UBI: Unsorted Block Images

UBI (Latin: "where?") manages multiple logical volumes on a single
flash device, specifically supporting NAND flash devices. UBI provides
a flexible partitioning concept which still allows for wear-levelling
across the whole flash device.

In a sense, UBI may be compared to the Logical Volume Manager
(LVM). Whereas LVM maps logical sector numbers to physical HDD sector
numbers, UBI maps logical eraseblocks to physical eraseblocks.

More information may be found at
http://www.linux-mtd.infradead.org/doc/ubi.html

Partitioning/Re-partitioning

  An UBI volume occupies a certain number of erase blocks. This is
  limited by a configured maximum volume size, which could also be
  viewed as the partition size. Each individual UBI volume's size can
  be changed independently of the other UBI volumes, provided that the
  sum of all volume sizes doesn't exceed a certain limit.

  UBI supports dynamic volumes and static volumes. Static volumes are
  read-only and their contents are protected by CRC check sums.

Bad eraseblocks handling

  UBI transparently handles bad eraseblocks. When a physical
  eraseblock becomes bad, it is substituted by a good physical
  eraseblock, and the user does not even notice this.

Scrubbing

  On a NAND flash bit flips can occur on any write operation,
  sometimes also on read. If bit flips persist on the device, at first
  they can still be corrected by ECC, but once they accumulate,
  correction will become impossible. Thus it is best to actively scrub
  the affected eraseblock, by first copying it to a free eraseblock
  and then erasing the original. The UBI layer performs this type of
  scrubbing under the covers, transparently to the UBI volume users.

Erase Counts

  UBI maintains an erase count header per eraseblock. This frees
  higher-level layers (like file systems) from doing this and allows
  for centralized erase count management instead. The erase counts are
  used by the wear-levelling algorithm in the UBI layer. The algorithm
  itself is exchangeable.

Booting from NAND

  For booting directly from NAND flash the hardware must at least be
  capable of fetching and executing a small portion of the NAND
  flash. Some NAND flash controllers have this kind of support. They
  usually limit the window to a few kilobytes in erase block 0. This
  "initial program loader" (IPL) must then contain sufficient logic to
  load and execute the next boot phase.

  Due to bad eraseblocks, which may be randomly scattered over the
  flash device, it is problematic to store the "secondary program
  loader" (SPL) statically. Also, due to bit-flips it may become
  corrupted over time. UBI allows to solve this problem gracefully by
  storing the SPL in a small static UBI volume.

UBI volumes vs. static partitions

  UBI volumes are still very similar to static MTD partitions:

    * both consist of eraseblocks (logical eraseblocks in case of UBI
      volumes, and physical eraseblocks in case of static partitions;
    * both support three basic operations - read, write, erase.

  But UBI volumes have the following advantages over traditional
  static MTD partitions:

    * there are no eraseblock wear-leveling constraints in case of UBI
      volumes, so the user should not care about this;
    * there are no bit-flips and bad eraseblocks in case of UBI volumes.

  So, UBI volumes may be considered as flash devices with relaxed
  restrictions.

Where can it be found?

  Documentation, kernel code and applications can be found in the MTD
  gits.

What are the applications for?

  The applications help to create binary flash images for two purposes: pfi
  files (partial flash images) for in-system update of UBI volumes, and plain
  binary images, with or without OOB data in case of NAND, for a manufacturing
  step. Furthermore some tools are/and will be created that allow flash content
  analysis after a system has crashed..

Who did UBI?

  The original ideas, where UBI is based on, were developed by Andreas
  Arnez, Frank Haverkamp and Thomas Gleixner. Josh W. Boyer and some others
  were involved too. The implementation of the kernel layer was done by Artem
  B. Bityutskiy. The user-space applications and tools were written by Oliver
  Lohmann with contributions from Frank Haverkamp, Andreas Arnez, and Artem.
  Joern Engel contributed a patch which modifies JFFS2 so that it can be run on
  a UBI volume. Thomas Gleixner did modifications to the NAND layer. Alexander
  Schmidt made some testing work as well as core functionality improvements.

Signed-off-by: Artem B. Bityutskiy <dedekind@linutronix.de>
Signed-off-by: Frank Haverkamp <haver@vnet.ibm.com>

drivers/mtd/Kconfig

@@ -292,5 +292,7 @@ source "drivers/mtd/nand/Kconfig"
 
 source "drivers/mtd/onenand/Kconfig"
 
+source "drivers/mtd/ubi/Kconfig"
+
 endmenu
 

drivers/mtd/Makefile

@@ -28,3 +28,5 @@ nftl-objs		:= nftlcore.o nftlmount.o
 inftl-objs		:= inftlcore.o inftlmount.o
 
 obj-y		+= chips/ maps/ devices/ nand/ onenand/
+
+obj-$(CONFIG_MTD_UBI)		+= ubi/

drivers/mtd/ubi/Kconfig

@@ -0,0 +1,58 @@
+# drivers/mtd/ubi/Kconfig
+
+menu "UBI - Unsorted block images"
+	depends on MTD
+
+config MTD_UBI
+	tristate "Enable UBI"
+	depends on MTD
+	select CRC32
+	help
+	  UBI is a software layer above MTD layer which admits of LVM-like
+	  logical volumes on top of MTD devices, hides some complexities of
+	  flash chips like wear and bad blocks and provides some other useful
+	  capabilities. Please, consult the MTD web site for more details
+	  (www.linux-mtd.infradead.org).
+
+config MTD_UBI_WL_THRESHOLD
+	int "UBI wear-leveling threshold"
+	default 4096
+	range 2 65536
+	depends on MTD_UBI
+	help
+	  This parameter defines the maximum difference between the highest
+	  erase counter value and the lowest erase counter value of eraseblocks
+	  of UBI devices. When this threshold is exceeded, UBI starts performing
+	  wear leveling by means of moving data from eraseblock with low erase
+	  counter to eraseblocks with high erase counter. Leave the default
+	  value if unsure.
+
+config MTD_UBI_BEB_RESERVE
+	int "Percentage of reserved eraseblocks for bad eraseblocks handling"
+	default 1
+	range 0 25
+	depends on MTD_UBI
+	help
+	  If the MTD device admits of bad eraseblocks (e.g. NAND flash), UBI
+	  reserves some amount of physical eraseblocks to handle new bad
+	  eraseblocks. For example, if a flash physical eraseblock becomes bad,
+	  UBI uses these reserved physical eraseblocks to relocate the bad one.
+	  This option specifies how many physical eraseblocks will be reserved
+	  for bad eraseblock handling (percents of total number of good flash
+	  eraseblocks). If the underlying flash does not admit of bad
+	  eraseblocks (e.g. NOR flash), this value is ignored and nothing is
+	  reserved. Leave the default value if unsure.
+
+config MTD_UBI_GLUEBI
+	bool "Emulate MTD devices"
+	default n
+	depends on MTD_UBI
+	help
+	   This option enables MTD devices emulation on top of UBI volumes: for
+	   each UBI volumes an MTD device is created, and all I/O to this MTD
+	   device is redirected to the UBI volume. This is handy to make
+	   MTD-oriented software (like JFFS2) work on top of UBI. Do not enable
+	   this if no legacy software will be used.
+
+source "drivers/mtd/ubi/Kconfig.debug"
+endmenu

drivers/mtd/ubi/Kconfig.debug

@@ -0,0 +1,104 @@
+comment "UBI debugging options"
+	depends on MTD_UBI
+
+config MTD_UBI_DEBUG
+	bool "UBI debugging"
+	depends on SYSFS
+	depends on MTD_UBI
+	select DEBUG_FS
+	select KALLSYMS_ALL
+	help
+	  This option enables UBI debugging.
+
+config MTD_UBI_DEBUG_MSG
+	bool "UBI debugging messages"
+	depends on MTD_UBI_DEBUG
+	default n
+	help
+	  This option enables UBI debugging messages.
+
+config MTD_UBI_DEBUG_PARANOID
+	bool "Extra self-checks"
+	default n
+	depends on MTD_UBI_DEBUG
+	help
+	  This option enables extra checks in UBI code. Note this slows UBI down
+	  significantly.
+
+config MTD_UBI_DEBUG_DISABLE_BGT
+	bool "Do not enable the UBI background thread"
+	depends on MTD_UBI_DEBUG
+	default n
+	help
+	  This option switches the background thread off by default. The thread
+	  may be also be enabled/disabled via UBI sysfs.
+
+config MTD_UBI_DEBUG_USERSPACE_IO
+	bool "Direct user-space write/erase support"
+	default n
+	depends on MTD_UBI_DEBUG
+	help
+	  By default, users cannot directly write and erase individual
+	  eraseblocks of dynamic volumes, and have to use update operation
+	  instead. This option enables this capability - it is very useful for
+	  debugging and testing.
+
+config MTD_UBI_DEBUG_EMULATE_BITFLIPS
+	bool "Emulate flash bit-flips"
+	depends on MTD_UBI_DEBUG
+	default n
+	help
+	  This option emulates bit-flips with probability 1/50, which in turn
+	  causes scrubbing. Useful for debugging and stressing UBI.
+
+config MTD_UBI_DEBUG_EMULATE_WRITE_FAILURES
+	bool "Emulate flash write failures"
+	depends on MTD_UBI_DEBUG
+	default n
+	help
+	  This option emulates write failures with probability 1/100. Useful for
+	  debugging and testing how UBI handlines errors.
+
+config MTD_UBI_DEBUG_EMULATE_ERASE_FAILURES
+	bool "Emulate flash erase failures"
+	depends on MTD_UBI_DEBUG
+	default n
+	help
+	  This option emulates erase failures with probability 1/100. Useful for
+	  debugging and testing how UBI handlines errors.
+
+menu "Additional UBI debugging messages"
+	depends on MTD_UBI_DEBUG
+
+config MTD_UBI_DEBUG_MSG_BLD
+	bool "Additional UBI initialization and build messages"
+	default n
+	depends on MTD_UBI_DEBUG
+	help
+	  This option enables detailed UBI initialization and device build
+	  debugging messages.
+
+config MTD_UBI_DEBUG_MSG_EBA
+	bool "Eraseblock association unit messages"
+	default n
+	depends on MTD_UBI_DEBUG
+	help
+	  This option enables debugging messages from the UBI eraseblock
+	  association unit.
+
+config MTD_UBI_DEBUG_MSG_WL
+	bool "Wear-leveling unit messages"
+	default n
+	depends on MTD_UBI_DEBUG
+	help
+	  This option enables debugging messages from the UBI wear-leveling
+	  unit.
+
+config MTD_UBI_DEBUG_MSG_IO
+	bool "Input/output unit messages"
+	default n
+	depends on MTD_UBI_DEBUG
+	help
+	  This option enables debugging messages from the UBI input/output unit.
+
+endmenu # UBI debugging messages

drivers/mtd/ubi/Makefile

@@ -0,0 +1,7 @@
+obj-$(CONFIG_MTD_UBI) += ubi.o
+
+ubi-y += vtbl.o vmt.o upd.o build.o cdev.o kapi.o eba.o io.o wl.o scan.o
+ubi-y += misc.o
+
+ubi-$(CONFIG_MTD_UBI_DEBUG) += debug.o
+ubi-$(CONFIG_MTD_UBI_GLUEBI) += gluebi.o

drivers/mtd/ubi/debug.c

@@ -0,0 +1,224 @@
+/*
+ * 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)
+ */
+
+/*
+ * Here we keep all the UBI debugging stuff which should normally be disabled
+ * and compiled-out, but it is extremely helpful when hunting bugs or doing big
+ * changes.
+ */
+
+#ifdef CONFIG_MTD_UBI_DEBUG_MSG
+
+#include "ubi.h"
+
+/**
+ * ubi_dbg_dump_ec_hdr - dump an erase counter header.
+ * @ec_hdr: the erase counter header to dump
+ */
+void ubi_dbg_dump_ec_hdr(const struct ubi_ec_hdr *ec_hdr)
+{
+	dbg_msg("erase counter header dump:");
+	dbg_msg("magic          %#08x", ubi32_to_cpu(ec_hdr->magic));
+	dbg_msg("version        %d",    (int)ec_hdr->version);
+	dbg_msg("ec             %llu",  (long long)ubi64_to_cpu(ec_hdr->ec));
+	dbg_msg("vid_hdr_offset %d",    ubi32_to_cpu(ec_hdr->vid_hdr_offset));
+	dbg_msg("data_offset    %d",    ubi32_to_cpu(ec_hdr->data_offset));
+	dbg_msg("hdr_crc        %#08x", ubi32_to_cpu(ec_hdr->hdr_crc));
+	dbg_msg("erase counter header hexdump:");
+	ubi_dbg_hexdump(ec_hdr, UBI_EC_HDR_SIZE);
+}
+
+/**
+ * ubi_dbg_dump_vid_hdr - dump a volume identifier header.
+ * @vid_hdr: the volume identifier header to dump
+ */
+void ubi_dbg_dump_vid_hdr(const struct ubi_vid_hdr *vid_hdr)
+{
+	dbg_msg("volume identifier header dump:");
+	dbg_msg("magic     %08x", ubi32_to_cpu(vid_hdr->magic));
+	dbg_msg("version   %d",   (int)vid_hdr->version);
+	dbg_msg("vol_type  %d",   (int)vid_hdr->vol_type);
+	dbg_msg("copy_flag %d",   (int)vid_hdr->copy_flag);
+	dbg_msg("compat    %d",   (int)vid_hdr->compat);
+	dbg_msg("vol_id    %d",   ubi32_to_cpu(vid_hdr->vol_id));
+	dbg_msg("lnum      %d",   ubi32_to_cpu(vid_hdr->lnum));
+	dbg_msg("leb_ver   %u",   ubi32_to_cpu(vid_hdr->leb_ver));
+	dbg_msg("data_size %d",   ubi32_to_cpu(vid_hdr->data_size));
+	dbg_msg("used_ebs  %d",   ubi32_to_cpu(vid_hdr->used_ebs));
+	dbg_msg("data_pad  %d",   ubi32_to_cpu(vid_hdr->data_pad));
+	dbg_msg("sqnum     %llu",
+		(unsigned long long)ubi64_to_cpu(vid_hdr->sqnum));
+	dbg_msg("hdr_crc   %08x", ubi32_to_cpu(vid_hdr->hdr_crc));
+	dbg_msg("volume identifier header hexdump:");
+}
+
+/**
+ * ubi_dbg_dump_vol_info- dump volume information.
+ * @vol: UBI volume description object
+ */
+void ubi_dbg_dump_vol_info(const struct ubi_volume *vol)
+{
+	dbg_msg("volume information dump:");
+	dbg_msg("vol_id          %d", vol->vol_id);
+	dbg_msg("reserved_pebs   %d", vol->reserved_pebs);
+	dbg_msg("alignment       %d", vol->alignment);
+	dbg_msg("data_pad        %d", vol->data_pad);
+	dbg_msg("vol_type        %d", vol->vol_type);
+	dbg_msg("name_len        %d", vol->name_len);
+	dbg_msg("usable_leb_size %d", vol->usable_leb_size);
+	dbg_msg("used_ebs        %d", vol->used_ebs);
+	dbg_msg("used_bytes      %lld", vol->used_bytes);
+	dbg_msg("last_eb_bytes   %d", vol->last_eb_bytes);
+	dbg_msg("corrupted       %d", vol->corrupted);
+	dbg_msg("upd_marker      %d", vol->upd_marker);
+
+	if (vol->name_len <= UBI_VOL_NAME_MAX &&
+	    strnlen(vol->name, vol->name_len + 1) == vol->name_len) {
+		dbg_msg("name          %s", vol->name);
+	} else {
+		dbg_msg("the 1st 5 characters of the name: %c%c%c%c%c",
+			vol->name[0], vol->name[1], vol->name[2],
+			vol->name[3], vol->name[4]);
+	}
+}
+
+/**
+ * ubi_dbg_dump_vtbl_record - dump a &struct ubi_vtbl_record object.
+ * @r: the object to dump
+ * @idx: volume table index
+ */
+void ubi_dbg_dump_vtbl_record(const struct ubi_vtbl_record *r, int idx)
+{
+	int name_len = ubi16_to_cpu(r->name_len);
+
+	dbg_msg("volume table record %d dump:", idx);
+	dbg_msg("reserved_pebs   %d", ubi32_to_cpu(r->reserved_pebs));
+	dbg_msg("alignment       %d", ubi32_to_cpu(r->alignment));
+	dbg_msg("data_pad        %d", ubi32_to_cpu(r->data_pad));
+	dbg_msg("vol_type        %d", (int)r->vol_type);
+	dbg_msg("upd_marker      %d", (int)r->upd_marker);
+	dbg_msg("name_len        %d", name_len);
+
+	if (r->name[0] == '\0') {
+		dbg_msg("name          NULL");
+		return;
+	}
+
+	if (name_len <= UBI_VOL_NAME_MAX &&
+	    strnlen(&r->name[0], name_len + 1) == name_len) {
+		dbg_msg("name          %s", &r->name[0]);
+	} else {
+		dbg_msg("1st 5 characters of the name: %c%c%c%c%c",
+			r->name[0], r->name[1], r->name[2], r->name[3],
+			r->name[4]);
+	}
+	dbg_msg("crc             %#08x", ubi32_to_cpu(r->crc));
+}
+
+/**
+ * ubi_dbg_dump_sv - dump a &struct ubi_scan_volume object.
+ * @sv: the object to dump
+ */
+void ubi_dbg_dump_sv(const struct ubi_scan_volume *sv)
+{
+	dbg_msg("volume scanning information dump:");
+	dbg_msg("vol_id         %d", sv->vol_id);
+	dbg_msg("highest_lnum   %d", sv->highest_lnum);
+	dbg_msg("leb_count      %d", sv->leb_count);
+	dbg_msg("compat         %d", sv->compat);
+	dbg_msg("vol_type       %d", sv->vol_type);
+	dbg_msg("used_ebs       %d", sv->used_ebs);
+	dbg_msg("last_data_size %d", sv->last_data_size);
+	dbg_msg("data_pad       %d", sv->data_pad);
+}
+
+/**
+ * ubi_dbg_dump_seb - dump a &struct ubi_scan_leb object.
+ * @seb: the object to dump
+ * @type: object type: 0 - not corrupted, 1 - corrupted
+ */
+void ubi_dbg_dump_seb(const struct ubi_scan_leb *seb, int type)
+{
+	dbg_msg("eraseblock scanning information dump:");
+	dbg_msg("ec       %d", seb->ec);
+	dbg_msg("pnum     %d", seb->pnum);
+	if (type == 0) {
+		dbg_msg("lnum     %d", seb->lnum);
+		dbg_msg("scrub    %d", seb->scrub);
+		dbg_msg("sqnum    %llu", seb->sqnum);
+		dbg_msg("leb_ver  %u", seb->leb_ver);
+	}
+}
+
+/**
+ * ubi_dbg_dump_mkvol_req - dump a &struct ubi_mkvol_req object.
+ * @req: the object to dump
+ */
+void ubi_dbg_dump_mkvol_req(const struct ubi_mkvol_req *req)
+{
+	char nm[17];
+
+	dbg_msg("volume creation request dump:");
+	dbg_msg("vol_id    %d",   req->vol_id);
+	dbg_msg("alignment %d",   req->alignment);
+	dbg_msg("bytes     %lld", (long long)req->bytes);
+	dbg_msg("vol_type  %d",   req->vol_type);
+	dbg_msg("name_len  %d",   req->name_len);
+
+	memcpy(nm, req->name, 16);
+	nm[16] = 0;
+	dbg_msg("the 1st 16 characters of the name: %s", nm);
+}
+
+#define BYTES_PER_LINE 32
+
+/**
+ * ubi_dbg_hexdump - dump a buffer.
+ * @ptr: the buffer to dump
+ * @size: buffer size which must be multiple of 4 bytes
+ */
+void ubi_dbg_hexdump(const void *ptr, int size)
+{
+	int i, k = 0, rows, columns;
+	const uint8_t *p = ptr;
+
+	size = ALIGN(size, 4);
+	rows = size/BYTES_PER_LINE + size % BYTES_PER_LINE;
+	for (i = 0; i < rows; i++) {
+		int j;
+
+		cond_resched();
+		columns = min(size - k, BYTES_PER_LINE) / 4;
+		if (columns == 0)
+			break;
+		printk(KERN_DEBUG "%5d:  ", i * BYTES_PER_LINE);
+		for (j = 0; j < columns; j++) {
+			int n, N;
+
+			N = size - k > 4 ? 4 : size - k;
+			for (n = 0; n < N; n++)
+				printk("%02x", p[k++]);
+			printk(" ");
+		}
+		printk("\n");
+	}
+}
+
+#endif /* CONFIG_MTD_UBI_DEBUG_MSG */

drivers/mtd/ubi/debug.h

@@ -0,0 +1,161 @@
+/*
+ * 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_DEBUG_H__
+#define __UBI_DEBUG_H__
+
+#ifdef CONFIG_MTD_UBI_DEBUG
+#include <linux/random.h>
+
+#define ubi_assert(expr)  BUG_ON(!(expr))
+#define dbg_err(fmt, ...) ubi_err(fmt, ##__VA_ARGS__)
+#else
+#define ubi_assert(expr)  ({})
+#define dbg_err(fmt, ...) ({})
+#endif
+
+#ifdef CONFIG_MTD_UBI_DEBUG_DISABLE_BGT
+#define DBG_DISABLE_BGT 1
+#else
+#define DBG_DISABLE_BGT 0
+#endif
+
+#ifdef CONFIG_MTD_UBI_DEBUG_MSG
+/* Generic debugging message */
+#define dbg_msg(fmt, ...) \
+	printk(KERN_DEBUG "UBI DBG: %s: " fmt "\n", __FUNCTION__, ##__VA_ARGS__)
+
+#define ubi_dbg_dump_stack() dump_stack()
+
+struct ubi_ec_hdr;
+struct ubi_vid_hdr;
+struct ubi_volume;
+struct ubi_vtbl_record;
+struct ubi_scan_volume;
+struct ubi_scan_leb;
+struct ubi_mkvol_req;
+
+void ubi_dbg_print(int type, const char *func, const char *fmt, ...);
+void ubi_dbg_dump_ec_hdr(const struct ubi_ec_hdr *ec_hdr);
+void ubi_dbg_dump_vid_hdr(const struct ubi_vid_hdr *vid_hdr);
+void ubi_dbg_dump_vol_info(const struct ubi_volume *vol);
+void ubi_dbg_dump_vtbl_record(const struct ubi_vtbl_record *r, int idx);
+void ubi_dbg_dump_sv(const struct ubi_scan_volume *sv);
+void ubi_dbg_dump_seb(const struct ubi_scan_leb *seb, int type);
+void ubi_dbg_dump_mkvol_req(const struct ubi_mkvol_req *req);
+void ubi_dbg_hexdump(const void *buf, int size);
+
+#else
+
+#define dbg_msg(fmt, ...)    ({})
+#define ubi_dbg_dump_stack() ({})
+#define ubi_dbg_print(func, fmt, ...)    ({})
+#define ubi_dbg_dump_ec_hdr(ec_hdr)      ({})
+#define ubi_dbg_dump_vid_hdr(vid_hdr)    ({})
+#define ubi_dbg_dump_vol_info(vol)       ({})
+#define ubi_dbg_dump_vtbl_record(r, idx) ({})
+#define ubi_dbg_dump_sv(sv)              ({})
+#define ubi_dbg_dump_seb(seb, type)      ({})
+#define ubi_dbg_dump_mkvol_req(req)      ({})
+#define ubi_dbg_hexdump(buf, size)       ({})
+
+#endif /* CONFIG_MTD_UBI_DEBUG_MSG */
+
+#ifdef CONFIG_MTD_UBI_DEBUG_MSG_EBA
+/* Messages from the eraseblock association unit */
+#define dbg_eba(fmt, ...) \
+	printk(KERN_DEBUG "UBI DBG eba: %s: " fmt "\n", __FUNCTION__, \
+	       ##__VA_ARGS__)
+#else
+#define dbg_eba(fmt, ...) ({})
+#endif
+
+#ifdef CONFIG_MTD_UBI_DEBUG_MSG_WL
+/* Messages from the wear-leveling unit */
+#define dbg_wl(fmt, ...) \
+	printk(KERN_DEBUG "UBI DBG wl: %s: " fmt "\n", __FUNCTION__, \
+	       ##__VA_ARGS__)
+#else
+#define dbg_wl(fmt, ...) ({})
+#endif
+
+#ifdef CONFIG_MTD_UBI_DEBUG_MSG_IO
+/* Messages from the input/output unit */
+#define dbg_io(fmt, ...) \
+	printk(KERN_DEBUG "UBI DBG io: %s: " fmt "\n", __FUNCTION__, \
+	       ##__VA_ARGS__)
+#else
+#define dbg_io(fmt, ...) ({})
+#endif
+
+#ifdef CONFIG_MTD_UBI_DEBUG_MSG_BLD
+/* Initialization and build messages */
+#define dbg_bld(fmt, ...) \
+	printk(KERN_DEBUG "UBI DBG bld: %s: " fmt "\n", __FUNCTION__, \
+	       ##__VA_ARGS__)
+#else
+#define dbg_bld(fmt, ...) ({})
+#endif
+
+#ifdef CONFIG_MTD_UBI_DEBUG_EMULATE_BITFLIPS
+/**
+ * ubi_dbg_is_bitflip - if it is time to emulate a bit-flip.
+ *
+ * Returns non-zero if a bit-flip should be emulated, otherwise returns zero.
+ */
+static inline int ubi_dbg_is_bitflip(void)
+{
+	return !(random32() % 200);
+}
+#else
+#define ubi_dbg_is_bitflip() 0
+#endif
+
+#ifdef CONFIG_MTD_UBI_DEBUG_EMULATE_WRITE_FAILURES
+/**
+ * ubi_dbg_is_write_failure - if it is time to emulate a write failure.
+ *
+ * Returns non-zero if a write failure should be emulated, otherwise returns
+ * zero.
+ */
+static inline int ubi_dbg_is_write_failure(void)
+{
+	return !(random32() % 500);
+}
+#else
+#define ubi_dbg_is_write_failure() 0
+#endif
+
+#ifdef CONFIG_MTD_UBI_DEBUG_EMULATE_ERASE_FAILURES
+/**
+ * ubi_dbg_is_erase_failure - if its time to emulate an erase failure.
+ *
+ * Returns non-zero if an erase failure should be emulated, otherwise returns
+ * zero.
+ */
+static inline int ubi_dbg_is_erase_failure(void)
+{
+		return !(random32() % 400);
+}
+#else
+#define ubi_dbg_is_erase_failure() 0
+#endif
+
+#endif /* !__UBI_DEBUG_H__ */

drivers/mtd/ubi/gluebi.c

@@ -0,0 +1,324 @@
+/*
+ * 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), Joern Engel
+ */
+
+/*
+ * This file includes implementation of fake MTD devices for each UBI volume.
+ * This sounds strange, but it is in fact quite useful to make MTD-oriented
+ * software (including all the legacy software) to work on top of UBI.
+ *
+ * Gluebi emulates MTD devices of "MTD_UBIVOLUME" type. Their minimal I/O unit
+ * size (mtd->writesize) is equivalent to the UBI minimal I/O unit. The
+ * eraseblock size is equivalent to the logical eraseblock size of the volume.
+ */
+
+#include <asm/div64.h>
+#include "ubi.h"
+
+/**
+ * gluebi_get_device - get MTD device reference.
+ * @mtd: the MTD device description object
+ *
+ * This function is called every time the MTD device is being opened and
+ * implements the MTD get_device() operation. Returns zero in case of success
+ * and a negative error code in case of failure.
+ */
+static int gluebi_get_device(struct mtd_info *mtd)
+{
+	struct ubi_volume *vol;
+
+	vol = container_of(mtd, struct ubi_volume, gluebi_mtd);
+
+	/*
+	 * We do not introduce locks for gluebi reference count because the
+	 * get_device()/put_device() calls are already serialized at MTD.
+	 */
+	if (vol->gluebi_refcount > 0) {
+		/*
+		 * The MTD device is already referenced and this is just one
+		 * more reference. MTD allows many users to open the same
+		 * volume simultaneously and do not distinguish between
+		 * readers/writers/exclusive openers as UBI does. So we do not
+		 * open the UBI volume again - just increase the reference
+		 * counter and return.
+		 */
+		vol->gluebi_refcount += 1;
+		return 0;
+	}
+
+	/*
+	 * This is the first reference to this UBI volume via the MTD device
+	 * interface. Open the corresponding volume in read-write mode.
+	 */
+	vol->gluebi_desc = ubi_open_volume(vol->ubi->ubi_num, vol->vol_id,
+					   UBI_READWRITE);
+	if (IS_ERR(vol->gluebi_desc))
+		return PTR_ERR(vol->gluebi_desc);
+	vol->gluebi_refcount += 1;
+	return 0;
+}
+
+/**
+ * gluebi_put_device - put MTD device reference.
+ * @mtd: the MTD device description object
+ *
+ * This function is called every time the MTD device is being put. Returns
+ * zero in case of success and a negative error code in case of failure.
+ */
+static void gluebi_put_device(struct mtd_info *mtd)
+{
+	struct ubi_volume *vol;
+
+	vol = container_of(mtd, struct ubi_volume, gluebi_mtd);
+	vol->gluebi_refcount -= 1;
+	ubi_assert(vol->gluebi_refcount >= 0);
+	if (vol->gluebi_refcount == 0)
+		ubi_close_volume(vol->gluebi_desc);
+}
+
+/**
+ * gluebi_read - read operation of emulated MTD devices.
+ * @mtd: MTD device description object
+ * @from: absolute offset from where to read
+ * @len: how many bytes to read
+ * @retlen: count of read bytes is returned here
+ * @buf: buffer to store the read data
+ *
+ * This function returns zero in case of success and a negative error code in
+ * case of failure.
+ */
+static int gluebi_read(struct mtd_info *mtd, loff_t from, size_t len,
+		       size_t *retlen, unsigned char *buf)
+{
+	int err = 0, lnum, offs, total_read;
+	struct ubi_volume *vol;
+	struct ubi_device *ubi;
+	uint64_t tmp = from;
+
+	dbg_msg("read %zd bytes from offset %lld", len, from);
+
+	if (len < 0 || from < 0 || from + len > mtd->size)
+		return -EINVAL;
+
+	vol = container_of(mtd, struct ubi_volume, gluebi_mtd);
+	ubi = vol->ubi;
+
+	offs = do_div(tmp, mtd->erasesize);
+	lnum = tmp;
+
+	total_read = len;
+	while (total_read) {
+		size_t to_read = mtd->erasesize - offs;
+
+		if (to_read > total_read)
+			to_read = total_read;
+
+		err = ubi_eba_read_leb(ubi, vol->vol_id, lnum, buf, offs,
+				       to_read, 0);
+		if (err)
+			break;
+
+		lnum += 1;
+		offs = 0;
+		total_read -= to_read;
+		buf += to_read;
+	}
+
+	*retlen = len - total_read;
+	return err;
+}
+
+/**
+ * gluebi_write - write operation of emulated MTD devices.
+ * @mtd: MTD device description object
+ * @to: absolute offset where to write
+ * @len: how many bytes to write
+ * @retlen: count of written bytes is returned here
+ * @buf: buffer with data to write
+ *
+ * This function returns zero in case of success and a negative error code in
+ * case of failure.
+ */
+static int gluebi_write(struct mtd_info *mtd, loff_t to, size_t len,
+		       size_t *retlen, const u_char *buf)
+{
+	int err = 0, lnum, offs, total_written;
+	struct ubi_volume *vol;
+	struct ubi_device *ubi;
+	uint64_t tmp = to;
+
+	dbg_msg("write %zd bytes to offset %lld", len, to);
+
+	if (len < 0 || to < 0 || len + to > mtd->size)
+		return -EINVAL;
+
+	vol = container_of(mtd, struct ubi_volume, gluebi_mtd);
+	ubi = vol->ubi;
+
+	if (ubi->ro_mode)
+		return -EROFS;
+
+	offs = do_div(tmp, mtd->erasesize);
+	lnum = tmp;
+
+	if (len % mtd->writesize || offs % mtd->writesize)
+		return -EINVAL;
+
+	total_written = len;
+	while (total_written) {
+		size_t to_write = mtd->erasesize - offs;
+
+		if (to_write > total_written)
+			to_write = total_written;
+
+		err = ubi_eba_write_leb(ubi, vol->vol_id, lnum, buf, offs,
+					to_write, UBI_UNKNOWN);
+		if (err)
+			break;
+
+		lnum += 1;
+		offs = 0;
+		total_written -= to_write;
+		buf += to_write;
+	}
+
+	*retlen = len - total_written;
+	return err;
+}
+
+/**
+ * gluebi_erase - erase operation of emulated MTD devices.
+ * @mtd: the MTD device description object
+ * @instr: the erase operation description
+ *
+ * This function calls the erase callback when finishes. Returns zero in case
+ * of success and a negative error code in case of failure.
+ */
+static int gluebi_erase(struct mtd_info *mtd, struct erase_info *instr)
+{
+	int err, i, lnum, count;
+	struct ubi_volume *vol;
+	struct ubi_device *ubi;
+
+	dbg_msg("erase %u bytes at offset %u", instr->len, instr->addr);
+
+	if (instr->addr < 0 || instr->addr > mtd->size - mtd->erasesize)
+		return -EINVAL;
+
+	if (instr->len < 0 || instr->addr + instr->len > mtd->size)
+		return -EINVAL;
+
+	if (instr->addr % mtd->writesize || instr->len % mtd->writesize)
+		return -EINVAL;
+
+	lnum = instr->addr / mtd->erasesize;
+	count = instr->len / mtd->erasesize;
+
+	vol = container_of(mtd, struct ubi_volume, gluebi_mtd);
+	ubi = vol->ubi;
+
+	if (ubi->ro_mode)
+		return -EROFS;
+
+	for (i = 0; i < count; i++) {
+		err = ubi_eba_unmap_leb(ubi, vol->vol_id, lnum + i);
+		if (err)
+			goto out_err;
+	}
+
+	/*
+	 * MTD erase operations are synchronous, so we have to make sure the
+	 * physical eraseblock is wiped out.
+	 */
+	err = ubi_wl_flush(ubi);
+	if (err)
+		goto out_err;
+
+        instr->state = MTD_ERASE_DONE;
+        mtd_erase_callback(instr);
+	return 0;
+
+out_err:
+	instr->state = MTD_ERASE_FAILED;
+	instr->fail_addr = lnum * mtd->erasesize;
+	return err;
+}
+
+/**
+ * ubi_create_gluebi - initialize gluebi for an UBI volume.
+ * @ubi: UBI device description object
+ * @vol: volume description object
+ *
+ * This function is called when an UBI volume is created in order to create
+ * corresponding fake MTD device. Returns zero in case of success and a
+ * negative error code in case of failure.
+ */
+int ubi_create_gluebi(struct ubi_device *ubi, struct ubi_volume *vol)
+{
+	int err;
+	struct mtd_info *mtd = &vol->gluebi_mtd;
+
+	mtd->name = kmemdup(vol->name, vol->name_len + 1, GFP_KERNEL);
+	if (!mtd->name)
+		return -ENOMEM;
+
+	mtd->type = MTD_UBIVOLUME;
+	if (!ubi->ro_mode)
+		mtd->flags = MTD_WRITEABLE;
+	mtd->writesize  = ubi->min_io_size;
+	mtd->owner      = THIS_MODULE;
+	mtd->size       = vol->usable_leb_size * vol->reserved_pebs;
+	mtd->erasesize  = vol->usable_leb_size;
+	mtd->read       = gluebi_read;
+	mtd->write      = gluebi_write;
+	mtd->erase      = gluebi_erase;
+	mtd->get_device = gluebi_get_device;
+	mtd->put_device = gluebi_put_device;
+
+	if (add_mtd_device(mtd)) {
+		ubi_err("cannot not add MTD device\n");
+		kfree(mtd->name);
+		return -ENFILE;
+	}
+
+	dbg_msg("added mtd%d (\"%s\"), size %u, EB size %u",
+		mtd->index, mtd->name, mtd->size, mtd->erasesize);
+	return 0;
+}
+
+/**
+ * ubi_destroy_gluebi - close gluebi for an UBI volume.
+ * @vol: volume description object
+ *
+ * This function is called when an UBI volume is removed in order to remove
+ * corresponding fake MTD device. Returns zero in case of success and a
+ * negative error code in case of failure.
+ */
+int ubi_destroy_gluebi(struct ubi_volume *vol)
+{
+	int err;
+	struct mtd_info *mtd = &vol->gluebi_mtd;
+
+	dbg_msg("remove mtd%d", mtd->index);
+	err = del_mtd_device(mtd);
+	if (err)
+		return err;
+	kfree(mtd->name);
+	return 0;
+}

drivers/mtd/ubi/misc.c

@@ -0,0 +1,105 @@
+/*
+ * 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)
+ */
+
+/* Here we keep miscellaneous functions which are used all over the UBI code */
+
+#include "ubi.h"
+
+/**
+ * calc_data_len - calculate how much real data is stored in a buffer.
+ * @ubi: UBI device description object
+ * @buf: a buffer with the contents of the physical eraseblock
+ * @length: the buffer length
+ *
+ * This function calculates how much "real data" is stored in @buf and returnes
+ * the length. Continuous 0xFF bytes at the end of the buffer are not
+ * considered as "real data".
+ */
+int ubi_calc_data_len(const struct ubi_device *ubi, const void *buf,
+		      int length)
+{
+	int i;
+
+	ubi_assert(length % ubi->min_io_size == 0);
+
+	for (i = length - 1; i >= 0; i--)
+		if (((const uint8_t *)buf)[i] != 0xFF)
+			break;
+
+	/* The resulting length must be aligned to the minimum flash I/O size */
+	length = ALIGN(i + 1, ubi->min_io_size);
+	return length;
+}
+
+/**
+ * ubi_check_volume - check the contents of a static volume.
+ * @ubi: UBI device description object
+ * @vol_id: ID of the volume to check
+ *
+ * This function checks if static volume @vol_id is corrupted by fully reading
+ * it and checking data CRC. This function returns %0 if the volume is not
+ * corrupted, %1 if it is corrupted and a negative error code in case of
+ * failure. Dynamic volumes are not checked and zero is returned immediately.
+ */
+int ubi_check_volume(struct ubi_device *ubi, int vol_id)
+{
+	void *buf;
+	int err = 0, i;
+	struct ubi_volume *vol = ubi->volumes[vol_id];
+
+	if (vol->vol_type != UBI_STATIC_VOLUME)
+		return 0;
+
+	buf = kmalloc(vol->usable_leb_size, GFP_KERNEL);
+	if (!buf)
+		return -ENOMEM;
+
+	for (i = 0; i < vol->used_ebs; i++) {
+		int size;
+
+		if (i == vol->used_ebs - 1)
+			size = vol->last_eb_bytes;
+		else
+			size = vol->usable_leb_size;
+
+		err = ubi_eba_read_leb(ubi, vol_id, i, buf, 0, size, 1);
+		if (err) {
+			if (err == -EBADMSG)
+				err = 1;
+			break;
+		}
+	}
+
+	kfree(buf);
+	return err;
+}
+
+/**
+ * ubi_calculate_rsvd_pool - calculate how many PEBs must be reserved for bad
+ * eraseblock handling.
+ * @ubi: UBI device description object
+ */
+void ubi_calculate_reserved(struct ubi_device *ubi)
+{
+	ubi->beb_rsvd_level = ubi->good_peb_count/100;
+	ubi->beb_rsvd_level *= CONFIG_MTD_UBI_BEB_RESERVE;
+	if (ubi->beb_rsvd_level < MIN_RESEVED_PEBS)
+		ubi->beb_rsvd_level = MIN_RESEVED_PEBS;
+}

drivers/mtd/ubi/scan.h

@@ -0,0 +1,167 @@
+/*
+ * 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_SCAN_H__
+#define __UBI_SCAN_H__
+
+/* The erase counter value for this physical eraseblock is unknown */
+#define UBI_SCAN_UNKNOWN_EC (-1)
+
+/**
+ * struct ubi_scan_leb - scanning information about a physical eraseblock.
+ * @ec: erase counter (%UBI_SCAN_UNKNOWN_EC if it is unknown)
+ * @pnum: physical eraseblock number
+ * @lnum: logical eraseblock number
+ * @scrub: if this physical eraseblock needs scrubbing
+ * @sqnum: sequence number
+ * @u: unions RB-tree or @list links
+ * @u.rb: link in the per-volume RB-tree of &struct ubi_scan_leb objects
+ * @u.list: link in one of the eraseblock lists
+ * @leb_ver: logical eraseblock version (obsolete)
+ *
+ * One object of this type is allocated for each physical eraseblock during
+ * scanning.
+ */
+struct ubi_scan_leb {
+	int ec;
+	int pnum;
+	int lnum;
+	int scrub;
+	unsigned long long sqnum;
+	union {
+		struct rb_node rb;
+		struct list_head list;
+	} u;
+	uint32_t leb_ver;
+};
+
+/**
+ * struct ubi_scan_volume - scanning information about a volume.
+ * @vol_id: volume ID
+ * @highest_lnum: highest logical eraseblock number in this volume
+ * @leb_count: number of logical eraseblocks in this volume
+ * @vol_type: volume type
+ * @used_ebs: number of used logical eraseblocks in this volume (only for
+ * static volumes)
+ * @last_data_size: amount of data in the last logical eraseblock of this
+ * volume (always equivalent to the usable logical eraseblock size in case of
+ * dynamic volumes)
+ * @data_pad: how many bytes at the end of logical eraseblocks of this volume
+ * are not used (due to volume alignment)
+ * @compat: compatibility flags of this volume
+ * @rb: link in the volume RB-tree
+ * @root: root of the RB-tree containing all the eraseblock belonging to this
+ * volume (&struct ubi_scan_leb objects)
+ *
+ * One object of this type is allocated for each volume during scanning.
+ */
+struct ubi_scan_volume {
+	int vol_id;
+	int highest_lnum;
+	int leb_count;
+	int vol_type;
+	int used_ebs;
+	int last_data_size;
+	int data_pad;
+	int compat;
+	struct rb_node rb;
+	struct rb_root root;
+};
+
+/**
+ * struct ubi_scan_info - UBI scanning information.
+ * @volumes: root of the volume RB-tree
+ * @corr: list of corrupted physical eraseblocks
+ * @free: list of free physical eraseblocks
+ * @erase: list of physical eraseblocks which have to be erased
+ * @alien: list of physical eraseblocks which should not be used by UBI (e.g.,
+ * @bad_peb_count: count of bad physical eraseblocks
+ * those belonging to "preserve"-compatible internal volumes)
+ * @vols_found: number of volumes found during scanning
+ * @highest_vol_id: highest volume ID
+ * @alien_peb_count: count of physical eraseblocks in the @alien list
+ * @is_empty: flag indicating whether the MTD device is empty or not
+ * @min_ec: lowest erase counter value
+ * @max_ec: highest erase counter value
+ * @max_sqnum: highest sequence number value
+ * @mean_ec: mean erase counter value
+ * @ec_sum: a temporary variable used when calculating @mean_ec
+ * @ec_count: a temporary variable used when calculating @mean_ec
+ *
+ * This data structure contains the result of scanning and may be used by other
+ * UBI units to build final UBI data structures, further error-recovery and so
+ * on.
+ */
+struct ubi_scan_info {
+	struct rb_root volumes;
+	struct list_head corr;
+	struct list_head free;
+	struct list_head erase;
+	struct list_head alien;
+	int bad_peb_count;
+	int vols_found;
+	int highest_vol_id;
+	int alien_peb_count;
+	int is_empty;
+	int min_ec;
+	int max_ec;
+	unsigned long long max_sqnum;
+	int mean_ec;
+	int ec_sum;
+	int ec_count;
+};
+
+struct ubi_device;
+struct ubi_vid_hdr;
+
+/*
+ * ubi_scan_move_to_list - move a physical eraseblock from the volume tree to a
+ * list.
+ *
+ * @sv: volume scanning information
+ * @seb: scanning eraseblock infprmation
+ * @list: the list to move to
+ */
+static inline void ubi_scan_move_to_list(struct ubi_scan_volume *sv,
+					 struct ubi_scan_leb *seb,
+					 struct list_head *list)
+{
+		rb_erase(&seb->u.rb, &sv->root);
+		list_add_tail(&seb->u.list, list);
+}
+
+int ubi_scan_add_to_list(struct ubi_scan_info *si, int pnum, int ec,
+			 struct list_head *list);
+int ubi_scan_add_used(const struct ubi_device *ubi, struct ubi_scan_info *si,
+		      int pnum, int ec, const struct ubi_vid_hdr *vid_hdr,
+		      int bitflips);
+struct ubi_scan_volume *ubi_scan_find_sv(const struct ubi_scan_info *si,
+					 int vol_id);
+struct ubi_scan_leb *ubi_scan_find_seb(const struct ubi_scan_volume *sv,
+				       int lnum);
+void ubi_scan_rm_volume(struct ubi_scan_info *si, struct ubi_scan_volume *sv);
+struct ubi_scan_leb *ubi_scan_get_free_peb(const struct ubi_device *ubi,
+					   struct ubi_scan_info *si);
+int ubi_scan_erase_peb(const struct ubi_device *ubi,
+		       const struct ubi_scan_info *si, int pnum, int ec);
+struct ubi_scan_info *ubi_scan(struct ubi_device *ubi);
+void ubi_scan_destroy_si(struct ubi_scan_info *si);
+
+#endif /* !__UBI_SCAN_H__ */

drivers/mtd/ubi/upd.c

@@ -0,0 +1,348 @@
+/*
+ * Copyright (c) International Business Machines Corp., 2006
+ * Copyright (c) Nokia Corporation, 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)
+ *
+ * Jan 2007: Alexander Schmidt, hacked per-volume update.
+ */
+
+/*
+ * This file contains implementation of the volume update functionality.
+ *
+ * The update operation is based on the per-volume update marker which is
+ * stored in the volume table. The update marker is set before the update
+ * starts, and removed after the update has been finished. So if the update was
+ * interrupted by an unclean re-boot or due to some other reasons, the update
+ * marker stays on the flash media and UBI finds it when it attaches the MTD
+ * device next time. If the update marker is set for a volume, the volume is
+ * treated as damaged and most I/O operations are prohibited. Only a new update
+ * operation is allowed.
+ *
+ * Note, in general it is possible to implement the update operation as a
+ * transaction with a roll-back capability.
+ */
+
+#include <linux/err.h>
+#include <asm/uaccess.h>
+#include <asm/div64.h>
+#include "ubi.h"
+
+/**
+ * set_update_marker - set update marker.
+ * @ubi: UBI device description object
+ * @vol_id: volume ID
+ *
+ * This function sets the update marker flag for volume @vol_id. Returns zero
+ * in case of success and a negative error code in case of failure.
+ */
+static int set_update_marker(struct ubi_device *ubi, int vol_id)
+{
+	int err;
+	struct ubi_vtbl_record vtbl_rec;
+	struct ubi_volume *vol = ubi->volumes[vol_id];
+
+	dbg_msg("set update marker for volume %d", vol_id);
+
+	if (vol->upd_marker) {
+		ubi_assert(ubi->vtbl[vol_id].upd_marker);
+		dbg_msg("already set");
+		return 0;
+	}
+
+	memcpy(&vtbl_rec, &ubi->vtbl[vol_id], sizeof(struct ubi_vtbl_record));
+	vtbl_rec.upd_marker = 1;
+
+	err = ubi_change_vtbl_record(ubi, vol_id, &vtbl_rec);
+	vol->upd_marker = 1;
+	return err;
+}
+
+/**
+ * clear_update_marker - clear update marker.
+ * @ubi: UBI device description object
+ * @vol_id: volume ID
+ * @bytes: new data size in bytes
+ *
+ * This function clears the update marker for volume @vol_id, sets new volume
+ * data size and clears the "corrupted" flag (static volumes only). Returns
+ * zero in case of success and a negative error code in case of failure.
+ */
+static int clear_update_marker(struct ubi_device *ubi, int vol_id, long long bytes)
+{
+	int err;
+	uint64_t tmp;
+	struct ubi_vtbl_record vtbl_rec;
+	struct ubi_volume *vol = ubi->volumes[vol_id];
+
+	dbg_msg("clear update marker for volume %d", vol_id);
+
+	memcpy(&vtbl_rec, &ubi->vtbl[vol_id], sizeof(struct ubi_vtbl_record));
+	ubi_assert(vol->upd_marker && vtbl_rec.upd_marker);
+	vtbl_rec.upd_marker = 0;
+
+	if (vol->vol_type == UBI_STATIC_VOLUME) {
+		vol->corrupted = 0;
+		vol->used_bytes = tmp = bytes;
+		vol->last_eb_bytes = do_div(tmp, vol->usable_leb_size);
+		vol->used_ebs = tmp;
+		if (vol->last_eb_bytes)
+			vol->used_ebs += 1;
+		else
+			vol->last_eb_bytes = vol->usable_leb_size;
+	}
+
+	err = ubi_change_vtbl_record(ubi, vol_id, &vtbl_rec);
+	vol->upd_marker = 0;
+	return err;
+}
+
+/**
+ * ubi_start_update - start volume update.
+ * @ubi: UBI device description object
+ * @vol_id: volume ID
+ * @bytes: update bytes
+ *
+ * This function starts volume update operation. If @bytes is zero, the volume
+ * is just wiped out. Returns zero in case of success and a negative error code
+ * in case of failure.
+ */
+int ubi_start_update(struct ubi_device *ubi, int vol_id, long long bytes)
+{
+	int i, err;
+	uint64_t tmp;
+	struct ubi_volume *vol = ubi->volumes[vol_id];
+
+	dbg_msg("start update of volume %d, %llu bytes", vol_id, bytes);
+	vol->updating = 1;
+
+	err = set_update_marker(ubi, vol_id);
+	if (err)
+		return err;
+
+	/* Before updating - wipe out the volume */
+	for (i = 0; i < vol->reserved_pebs; i++) {
+		err = ubi_eba_unmap_leb(ubi, vol_id, i);
+		if (err)
+			return err;
+	}
+
+	if (bytes == 0) {
+		err = clear_update_marker(ubi, vol_id, 0);
+		if (err)
+			return err;
+		err = ubi_wl_flush(ubi);
+		if (!err)
+			vol->updating = 0;
+	}
+
+	vol->upd_buf = kmalloc(ubi->leb_size, GFP_KERNEL);
+	if (!vol->upd_buf)
+		return -ENOMEM;
+
+	tmp = bytes;
+	vol->upd_ebs = !!do_div(tmp, vol->usable_leb_size);
+	vol->upd_ebs += tmp;
+	vol->upd_bytes = bytes;
+	vol->upd_received = 0;
+	return 0;
+}
+
+/**
+ * write_leb - write update data.
+ * @ubi: UBI device description object
+ * @vol_id: volume ID
+ * @lnum: logical eraseblock number
+ * @buf: data to write
+ * @len: data size
+ * @used_ebs: how many logical eraseblocks will this volume contain (static
+ * volumes only)
+ *
+ * This function writes update data to corresponding logical eraseblock. In
+ * case of dynamic volume, this function checks if the data contains 0xFF bytes
+ * at the end. If yes, the 0xFF bytes are cut and not written. So if the whole
+ * buffer contains only 0xFF bytes, the LEB is left unmapped.
+ *
+ * The reason why we skip the trailing 0xFF bytes in case of dynamic volume is
+ * that we want to make sure that more data may be appended to the logical
+ * eraseblock in future. Indeed, writing 0xFF bytes may have side effects and
+ * this PEB won't be writable anymore. So if one writes the file-system image
+ * to the UBI volume where 0xFFs mean free space - UBI makes sure this free
+ * space is writable after the update.
+ *
+ * We do not do this for static volumes because they are read-only. But this
+ * also cannot be done because we have to store per-LEB CRC and the correct
+ * data length.
+ *
+ * This function returns zero in case of success and a negative error code in
+ * case of failure.
+ */
+static int write_leb(struct ubi_device *ubi, int vol_id, int lnum, void *buf,
+		     int len, int used_ebs)
+{
+	int err, l;
+	struct ubi_volume *vol = ubi->volumes[vol_id];
+
+	if (vol->vol_type == UBI_DYNAMIC_VOLUME) {
+		l = ALIGN(len, ubi->min_io_size);
+		memset(buf + len, 0xFF, l - len);
+
+		l = ubi_calc_data_len(ubi, buf, l);
+		if (l == 0) {
+			dbg_msg("all %d bytes contain 0xFF - skip", len);
+			return 0;
+		}
+		if (len != l)
+			dbg_msg("skip last %d bytes (0xFF)", len - l);
+
+		err = ubi_eba_write_leb(ubi, vol_id, lnum, buf, 0, l,
+					UBI_UNKNOWN);
+	} else {
+		/*
+		 * When writing static volume, and this is the last logical
+		 * eraseblock, the length (@len) does not have to be aligned to
+		 * the minimal flash I/O unit. The 'ubi_eba_write_leb_st()'
+		 * function accepts exact (unaligned) length and stores it in
+		 * the VID header. And it takes care of proper alignment by
+		 * padding the buffer. Here we just make sure the padding will
+		 * contain zeros, not random trash.
+		 */
+		memset(buf + len, 0, vol->usable_leb_size - len);
+		err = ubi_eba_write_leb_st(ubi, vol_id, lnum, buf, len,
+					   UBI_UNKNOWN, used_ebs);
+	}
+
+	return err;
+}
+
+/**
+ * ubi_more_update_data - write more update data.
+ * @vol: volume description object
+ * @buf: write data (user-space memory buffer)
+ * @count: how much bytes to write
+ *
+ * This function writes more data to the volume which is being updated. It may
+ * be called arbitrary number of times until all of the update data arrive.
+ * This function returns %0 in case of success, number of bytes written during
+ * the last call if the whole volume update was successfully finished, and a
+ * negative error code in case of failure.
+ */
+int ubi_more_update_data(struct ubi_device *ubi, int vol_id,
+			 const void __user *buf, int count)
+{
+	uint64_t tmp;
+	struct ubi_volume *vol = ubi->volumes[vol_id];
+	int lnum, offs, err = 0, len, to_write = count;
+
+	dbg_msg("write %d of %lld bytes, %lld already passed",
+		count, vol->upd_bytes, vol->upd_received);
+
+	if (ubi->ro_mode)
+		return -EROFS;
+
+	tmp = vol->upd_received;
+	offs = do_div(tmp, vol->usable_leb_size);
+	lnum = tmp;
+
+	if (vol->upd_received + count > vol->upd_bytes)
+		to_write = count = vol->upd_bytes - vol->upd_received;
+
+	/*
+	 * When updating volumes, we accumulate whole logical eraseblock of
+	 * data and write it at once.
+	 */
+	if (offs != 0) {
+		/*
+		 * This is a write to the middle of the logical eraseblock. We
+		 * copy the data to our update buffer and wait for more data or
+		 * flush it if the whole eraseblock is written or the update
+		 * is finished.
+		 */
+
+		len = vol->usable_leb_size - offs;
+		if (len > count)
+			len = count;
+
+		err = copy_from_user(vol->upd_buf + offs, buf, len);
+		if (err)
+			return -EFAULT;
+
+		if (offs + len == vol->usable_leb_size ||
+		    vol->upd_received + len == vol->upd_bytes) {
+			int flush_len = offs + len;
+
+			/*
+			 * OK, we gathered either the whole eraseblock or this
+			 * is the last chunk, it's time to flush the buffer.
+			 */
+			ubi_assert(flush_len <= vol->usable_leb_size);
+			err = write_leb(ubi, vol_id, lnum, vol->upd_buf,
+					flush_len, vol->upd_ebs);
+			if (err)
+				return err;
+		}
+
+		vol->upd_received += len;
+		count -= len;
+		buf += len;
+		lnum += 1;
+	}
+
+	/*
+	 * If we've got more to write, let's continue. At this point we know we
+	 * are starting from the beginning of an eraseblock.
+	 */
+	while (count) {
+		if (count > vol->usable_leb_size)
+			len = vol->usable_leb_size;
+		else
+			len = count;
+
+		err = copy_from_user(vol->upd_buf, buf, len);
+		if (err)
+			return -EFAULT;
+
+		if (len == vol->usable_leb_size ||
+		    vol->upd_received + len == vol->upd_bytes) {
+			err = write_leb(ubi, vol_id, lnum, vol->upd_buf, len,
+					vol->upd_ebs);
+			if (err)
+				break;
+		}
+
+		vol->upd_received += len;
+		count -= len;
+		lnum += 1;
+		buf += len;
+	}
+
+	ubi_assert(vol->upd_received <= vol->upd_bytes);
+	if (vol->upd_received == vol->upd_bytes) {
+		/* The update is finished, clear the update marker */
+		err = clear_update_marker(ubi, vol_id, vol->upd_bytes);
+		if (err)
+			return err;
+		err = ubi_wl_flush(ubi);
+		if (err == 0) {
+			err = to_write;
+			kfree(vol->upd_buf);
+			vol->updating = 0;
+		}
+	}
+
+	return err;
+}