2020-12-22 18:25:01

by paulo.alvarez

[permalink] [raw]
Subject: e2fsprogs: Add windows IO manager

This patch series add an IO manager for Window, it is a first step in
a work that has allowed us to format, read and write to ext4 filesystems
on Windows. Any comments are appreciated.



2020-12-22 18:25:11

by paulo.alvarez

[permalink] [raw]
Subject: [PATCH 2/3] Code adaptation to use the Windows IO manager

From: Paulo Antonio Alvarez <[email protected]>

---
include/mingw/unistd.h | 51 ++++++++++++++++++++++++++++++++++++----
lib/ext2fs/ext2_io.h | 5 ++++
lib/ext2fs/getsectsize.c | 15 ++++++++++++
lib/ext2fs/getsize.c | 16 +++++++++----
lib/support/plausible.c | 7 +++++-
util/subst.c | 4 ++++
6 files changed, 88 insertions(+), 10 deletions(-)

diff --git a/include/mingw/unistd.h b/include/mingw/unistd.h
index 9c0dc81a..b2018584 100644
--- a/include/mingw/unistd.h
+++ b/include/mingw/unistd.h
@@ -1,13 +1,54 @@
-
#pragma once

+// Copyright transferred from Raider Solutions, Inc to
+// Kern Sibbald and John Walker by express permission.
+//
+// Copyright (C) 2004-2006 Kern Sibbald
+// Copyright (C) 2014 Adam Kropelin
+//
+// This program is free software; you can redistribute it and/or
+// modify it under the terms of the GNU General Public License as
+// published by the Free Software Foundation; either version 2 of
+// the License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+// General Public License for more details.
+//
+// You should have received a copy of the GNU General Public
+// License along with this program; if not, write to the Free
+// Software Foundation, Inc., 59 Temple Place - Suite 330, Boston,
+// MA 02111-1307, USA.
+
+#ifndef __COMPAT_UNISTD_H_
+#define __COMPAT_UNISTD_H_
+
#include_next <unistd.h>

-__inline __uid_t getuid(void){return 0;}
-__inline int geteuid(void){return 1;}
+#define _PC_PATH_MAX 1
+#define _PC_NAME_MAX 2
+
+#ifdef __cplusplus
+extern "C" {
+#endif

-__inline __gid_t getgid(void){return 0;}
-__inline __gid_t getegid(void){return 0;}
+long pathconf(const char *, int);
+#define getpid _getpid
+#define getppid() 0
+
+unsigned int sleep(unsigned int seconds);
+
+#define getuid() 0
+#define getgid() 0
+#define geteuid() 1
+#define getegid() 0

// no-oped sync
__inline void sync(void){};
+
+#ifdef __cplusplus
+};
+#endif
+
+#endif /* __COMPAT_UNISTD_H_ */
\ No newline at end of file
diff --git a/lib/ext2fs/ext2_io.h b/lib/ext2fs/ext2_io.h
index 5540900a..f605f901 100644
--- a/lib/ext2fs/ext2_io.h
+++ b/lib/ext2fs/ext2_io.h
@@ -139,9 +139,14 @@ extern errcode_t io_channel_cache_readahead(io_channel io,
unsigned long long block,
unsigned long long count);

+#ifdef _WIN32
+/* windows_io.c */
+extern io_manager windows_io_manager;
+#else
/* unix_io.c */
extern io_manager unix_io_manager;
extern io_manager unixfd_io_manager;
+#endif

/* sparse_io.c */
extern io_manager sparse_io_manager;
diff --git a/lib/ext2fs/getsectsize.c b/lib/ext2fs/getsectsize.c
index d6bc3767..3a461eb9 100644
--- a/lib/ext2fs/getsectsize.c
+++ b/lib/ext2fs/getsectsize.c
@@ -51,6 +51,11 @@
*/
errcode_t ext2fs_get_device_sectsize(const char *file, int *sectsize)
{
+#ifdef _WIN64
+ *sectsize = 512; // just guessing
+ return 0;
+#else // not _WIN64
+
int fd;

fd = ext2fs_open_file(file, O_RDONLY, 0);
@@ -72,6 +77,8 @@ errcode_t ext2fs_get_device_sectsize(const char *file, int *sectsize)
*sectsize = 0;
close(fd);
return 0;
+
+#endif // ifdef _WIN64
}

/*
@@ -110,6 +117,12 @@ int ext2fs_get_dio_alignment(int fd)
*/
errcode_t ext2fs_get_device_phys_sectsize(const char *file, int *sectsize)
{
+#ifdef _WIN64
+
+ return ext2fs_get_device_sectsize(file, sectsize);
+
+#else // not _WIN64
+
int fd;

fd = ext2fs_open_file(file, O_RDONLY, 0);
@@ -133,4 +146,6 @@ errcode_t ext2fs_get_device_phys_sectsize(const char *file, int *sectsize)
*sectsize = 0;
close(fd);
return 0;
+
+#endif // ifdef _WIN64
}
diff --git a/lib/ext2fs/getsize.c b/lib/ext2fs/getsize.c
index be067755..72656821 100644
--- a/lib/ext2fs/getsize.c
+++ b/lib/ext2fs/getsize.c
@@ -71,6 +71,8 @@
#define HAVE_GET_FILE_SIZE_EX 1
#endif

+HANDLE windows_get_handle(io_channel channel);
+
errcode_t ext2fs_get_device_size2(const char *file, int blocksize,
blk64_t *retblocks)
{
@@ -84,12 +86,17 @@ errcode_t ext2fs_get_device_size2(const char *file, int blocksize,
DWORD filesize;
#endif /* HAVE_GET_FILE_SIZE_EX */

- dev = CreateFile(file, GENERIC_READ,
- FILE_SHARE_READ | FILE_SHARE_WRITE ,
- NULL, OPEN_EXISTING, FILE_ATTRIBUTE_NORMAL, NULL);
+ io_channel data_io = 0;
+ int retval;
+
+ retval = windows_io_manager->open(file, 0, &data_io);
+ if (retval)
+ return retval;

+ dev = windows_get_handle(data_io);
if (dev == INVALID_HANDLE_VALUE)
return EBADF;
+
if (DeviceIoControl(dev, IOCTL_DISK_GET_PARTITION_INFO,
&pi, sizeof(PARTITION_INFORMATION),
&pi, sizeof(PARTITION_INFORMATION),
@@ -120,7 +127,8 @@ errcode_t ext2fs_get_device_size2(const char *file, int blocksize,
}
#endif /* HAVE_GET_FILE_SIZE_EX */

- CloseHandle(dev);
+ windows_io_manager->close(data_io);
+
return 0;
}

diff --git a/lib/support/plausible.c b/lib/support/plausible.c
index 024f205e..2a3ae140 100644
--- a/lib/support/plausible.c
+++ b/lib/support/plausible.c
@@ -103,7 +103,12 @@ static void print_ext2_info(const char *device)
time_t tm;

retval = ext2fs_open2(device, 0, EXT2_FLAG_64BITS, 0, 0,
- unix_io_manager, &fs);
+#ifdef _WIN64
+ windows_io_manager,
+#else
+ unix_io_manager,
+#endif
+ &fs);
if (retval)
return;
sb = fs->super;
diff --git a/util/subst.c b/util/subst.c
index 66d7d9a9..c0eda5cf 100644
--- a/util/subst.c
+++ b/util/subst.c
@@ -434,16 +434,20 @@ int main(int argc, char **argv)
printf("Using original atime\n");
set_utimes(outfn, fileno(old), tv);
}
+#ifndef _WIN64
if (ofd >= 0)
(void) fchmod(ofd, 0444);
+#endif
fclose(out);
if (unlink(newfn) < 0)
perror("unlink");
} else {
if (verbose)
printf("Creating or replacing %s.\n", outfn);
+#ifndef _WIN64
if (ofd >= 0)
(void) fchmod(ofd, 0444);
+#endif
fclose(out);
if (old)
fclose(old);
--
2.17.1

2020-12-22 18:25:40

by paulo.alvarez

[permalink] [raw]
Subject: [PATCH 1/3] Add a Windows implementation of the IO manager

From: Paulo Antonio Alvarez <[email protected]>

---
lib/ext2fs/windows_io.c | 1041 +++++++++++++++++++++++++++++++++++++++
1 file changed, 1041 insertions(+)
create mode 100644 lib/ext2fs/windows_io.c

diff --git a/lib/ext2fs/windows_io.c b/lib/ext2fs/windows_io.c
new file mode 100644
index 00000000..68b5571b
--- /dev/null
+++ b/lib/ext2fs/windows_io.c
@@ -0,0 +1,1041 @@
+/*
+ * windows_io.c --- This is the Windows implementation of the I/O manager.
+ *
+ * Implements a one-block write-through cache.
+ *
+ * Copyright (C) 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001,
+ * 2002 by Theodore Ts'o.
+ *
+ * %Begin-Header%
+ * This file may be redistributed under the terms of the GNU Library
+ * General Public License, version 2.
+ * %End-Header%
+ */
+
+#include <windows.h>
+#include <winioctl.h>
+#include <io.h>
+
+#if !defined(__FreeBSD__) && !defined(__NetBSD__) && !defined(__OpenBSD__)
+#define _XOPEN_SOURCE 600
+#define _DARWIN_C_SOURCE
+#define _FILE_OFFSET_BITS 64
+#ifndef _LARGEFILE_SOURCE
+#define _LARGEFILE_SOURCE
+#endif
+#ifndef _LARGEFILE64_SOURCE
+#define _LARGEFILE64_SOURCE
+#endif
+#ifndef _GNU_SOURCE
+#define _GNU_SOURCE
+#endif
+#endif
+
+#include "config.h"
+#include <stdio.h>
+#include <string.h>
+#if HAVE_UNISTD_H
+#include <unistd.h>
+#endif
+#if HAVE_ERRNO_H
+#include <errno.h>
+#endif
+#include <fcntl.h>
+#include <time.h>
+#if HAVE_SYS_TYPES_H
+#include <sys/types.h>
+#endif
+#define PR_GET_DUMPABLE 3
+#if HAVE_SYS_STAT_H
+#include <sys/stat.h>
+#endif
+#include <fcntl.h>
+
+#undef ALIGN_DEBUG
+
+//#define DEBUG
+#ifdef DEBUG
+#define TRACE(...) {\
+ char __log[256];\
+ snprintf(__log, sizeof(__log), __VA_ARGS__);\
+ __log[sizeof(__log)-1] = 0;\
+ OutputDebugString(__log);\
+ }
+#else
+#define TRACE(...) do { } while (0);
+#endif
+
+#include "ext2_fs.h"
+#include "ext2fs.h"
+#include "ext2fsP.h"
+
+/*
+ * For checking structure magic numbers...
+ */
+
+#define EXT2_CHECK_MAGIC(struct, code) \
+ if ((struct)->magic != (code)) return (code)
+#define EXT2_CHECK_MAGIC_RETURN(struct, code, ret) \
+ if ((struct)->magic != (code)) return (ret)
+
+#define EXT2_ET_MAGIC_WINDOWS_IO_CHANNEL 0x10ed
+
+struct windows_cache {
+ char *buf;
+ unsigned long long block;
+ int access_time;
+ unsigned dirty:1;
+ unsigned in_use:1;
+};
+
+#define CACHE_SIZE 8
+#define WRITE_DIRECT_SIZE 4 /* Must be smaller than CACHE_SIZE */
+#define READ_DIRECT_SIZE 4 /* Should be smaller than CACHE_SIZE */
+
+struct windows_private_data {
+ int magic;
+ char name[MAX_PATH];
+ HANDLE handle;
+ char dos_device[MAX_PATH];
+ char cf_device[MAX_PATH];
+ int dev;
+ int flags;
+ int align;
+ int access_time;
+ ext2_loff_t offset;
+ struct windows_cache cache[CACHE_SIZE];
+ void *bounce;
+ struct struct_io_stats io_stats;
+};
+
+#define IS_ALIGNED(n, align) ((((uintptr_t) n) & \
+ ((uintptr_t) ((align)-1))) == 0)
+
+static int fake_dos_name_for_device(struct windows_private_data *data)
+{
+ if (strncmp(data->name, "\\\\", 2) == 0) {
+ data->dos_device[0] = 0;
+ strcpy(data->cf_device, data->name);
+ return 0;
+ }
+
+ _snprintf(data->dos_device, MAX_PATH, "fakedevice%lu", GetCurrentProcessId());
+
+ if (!DefineDosDevice(DDD_RAW_TARGET_PATH, data->dos_device, data->name))
+ return 1;
+
+ _snprintf(data->cf_device, MAX_PATH, "\\\\.\\%s", data->dos_device);
+ TRACE("e2fsprogs::fake_dos_name_for_device::DefineDosDevice(\"%s\")", data->dos_device);
+
+ return 0;
+}
+
+static void remove_fake_dos_name(struct windows_private_data *data)
+{
+ if (*data->dos_device) {
+ TRACE("e2fsprogs::remove_fake_dos_name::DefineDosDevice(\"%s\")", data->dos_device);
+ DefineDosDevice(DDD_RAW_TARGET_PATH | DDD_EXACT_MATCH_ON_REMOVE | DDD_REMOVE_DEFINITION, data->dos_device, data->name);
+ }
+}
+
+static errcode_t windows_get_stats(io_channel channel, io_stats *stats)
+{
+ errcode_t retval = 0;
+
+ struct windows_private_data *data;
+
+ EXT2_CHECK_MAGIC(channel, EXT2_ET_MAGIC_IO_CHANNEL);
+ data = (struct windows_private_data *) channel->private_data;
+ EXT2_CHECK_MAGIC(data, EXT2_ET_MAGIC_WINDOWS_IO_CHANNEL);
+
+ if (stats)
+ *stats = &data->io_stats;
+
+ return retval;
+}
+
+/*
+ * Here are the raw I/O functions
+ */
+static errcode_t raw_read_blk(io_channel channel,
+ struct windows_private_data *data,
+ unsigned long long block,
+ int count, void *bufv)
+{
+ errcode_t retval;
+ ssize_t size;
+ ext2_loff_t location;
+ DWORD actual = 0;
+ unsigned char *buf = bufv;
+ ssize_t really_read = 0;
+
+ size = (count < 0) ? -count : count * channel->block_size;
+ data->io_stats.bytes_read += size;
+ location = ((ext2_loff_t) block * channel->block_size) + data->offset;
+
+ if (data->flags & IO_FLAG_FORCE_BOUNCE) {
+ if (SetFilePointer(data->handle, location, NULL, FILE_BEGIN) == INVALID_SET_FILE_POINTER) {
+ retval = GetLastError();
+ goto error_out;
+ }
+ goto bounce_read;
+ }
+
+ if (SetFilePointer(data->handle, location, NULL, FILE_BEGIN) == INVALID_SET_FILE_POINTER) {
+ retval = GetLastError();
+ goto error_out;
+ }
+ if ((channel->align == 0) || (IS_ALIGNED(buf, channel->align) && IS_ALIGNED(size, channel->align))) {
+ if (!ReadFile(data->handle, buf, size, &actual, NULL)) {
+ retval = GetLastError();
+ goto error_out;
+ }
+ if (actual != size) {
+ short_read:
+ if (actual < 0) {
+ retval = GetLastError();
+ actual = 0;
+ } else
+ retval = EXT2_ET_SHORT_READ;
+ goto error_out;
+ }
+ return 0;
+ }
+
+ /*
+ * The buffer or size which we're trying to read isn't aligned
+ * to the O_DIRECT rules, so we need to do this the hard way...
+ */
+bounce_read:
+ while (size > 0) {
+ if (!ReadFile(data->handle, data->bounce, channel->block_size, &actual, NULL)) {
+ retval = GetLastError();
+ goto error_out;
+ }
+ if (actual != channel->block_size) {
+ actual = really_read;
+ buf -= really_read;
+ size += really_read;
+ goto short_read;
+ }
+ actual = size;
+ if (size > channel->block_size)
+ actual = channel->block_size;
+ memcpy(buf, data->bounce, actual);
+ really_read += actual;
+ size -= actual;
+ buf += actual;
+ }
+ return 0;
+
+error_out:
+ if (actual >= 0 && actual < size)
+ memset((char *) buf+actual, 0, size-actual);
+ if (channel->read_error)
+ retval = (channel->read_error)(channel, block, count, buf,
+ size, actual, retval);
+ return retval;
+}
+
+static errcode_t raw_write_blk(io_channel channel,
+ struct windows_private_data *data,
+ unsigned long long block,
+ int count, const void *bufv)
+{
+ ssize_t size;
+ ext2_loff_t location;
+ DWORD actual = 0;
+ errcode_t retval;
+ const unsigned char *buf = bufv;
+
+ if (count == 1)
+ size = channel->block_size;
+ else {
+ if (count < 0)
+ size = -count;
+ else
+ size = count * channel->block_size;
+ }
+ data->io_stats.bytes_written += size;
+
+ location = ((ext2_loff_t) block * channel->block_size) + data->offset;
+
+ if (data->flags & IO_FLAG_FORCE_BOUNCE) {
+ if (SetFilePointer(data->handle, location, NULL, FILE_BEGIN) == INVALID_SET_FILE_POINTER) {
+ retval = GetLastError();
+ goto error_out;
+ }
+ goto bounce_write;
+ }
+
+ if (SetFilePointer(data->handle, location, NULL, FILE_BEGIN) == INVALID_SET_FILE_POINTER) {
+ retval = GetLastError();
+ goto error_out;
+ }
+
+ SetLastError(0);
+
+ if ((channel->align == 0) || (IS_ALIGNED(buf, channel->align) && IS_ALIGNED(size, channel->align))) {
+ if (!WriteFile(data->handle, buf, size, &actual, NULL)) {
+ retval = GetLastError();
+ goto error_out;
+ }
+
+ if (actual != size) {
+ short_write:
+ retval = EXT2_ET_SHORT_WRITE;
+ goto error_out;
+ }
+ return 0;
+ }
+
+ /*
+ * The buffer or size which we're trying to write isn't aligned
+ * to the O_DIRECT rules, so we need to do this the hard way...
+ */
+bounce_write:
+ while (size > 0) {
+ if (size < channel->block_size) {
+ if (!ReadFile(data->handle, data->bounce, channel->block_size, &actual, NULL)) {
+ retval = GetLastError();
+ goto error_out;
+ }
+ if (actual != channel->block_size) {
+ if (actual < 0) {
+ retval = GetLastError();
+ goto error_out;
+ }
+ memset((char *) data->bounce + actual, 0,
+ channel->block_size - actual);
+ }
+ }
+ actual = size;
+ if (size > channel->block_size)
+ actual = channel->block_size;
+ memcpy(data->bounce, buf, actual);
+ if (SetFilePointer(data->handle, location, NULL, FILE_BEGIN) == INVALID_SET_FILE_POINTER) {
+ retval = GetLastError();
+ goto error_out;
+ }
+ if (!WriteFile(data->handle, data->bounce, channel->block_size, &actual, NULL)) {
+ retval = GetLastError();
+ goto error_out;
+ }
+
+ if (actual != channel->block_size)
+ goto short_write;
+ size -= actual;
+ buf += actual;
+ location += actual;
+ }
+ return 0;
+
+error_out:
+ if (channel->write_error)
+ retval = (channel->write_error)(channel, block, count, buf,
+ size, actual, retval);
+ return retval;
+}
+
+
+/*
+ * Here we implement the cache functions
+ */
+
+/* Allocate the cache buffers */
+static errcode_t alloc_cache(io_channel channel,
+ struct windows_private_data *data)
+{
+ errcode_t retval;
+ struct windows_cache *cache;
+ int i;
+
+ data->access_time = 0;
+ for (i=0, cache = data->cache; i < CACHE_SIZE; i++, cache++) {
+ cache->block = 0;
+ cache->access_time = 0;
+ cache->dirty = 0;
+ cache->in_use = 0;
+ if (cache->buf)
+ ext2fs_free_mem(&cache->buf);
+ retval = io_channel_alloc_buf(channel, 0, &cache->buf);
+ if (retval)
+ return retval;
+ }
+ if (channel->align || data->flags & IO_FLAG_FORCE_BOUNCE) {
+ if (data->bounce)
+ ext2fs_free_mem(&data->bounce);
+ retval = io_channel_alloc_buf(channel, 0, &data->bounce);
+ }
+ return retval;
+}
+
+/* Free the cache buffers */
+static void free_cache(struct windows_private_data *data)
+{
+ struct windows_cache *cache;
+ int i;
+
+ data->access_time = 0;
+ for (i=0, cache = data->cache; i < CACHE_SIZE; i++, cache++) {
+ cache->block = 0;
+ cache->access_time = 0;
+ cache->dirty = 0;
+ cache->in_use = 0;
+ if (cache->buf)
+ ext2fs_free_mem(&cache->buf);
+ }
+ if (data->bounce)
+ ext2fs_free_mem(&data->bounce);
+}
+
+#ifndef NO_IO_CACHE
+/*
+ * Try to find a block in the cache. If the block is not found, and
+ * eldest is a non-zero pointer, then fill in eldest with the cache
+ * entry to that should be reused.
+ */
+static struct windows_cache *find_cached_block(struct windows_private_data *data,
+ unsigned long long block,
+ struct windows_cache **eldest)
+{
+ struct windows_cache *cache, *unused_cache, *oldest_cache;
+ int i;
+
+ unused_cache = oldest_cache = 0;
+ for (i=0, cache = data->cache; i < CACHE_SIZE; i++, cache++) {
+ if (!cache->in_use) {
+ if (!unused_cache)
+ unused_cache = cache;
+ continue;
+ }
+ if (cache->block == block) {
+ cache->access_time = ++data->access_time;
+ return cache;
+ }
+ if (!oldest_cache ||
+ (cache->access_time < oldest_cache->access_time))
+ oldest_cache = cache;
+ }
+ if (eldest)
+ *eldest = (unused_cache) ? unused_cache : oldest_cache;
+ return 0;
+}
+
+/*
+ * Reuse a particular cache entry for another block.
+ */
+static void reuse_cache(io_channel channel, struct windows_private_data *data,
+ struct windows_cache *cache, unsigned long long block)
+{
+ if (cache->dirty && cache->in_use)
+ raw_write_blk(channel, data, cache->block, 1, cache->buf);
+
+ cache->in_use = 1;
+ cache->dirty = 0;
+ cache->block = block;
+ cache->access_time = ++data->access_time;
+}
+
+/*
+ * Flush all of the blocks in the cache
+ */
+static errcode_t flush_cached_blocks(io_channel channel,
+ struct windows_private_data *data,
+ int invalidate)
+{
+ struct windows_cache *cache;
+ errcode_t retval, retval2;
+ int i;
+
+ retval2 = 0;
+ for (i=0, cache = data->cache; i < CACHE_SIZE; i++, cache++) {
+ if (!cache->in_use)
+ continue;
+
+ if (invalidate)
+ cache->in_use = 0;
+
+ if (!cache->dirty)
+ continue;
+
+ retval = raw_write_blk(channel, data,
+ cache->block, 1, cache->buf);
+ if (retval)
+ retval2 = retval;
+ else
+ cache->dirty = 0;
+ }
+ return retval2;
+}
+#endif /* NO_IO_CACHE */
+
+static errcode_t windows_open_channel(struct windows_private_data *data,
+ int flags, io_channel *channel,
+ io_manager io_mgr)
+{
+ io_channel io = NULL;
+ errcode_t retval;
+ ext2fs_struct_stat st;
+
+ retval = ext2fs_get_mem(sizeof(struct struct_io_channel), &io);
+ if (retval)
+ goto cleanup;
+ memset(io, 0, sizeof(struct struct_io_channel));
+ io->magic = EXT2_ET_MAGIC_IO_CHANNEL;
+
+ io->manager = io_mgr;
+ retval = ext2fs_get_mem(strlen(data->name)+1, &io->name);
+ if (retval)
+ goto cleanup;
+
+ strcpy(io->name, data->name);
+ io->private_data = data;
+ io->block_size = 1024;
+ io->read_error = 0;
+ io->write_error = 0;
+ io->refcount = 1;
+
+#if defined(O_DIRECT)
+ if (flags & IO_FLAG_DIRECT_IO)
+ io->align = ext2fs_get_dio_alignment(data->dev);
+#elif defined(F_NOCACHE)
+ if (flags & IO_FLAG_DIRECT_IO)
+ io->align = 4096;
+#endif
+
+ /*
+ * If the device is really a block device, then set the
+ * appropriate flag, otherwise we can set DISCARD_ZEROES flag
+ * because we are going to use punch hole instead of discard
+ * and if it succeed, subsequent read from sparse area returns
+ * zero.
+ */
+ if (ext2fs_fstat(data->dev, &st) == 0) {
+ if (ext2fsP_is_disk_device(st.st_mode))
+ io->flags |= CHANNEL_FLAGS_BLOCK_DEVICE;
+ else
+ io->flags |= CHANNEL_FLAGS_DISCARD_ZEROES;
+ }
+
+ if ((retval = alloc_cache(io, data)))
+ goto cleanup;
+
+#ifdef BLKROGET
+ if (flags & IO_FLAG_RW) {
+ int error;
+ int readonly = 0;
+
+ /* Is the block device actually writable? */
+ error = ioctl(data->dev, BLKROGET, &readonly);
+ if (!error && readonly) {
+ retval = EPERM;
+ goto cleanup;
+ }
+ }
+#endif
+
+ *channel = io;
+ return 0;
+
+cleanup:
+ if (data) {
+ if (data->dev >= 0)
+ close(data->dev);
+ free_cache(data);
+ ext2fs_free_mem(&data);
+ }
+ if (io) {
+ if (io->name) {
+ ext2fs_free_mem(&io->name);
+ }
+ ext2fs_free_mem(&io);
+ }
+ return retval;
+}
+
+static DWORD windows_open_device(struct windows_private_data *data, int open_flags)
+{
+ DWORD ret = 0;
+
+ if (*data->name != '\\')
+ strcpy(data->cf_device, data->name);
+ else if (fake_dos_name_for_device(data))
+ return -1;
+
+ DWORD desired_access = GENERIC_READ | ((open_flags & O_RDWR) ? GENERIC_WRITE : 0);
+ DWORD share_mode = (open_flags & O_EXCL) ? 0 : FILE_SHARE_READ | ((open_flags & O_RDWR) ? FILE_SHARE_WRITE : 0);
+ DWORD flags_and_attributes =
+#if defined(O_DIRECT)
+ (open_flags & O_DIRECT) ? (FILE_FLAG_NO_BUFFERING | FILE_FLAG_WRITE_THROUGH) : FILE_ATTRIBUTE_NORMAL;
+#else
+ FILE_FLAG_NO_BUFFERING | FILE_FLAG_WRITE_THROUGH;
+#endif
+
+ data->handle = CreateFile(data->cf_device, desired_access, share_mode, NULL, OPEN_EXISTING,
+ flags_and_attributes, NULL);
+
+ if (data->handle == INVALID_HANDLE_VALUE) {
+ ret = GetLastError();
+ goto invalid_handle;
+ }
+
+ TRACE("e2fsprogs::windows_open_device::CreateFile(\"%s\") = %p", data->cf_device, data->handle);
+
+ data->dev = _open_osfhandle((intptr_t)data->handle, 0);
+ if (data->dev < 0) {
+ ret = GetLastError() ? GetLastError() : 9999;
+ goto osfhandle_error;
+ }
+
+ TRACE("e2fsprogs::windows_open_device::_open_osfhandle(%p) = %d", data->handle, data->dev);
+
+ return 0;
+
+osfhandle_error:
+ TRACE("e2fsprogs::windows_open_device::CloseHandle(%p)", data->handle);
+ CloseHandle(data->handle);
+invalid_handle:
+ remove_fake_dos_name(data);
+ TRACE("e2fsprogs::windows_open_device() = %lu, errno = %d", ret, errno);
+ return ret;
+}
+
+static struct windows_private_data *init_private_data(const char *name, int flags)
+{
+ struct windows_private_data *data = NULL;
+
+ if (ext2fs_get_mem(sizeof(struct windows_private_data), &data))
+ return NULL;
+
+ memset(data, 0, sizeof(struct windows_private_data));
+ strncpy(data->name, name, sizeof(data->name));
+ data->magic = EXT2_ET_MAGIC_WINDOWS_IO_CHANNEL;
+ data->io_stats.num_fields = 2;
+ data->flags = flags;
+ data->handle = INVALID_HANDLE_VALUE;
+
+ return data;
+}
+
+static errcode_t windows_open(const char *name, int flags, io_channel *channel)
+{
+ int fd = -1;
+ int open_flags;
+ struct windows_private_data *data;
+
+ if (name == 0)
+ return EXT2_ET_BAD_DEVICE_NAME;
+
+ data = init_private_data(name, flags);
+ if (!data)
+ return EXT2_ET_NO_MEMORY;
+
+ open_flags = (flags & IO_FLAG_RW) ? O_RDWR : O_RDONLY;
+ if (flags & IO_FLAG_EXCLUSIVE)
+ open_flags |= O_EXCL;
+#if defined(O_DIRECT)
+ if (flags & IO_FLAG_DIRECT_IO)
+ open_flags |= O_DIRECT;
+#endif
+
+ if (windows_open_device(data, open_flags)) {
+ ext2fs_free_mem(&data);
+ return EXT2_ET_BAD_DEVICE_NAME;
+ }
+
+#if defined(F_NOCACHE) && !defined(IO_DIRECT)
+ if (flags & IO_FLAG_DIRECT_IO) {
+ if (fcntl(fd, F_NOCACHE, 1) < 0)
+ return errno;
+ }
+#endif
+ return windows_open_channel(data, flags, channel, windows_io_manager);
+}
+
+static errcode_t windows_close(io_channel channel)
+{
+ struct windows_private_data *data;
+ errcode_t retval = 0;
+
+ EXT2_CHECK_MAGIC(channel, EXT2_ET_MAGIC_IO_CHANNEL);
+ data = (struct windows_private_data *) channel->private_data;
+ EXT2_CHECK_MAGIC(data, EXT2_ET_MAGIC_WINDOWS_IO_CHANNEL);
+
+ if (--channel->refcount > 0)
+ return 0;
+
+#ifndef NO_IO_CACHE
+ retval = flush_cached_blocks(channel, data, 0);
+#endif
+
+ remove_fake_dos_name(data);
+
+ if (_close(data->dev) != 0)
+ retval = errno;
+
+ TRACE("e2fsprogs::windows_close::_close(%d)", data->dev);
+
+ free_cache(data);
+
+ ext2fs_free_mem(&channel->private_data);
+ if (channel->name)
+ ext2fs_free_mem(&channel->name);
+ ext2fs_free_mem(&channel);
+ return retval;
+}
+
+static errcode_t windows_set_blksize(io_channel channel, int blksize)
+{
+ struct windows_private_data *data;
+ errcode_t retval;
+
+ EXT2_CHECK_MAGIC(channel, EXT2_ET_MAGIC_IO_CHANNEL);
+ data = (struct windows_private_data *) channel->private_data;
+ EXT2_CHECK_MAGIC(data, EXT2_ET_MAGIC_WINDOWS_IO_CHANNEL);
+
+ if (channel->block_size != blksize) {
+#ifndef NO_IO_CACHE
+ if ((retval = flush_cached_blocks(channel, data, 0)))
+ return retval;
+#endif
+
+ channel->block_size = blksize;
+ free_cache(data);
+ if ((retval = alloc_cache(channel, data)))
+ return retval;
+ }
+ return 0;
+}
+
+static errcode_t windows_read_blk64(io_channel channel, unsigned long long block, int count, void *buf)
+{
+ struct windows_private_data *data;
+ struct windows_cache *cache, *reuse[READ_DIRECT_SIZE];
+ errcode_t retval;
+ char *cp;
+ int i, j;
+
+ EXT2_CHECK_MAGIC(channel, EXT2_ET_MAGIC_IO_CHANNEL);
+ data = (struct windows_private_data *) channel->private_data;
+ EXT2_CHECK_MAGIC(data, EXT2_ET_MAGIC_WINDOWS_IO_CHANNEL);
+
+#ifdef NO_IO_CACHE
+ return raw_read_blk(channel, data, block, count, buf);
+#else
+ /*
+ * If we're doing an odd-sized read or a very large read,
+ * flush out the cache and then do a direct read.
+ */
+ if (count < 0 || count > WRITE_DIRECT_SIZE) {
+ if ((retval = flush_cached_blocks(channel, data, 0)))
+ return retval;
+ return raw_read_blk(channel, data, block, count, buf);
+ }
+
+ cp = buf;
+ while (count > 0) {
+ /* If it's in the cache, use it! */
+ if ((cache = find_cached_block(data, block, &reuse[0]))) {
+#ifdef DEBUG
+ printf("Using cached block %lu\n", block);
+#endif
+ memcpy(cp, cache->buf, channel->block_size);
+ count--;
+ block++;
+ cp += channel->block_size;
+ continue;
+ }
+ if (count == 1) {
+ /*
+ * Special case where we read directly into the
+ * cache buffer; important in the O_DIRECT case
+ */
+ cache = reuse[0];
+ reuse_cache(channel, data, cache, block);
+ if ((retval = raw_read_blk(channel, data, block, 1,
+ cache->buf))) {
+ cache->in_use = 0;
+ return retval;
+ }
+ memcpy(cp, cache->buf, channel->block_size);
+ return 0;
+ }
+
+ /*
+ * Find the number of uncached blocks so we can do a
+ * single read request
+ */
+ for (i=1; i < count; i++)
+ if (find_cached_block(data, block+i, &reuse[i]))
+ break;
+#ifdef DEBUG
+ printf("Reading %d blocks starting at %lu\n", i, block);
+#endif
+ if ((retval = raw_read_blk(channel, data, block, i, cp)))
+ return retval;
+
+ /* Save the results in the cache */
+ for (j=0; j < i; j++) {
+ count--;
+ cache = reuse[j];
+ reuse_cache(channel, data, cache, block++);
+ memcpy(cache->buf, cp, channel->block_size);
+ cp += channel->block_size;
+ }
+ }
+ return 0;
+#endif /* NO_IO_CACHE */
+}
+
+static errcode_t windows_read_blk(io_channel channel, unsigned long block,
+ int count, void *buf)
+{
+ return windows_read_blk64(channel, block, count, buf);
+}
+
+static errcode_t windows_write_blk64(io_channel channel,
+ unsigned long long block,
+ int count, const void *buf)
+{
+ struct windows_private_data *data;
+ struct windows_cache *cache, *reuse;
+ errcode_t retval = 0;
+ const char *cp;
+ int writethrough;
+
+ EXT2_CHECK_MAGIC(channel, EXT2_ET_MAGIC_IO_CHANNEL);
+ data = (struct windows_private_data *) channel->private_data;
+ EXT2_CHECK_MAGIC(data, EXT2_ET_MAGIC_WINDOWS_IO_CHANNEL);
+
+#ifdef NO_IO_CACHE
+ return raw_write_blk(channel, data, block, count, buf);
+#else
+ /*
+ * If we're doing an odd-sized write or a very large write,
+ * flush out the cache completely and then do a direct write.
+ */
+ if (count < 0 || count > WRITE_DIRECT_SIZE) {
+ if ((retval = flush_cached_blocks(channel, data, 1)))
+ return retval;
+ return raw_write_blk(channel, data, block, count, buf);
+ }
+
+ /*
+ * For a moderate-sized multi-block write, first force a write
+ * if we're in write-through cache mode, and then fill the
+ * cache with the blocks.
+ */
+ writethrough = channel->flags & CHANNEL_FLAGS_WRITETHROUGH;
+ if (writethrough)
+ retval = raw_write_blk(channel, data, block, count, buf);
+
+ cp = buf;
+ while (count > 0) {
+ cache = find_cached_block(data, block, &reuse);
+ if (!cache) {
+ cache = reuse;
+ reuse_cache(channel, data, cache, block);
+ }
+ if (cache->buf != cp)
+ memcpy(cache->buf, cp, channel->block_size);
+ cache->dirty = !writethrough;
+ count--;
+ block++;
+ cp += channel->block_size;
+ }
+ return retval;
+#endif /* NO_IO_CACHE */
+}
+
+static errcode_t windows_cache_readahead(io_channel channel,
+ unsigned long long block,
+ unsigned long long count)
+{
+ return EXT2_ET_OP_NOT_SUPPORTED;
+}
+
+static errcode_t windows_write_blk(io_channel channel, unsigned long block,
+ int count, const void *buf)
+{
+ return windows_write_blk64(channel, block, count, buf);
+}
+
+static errcode_t windows_write_byte(io_channel channel, unsigned long offset,
+ int size, const void *buf)
+{
+ return EXT2_ET_UNIMPLEMENTED;
+}
+
+HANDLE windows_get_handle(io_channel channel)
+{
+ struct windows_private_data *data;
+
+ EXT2_CHECK_MAGIC_RETURN(channel, EXT2_ET_MAGIC_IO_CHANNEL, INVALID_HANDLE_VALUE);
+ data = (struct windows_private_data *) channel->private_data;
+ EXT2_CHECK_MAGIC_RETURN(data, EXT2_ET_MAGIC_WINDOWS_IO_CHANNEL, INVALID_HANDLE_VALUE);
+
+ return data->handle;
+}
+
+/*
+ * Flush data buffers to disk.
+ */
+static errcode_t windows_flush(io_channel channel)
+{
+ struct windows_private_data *data;
+ errcode_t retval = 0;
+
+ EXT2_CHECK_MAGIC(channel, EXT2_ET_MAGIC_IO_CHANNEL);
+ data = (struct windows_private_data *) channel->private_data;
+ EXT2_CHECK_MAGIC(data, EXT2_ET_MAGIC_WINDOWS_IO_CHANNEL);
+
+#ifndef NO_IO_CACHE
+ retval = flush_cached_blocks(channel, data, 0);
+#endif
+
+ return retval;
+}
+
+static errcode_t windows_set_option(io_channel channel, const char *option,
+ const char *arg)
+{
+ struct windows_private_data *data;
+ unsigned long long tmp;
+ char *end;
+
+ EXT2_CHECK_MAGIC(channel, EXT2_ET_MAGIC_IO_CHANNEL);
+ data = (struct windows_private_data *) channel->private_data;
+ EXT2_CHECK_MAGIC(data, EXT2_ET_MAGIC_WINDOWS_IO_CHANNEL);
+
+ if (!strcmp(option, "offset")) {
+ if (!arg)
+ return EXT2_ET_INVALID_ARGUMENT;
+
+ tmp = strtoull(arg, &end, 0);
+ if (*end)
+ return EXT2_ET_INVALID_ARGUMENT;
+ data->offset = tmp;
+ if (data->offset < 0)
+ return EXT2_ET_INVALID_ARGUMENT;
+ return 0;
+ }
+ return EXT2_ET_INVALID_ARGUMENT;
+}
+
+static errcode_t windows_discard(io_channel channel, unsigned long long block,
+ unsigned long long count)
+{
+ TRACE("e2fsprogs::windows_discard::EXT2_ET_UNIMPLEMENTED");
+ return EXT2_ET_UNIMPLEMENTED;
+}
+
+/* parameters might not be used if OS doesn't support zeroout */
+#if __GNUC_PREREQ (4, 6)
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#endif
+static errcode_t windows_zeroout(io_channel channel, unsigned long long block,
+ unsigned long long count)
+{
+ struct windows_private_data *data;
+ int ret;
+
+ EXT2_CHECK_MAGIC(channel, EXT2_ET_MAGIC_IO_CHANNEL);
+ data = (struct windows_private_data *) channel->private_data;
+ EXT2_CHECK_MAGIC(data, EXT2_ET_MAGIC_WINDOWS_IO_CHANNEL);
+
+ if (channel->flags & CHANNEL_FLAGS_BLOCK_DEVICE) {
+ /* Not implemented until the BLKZEROOUT mess is fixed */
+ goto unimplemented;
+ } else {
+ /* Regular file, try to use truncate/punch/zero. */
+ struct stat statbuf;
+
+ if (count == 0)
+ return 0;
+ /*
+ * If we're trying to zero a range past the end of the file,
+ * extend the file size, then truncate everything.
+ */
+ ret = fstat(data->dev, &statbuf);
+ if (ret)
+ goto err;
+ if ((unsigned long long) statbuf.st_size <
+ (block + count) * channel->block_size + data->offset) {
+ ret = ftruncate(data->dev,
+ (block + count) * channel->block_size + data->offset);
+ if (ret)
+ goto err;
+ }
+ goto unimplemented;
+ }
+err:
+ if (ret < 0) {
+ if (errno == EOPNOTSUPP)
+ goto unimplemented;
+ return errno;
+ }
+ return 0;
+unimplemented:
+ return EXT2_ET_UNIMPLEMENTED;
+}
+
+int ext2fs_open_file(const char *pathname, int flags, mode_t mode)
+{
+ flags |= O_BINARY;
+
+ if (mode)
+#if defined(HAVE_OPEN64) && !defined(__OSX_AVAILABLE_BUT_DEPRECATED)
+ return open64(pathname, flags, mode);
+ else
+ return open64(pathname, flags);
+#else
+ return open(pathname, flags, mode);
+ else
+ return open(pathname, flags);
+#endif
+}
+
+int ext2fs_stat(const char *path, ext2fs_struct_stat *buf)
+{
+#if defined(HAVE_FSTAT64) && !defined(__OSX_AVAILABLE_BUT_DEPRECATED)
+ return stat64(path, buf);
+#else
+ return stat(path, buf);
+#endif
+}
+
+int ext2fs_fstat(int fd, ext2fs_struct_stat *buf)
+{
+#if defined(HAVE_FSTAT64) && !defined(__OSX_AVAILABLE_BUT_DEPRECATED)
+ return fstat64(fd, buf);
+#else
+ return fstat(fd, buf);
+#endif
+}
+
+#if __GNUC_PREREQ (4, 6)
+#pragma GCC diagnostic pop
+#endif
+
+static struct struct_io_manager struct_windows_manager = {
+ .magic = EXT2_ET_MAGIC_IO_MANAGER,
+ .name = "Windows I/O Manager",
+ .open = windows_open,
+ .close = windows_close,
+ .set_blksize = windows_set_blksize,
+ .read_blk = windows_read_blk,
+ .write_blk = windows_write_blk,
+ .flush = windows_flush,
+ .write_byte = windows_write_byte,
+ .set_option = windows_set_option,
+ .get_stats = windows_get_stats,
+ .read_blk64 = windows_read_blk64,
+ .write_blk64 = windows_write_blk64,
+ .discard = windows_discard,
+ .cache_readahead = windows_cache_readahead,
+ .zeroout = windows_zeroout,
+};
+
+io_manager windows_io_manager = &struct_windows_manager;
--
2.17.1

2021-04-02 18:34:43

by Theodore Ts'o

[permalink] [raw]
Subject: Re: [PATCH 1/3] Add a Windows implementation of the IO manager

Thanks, I've applied this series to e2fsprogs.

- Ted