提示：扫一扫查出行【扫一扫了解最新限行尾号】复制提示原文地址:
http://opendevkit.com/?e=37
1. ubifs号称性能比yaffs2 好，同时压缩可读写，文件系统image体较小同时可写，相当不错
2. ubifs制作
(1) mkfs.ubifs
mkfs.ubifs -r system -m 2048 -e 126976 -c 1057 -x zlib -o system.ubifs
-r说的目录
-m说的是nand flash的页大小，一般都是2048或者4096，当然也有些是512，硬件特性决定好了，或者说驱动规定好了，有些平台4096页当做2048用也是用的
-e说的是逻辑擦除块大小，大家知道nand flash页读页写块擦，一个设备多个块，一个块多个页，一般也都是一个块是64个页，这样算一下无论擦除块大小就是2048*64=131072，-e的算法是物理擦除块大小-2*页大小，这里就是131072-2*2048=126976
-c说的是最大逻辑块编号，这个很重要，不能大也不能小，太小也要大于image大小，太大mount有问题，计算起点是分区的物理块数量，比如128MiB的mtd分区，物理块数量是128MiB/2048/64 = 1024个，需要减去2个坏块保留块，减去1个wear-leveling块，还要减去1个eba的块，等等，比如最终的值是1022，注意，如果物理上这个分区有坏块的话，kernel会扫描到的，这时候，我们计算的这个值就要减去坏块数了，否则会有逻辑块大于物理块数的内核问题mount失败，确切知道坏块数是比较困难的，一般做法是做一个坏块容忍数，比如20个，这样我们再减去20个坏块，不要担心这个会浪费空间，ubinize的autoresize选项就是解决这个问题的。具体的这个值需要计算。！！！！
-o说的image名字
-x说的是压缩方法，默认是lzo，还支持zlib，zlib压缩率高些，但是lzo压缩解压速度快
(2) ubinize
ubinize -o system.ubi -m 2048 -p 131072 ubinize.cfg
-o说的是输出image
-m还是页大小
-p是物理擦除块大小
ubinize.cfg是volume配置文件，例子如下：
- [ubifs]
- mode=ubi
- image=system.ubifs #说的是mkfs.ubifs的结果
- vol_id=0
- vol_size=100MiB #说的是volume大小，用-e和-c的值做乘法计算，一般不用写，autoresize会自动根据mtd分区大小适应，默认值是image大小，写了这个作用是帮助检查image是否超过了分区限制，制作时候就提示，否则mount会出错。-c的值是经过计算的最大值了，不过autoresize参数会自适应大小，不会浪费空间的。
- vol_type=dynamic
- vol_alignment=1
- vol_name=system #说的是分区名字
- vol_flags=autoresize
(4) uboot支持
-0- 需要打开的配置 - 需要烧写mkfs.ubifs结果时候需要打开，ubinize处理过的不需要
CONFIG_CMD_UBI
CONFIG_CMD_UBIFS
CONFIG_LZO
CONFIG_RBTREE
CONFIG_ZLIB
CONFIG_GZIP
-1- ubifs烧写
mkfs.ubifs工具制作的结果，就是ubifs镜像，不包含volume信息，需要用
nand erase.part system
ubi part system - 激活分区
ubi create system - 创建分区
ubi write 84000000 system xxxxx - xxxxx表示镜像实际大小
-2- ubi volume bin烧写
mkfs.ubifs后，使用ubinize处理了ubifs的镜像后，镜像含有了volume信息，直接
nand write 84000000 system xxxxx - xxxxx表示镜像实际大小
(5) kernel支持
Device Drivers --->Memory Technology Device(MTD) support --->UBI - Unsorted block images ---><*> Enable UBI - Unsorted block imagesFile systems --->Miscellaneous filesystems ---><*> UBIFS file system support[*] Advanced compression options[*] LZO compression support[*] ZLIB compression support(6) android支持
-1- andriod/system/core/rootdir/init.rc
mount yaffs2 mtd@userdata /data nosuid nodev
改为：
mount ubifs ubi@userdata /data nosuid nodev
-2- 增加对ubifs的mount支持
andriod/system/core/init/builtins.c在 } else if (!strncmp(source, "loop@", 5)) {之前加上+ }else if (!strncmp(source, "ubi@", 4)) {+ n = ubi_attach_mtd(source + 4);+ if (n < 0) {+ return -1;+ }+ sprintf(tmp, "/dev/ubi%d_0", n);+ if (wait)+ wait_for_file(tmp, COMMAND_RETRY_TIMEOUT);+ if (mount(tmp, target, system, flags, options) < 0) {+ ubi_detach_dev(n);+ return -1;+ }+ return 0;+ }else if (!strncmp(source, "loop@", 5)) {3. andriod/system/core/init/init.c在static int property_triggers_enabled = 0;之后加上+static unsigned ubifs_enabled = 1;static int set_init_properties_action(int nargs, char **args){property_set("ro.revision", tmp);+ property_set("ro.ubifs",ubifs_enabled ? "1" : "0");return 0;}int main(int argc, char **argv){action_for_each_trigger("post-fs", action_add_queue_tail);+ action_for_each_trigger("ubi-fs", action_add_queue_tail);}4. andriod/system/core/init/util.hvoid get_hardware_name(char *hardware, unsigned int *revision);+int ubi_attach_mtd(const char *name);+int ubi_detach_dev(int dev);5. andriod/system/core/init/util.c#include <sys/un.h>+#include <sys/ioctl.h>#include "util.h"+#include "ubi-user.h"+#define UBI_CTRL_DEV "/dev/ubi_ctrl"+#define UBI_SYS_PATH "/sys/class/ubi"在最后添加下面三个函数static int ubi_dev_read_int(int dev, const char *file, int def){int fd, val = def;char path[128], buf[64];sprintf(path, UBI_SYS_PATH "/ubi%d/%s", dev, file);wait_for_file(path, 5);fd = open(path, O_RDONLY);if (fd == -1) {return val;}if (read(fd, buf, 64) > 0) {val = atoi(buf);}close(fd);return val;}int ubi_attach_mtd(const char *name){int ret;int mtd_num, ubi_num;int ubi_ctrl, ubi_dev;int vols, avail_lebs, leb_size;char path[128];struct ubi_attach_req attach_req;struct ubi_mkvol_req mkvol_req;mtd_num = mtd_name_to_number(name);if (mtd_num == -1) {return -1;}ubi_ctrl = open(UBI_CTRL_DEV, O_RDONLY);if (ubi_ctrl == -1) {return -1;}memset(&attach_req, 0, sizeof(struct ubi_attach_req));attach_req.ubi_num = UBI_DEV_NUM_AUTO;attach_req.mtd_num = mtd_num;ret = ioctl(ubi_ctrl, UBI_IOCATT, &attach_req);if (ret == -1) {close(ubi_ctrl);return -1;}ubi_num = attach_req.ubi_num;vols = ubi_dev_read_int(ubi_num, "volumes_count", -1);if (vols == 0) {sprintf(path, "/dev/ubi%d", ubi_num);ubi_dev = open(path, O_RDONLY);if (ubi_dev == -1) {close(ubi_ctrl);return ubi_num;}avail_lebs = ubi_dev_read_int(ubi_num, "avail_eraseblocks", 0);leb_size = ubi_dev_read_int(ubi_num, "eraseblock_size", 0);memset(&mkvol_req, 0, sizeof(struct ubi_mkvol_req));mkvol_req.vol_id = UBI_VOL_NUM_AUTO;mkvol_req.alignment = 1;mkvol_req.bytes = (long long)avail_lebs * leb_size;mkvol_req.vol_type = UBI_DYNAMIC_VOLUME;ret = snprintf(mkvol_req.name, UBI_MAX_VOLUME_NAME + 1, "%s", name);mkvol_req.name_len = ret;ioctl(ubi_dev, UBI_IOCMKVOL, &mkvol_req);close(ubi_dev);}close(ubi_ctrl);return ubi_num;}int ubi_detach_dev(int dev){int ret, ubi_ctrl;ubi_ctrl = open(UBI_CTRL_DEV, O_RDONLY);if (ubi_ctrl == -1) {return -1;}ret = ioctl(ubi_ctrl, UBI_IOCDET, &dev);close(ubi_ctrl);return ret;}6.增加文件andriod/system/core/init/ubi-user.h/** Copyright (c) International Business Machines Corp., 2006** 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** Author: Artem Bityutskiy (Битюцкий Артём)*/#ifndef __UBI_USER_H__#define __UBI_USER_H__/** UBI device creation (the same as MTD device attachment)* ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~** MTD devices may be attached using %UBI_IOCATT ioctl command of the UBI* control device. The caller has to properly fill and pass* &struct ubi_attach_req object - UBI will attach the MTD device specified in* the request and return the newly created UBI device number as the ioctl* return value.** UBI device deletion (the same as MTD device detachment)* ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~** An UBI device maybe deleted with %UBI_IOCDET ioctl command of the UBI* control device.** UBI volume creation* ~~~~~~~~~~~~~~~~~~~** UBI volumes are created via the %UBI_IOCMKVOL IOCTL command of UBI character* device. A &struct ubi_mkvol_req object has to be properly filled and a* pointer to it has to be passed to the IOCTL.** UBI volume deletion* ~~~~~~~~~~~~~~~~~~~** To delete a volume, the %UBI_IOCRMVOL IOCTL command of the UBI character* device should be used. A pointer to the 32-bit volume ID hast to be passed* to the IOCTL.** UBI volume re-size* ~~~~~~~~~~~~~~~~~~** To re-size a volume, the %UBI_IOCRSVOL IOCTL command of the UBI character* device should be used. A &struct ubi_rsvol_req object has to be properly* filled and a pointer to it has to be passed to the IOCTL.** UBI volume update* ~~~~~~~~~~~~~~~~~** Volume update should be done via the %UBI_IOCVOLUP IOCTL command of the* corresponding UBI volume character device. A pointer to a 64-bit update* size should be passed to the IOCTL. After this, UBI expects user to write* this number of bytes to the volume character device. The update is finished* when the claimed number of bytes is passed. So, the volume update sequence* is something like:** fd = open("/dev/my_volume");* ioctl(fd, UBI_IOCVOLUP, &image_size);* write(fd, buf, image_size);* close(fd);** Atomic eraseblock change* ~~~~~~~~~~~~~~~~~~~~~~~~** Atomic eraseblock change operation is done via the %UBI_IOCEBCH IOCTL* command of the corresponding UBI volume character device. A pointer to* &struct ubi_leb_change_req has to be passed to the IOCTL. Then the user is* expected to write the requested amount of bytes. This is similar to the* "volume update" IOCTL.*//** When a new UBI volume or UBI device is created, users may either specify the* volume/device number they want to create or to let UBI automatically assign* the number using these constants.*/#define UBI_VOL_NUM_AUTO (-1)#define UBI_DEV_NUM_AUTO (-1)/* Maximum volume name length */#define UBI_MAX_VOLUME_NAME 127/* IOCTL commands of UBI character devices */#define UBI_IOC_MAGIC 'o'/* Create an UBI volume */#define UBI_IOCMKVOL _IOW(UBI_IOC_MAGIC, 0, struct ubi_mkvol_req)/* Remove an UBI volume */#define UBI_IOCRMVOL _IOW(UBI_IOC_MAGIC, 1, int32_t)/* Re-size an UBI volume */#define UBI_IOCRSVOL _IOW(UBI_IOC_MAGIC, 2, struct ubi_rsvol_req)/* IOCTL commands of the UBI control character device */#define UBI_CTRL_IOC_MAGIC 'o'/* Attach an MTD device */#define UBI_IOCATT _IOW(UBI_CTRL_IOC_MAGIC, 64, struct ubi_attach_req)/* Detach an MTD device */#define UBI_IOCDET _IOW(UBI_CTRL_IOC_MAGIC, 65, int32_t)/* IOCTL commands of UBI volume character devices */#define UBI_VOL_IOC_MAGIC 'O'/* Start UBI volume update */#define UBI_IOCVOLUP _IOW(UBI_VOL_IOC_MAGIC, 0, int64_t)/* An eraseblock erasure command, used for debugging, disabled by default */#define UBI_IOCEBER _IOW(UBI_VOL_IOC_MAGIC, 1, int32_t)/* An atomic eraseblock change command */#define UBI_IOCEBCH _IOW(UBI_VOL_IOC_MAGIC, 2, int32_t)/* Maximum MTD device name length supported by UBI */#define MAX_UBI_MTD_NAME_LEN 127/** UBI data type hint constants.** UBI_LONGTERM: long-term data* UBI_SHORTTERM: short-term data* UBI_UNKNOWN: data persistence is unknown** These constants are used when data is written to UBI volumes in order to* help the UBI wear-leveling unit to find more appropriate physical* eraseblocks.*/enum {UBI_LONGTERM = 1,UBI_SHORTTERM = 2,UBI_UNKNOWN = 3,};/** UBI volume type constants.** @UBI_DYNAMIC_VOLUME: dynamic volume* @UBI_STATIC_VOLUME: static volume*/enum {UBI_DYNAMIC_VOLUME = 3,UBI_STATIC_VOLUME = 4,};/*** struct ubi_attach_req - attach MTD device request.* @ubi_num: UBI device number to create* @mtd_num: MTD device number to attach* @vid_hdr_offset: VID header offset (use defaults if %0)* @padding: reserved for future, not used, has to be zeroed** This data structure is used to specify MTD device UBI has to attach and the* parameters it has to use. The number which should be assigned to the new UBI* device is passed in @ubi_num. UBI may automatically assign the number if* @UBI_DEV_NUM_AUTO is passed. In this case, the device number is returned in* @ubi_num.** Most applications should pass %0 in @vid_hdr_offset to make UBI use default* offset of the VID header within physical eraseblocks. The default offset is* the next min. I/O unit after the EC header. For example, it will be offset* 512 in case of a 512 bytes page NAND flash with no sub-page support. Or* it will be 512 in case of a 2KiB page NAND flash with 4 512-byte sub-pages.** But in rare cases, if this optimizes things, the VID header may be placed to* a different offset. For example, the boot-loader might do things faster if the* VID header sits at the end of the first 2KiB NAND page with 4 sub-pages. As* the boot-loader would not normally need to read EC headers (unless it needs* UBI in RW mode), it might be faster to calculate ECC. This is weird example,* but it real-life example. So, in this example, @vid_hdr_offer would be* 2KiB-64 bytes = 1984. Note, that this position is not even 512-bytes* aligned, which is OK, as UBI is clever enough to realize this is 4th sub-page* of the first page and add needed padding.*/struct ubi_attach_req {int32_t ubi_num;int32_t mtd_num;int32_t vid_hdr_offset;uint8_t padding[12];};/*** struct ubi_mkvol_req - volume description data structure used in* volume creation requests.* @vol_id: volume number* @alignment: volume alignment* @bytes: volume size in bytes* @vol_type: volume type (%UBI_DYNAMIC_VOLUME or %UBI_STATIC_VOLUME)* @padding1: reserved for future, not used, has to be zeroed* @name_len: volume name length* @padding2: reserved for future, not used, has to be zeroed* @name: volume name** This structure is used by user-space programs when creating new volumes. The* @used_bytes field is only necessary when creating static volumes.** The @alignment field specifies the required alignment of the volume logical* eraseblock. This means, that the size of logical eraseblocks will be aligned* to this number, i.e.,* (UBI device logical eraseblock size) mod (@alignment) = 0.** To put it differently, the logical eraseblock of this volume may be slightly* shortened in order to make it properly aligned. The alignment has to be* multiple of the flash minimal input/output unit, or %1 to utilize the entire* available space of logical eraseblocks.** The @alignment field may be useful, for example, when one wants to maintain* a block device on top of an UBI volume. In this case, it is desirable to fit* an integer number of blocks in logical eraseblocks of this UBI volume. With* alignment it is possible to update this volume using plane UBI volume image* BLOBs, without caring about how to properly align them.*/struct ubi_mkvol_req {int32_t vol_id;int32_t alignment;int64_t bytes;int8_t vol_type;int8_t padding1;int16_t name_len;int8_t padding2[4];char name[UBI_MAX_VOLUME_NAME + 1];} __attribute__ ((packed));/*** struct ubi_rsvol_req - a data structure used in volume re-size requests.* @vol_id: ID of the volume to re-size* @bytes: new size of the volume in bytes** Re-sizing is possible for both dynamic and static volumes. But while dynamic* volumes may be re-sized arbitrarily, static volumes cannot be made to be* smaller then the number of bytes they bear. To arbitrarily shrink a static* volume, it must be wiped out first (by means of volume update operation with* zero number of bytes).*/struct ubi_rsvol_req {int64_t bytes;int32_t vol_id;} __attribute__ ((packed));/*** struct ubi_leb_change_req - a data structure used in atomic logical* eraseblock change requests.* @lnum: logical eraseblock number to change* @bytes: how many bytes will be written to the logical eraseblock* @dtype: data type (%UBI_LONGTERM, %UBI_SHORTTERM, %UBI_UNKNOWN)* @padding: reserved for future, not used, has to be zeroed*/struct ubi_leb_change_req {int32_t lnum;int32_t bytes;uint8_t dtype;uint8_t padding[7];} __attribute__ ((packed));#endif /* __UBI_USER_H__ */3. 常见问题
(1) ubifs_check_node: bad CRC: calculated 0xca82b3d7, read 0x9be0e26
ubifs_check_node: bad node at LEB 51:45312
注意这个crc错误，说的错误并不是LED 0的问题，也就是说第一个逻辑块没问题，CRC是正确的，这个问题的一个解决办法是要精确mkfs.ubifs时候的-c参数的值，不能太大了
(2)
[ 5.433349] UBIFS error (pid 71): ubifs_read_node: bad node type (150 but expected 1)
[ 5.434204] UBIFS error (pid 71): ubifs_read_node: bad node at LEB 524:4072, LEB mapping status 1
[ 5.435241] Not a node, first 24 bytes:
[ 5.435729] 00000000: 34 fb 21 ee 84 18 69 2d 60 b0 33 e6 74 f8 1c 15 da ca a1 c9 96 e3 ac 51 4.!...i-`.3.t..........Q
这个问题是，ubifs给nand驱动的buffer不一定是按照硬件对齐要求的，所以要驱动来判断，dma等地址需要页对齐等特性
(3) crc错误，但是LED 0:0
可能镜像没烧对，或者mtd-utils的版本和内核版本相差较远
(4) 可以擦掉一个分区，并不需要非要烧写image到那个分区，也可以直接mount的，这样可以做实验验证ubi的性能，前提是kernel配置好了，mount正确了
(5) 如果是跟文件系统或者要手动mount，以下步骤供参考:
uboot里, mtd命令后，看到需要用ubifs的mtd分区的编号，比如
device nand0 <rda_nand>, # parts = 10
#: name size offset mask_flags
0: bootloader 0x00200000 0x00000000 0
1: boot 0x00800000 0x00800000 0
2: system 0x08000000 0x02000000 0
system分区mtd编号是2，命令如下:
nand erase.part system
命令行参数增加
ubi.mtd=2，如果有多个，可以增加，如ubi.mtd=2,ubi.mtd=1，这样传递后，内核启动后会做attach的操作，类似android的init中的attach那样，attach之后，/dev/下就建立好设备了，进入系统后用:
内核控制台里，mount -t ubifs /dev/ubi0_0 /mnt 即可mount
