mail archive of the barebox mailing list
 help / color / mirror / Atom feed
* [PATCH] ext4 filesystem support
@ 2012-12-03 10:27 Sascha Hauer
  2012-12-03 10:27 ` [PATCH 1/4] ls command: call stat() only when necessary Sascha Hauer
                   ` (3 more replies)
  0 siblings, 4 replies; 5+ messages in thread
From: Sascha Hauer @ 2012-12-03 10:27 UTC (permalink / raw)
  To: barebox

Hi All,

The following adds ext4 readonly filesystem support to barebox.
The implementation is from U-Boot which recently gained ext4
support.
ext filesystems can automatically be detected, so the -t option
to mount is not necessary. Some preparation is needed because
ext filesystems have their magic at byte offset 1080. Currently
we do not pass in such a big buffer into the filesystem detection
code. To fix this we now pass the buffer size to the file detection
code so that we do not read past the buffer.

Sascha

----------------------------------------------------------------
Sascha Hauer (4):
      ls command: call stat() only when necessary
      filetype: Pass bufsize
      add ext fs detection support
      fs: implement initial ext4 support from U-Boot

 arch/arm/lib/bootm.c                 |    2 +-
 arch/arm/mach-imx/imx-bbu-internal.c |    4 +-
 commands/bootm.c                     |    2 +-
 commands/ls.c                        |    9 +-
 common/filetype.c                    |   47 ++-
 common/uimage.c                      |    2 +-
 fs/Kconfig                           |    2 +
 fs/Makefile                          |    1 +
 fs/ext4/Kconfig                      |    3 +
 fs/ext4/Makefile                     |    1 +
 fs/ext4/ext4_common.c                |  551 ++++++++++++++++++++++++++++++++++
 fs/ext4/ext4_common.h                |   58 ++++
 fs/ext4/ext4fs.c                     |  153 ++++++++++
 fs/ext4/ext4fs.h                     |  127 ++++++++
 fs/ext4/ext_barebox.c                |  293 ++++++++++++++++++
 fs/ext4/ext_common.h                 |  195 ++++++++++++
 include/filetype.h                   |    5 +-
 lib/gui/image_renderer.c             |    6 +-
 lib/uncompress.c                     |    4 +-
 19 files changed, 1438 insertions(+), 27 deletions(-)
 create mode 100644 fs/ext4/Kconfig
 create mode 100644 fs/ext4/Makefile
 create mode 100644 fs/ext4/ext4_common.c
 create mode 100644 fs/ext4/ext4_common.h
 create mode 100644 fs/ext4/ext4fs.c
 create mode 100644 fs/ext4/ext4fs.h
 create mode 100644 fs/ext4/ext_barebox.c
 create mode 100644 fs/ext4/ext_common.h

_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

^ permalink raw reply	[flat|nested] 5+ messages in thread
* [PATCH 1/4] ls command: call stat() only when necessary
  2012-12-03 10:27 [PATCH] ext4 filesystem support Sascha Hauer
@ 2012-12-03 10:27 ` Sascha Hauer
  2012-12-03 10:27 ` [PATCH 2/4] filetype: Pass bufsize Sascha Hauer
                   ` (2 subsequent siblings)
  3 siblings, 0 replies; 5+ messages in thread
From: Sascha Hauer @ 2012-12-03 10:27 UTC (permalink / raw)
  To: barebox

When calling ls in short mode we do not have to call stat()
for additional informations because we do not use them. This
speeds up ls on filesystems on which stat() is expensive
because the barebox filesystem support always has to iterate
over the directory tree.

Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
---
 commands/ls.c |    9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/commands/ls.c b/commands/ls.c
index 1fdb244..f2d9903 100644
--- a/commands/ls.c
+++ b/commands/ls.c
@@ -73,12 +73,13 @@ int ls(const char *path, ulong flags)
 
 	while ((d = readdir(dir))) {
 		sprintf(tmp, "%s/%s", path, d->d_name);
-		if (lstat(tmp, &s))
-			goto out;
-		if (flags & LS_COLUMN)
+		if (flags & LS_COLUMN) {
 			string_list_add_sorted(&sl, d->d_name);
-		else
+		} else {
+			if (lstat(tmp, &s))
+				goto out;
 			ls_one(d->d_name, tmp, &s);
+		}
 	}
 
 	closedir(dir);
-- 
1.7.10.4


_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

^ permalink raw reply	[flat|nested] 5+ messages in thread
* [PATCH 2/4] filetype: Pass bufsize
  2012-12-03 10:27 [PATCH] ext4 filesystem support Sascha Hauer
  2012-12-03 10:27 ` [PATCH 1/4] ls command: call stat() only when necessary Sascha Hauer
@ 2012-12-03 10:27 ` Sascha Hauer
  2012-12-03 10:27 ` [PATCH 3/4] add ext fs detection support Sascha Hauer
  2012-12-03 10:27 ` [PATCH 4/4] fs: implement initial ext4 support from U-Boot Sascha Hauer
  3 siblings, 0 replies; 5+ messages in thread
From: Sascha Hauer @ 2012-12-03 10:27 UTC (permalink / raw)
  To: barebox

Pass the buffer size to the file detection code. This makes sure we do not
read past the buffer. This is especially useful for ext filesystem detection
as the magic is at byte offset 1080. Also introduce a FILE_TYPE_SAFE_BUFSIZE
define which is set to the minimum bufsize the detection code needs to detect
all known filetypes.

Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
---
 arch/arm/lib/bootm.c                 |    2 +-
 arch/arm/mach-imx/imx-bbu-internal.c |    4 ++--
 commands/bootm.c                     |    2 +-
 common/filetype.c                    |   39 +++++++++++++++++++++++-----------
 common/uimage.c                      |    2 +-
 include/filetype.h                   |    4 +++-
 lib/gui/image_renderer.c             |    6 +++---
 lib/uncompress.c                     |    4 ++--
 8 files changed, 40 insertions(+), 23 deletions(-)

diff --git a/arch/arm/lib/bootm.c b/arch/arm/lib/bootm.c
index 288c0b2..51ac9af 100644
--- a/arch/arm/lib/bootm.c
+++ b/arch/arm/lib/bootm.c
@@ -143,7 +143,7 @@ static int do_bootz_linux_fdt(int fd, struct image_data *data)
 	if (ret < sizeof(*header))
 		return ret;
 
-	if (file_detect_type(header) != filetype_oftree)
+	if (file_detect_type(header, sizeof(*header)) != filetype_oftree)
 		return -ENXIO;
 
 	end = be32_to_cpu(header->totalsize);
diff --git a/arch/arm/mach-imx/imx-bbu-internal.c b/arch/arm/mach-imx/imx-bbu-internal.c
index c34f86f..881c20a 100644
--- a/arch/arm/mach-imx/imx-bbu-internal.c
+++ b/arch/arm/mach-imx/imx-bbu-internal.c
@@ -110,7 +110,7 @@ static int imx_bbu_internal_v1_update(struct bbu_handler *handler, struct bbu_da
 	int ret, image_len;
 	void *buf;
 
-	if (file_detect_type(data->image) != filetype_arm_barebox) {
+	if (file_detect_type(data->image, data->len) != filetype_arm_barebox) {
 		if (!bbu_force(data, "Not an ARM barebox image"))
 			return -EINVAL;
 	}
@@ -332,7 +332,7 @@ static int imx_bbu_internal_v2_update(struct bbu_handler *handler, struct bbu_da
 	int ret, image_len;
 	void *buf;
 
-	if (file_detect_type(data->image) != filetype_arm_barebox) {
+	if (file_detect_type(data->image, data->len) != filetype_arm_barebox) {
 		if (!bbu_force(data, "Not an ARM barebox image"))
 			return -EINVAL;
 	}
diff --git a/commands/bootm.c b/commands/bootm.c
index 98d2e4f..483e6a1 100644
--- a/commands/bootm.c
+++ b/commands/bootm.c
@@ -184,7 +184,7 @@ static int bootm_open_oftree(struct image_data *data, const char *oftree, int nu
 		}
 	}
 
-	ft = file_detect_type(fdt);
+	ft = file_detect_type(fdt, size);
 	if (ft != filetype_oftree) {
 		printf("%s is not an oftree but %s\n", oftree,
 				file_type_to_string(ft));
diff --git a/common/filetype.c b/common/filetype.c
index b8d54f7..c1bd11d 100644
--- a/common/filetype.c
+++ b/common/filetype.c
@@ -105,19 +105,24 @@ enum filetype is_fat_or_mbr(const unsigned char *sector, unsigned long *bootsec)
 	return filetype_mbr;
 }
 
-enum filetype file_detect_type(void *_buf)
+enum filetype file_detect_type(void *_buf, size_t bufsize)
 {
 	u32 *buf = _buf;
 	u64 *buf64 = _buf;
 	u8 *buf8 = _buf;
 	enum filetype type;
 
+	if (bufsize < 9)
+		return filetype_unknown;
+
 	if (strncmp(buf8, "#!/bin/sh", 9) == 0)
 		return filetype_sh;
-	if (is_barebox_arm_head(_buf))
-		return filetype_arm_barebox;
-	if (buf[9] == 0x016f2818 || buf[9] == 0x18286f01)
-		return filetype_arm_zimage;
+
+	if (bufsize < 32)
+		return filetype_unknown;
+
+	if (strncmp(buf8, "BM", 2) == 0)
+		return filetype_bmp;
 	if (buf8[0] == 0x89 && buf8[1] == 0x4c && buf8[2] == 0x5a &&
 			buf8[3] == 0x4f)
 		return filetype_lzo_compressed;
@@ -136,15 +141,25 @@ enum filetype file_detect_type(void *_buf)
 		return filetype_oftree;
 	if (strncmp(buf8, "ANDROID!", 8) == 0)
 		return filetype_aimage;
+	if (buf64[0] == le64_to_cpu(0x0a1a0a0d474e5089ull))
+		return filetype_png;
 	if (strncmp(buf8 + 0x10, "barebox", 7) == 0)
 		return filetype_mips_barebox;
+
+	if (bufsize < 64)
+		return filetype_unknown;
+
+	if (is_barebox_arm_head(_buf))
+		return filetype_arm_barebox;
+	if (buf[9] == 0x016f2818 || buf[9] == 0x18286f01)
+		return filetype_arm_zimage;
+
+	if (bufsize < 512)
+		return filetype_unknown;
+
 	type = is_fat_or_mbr(buf8, NULL);
 	if (type != filetype_unknown)
 		return type;
-	if (strncmp(buf8, "BM", 2) == 0)
-		return filetype_bmp;
-	if (buf64[0] == le64_to_cpu(0x0a1a0a0d474e5089ull))
-		return filetype_png;
 
 	return filetype_unknown;
 }
@@ -160,13 +175,13 @@ enum filetype file_name_detect_type(const char *filename)
 	if (fd < 0)
 		return fd;
 
-	buf = xzalloc(512);
+	buf = xzalloc(FILE_TYPE_SAFE_BUFSIZE);
 
-	ret = read(fd, buf, 512);
+	ret = read(fd, buf, FILE_TYPE_SAFE_BUFSIZE);
 	if (ret < 0)
 		goto err_out;
 
-	type = file_detect_type(buf);
+	type = file_detect_type(buf, ret);
 
 	if (type == filetype_mbr) {
 		/*
diff --git a/common/uimage.c b/common/uimage.c
index 3f5a3d5..3bec6b3 100644
--- a/common/uimage.c
+++ b/common/uimage.c
@@ -516,7 +516,7 @@ void *uimage_load_to_buf(struct uimage_handle *handle, int image_no,
 	if (ret < 0)
 		return NULL;
 
-	ft = file_detect_type(ftbuf);
+	ft = file_detect_type(ftbuf, 128);
 	if ((int)ft < 0)
 		return NULL;
 
diff --git a/include/filetype.h b/include/filetype.h
index 0a722a0..5fac531 100644
--- a/include/filetype.h
+++ b/include/filetype.h
@@ -25,9 +25,11 @@ enum filetype {
 	filetype_max,
 };
 
+#define FILE_TYPE_SAFE_BUFSIZE		2048
+
 const char *file_type_to_string(enum filetype f);
 const char *file_type_to_short_string(enum filetype f);
-enum filetype file_detect_type(void *_buf);
+enum filetype file_detect_type(void *_buf, size_t bufsize);
 enum filetype file_name_detect_type(const char *filename);
 enum filetype is_fat_or_mbr(const unsigned char *sector, unsigned long *bootsec);
 
diff --git a/lib/gui/image_renderer.c b/lib/gui/image_renderer.c
index 41dc43b..78e5777 100644
--- a/lib/gui/image_renderer.c
+++ b/lib/gui/image_renderer.c
@@ -13,10 +13,10 @@
 
 static LIST_HEAD(image_renderers);
 
-static struct image_renderer *get_renderer(void* buf)
+static struct image_renderer *get_renderer(void* buf, size_t bufsize)
 {
 	struct image_renderer *ir;
-	enum filetype type = file_detect_type(buf);
+	enum filetype type = file_detect_type(buf, bufsize);
 
 	list_for_each_entry(ir, &image_renderers, list) {
 		if (ir->type == type)
@@ -40,7 +40,7 @@ struct image *image_renderer_open(const char* file)
 		return ERR_PTR(-ENOMEM);
 	}
 
-	ir = get_renderer(data);
+	ir = get_renderer(data, size);
 	if (!ir) {
 		ret = -ENOENT;
 		goto out;
diff --git a/lib/uncompress.c b/lib/uncompress.c
index 8e4d3a1..e0a69df 100644
--- a/lib/uncompress.c
+++ b/lib/uncompress.c
@@ -78,7 +78,7 @@ int uncompress(unsigned char *inbuf, int len,
 	char *err;
 
 	if (inbuf) {
-		ft = file_detect_type(inbuf);
+		ft = file_detect_type(inbuf, len);
 		uncompress_buf = NULL;
 		uncompress_size = 0;
 	} else {
@@ -93,7 +93,7 @@ int uncompress(unsigned char *inbuf, int len,
 		if (ret < 0)
 			goto err;
 
-		ft = file_detect_type(uncompress_buf);
+		ft = file_detect_type(uncompress_buf, 32);
 	}
 
 	switch (ft) {
-- 
1.7.10.4


_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

^ permalink raw reply	[flat|nested] 5+ messages in thread
* [PATCH 3/4] add ext fs detection support
  2012-12-03 10:27 [PATCH] ext4 filesystem support Sascha Hauer
  2012-12-03 10:27 ` [PATCH 1/4] ls command: call stat() only when necessary Sascha Hauer
  2012-12-03 10:27 ` [PATCH 2/4] filetype: Pass bufsize Sascha Hauer
@ 2012-12-03 10:27 ` Sascha Hauer
  2012-12-03 10:27 ` [PATCH 4/4] fs: implement initial ext4 support from U-Boot Sascha Hauer
  3 siblings, 0 replies; 5+ messages in thread
From: Sascha Hauer @ 2012-12-03 10:27 UTC (permalink / raw)
  To: barebox

Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
---
 common/filetype.c  |    8 ++++++++
 include/filetype.h |    1 +
 2 files changed, 9 insertions(+)

diff --git a/common/filetype.c b/common/filetype.c
index c1bd11d..748e364 100644
--- a/common/filetype.c
+++ b/common/filetype.c
@@ -47,6 +47,7 @@ static const struct filetype_str filetype_str[] = {
 	[filetype_mbr] = { "MBR sector", "mbr" },
 	[filetype_bmp] = { "BMP image", "bmp" },
 	[filetype_png] = { "PNG image", "png" },
+	[filetype_ext] = { "ext filesystem", "ext" },
 };
 
 const char *file_type_to_string(enum filetype f)
@@ -110,6 +111,7 @@ enum filetype file_detect_type(void *_buf, size_t bufsize)
 	u32 *buf = _buf;
 	u64 *buf64 = _buf;
 	u8 *buf8 = _buf;
+	u16 *buf16 = _buf;
 	enum filetype type;
 
 	if (bufsize < 9)
@@ -161,6 +163,12 @@ enum filetype file_detect_type(void *_buf, size_t bufsize)
 	if (type != filetype_unknown)
 		return type;
 
+	if (bufsize < 1536)
+		return filetype_unknown;
+
+	if (buf16[512 + 28] == le16_to_cpu(0xef53))
+		return filetype_ext;
+
 	return filetype_unknown;
 }
 
diff --git a/include/filetype.h b/include/filetype.h
index 5fac531..91139db 100644
--- a/include/filetype.h
+++ b/include/filetype.h
@@ -22,6 +22,7 @@ enum filetype {
 	filetype_mbr,
 	filetype_bmp,
 	filetype_png,
+	filetype_ext,
 	filetype_max,
 };
 
-- 
1.7.10.4


_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

^ permalink raw reply	[flat|nested] 5+ messages in thread
* [PATCH 4/4] fs: implement initial ext4 support from U-Boot
  2012-12-03 10:27 [PATCH] ext4 filesystem support Sascha Hauer
                   ` (2 preceding siblings ...)
  2012-12-03 10:27 ` [PATCH 3/4] add ext fs detection support Sascha Hauer
@ 2012-12-03 10:27 ` Sascha Hauer
  3 siblings, 0 replies; 5+ messages in thread
From: Sascha Hauer @ 2012-12-03 10:27 UTC (permalink / raw)
  To: barebox

The ext4 implementation has been taken from U-Boot with some
changes:

- No global variables to allow for multiple filesystems to
  be mounted and multiple files to be open.
- remove fs internal link following and use the barebox link
  implementation.
- remove write support. This is incomplete in U-Boot, so I decided
  to skip this for now.

Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
---
 fs/Kconfig            |    2 +
 fs/Makefile           |    1 +
 fs/ext4/Kconfig       |    3 +
 fs/ext4/Makefile      |    1 +
 fs/ext4/ext4_common.c |  551 +++++++++++++++++++++++++++++++++++++++++++++++++
 fs/ext4/ext4_common.h |   58 ++++++
 fs/ext4/ext4fs.c      |  153 ++++++++++++++
 fs/ext4/ext4fs.h      |  127 ++++++++++++
 fs/ext4/ext_barebox.c |  293 ++++++++++++++++++++++++++
 fs/ext4/ext_common.h  |  195 +++++++++++++++++
 10 files changed, 1384 insertions(+)
 create mode 100644 fs/ext4/Kconfig
 create mode 100644 fs/ext4/Makefile
 create mode 100644 fs/ext4/ext4_common.c
 create mode 100644 fs/ext4/ext4_common.h
 create mode 100644 fs/ext4/ext4fs.c
 create mode 100644 fs/ext4/ext4fs.h
 create mode 100644 fs/ext4/ext_barebox.c
 create mode 100644 fs/ext4/ext_common.h

diff --git a/fs/Kconfig b/fs/Kconfig
index 0ab69d7..c31c0cd 100644
--- a/fs/Kconfig
+++ b/fs/Kconfig
@@ -14,6 +14,8 @@ config FS_CRAMFS
 	select ZLIB
 	prompt "cramfs support"
 
+source fs/ext4/Kconfig
+
 config FS_RAMFS
 	bool
 	default y
diff --git a/fs/Makefile b/fs/Makefile
index ad745d9..cc59da7 100644
--- a/fs/Makefile
+++ b/fs/Makefile
@@ -1,4 +1,5 @@
 obj-$(CONFIG_FS_CRAMFS)	+= cramfs/
+obj-$(CONFIG_FS_EXT4)	+= ext4/
 obj-$(CONFIG_FS_RAMFS)	+= ramfs.o
 obj-y			+= devfs-core.o
 obj-$(CONFIG_FS_DEVFS)	+= devfs.o
diff --git a/fs/ext4/Kconfig b/fs/ext4/Kconfig
new file mode 100644
index 0000000..f36043d
--- /dev/null
+++ b/fs/ext4/Kconfig
@@ -0,0 +1,3 @@
+config FS_EXT4
+	bool
+	prompt "ext4 filesystem support"
diff --git a/fs/ext4/Makefile b/fs/ext4/Makefile
new file mode 100644
index 0000000..5084e3f
--- /dev/null
+++ b/fs/ext4/Makefile
@@ -0,0 +1 @@
+obj-$(CONFIG_FS_EXT4) += ext4fs.o ext4_common.o ext_barebox.o
diff --git a/fs/ext4/ext4_common.c b/fs/ext4/ext4_common.c
new file mode 100644
index 0000000..f426118
--- /dev/null
+++ b/fs/ext4/ext4_common.c
@@ -0,0 +1,551 @@
+/*
+ * (C) Copyright 2011 - 2012 Samsung Electronics
+ * EXT4 filesystem implementation in Uboot by
+ * Uma Shankar <uma.shankar@samsung.com>
+ * Manjunatha C Achar <a.manjunatha@samsung.com>
+ *
+ * ext4ls and ext4load : Based on ext2 ls load support in Uboot.
+ *
+ * (C) Copyright 2004
+ * esd gmbh <www.esd-electronics.com>
+ * Reinhard Arlt <reinhard.arlt@esd-electronics.com>
+ *
+ * based on code from grub2 fs/ext2.c and fs/fshelp.c by
+ * GRUB  --  GRand Unified Bootloader
+ * Copyright (C) 2003, 2004  Free Software Foundation, Inc.
+ *
+ * ext4write : Based on generic ext4 protocol.
+ *
+ * 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.
+ *
+ */
+
+#include <common.h>
+#include <malloc.h>
+#include <stddef.h>
+#include <linux/stat.h>
+#include <linux/time.h>
+#include <asm/byteorder.h>
+#include <dma.h>
+
+#include "ext4_common.h"
+
+static struct ext4_extent_header *ext4fs_get_extent_block(struct ext2_data *data,
+		char *buf, struct ext4_extent_header *ext_block,
+		uint32_t fileblock, int log2_blksz)
+{
+	struct ext4_extent_idx *index;
+	unsigned long long block;
+	struct ext_filesystem *fs = data->fs;
+	int i, ret;
+
+	while (1) {
+		index = (struct ext4_extent_idx *)(ext_block + 1);
+
+		if (le32_to_cpu(ext_block->eh_magic) != EXT4_EXT_MAGIC)
+			return 0;
+
+		if (ext_block->eh_depth == 0)
+			return ext_block;
+		i = -1;
+		do {
+			i++;
+			if (i >= le32_to_cpu(ext_block->eh_entries))
+				break;
+		} while (fileblock > le32_to_cpu(index[i].ei_block));
+
+		if (--i < 0)
+			return 0;
+
+		block = le32_to_cpu(index[i].ei_leaf_hi);
+		block = (block << 32) + le32_to_cpu(index[i].ei_leaf_lo);
+
+		ret = ext4fs_devread(fs, block << log2_blksz, 0, fs->blksz, buf);
+		if (ret)
+			return NULL;
+		else
+			ext_block = (struct ext4_extent_header *)buf;
+	}
+}
+
+static int ext4fs_blockgroup(struct ext2_data *data, int group,
+		struct ext2_block_group *blkgrp)
+{
+	long int blkno;
+	unsigned int blkoff, desc_per_blk;
+	struct ext_filesystem *fs = data->fs;
+
+	desc_per_blk = EXT2_BLOCK_SIZE(data) / sizeof(struct ext2_block_group);
+
+	blkno = __le32_to_cpu(data->sblock.first_data_block) + 1 +
+			group / desc_per_blk;
+	blkoff = (group % desc_per_blk) * sizeof(struct ext2_block_group);
+
+	debug("ext4fs read %d group descriptor (blkno %ld blkoff %u)\n",
+	      group, blkno, blkoff);
+
+	return ext4fs_devread(fs, blkno << LOG2_EXT2_BLOCK_SIZE(data),
+			      blkoff, sizeof(struct ext2_block_group),
+			      (char *)blkgrp);
+}
+
+int ext4fs_read_inode(struct ext2_data *data, int ino, struct ext2_inode *inode)
+{
+	struct ext2_block_group blkgrp;
+	struct ext2_sblock *sblock = &data->sblock;
+	struct ext_filesystem *fs = data->fs;
+	int inodes_per_block, ret;
+	long int blkno;
+	unsigned int blkoff;
+
+	/* It is easier to calculate if the first inode is 0. */
+	ino--;
+	ret = ext4fs_blockgroup(data, ino / __le32_to_cpu
+				   (sblock->inodes_per_group), &blkgrp);
+	if (ret)
+		return ret;
+
+	inodes_per_block = EXT2_BLOCK_SIZE(data) / fs->inodesz;
+	blkno = __le32_to_cpu(blkgrp.inode_table_id) +
+	    (ino % __le32_to_cpu(sblock->inodes_per_group)) / inodes_per_block;
+	blkoff = (ino % inodes_per_block) * fs->inodesz;
+	/* Read the inode. */
+	ret = ext4fs_devread(fs, blkno << LOG2_EXT2_BLOCK_SIZE(data), blkoff,
+				sizeof(struct ext2_inode), (char *)inode);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+int ext4fs_get_indir_block(struct ext2fs_node *node, struct ext4fs_indir_block *indir, int blkno)
+{
+	struct ext_filesystem *fs = node->data->fs;
+	int blksz;
+	int ret;
+
+	blksz = EXT2_BLOCK_SIZE(node->data);
+
+	if (indir->blkno == blkno)
+		return 0;
+
+	ret = ext4fs_devread(fs, blkno, 0, blksz, (void *)indir->data);
+	if (ret) {
+		printf("** SI ext2fs read block (indir 1)"
+			"failed. **\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+long int read_allocated_block(struct ext2fs_node *node, int fileblock)
+{
+	long int blknr;
+	int blksz;
+	int log2_blksz;
+	long int rblock;
+	long int perblock_parent;
+	long int perblock_child;
+	unsigned long long start;
+	struct ext2_inode *inode = &node->inode;
+	struct ext2_data *data = node->data;
+	int ret;
+
+	/* get the blocksize of the filesystem */
+	blksz = EXT2_BLOCK_SIZE(node->data);
+	log2_blksz = LOG2_EXT2_BLOCK_SIZE(node->data);
+
+	if (le32_to_cpu(inode->flags) & EXT4_EXTENTS_FL) {
+		char *buf = zalloc(blksz);
+		struct ext4_extent_header *ext_block;
+		struct ext4_extent *extent;
+		int i = -1;
+
+		if (!buf)
+			return -ENOMEM;
+
+		ext_block = ext4fs_get_extent_block(node->data, buf,
+				(struct ext4_extent_header *)inode->b.blocks.dir_blocks,
+				fileblock, log2_blksz);
+		if (!ext_block) {
+			printf("invalid extent block\n");
+			free(buf);
+			return -EINVAL;
+		}
+
+		extent = (struct ext4_extent *)(ext_block + 1);
+
+		do {
+			i++;
+			if (i >= le32_to_cpu(ext_block->eh_entries))
+				break;
+		} while (fileblock >= le32_to_cpu(extent[i].ee_block));
+
+		if (--i >= 0) {
+			fileblock -= le32_to_cpu(extent[i].ee_block);
+			if (fileblock >= le32_to_cpu(extent[i].ee_len)) {
+				free(buf);
+				return 0;
+			}
+
+			start = le32_to_cpu(extent[i].ee_start_hi);
+			start = (start << 32) +
+					le32_to_cpu(extent[i].ee_start_lo);
+			free(buf);
+			return fileblock + start;
+		}
+
+		free(buf);
+		return -EIO;
+	}
+
+	if (fileblock < INDIRECT_BLOCKS) {
+		/* Direct blocks. */
+		blknr = __le32_to_cpu(inode->b.blocks.dir_blocks[fileblock]);
+	} else if (fileblock < (INDIRECT_BLOCKS + (blksz / 4))) {
+		/* Indirect. */
+		ret = ext4fs_get_indir_block(node, &data->indir1,
+				__le32_to_cpu(inode->b.blocks.indir_block) << log2_blksz);
+		if (ret)
+			return ret;
+		blknr = __le32_to_cpu(data->indir1.data[fileblock - INDIRECT_BLOCKS]);
+	} else if (fileblock < (INDIRECT_BLOCKS + (blksz / 4 *
+					(blksz / 4 + 1)))) {
+		/* Double indirect. */
+		long int perblock = blksz / 4;
+		long int rblock = fileblock - (INDIRECT_BLOCKS + blksz / 4);
+
+		ret = ext4fs_get_indir_block(node, &data->indir1,
+				__le32_to_cpu(inode->b.blocks.double_indir_block) << log2_blksz);
+		if (ret)
+			return ret;
+
+		ret = ext4fs_get_indir_block(node, &data->indir2,
+				__le32_to_cpu(data->indir1.data[rblock / perblock]) << log2_blksz);
+		if (ret)
+			return ret;
+
+		blknr = __le32_to_cpu(data->indir2.data[rblock % perblock]);
+	} else {
+		/* Triple indirect. */
+		rblock = fileblock - (INDIRECT_BLOCKS + blksz / 4 +
+				      (blksz / 4 * blksz / 4));
+		perblock_child = blksz / 4;
+		perblock_parent = ((blksz / 4) * (blksz / 4));
+
+		ret = ext4fs_get_indir_block(node, &data->indir1,
+				__le32_to_cpu(inode->b.blocks.triple_indir_block) << log2_blksz);
+		if (ret)
+			return ret;
+
+		ret = ext4fs_get_indir_block(node, &data->indir2,
+				__le32_to_cpu(data->indir1.data[rblock / perblock_parent]) << log2_blksz);
+		if (ret)
+			return ret;
+
+		ret = ext4fs_get_indir_block(node, &data->indir3,
+				__le32_to_cpu(data->indir2.data[rblock / perblock_child]) << log2_blksz);
+		if (ret)
+			return ret;
+
+		blknr = __le32_to_cpu(data->indir3.data[rblock % perblock_child]);
+	}
+
+	return blknr;
+}
+
+int ext4fs_iterate_dir(struct ext2fs_node *dir, char *name,
+				struct ext2fs_node **fnode, int *ftype)
+{
+	unsigned int fpos = 0;
+	int status, ret;
+	struct ext2fs_node *diro = (struct ext2fs_node *) dir;
+
+
+	if (name != NULL)
+		debug("Iterate dir %s\n", name);
+
+	if (!diro->inode_read) {
+		ret = ext4fs_read_inode(diro->data, diro->ino, &diro->inode);
+		if (ret)
+			return ret;
+	}
+	/* Search the file.  */
+	while (fpos < __le32_to_cpu(diro->inode.size)) {
+		struct ext2_dirent dirent;
+
+		status = ext4fs_read_file(diro, fpos,
+					   sizeof(struct ext2_dirent),
+					   (char *) &dirent);
+		if (status < 1)
+			return -EINVAL;
+
+		if (dirent.namelen != 0) {
+			char filename[dirent.namelen + 1];
+			struct ext2fs_node *fdiro;
+			int type = FILETYPE_UNKNOWN;
+
+			status = ext4fs_read_file(diro,
+						  fpos +
+						  sizeof(struct ext2_dirent),
+						  dirent.namelen, filename);
+			if (status < 1)
+				return -EINVAL;
+
+			fdiro = zalloc(sizeof(struct ext2fs_node));
+			if (!fdiro)
+				return -ENOMEM;
+
+			fdiro->data = diro->data;
+			fdiro->ino = __le32_to_cpu(dirent.inode);
+
+			filename[dirent.namelen] = '\0';
+
+			if (dirent.filetype != FILETYPE_UNKNOWN) {
+				fdiro->inode_read = 0;
+
+				if (dirent.filetype == FILETYPE_DIRECTORY)
+					type = FILETYPE_DIRECTORY;
+				else if (dirent.filetype == FILETYPE_SYMLINK)
+					type = FILETYPE_SYMLINK;
+				else if (dirent.filetype == FILETYPE_REG)
+					type = FILETYPE_REG;
+			} else {
+				ret = ext4fs_read_inode(diro->data,
+							   __le32_to_cpu
+							   (dirent.inode),
+							   &fdiro->inode);
+				if (ret) {
+					free(fdiro);
+					return ret;
+				}
+				fdiro->inode_read = 1;
+
+				if ((__le16_to_cpu(fdiro->inode.mode) &
+				     FILETYPE_INO_MASK) ==
+				    FILETYPE_INO_DIRECTORY) {
+					type = FILETYPE_DIRECTORY;
+				} else if ((__le16_to_cpu(fdiro->inode.mode)
+					    & FILETYPE_INO_MASK) ==
+					   FILETYPE_INO_SYMLINK) {
+					type = FILETYPE_SYMLINK;
+				} else if ((__le16_to_cpu(fdiro->inode.mode)
+					    & FILETYPE_INO_MASK) ==
+					   FILETYPE_INO_REG) {
+					type = FILETYPE_REG;
+				}
+			}
+
+			debug("iterate >%s<\n", filename);
+
+			if (strcmp(filename, name) == 0) {
+				*ftype = type;
+				*fnode = fdiro;
+				return 0;
+			}
+
+			free(fdiro);
+		}
+		fpos += __le16_to_cpu(dirent.direntlen);
+	}
+	return -ENOENT;
+}
+
+char *ext4fs_read_symlink(struct ext2fs_node *node)
+{
+	char *symlink;
+	struct ext2fs_node *diro = node;
+	int status, ret;
+
+	if (!diro->inode_read) {
+		ret = ext4fs_read_inode(diro->data, diro->ino, &diro->inode);
+		if (ret)
+			return NULL;
+	}
+	symlink = zalloc(__le32_to_cpu(diro->inode.size) + 1);
+	if (!symlink)
+		return 0;
+
+	if (__le32_to_cpu(diro->inode.size) <= 60) {
+		strncpy(symlink, diro->inode.b.symlink,
+			 __le32_to_cpu(diro->inode.size));
+	} else {
+		status = ext4fs_read_file(diro, 0,
+					   __le32_to_cpu(diro->inode.size),
+					   symlink);
+		if (status == 0) {
+			free(symlink);
+			return NULL;
+		}
+	}
+
+	symlink[__le32_to_cpu(diro->inode.size)] = '\0';
+
+	return symlink;
+}
+
+int ext4fs_find_file(const char *currpath,
+			     struct ext2fs_node *currroot,
+			     struct ext2fs_node **currfound, int *foundtype)
+{
+	char fpath[strlen(currpath) + 1];
+	char *name = fpath;
+	char *next;
+	int type = FILETYPE_DIRECTORY;
+	struct ext2fs_node *currnode = currroot;
+	struct ext2fs_node *oldnode = currroot;
+	int ret = 0;
+
+	strncpy(fpath, currpath, strlen(currpath) + 1);
+
+	/* Remove all leading slashes. */
+	while (*name == '/')
+		name++;
+
+	if (!*name) {
+		*currfound = currnode;
+		goto out;
+	}
+
+	for (;;) {
+		/* Extract the actual part from the pathname. */
+		next = strchr(name, '/');
+		if (next) {
+			/* Remove all leading slashes. */
+			while (*next == '/')
+				*(next++) = '\0';
+		}
+
+		if (type != FILETYPE_DIRECTORY) {
+			ext4fs_free_node(currnode, currroot);
+			return -ENOENT;
+		}
+
+		oldnode = currnode;
+
+		/* Iterate over the directory. */
+		ret = ext4fs_iterate_dir(currnode, name, &currnode, &type);
+		if (ret)
+			return ret;
+
+		ext4fs_free_node(oldnode, currroot);
+
+		/* Found the node! */
+		if (!next || *next == '\0') {
+			*currfound = currnode;
+			goto out;
+		}
+		name = next;
+	}
+
+out:
+	if (foundtype)
+		*foundtype = type;
+
+	return ret;
+}
+
+int ext4fs_open(struct ext2_data *data, const char *filename, struct ext2fs_node **inode)
+{
+	struct ext2fs_node *fdiro = NULL;
+	int status, ret;
+	int type;
+
+	status = ext4fs_find_file(filename, &data->diropen, &fdiro, &type);
+	if (status)
+		goto fail;
+
+	if (type != FILETYPE_REG)
+		return -EINVAL;
+
+	if (!fdiro->inode_read) {
+		ret = ext4fs_read_inode(fdiro->data, fdiro->ino,
+				&fdiro->inode);
+		if (ret)
+			goto fail;
+	}
+
+	*inode = fdiro;
+
+	return 0;
+fail:
+	ext4fs_free_node(fdiro, &data->diropen);
+
+	return -ENOENT;
+}
+
+int ext4fs_mount(struct ext_filesystem *fs)
+{
+	struct ext2_data *data;
+	int ret, blksz;
+
+	data = zalloc(sizeof(struct ext2_data));
+	if (!data)
+		return -ENOMEM;
+
+	/* Read the superblock. */
+	ret = ext4fs_devread(fs, 1 * 2, 0, sizeof(struct ext2_sblock),
+				(char *)&data->sblock);
+	if (ret)
+		goto fail;
+
+	/* Make sure this is an ext2 filesystem. */
+	if (__le16_to_cpu(data->sblock.magic) != EXT2_MAGIC) {
+		ret = -EINVAL;
+		goto fail;
+	}
+
+	if (__le32_to_cpu(data->sblock.revision_level == 0))
+		fs->inodesz = 128;
+	else
+		fs->inodesz = __le16_to_cpu(data->sblock.inode_size);
+
+	dev_info(fs->dev, "EXT2 rev %d, inode_size %d\n",
+	       __le32_to_cpu(data->sblock.revision_level), fs->inodesz);
+
+	data->diropen.data = data;
+	data->diropen.ino = 2;
+	data->diropen.inode_read = 1;
+	data->inode = &data->diropen.inode;
+	data->fs = fs;
+	fs->data = data;
+
+	blksz = EXT2_BLOCK_SIZE(data);
+
+	fs->data->indir1.data = malloc(blksz);
+	fs->data->indir2.data = malloc(blksz);
+	fs->data->indir3.data = malloc(blksz);
+
+	if (!fs->data->indir1.data || !fs->data->indir2.data ||
+			!fs->data->indir3.data) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	ret = ext4fs_read_inode(data, 2, data->inode);
+	if (ret)
+		goto fail;
+
+	return 0;
+fail:
+	free(data);
+
+	return ret;
+}
+
+void ext4fs_umount(struct ext_filesystem *fs)
+{
+	free(fs->data->indir1.data);
+	free(fs->data->indir2.data);
+	free(fs->data->indir3.data);
+	free(fs->data);
+}
diff --git a/fs/ext4/ext4_common.h b/fs/ext4/ext4_common.h
new file mode 100644
index 0000000..81fb67e
--- /dev/null
+++ b/fs/ext4/ext4_common.h
@@ -0,0 +1,58 @@
+/*
+ * (C) Copyright 2011 - 2012 Samsung Electronics
+ * EXT4 filesystem implementation in Uboot by
+ * Uma Shankar <uma.shankar@samsung.com>
+ * Manjunatha C Achar <a.manjunatha@samsung.com>
+ *
+ * ext4ls and ext4load :  based on ext2 ls load support in Uboot.
+ *
+ * (C) Copyright 2004
+ * esd gmbh <www.esd-electronics.com>
+ * Reinhard Arlt <reinhard.arlt@esd-electronics.com>
+ *
+ * based on code from grub2 fs/ext2.c and fs/fshelp.c by
+ * GRUB  --  GRand Unified Bootloader
+ * Copyright (C) 2003, 2004  Free Software Foundation, Inc.
+ *
+ * ext4write : Based on generic ext4 protocol.
+ *
+ * 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.
+ *
+ */
+
+#ifndef __EXT4_COMMON__
+#define __EXT4_COMMON__
+#include <malloc.h>
+#include <errno.h>
+#include <dma.h>
+#include "ext4fs.h"
+#include "ext_common.h"
+
+static inline void *zalloc(size_t size)
+{
+	void *p = dma_alloc(size);
+
+	if (p)
+		memset(p, 0, size);
+
+	return p;
+}
+
+int ext4fs_read_inode(struct ext2_data *data, int ino,
+		      struct ext2_inode *inode);
+int ext4fs_read_file(struct ext2fs_node *node, int pos,
+		unsigned int len, char *buf);
+int ext4fs_find_file(const char *path, struct ext2fs_node *rootnode,
+			struct ext2fs_node **foundnode, int *foundtype);
+int ext4fs_iterate_dir(struct ext2fs_node *dir, char *name,
+			struct ext2fs_node **fnode, int *ftype);
+
+#endif
diff --git a/fs/ext4/ext4fs.c b/fs/ext4/ext4fs.c
new file mode 100644
index 0000000..1b9af80
--- /dev/null
+++ b/fs/ext4/ext4fs.c
@@ -0,0 +1,153 @@
+/*
+ * (C) Copyright 2011 - 2012 Samsung Electronics
+ * EXT4 filesystem implementation in Uboot by
+ * Uma Shankar <uma.shankar@samsung.com>
+ * Manjunatha C Achar <a.manjunatha@samsung.com>
+ *
+ * ext4ls and ext4load : Based on ext2 ls and load support in Uboot.
+ *		       Ext4 read optimization taken from Open-Moko
+ *		       Qi bootloader
+ *
+ * (C) Copyright 2004
+ * esd gmbh <www.esd-electronics.com>
+ * Reinhard Arlt <reinhard.arlt@esd-electronics.com>
+ *
+ * based on code from grub2 fs/ext2.c and fs/fshelp.c by
+ * GRUB  --  GRand Unified Bootloader
+ * Copyright (C) 2003, 2004  Free Software Foundation, Inc.
+ *
+ *
+ * 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.
+ *
+ */
+
+#include <common.h>
+#include <malloc.h>
+#include <linux/stat.h>
+#include <linux/time.h>
+#include <asm/byteorder.h>
+#include "ext4_common.h"
+
+void ext4fs_free_node(struct ext2fs_node *node, struct ext2fs_node *currroot)
+{
+	if ((node != &node->data->diropen) && (node != currroot))
+		free(node);
+}
+
+/*
+ * Taken from openmoko-kernel mailing list: By Andy green
+ * Optimized read file API : collects and defers contiguous sector
+ * reads into one potentially more efficient larger sequential read action
+ */
+int ext4fs_read_file(struct ext2fs_node *node, int pos,
+		unsigned int len, char *buf)
+{
+	int i;
+	int blockcnt;
+	int log2blocksize = LOG2_EXT2_BLOCK_SIZE(node->data);
+	int blocksize = 1 << (log2blocksize + DISK_SECTOR_BITS);
+	unsigned int filesize = __le32_to_cpu(node->inode.size);
+	int previous_block_number = -1;
+	int delayed_start = 0;
+	int delayed_extent = 0;
+	int delayed_skipfirst = 0;
+	int delayed_next = 0;
+	char *delayed_buf = NULL;
+	short ret;
+	struct ext_filesystem *fs = node->data->fs;
+
+	/* Adjust len so it we can't read past the end of the file. */
+	if (len > filesize)
+		len = filesize;
+
+	blockcnt = ((len + pos) + blocksize - 1) / blocksize;
+
+	for (i = pos / blocksize; i < blockcnt; i++) {
+		int blknr;
+		int blockoff = pos % blocksize;
+		int blockend = blocksize;
+		int skipfirst = 0;
+		blknr = read_allocated_block(node, i);
+		if (blknr < 0)
+			return blknr;
+
+		blknr = blknr << log2blocksize;
+
+		/* Last block.  */
+		if (i == blockcnt - 1) {
+			blockend = (len + pos) % blocksize;
+
+			/* The last portion is exactly blocksize. */
+			if (!blockend)
+				blockend = blocksize;
+		}
+
+		/* First block. */
+		if (i == pos / blocksize) {
+			skipfirst = blockoff;
+			blockend -= skipfirst;
+		}
+		if (blknr) {
+			if (previous_block_number != -1) {
+				if (delayed_next == blknr) {
+					delayed_extent += blockend;
+					delayed_next += blockend >> SECTOR_BITS;
+				} else {	/* spill */
+					ret = ext4fs_devread(fs, delayed_start,
+							delayed_skipfirst,
+							delayed_extent,
+							delayed_buf);
+					if (ret)
+						return ret;
+					previous_block_number = blknr;
+					delayed_start = blknr;
+					delayed_extent = blockend;
+					delayed_skipfirst = skipfirst;
+					delayed_buf = buf;
+					delayed_next = blknr +
+						(blockend >> SECTOR_BITS);
+				}
+			} else {
+				previous_block_number = blknr;
+				delayed_start = blknr;
+				delayed_extent = blockend;
+				delayed_skipfirst = skipfirst;
+				delayed_buf = buf;
+				delayed_next = blknr +
+					(blockend >> SECTOR_BITS);
+			}
+		} else {
+			if (previous_block_number != -1) {
+				/* spill */
+				ret = ext4fs_devread(fs, delayed_start,
+							delayed_skipfirst,
+							delayed_extent,
+							delayed_buf);
+				if (ret)
+					return ret;
+				previous_block_number = -1;
+			}
+			memset(buf, 0, blocksize - skipfirst);
+		}
+		buf += blocksize - skipfirst;
+	}
+	if (previous_block_number != -1) {
+		/* spill */
+		ret = ext4fs_devread(fs, delayed_start,
+					delayed_skipfirst, delayed_extent,
+					delayed_buf);
+		if (ret)
+			return ret;
+		previous_block_number = -1;
+	}
+
+	return len;
+}
diff --git a/fs/ext4/ext4fs.h b/fs/ext4/ext4fs.h
new file mode 100644
index 0000000..ead212d
--- /dev/null
+++ b/fs/ext4/ext4fs.h
@@ -0,0 +1,127 @@
+/*
+ * (C) Copyright 2011 - 2012 Samsung Electronics
+ * EXT4 filesystem implementation in Uboot by
+ * Uma Shankar <uma.shankar@samsung.com>
+ * Manjunatha C Achar <a.manjunatha@samsung.com>
+ *
+ * Ext4 Extent data structures are taken from  original ext4 fs code
+ * as found in the linux kernel.
+ *
+ * Copyright (c) 2003-2006, Cluster File Systems, Inc, info@clusterfs.com
+ * Written by Alex Tomas <alex@clusterfs.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * 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.
+ *
+ */
+
+#ifndef __EXT4__
+#define __EXT4__
+
+#define EXT4_EXTENTS_FL		0x00080000 /* Inode uses extents */
+#define EXT4_EXT_MAGIC			0xf30a
+#define EXT4_FEATURE_RO_COMPAT_GDT_CSUM	0x0010
+#define EXT4_FEATURE_INCOMPAT_EXTENTS	0x0040
+#define EXT4_INDIRECT_BLOCKS		12
+
+#define EXT4_BG_INODE_UNINIT		0x0001
+#define EXT4_BG_BLOCK_UNINIT		0x0002
+#define EXT4_BG_INODE_ZEROED		0x0004
+
+/*
+ * ext4_inode has i_block array (60 bytes total).
+ * The first 12 bytes store ext4_extent_header;
+ * the remainder stores an array of ext4_extent.
+ */
+
+/*
+ * This is the extent on-disk structure.
+ * It's used at the bottom of the tree.
+ */
+struct ext4_extent {
+	__le32	ee_block;	/* first logical block extent covers */
+	__le16	ee_len;		/* number of blocks covered by extent */
+	__le16	ee_start_hi;	/* high 16 bits of physical block */
+	__le32	ee_start_lo;	/* low 32 bits of physical block */
+};
+
+/*
+ * This is index on-disk structure.
+ * It's used at all the levels except the bottom.
+ */
+struct ext4_extent_idx {
+	__le32	ei_block;	/* index covers logical blocks from 'block' */
+	__le32	ei_leaf_lo;	/* pointer to the physical block of the next *
+				 * level. leaf or next index could be there */
+	__le16	ei_leaf_hi;	/* high 16 bits of physical block */
+	__u16	ei_unused;
+};
+
+/* Each block (leaves and indexes), even inode-stored has header. */
+struct ext4_extent_header {
+	__le16	eh_magic;	/* probably will support different formats */
+	__le16	eh_entries;	/* number of valid entries */
+	__le16	eh_max;		/* capacity of store in entries */
+	__le16	eh_depth;	/* has tree real underlying blocks? */
+	__le32	eh_generation;	/* generation of the tree */
+};
+
+struct ext_filesystem {
+	/* Total Sector of partition */
+	uint64_t total_sect;
+	/* Block size  of partition */
+	uint32_t blksz;
+	/* Inode size of partition */
+	uint32_t inodesz;
+	/* Sectors per Block */
+	uint32_t sect_perblk;
+	/* Group Descriptor Block Number */
+	uint32_t gdtable_blkno;
+	/* Total block groups of partition */
+	uint32_t no_blkgrp;
+	/* No of blocks required for bgdtable */
+	uint32_t no_blk_pergdt;
+	/* Superblock */
+	struct ext2_sblock *sb;
+	/* Block group descritpor table */
+	struct ext2_block_group *bgd;
+	char *gdtable;
+
+	/* Block Bitmap Related */
+	unsigned char **blk_bmaps;
+	long int curr_blkno;
+	uint16_t first_pass_bbmap;
+
+	/* Inode Bitmap Related */
+	unsigned char **inode_bmaps;
+	int curr_inode_no;
+	uint16_t first_pass_ibmap;
+
+	/* Journal Related */
+
+	/* Block Device Descriptor */
+	struct cdev *cdev;
+
+	struct ext2_data *data;
+
+	struct device_d *dev;
+};
+
+struct ext2fs_node;
+
+int ext4fs_open(struct ext2_data *data, const char *filename, struct ext2fs_node **inode);
+int ext4fs_read(char *buf, unsigned len);
+int ext4fs_mount(struct ext_filesystem *fs);
+void ext4fs_umount(struct ext_filesystem *fs);
+char *ext4fs_read_symlink(struct ext2fs_node *node);
+void ext4fs_free_node(struct ext2fs_node *node, struct ext2fs_node *currroot);
+int ext4fs_devread(struct ext_filesystem *fs, int sector, int byte_offset, int byte_len, char *buf);
+long int read_allocated_block(struct ext2fs_node *node, int fileblock);
+
+#endif
diff --git a/fs/ext4/ext_barebox.c b/fs/ext4/ext_barebox.c
new file mode 100644
index 0000000..adc8f75
--- /dev/null
+++ b/fs/ext4/ext_barebox.c
@@ -0,0 +1,293 @@
+/*
+ * barebox ext4 support
+ *
+ * Copyright (c) 2012 Sascha Hauer <s.hauer@pengutronix.de>, Pengutronix
+ *
+ * See file CREDITS for list of people who contributed to this
+ * project.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2
+ * as published by the Free Software Foundation.
+ *
+ * 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.
+ *
+ */
+
+#include <common.h>
+#include <driver.h>
+#include <init.h>
+#include <malloc.h>
+#include <fs.h>
+#include <command.h>
+#include <errno.h>
+#include <linux/stat.h>
+#include <linux/ctype.h>
+#include <xfuncs.h>
+#include <fcntl.h>
+#include "ext4_common.h"
+
+int ext4fs_devread(struct ext_filesystem *fs, int __sector, int byte_offset,
+		int byte_len, char *buf)
+{
+	ssize_t size;
+	uint64_t sector = __sector;
+
+	size = cdev_read(fs->cdev, buf, byte_len, sector * SECTOR_SIZE + byte_offset, 0);
+	if (size < 0) {
+		dev_err(fs->dev, "read error at sector %d: %s\n", __sector,
+				strerror(-size));
+		return size;
+	}
+
+	return 0;
+}
+
+static int ext_open(struct device_d *dev, FILE *file, const char *filename)
+{
+	struct ext_filesystem *fs = dev->priv;
+	struct ext2fs_node *inode;
+	int ret;
+
+	ret = ext4fs_open(fs->data, filename, &inode);
+	if (ret)
+		return ret;
+
+	file->size = __le32_to_cpu(inode->inode.size);
+	file->inode = inode;
+
+	return 0;
+}
+
+static int ext_close(struct device_d *dev, FILE *f)
+{
+	struct ext_filesystem *fs = dev->priv;
+
+	ext4fs_free_node(f->inode, &fs->data->diropen);
+
+	return 0;
+}
+
+static int ext_read(struct device_d *_dev, FILE *f, void *buf, size_t insize)
+{
+	return ext4fs_read_file(f->inode, f->pos, insize, buf);
+}
+
+static loff_t ext_lseek(struct device_d *dev, FILE *f, loff_t pos)
+{
+	f->pos = pos;
+
+	return f->pos;
+}
+
+struct ext4fs_dir {
+	struct ext2fs_node *dirnode;
+	int fpos;
+	DIR dir;
+};
+
+static DIR *ext_opendir(struct device_d *dev, const char *pathname)
+{
+	struct ext_filesystem *fs = dev->priv;
+	struct ext4fs_dir *ext4_dir;
+	int type, ret;
+
+	ext4_dir = xzalloc(sizeof(*ext4_dir));
+
+	ret = ext4fs_find_file(pathname, &fs->data->diropen, &ext4_dir->dirnode,
+				  &type);
+	if (ret) {
+		free(ext4_dir);
+		return NULL;
+	}
+
+	if (type != FILETYPE_DIRECTORY)
+		return NULL;
+
+	ext4_dir->dir.priv = ext4_dir;
+
+	ret = ext4fs_read_inode(ext4_dir->dirnode->data, ext4_dir->dirnode->ino,
+			&ext4_dir->dirnode->inode);
+	if (ret) {
+		ext4fs_free_node(ext4_dir->dirnode, &fs->data->diropen);
+		free(ext4_dir);
+
+		return NULL;
+	}
+
+	return &ext4_dir->dir;
+}
+
+static struct dirent *ext_readdir(struct device_d *dev, DIR *dir)
+{
+	struct ext4fs_dir *ext4_dir = dir->priv;
+	struct ext2_dirent dirent;
+	struct ext2fs_node *diro = ext4_dir->dirnode;
+	int ret;
+	char *filename;
+
+	if (ext4_dir->fpos >= __le32_to_cpu(diro->inode.size))
+		return NULL;
+
+	ret = ext4fs_read_file(diro, ext4_dir->fpos, sizeof(struct ext2_dirent),
+			(char *) &dirent);
+	if (ret < 0)
+		return NULL;
+
+	if (dirent.namelen == 0)
+		return NULL;
+
+	filename = xzalloc(dirent.namelen + 1);
+
+	ret = ext4fs_read_file(diro, ext4_dir->fpos + sizeof(struct ext2_dirent),
+			dirent.namelen, filename);
+	if (ret < 0) {
+		free(filename);
+		return NULL;
+	}
+
+	filename[dirent.namelen] = '\0';
+
+	ext4_dir->fpos += __le16_to_cpu(dirent.direntlen);
+
+	strcpy(dir->d.d_name, filename);
+
+	free(filename);
+
+	return &dir->d;
+}
+
+static int ext_closedir(struct device_d *dev, DIR *dir)
+{
+	struct ext_filesystem *fs = dev->priv;
+	struct ext4fs_dir *ext4_dir = dir->priv;
+
+	ext4fs_free_node(ext4_dir->dirnode, &fs->data->diropen);
+
+	free(ext4_dir);
+
+	return 0;
+}
+
+static int ext_stat(struct device_d *dev, const char *filename, struct stat *s)
+{
+	struct ext_filesystem *fs = dev->priv;
+	struct ext2fs_node *node;
+	int status, ret;
+
+	status = ext4fs_find_file(filename, &fs->data->diropen, &node, NULL);
+	if (status)
+		return -ENOENT;
+
+	ret = ext4fs_read_inode(node->data, node->ino, &node->inode);
+	if (ret)
+		return ret;
+
+	s->st_size = __le32_to_cpu(node->inode.size);
+	s->st_mode = __le16_to_cpu(node->inode.mode);
+
+	ext4fs_free_node(node, &fs->data->diropen);
+
+	return 0;
+}
+
+static int ext_readlink(struct device_d *dev, const char *pathname,
+		char *buf, size_t bufsiz)
+{
+	struct ext_filesystem *fs = dev->priv;
+	struct ext2fs_node *node;
+	char *symlink;
+	int ret, len, type;
+
+	ret = ext4fs_find_file(pathname, &fs->data->diropen, &node, &type);
+	if (ret)
+		return ret;
+
+	if (type != FILETYPE_SYMLINK)
+		return -EINVAL;
+
+	symlink = ext4fs_read_symlink(node);
+	if (!symlink)
+		return -ENOENT;
+
+	len = min(bufsiz, strlen(symlink));
+
+	memcpy(buf, symlink, len);
+
+	free(symlink);
+
+	return 0;
+}
+
+static int ext_probe(struct device_d *dev)
+{
+	struct fs_device_d *fsdev = dev_to_fs_device(dev);
+	char *backingstore = fsdev->backingstore;
+	int ret;
+	struct ext_filesystem *fs;
+
+	fs = xzalloc(sizeof(*fs));
+
+	dev->priv = fs;
+	fs->dev = dev;
+
+	if (!strncmp(backingstore , "/dev/", 5))
+		backingstore += 5;
+
+	fs->cdev = cdev_open(backingstore, O_RDWR);
+	if (!fs->cdev) {
+		ret = -ENOENT;
+		goto err_open;
+	}
+
+	ret = ext4fs_mount(fs);
+	if (ret)
+		goto err_mount;
+
+	return 0;
+
+err_mount:
+	cdev_close(fs->cdev);
+err_open:
+	free(fs);
+
+	return ret;
+}
+
+static void ext_remove(struct device_d *dev)
+{
+	struct ext_filesystem *fs = dev->priv;
+
+	ext4fs_umount(fs);
+	cdev_close(fs->cdev);
+	free(fs);
+}
+
+static struct fs_driver_d ext_driver = {
+	.open      = ext_open,
+	.close     = ext_close,
+	.read      = ext_read,
+	.lseek     = ext_lseek,
+	.opendir   = ext_opendir,
+	.readdir   = ext_readdir,
+	.closedir  = ext_closedir,
+	.stat      = ext_stat,
+	.readlink  = ext_readlink,
+	.type      = filetype_ext,
+	.flags     = 0,
+	.drv = {
+		.probe  = ext_probe,
+		.remove = ext_remove,
+		.name = "ext4",
+	}
+};
+
+static int ext_init(void)
+{
+	return register_fs_driver(&ext_driver);
+}
+
+coredevice_initcall(ext_init);
diff --git a/fs/ext4/ext_common.h b/fs/ext4/ext_common.h
new file mode 100644
index 0000000..517a1c1
--- /dev/null
+++ b/fs/ext4/ext_common.h
@@ -0,0 +1,195 @@
+/*
+ * (C) Copyright 2011 - 2012 Samsung Electronics
+ * EXT4 filesystem implementation in Uboot by
+ * Uma Shankar <uma.shankar@samsung.com>
+ * Manjunatha C Achar <a.manjunatha@samsung.com>
+ *
+ * Data structures and headers for ext4 support have been taken from
+ * ext2 ls load support in Uboot
+ *
+ * (C) Copyright 2004
+ * esd gmbh <www.esd-electronics.com>
+ * Reinhard Arlt <reinhard.arlt@esd-electronics.com>
+ *
+ * based on code from grub2 fs/ext2.c and fs/fshelp.c by
+ * GRUB  --  GRand Unified Bootloader
+ * Copyright (C) 2003, 2004  Free Software Foundation, Inc.
+ *
+ * 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.
+ *
+ */
+
+#ifndef __EXT_COMMON__
+#define __EXT_COMMON__
+
+#define SECTOR_SIZE		0x200
+#define SECTOR_BITS		9
+
+/* Magic value used to identify an ext2 filesystem.  */
+#define	EXT2_MAGIC			0xEF53
+/* Amount of indirect blocks in an inode.  */
+#define INDIRECT_BLOCKS			12
+/* Maximum lenght of a pathname.  */
+#define EXT2_PATH_MAX				4096
+/* Maximum nesting of symlinks, used to prevent a loop.  */
+#define	EXT2_MAX_SYMLINKCNT		8
+
+/* Filetype used in directory entry.  */
+#define	FILETYPE_UNKNOWN		0
+#define	FILETYPE_REG			1
+#define	FILETYPE_DIRECTORY		2
+#define	FILETYPE_SYMLINK		7
+
+/* Filetype information as used in inodes.  */
+#define FILETYPE_INO_MASK		0170000
+#define FILETYPE_INO_REG		0100000
+#define FILETYPE_INO_DIRECTORY		0040000
+#define FILETYPE_INO_SYMLINK		0120000
+#define EXT2_ROOT_INO			2 /* Root inode */
+
+/* Bits used as offset in sector */
+#define DISK_SECTOR_BITS		9
+/* The size of an ext2 block in bytes.  */
+#define EXT2_BLOCK_SIZE(data)	   (1 << LOG2_BLOCK_SIZE(data))
+
+/* Log2 size of ext2 block in 512 blocks.  */
+#define LOG2_EXT2_BLOCK_SIZE(data) (__le32_to_cpu \
+				(data->sblock.log2_block_size) + 1)
+
+/* Log2 size of ext2 block in bytes.  */
+#define LOG2_BLOCK_SIZE(data)	   (__le32_to_cpu \
+		(data->sblock.log2_block_size) + 10)
+#define INODE_SIZE_FILESYSTEM(data)	(__le32_to_cpu \
+			(data->sblock.inode_size))
+
+#define EXT2_FT_DIR	2
+#define SUCCESS	1
+
+/* Macro-instructions used to manage several block sizes  */
+#define EXT2_MIN_BLOCK_LOG_SIZE	10 /* 1024 */
+#define EXT2_MAX_BLOCK_LOG_SIZE	16 /* 65536 */
+#define EXT2_MIN_BLOCK_SIZE		(1 << EXT2_MIN_BLOCK_LOG_SIZE)
+#define EXT2_MAX_BLOCK_SIZE		(1 << EXT2_MAX_BLOCK_LOG_SIZE)
+
+/* The ext2 superblock.  */
+struct ext2_sblock {
+	uint32_t total_inodes;
+	uint32_t total_blocks;
+	uint32_t reserved_blocks;
+	uint32_t free_blocks;
+	uint32_t free_inodes;
+	uint32_t first_data_block;
+	uint32_t log2_block_size;
+	uint32_t log2_fragment_size;
+	uint32_t blocks_per_group;
+	uint32_t fragments_per_group;
+	uint32_t inodes_per_group;
+	uint32_t mtime;
+	uint32_t utime;
+	uint16_t mnt_count;
+	uint16_t max_mnt_count;
+	uint16_t magic;
+	uint16_t fs_state;
+	uint16_t error_handling;
+	uint16_t minor_revision_level;
+	uint32_t lastcheck;
+	uint32_t checkinterval;
+	uint32_t creator_os;
+	uint32_t revision_level;
+	uint16_t uid_reserved;
+	uint16_t gid_reserved;
+	uint32_t first_inode;
+	uint16_t inode_size;
+	uint16_t block_group_number;
+	uint32_t feature_compatibility;
+	uint32_t feature_incompat;
+	uint32_t feature_ro_compat;
+	uint32_t unique_id[4];
+	char volume_name[16];
+	char last_mounted_on[64];
+	uint32_t compression_info;
+};
+
+struct ext2_block_group {
+	__u32 block_id;	/* Blocks bitmap block */
+	__u32 inode_id;	/* Inodes bitmap block */
+	__u32 inode_table_id;	/* Inodes table block */
+	__u16 free_blocks;	/* Free blocks count */
+	__u16 free_inodes;	/* Free inodes count */
+	__u16 used_dir_cnt;	/* Directories count */
+	__u16 bg_flags;
+	__u32 bg_reserved[2];
+	__u16 bg_itable_unused; /* Unused inodes count */
+	__u16 bg_checksum;	/* crc16(s_uuid+grouo_num+group_desc)*/
+};
+
+/* The ext2 inode. */
+struct ext2_inode {
+	uint16_t mode;
+	uint16_t uid;
+	uint32_t size;
+	uint32_t atime;
+	uint32_t ctime;
+	uint32_t mtime;
+	uint32_t dtime;
+	uint16_t gid;
+	uint16_t nlinks;
+	uint32_t blockcnt;	/* Blocks of 512 bytes!! */
+	uint32_t flags;
+	uint32_t osd1;
+	union {
+		struct datablocks {
+			uint32_t dir_blocks[INDIRECT_BLOCKS];
+			uint32_t indir_block;
+			uint32_t double_indir_block;
+			uint32_t triple_indir_block;
+		} blocks;
+		char symlink[60];
+	} b;
+	uint32_t version;
+	uint32_t acl;
+	uint32_t dir_acl;
+	uint32_t fragment_addr;
+	uint32_t osd2[3];
+};
+
+/* The header of an ext2 directory entry. */
+struct ext2_dirent {
+	uint32_t inode;
+	uint16_t direntlen;
+	uint8_t namelen;
+	uint8_t filetype;
+};
+
+struct ext2fs_node {
+	struct ext2_data *data;
+	struct ext2_inode inode;
+	int ino;
+	int inode_read;
+};
+
+struct ext4fs_indir_block {
+	int size;
+	int blkno;
+	uint32_t *data;
+};
+
+/* Information about a "mounted" ext2 filesystem. */
+struct ext2_data {
+	struct ext2_sblock sblock;
+	struct ext2_inode *inode;
+	struct ext2fs_node diropen;
+	struct ext_filesystem *fs;
+	struct ext4fs_indir_block indir1, indir2, indir3;
+};
+
+extern unsigned long part_offset;
+#endif
-- 
1.7.10.4


_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

^ permalink raw reply	[flat|nested] 5+ messages in thread
end of thread, other threads:[~2012-12-03 10:27 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2012-12-03 10:27 [PATCH] ext4 filesystem support Sascha Hauer
2012-12-03 10:27 ` [PATCH 1/4] ls command: call stat() only when necessary Sascha Hauer
2012-12-03 10:27 ` [PATCH 2/4] filetype: Pass bufsize Sascha Hauer
2012-12-03 10:27 ` [PATCH 3/4] add ext fs detection support Sascha Hauer
2012-12-03 10:27 ` [PATCH 4/4] fs: implement initial ext4 support from U-Boot Sascha Hauer

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox