linux读写二进制大文件

https://raw.githubusercontent.com/ceph/ceph/master/src/common/safe_io.c

一般而言，用pwrite和pread读写磁盘文件不需要用循环，但是在读写超大的文件时就一定需要循环，保证正确读写

// -*- mode:C++; tab-width:8; c-basic-offset:2; indent-tabs-mode:t -*-
// vim: ts=8 sw=2 smarttab
/*
 * Ceph - scalable distributed file system
 *
 * Copyright (C) 2011 New Dream Network
 *
 * This is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License version 2.1, as published by the Free Software
 * Foundation.  See file COPYING.
 *
 */

#define _XOPEN_SOURCE 500

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

#include "common/safe_io.h"
#include "include/compat.h"

ssize_t safe_read(int fd, void *buf, size_t count)
{
	size_t cnt = 0;

	while (cnt < count) {
		ssize_t r = read(fd, buf, count - cnt);
		if (r <= 0) {
			if (r == 0) {
				// EOF
				return cnt;
			}
			if (errno == EINTR)
				continue;
			return -errno;
		}
		cnt += r;
		buf = (char *)buf + r;
	}
	return cnt;
}

ssize_t safe_read_exact(int fd, void *buf, size_t count)
{
        ssize_t ret = safe_read(fd, buf, count);
	if (ret < 0)
		return ret;
	if ((size_t)ret != count)
		return -EDOM;
	return 0;
}
 
ssize_t safe_write(int fd, const void *buf, size_t count)
{
	while (count > 0) {
		ssize_t r = write(fd, buf, count);
		if (r < 0) {
			if (errno == EINTR)
				continue;
			return -errno;
		}
		count -= r;
		buf = (char *)buf + r;
	}
	return 0;
}

ssize_t safe_pread(int fd, void *buf, size_t count, off_t offset)
{
	size_t cnt = 0;
	char *b = (char*)buf;

	while (cnt < count) {
		ssize_t r = pread(fd, b + cnt, count - cnt, offset + cnt);
		if (r <= 0) {
			if (r == 0) {
				// EOF
				return cnt;
			}
			if (errno == EINTR)
				continue;
			return -errno;
		}

		cnt += r;
	}
	return cnt;
}

ssize_t safe_pread_exact(int fd, void *buf, size_t count, off_t offset)
{
	ssize_t ret = safe_pread(fd, buf, count, offset);
	if (ret < 0)
		return ret;
	if ((size_t)ret != count)
		return -EDOM;
	return 0;
}

ssize_t safe_pwrite(int fd, const void *buf, size_t count, off_t offset)
{
	while (count > 0) {
		ssize_t r = pwrite(fd, buf, count, offset);
		if (r < 0) {
			if (errno == EINTR)
				continue;
			return -errno;
		}
		count -= r;
		buf = (char *)buf + r;
		offset += r;
	}
	return 0;
}

#ifdef CEPH_HAVE_SPLICE
ssize_t safe_splice(int fd_in, loff_t *off_in, int fd_out, loff_t *off_out,
		    size_t len, unsigned int flags)
{
  size_t cnt = 0;

  while (cnt < len) {
    ssize_t r = splice(fd_in, off_in, fd_out, off_out, len - cnt, flags);
    if (r <= 0) {
      if (r == 0) {
	// EOF
	return cnt;
      }
      if (errno == EINTR)
	continue;
      return -errno;
    }
    cnt += r;
  }
  return cnt;
}

ssize_t safe_splice_exact(int fd_in, loff_t *off_in, int fd_out,
			  loff_t *off_out, size_t len, unsigned int flags)
{
  ssize_t ret = safe_splice(fd_in, off_in, fd_out, off_out, len, flags);
  if (ret < 0)
    return ret;
  if ((size_t)ret != len)
    return -EDOM;
  return 0;
}
#endif

int safe_write_file(const char *base, const char *file,
		    const char *val, size_t vallen)
{
  int ret;
  char fn[PATH_MAX];
  char tmp[PATH_MAX];
  int fd;

  // does the file already have correct content?
  char oldval[80];
  ret = safe_read_file(base, file, oldval, sizeof(oldval));
  if (ret == (int)vallen && memcmp(oldval, val, vallen) == 0)
    return 0;  // yes.

  snprintf(fn, sizeof(fn), "%s/%s", base, file);
  snprintf(tmp, sizeof(tmp), "%s/%s.tmp", base, file);
  fd = open(tmp, O_WRONLY|O_CREAT|O_TRUNC, 0644);
  if (fd < 0) {
    ret = errno;
    return -ret;
  }
  ret = safe_write(fd, val, vallen);
  if (ret) {
    VOID_TEMP_FAILURE_RETRY(close(fd));
    return ret;
  }

  ret = fsync(fd);
  if (ret < 0) ret = -errno;
  VOID_TEMP_FAILURE_RETRY(close(fd));
  if (ret < 0) {
    unlink(tmp);
    return ret;
  }
  ret = rename(tmp, fn);
  if (ret < 0) {
    ret = -errno;
    unlink(tmp);
    return ret;
  }

  fd = open(base, O_RDONLY);
  if (fd < 0) {
    ret = -errno;
    return ret;
  }
  ret = fsync(fd);
  if (ret < 0) ret = -errno;
  VOID_TEMP_FAILURE_RETRY(close(fd));

  return ret;
}

int safe_read_file(const char *base, const char *file,
		   char *val, size_t vallen)
{
  char fn[PATH_MAX];
  int fd, len;

  snprintf(fn, sizeof(fn), "%s/%s", base, file);
  fd = open(fn, O_RDONLY);
  if (fd < 0) {
    return -errno;
  }
  len = safe_read(fd, val, vallen);
  if (len < 0) {
    VOID_TEMP_FAILURE_RETRY(close(fd));
    return len;
  }
  // close sometimes returns errors, but only after write()
  VOID_TEMP_FAILURE_RETRY(close(fd));

  return len;
}

下面是自己写的代码，为了验证read/writeBlock的正确性，分两批写入文件，然后又分两批读入文件，实际工程中只要一次读写就行了

#include <unistd.h>
#include <iostream>
#include <stdlib.h>
#include <string.h>
#include <stdint.h>
#include <string>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <fstream>
#include <vector>
#include <sstream>
#include <errno.h>

using namespace std;



bool writeBlock(int fd, int64_t &fdOft,int8_t *arr, int64_t len) {
    int64_t num = 0;
    int64_t oft = 0;
    int8_t *arr1 = (int8_t*)arr;
    while (oft < len) {
        num = pwrite(fd, arr1+oft, len-oft, fdOft+oft);
        if (num == -1) {
            return false;
        }
        oft += num;
    }
    fdOft += oft;

    return true;
}

bool readBlock(int fd, int64_t &fdOft, void *arr, int64_t len) {
    int64_t num = 0;
    int64_t oft = 0;
    int8_t *b = (int8_t*)arr;
    while (oft < len) {
        num = pread(fd, b + oft, len - oft, fdOft + oft);
        if (num <= 0) {
            if (num == 0) {
                if (oft != len) return false;
                else return true;
            }
            if (errno == EINTR) {
                continue;//the call was interrupted by a signal befor any data was read
            }
            return false;
        }

        oft += num;
    }
    fdOft += oft;

    return true;
}

int main(int argc, char *argv[]) {
    int64_t len = atol(argv[1]);
    cout << "len = "<<len <<endl;
    int fd = open(argv[2], O_RDWR|O_CREAT|O_TRUNC, S_IRWXU);
    if (fd == -1) {
        cout << "not open\n";
        return -1;
    }

    int64_t total = sizeof(double)*len;
    cout << "total = " << total <<endl;
    double *arr = (double*)malloc(total);
    if (arr == NULL) {
        cout << "malloc err\n";
        return -1;
    }

    arr[0] = 0.001;
    for (int64_t i = 1; i < len; ++i) {
        arr[i] = arr[i-1] + 0.001;
    }

    //int64_t num = pwrite(fd,arr,total,0);
    int64_t fdOft = 0;
    int64_t first = total/2;
    int64_t second = total - first;
    bool flag = writeBlock(fd, fdOft,(int8_t*)arr, first);
    if (!flag) {
        cout << "first no\n";
    }
    else {
        cout << "first yes\n";
    }
	
	flag = writeBlock(fd, fdOft, (int8_t*)arr + first, second);
    //cout << num<<endl;
	if (!flag) {
        cout << "second no\n";
    }
    else {
        cout << "second yes\n";
    }

	close(fd);

    int fdr = open(argv[2], O_RDONLY);
    double *arr1 = (double*)malloc(total);
    if (arr1 == NULL) {
        cout << "malloc error\n";
        return -1;
    }
    
    int64_t readOft = 0;
    if (readBlock(fdr, readOft, arr1, first)) {
        cout << "first read yes\n";
    }

    if (readBlock(fdr, readOft, (int8_t*)arr1 + first, second)) {
        cout << "second read yes\n";
    }

    flag = true;
    for (int64_t i = 0; i < len; ++i) {
        if (arr1[i] != arr[i]) {
            flag = false;
            cout << i << " no equal\n";
        }
      
    }

    if (flag) cout << "equal\n";

    close(fdr);

    return 1;
}