        fprintf(stderr, "Error details:\n");
        fprintf(stderr, "  File: %s\n", file);
        fprintf(stderr, "  Line: %d\n", line);
        fprintf(stderr, "  Size: %zu bytes\n", size);
        exit(EXIT_FAILURE);
    }
    return ptr;
}

#define mallocCheck(size) malloc_check(size, __FILE__, __LINE__)

#endif