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 in the UBI design documentation:
ubidesign.pdf. Which can be found here:
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 and also
  some to JFFS2 to make it work.

Signed-off-by: Frank Haverkamp <haver@vnet.ibm.com>

1 files changed, 617 insertions, 0 deletions

diff --git a/ubi-utils/src/libpfiflash/pfiflash.c b/ubi-utils/src/libpfiflash/pfiflash.c
new file mode 100644
index 0000000..0859a22
--- /dev/null
+++ b/ubi-utils/src/libpfiflash/pfiflash.c
@@ -0,0 +1,617 @@
+/*
+ * 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., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+/**
+ * @file pfiflash.c
+ *
+ * @author Oliver Lohmann <oliloh@de.ibm.com>
+ *
+ * @brief This library is provides an interface to the pfiflash utility.
+ *
+ * <oliloh@de.ibm.com> Wed Mar 15 11:39:19 CET 2006 Initial creation.
+ *
+ * @TODO Comare data before writing it. This implies that the volume
+ * parameters are compared first: size, alignment, name, type, ...,
+ * this is the same, compare the data. Volume deletion is deffered
+ * until the difference has been found out.
+ */
+
+#include <stdio.h>
+#include <errno.h>
+#include <fcntl.h>
+#define __USE_GNU
+#include <string.h>
+
+#include <libubi.h>
+#include <pfiflash.h>
+
+#include <mtd/ubi-user.h>	/* FIXME Is this ok here!!?? */
+
+#include "ubimirror.h"
+#include "error.h"
+#include "reader.h"
+#include "example_ubi.h"
+#include "bootenv.h"
+
+static const char copyright [] __attribute__((unused)) =
+	"Copyright (c) International Business Machines Corp., 2006";
+
+#define EBUF(fmt...) do {			\
+		snprintf(err_buf, err_buf_size, fmt);	\
+	} while (0)
+
+static pdd_func_t pdd_funcs[PDD_HANDLING_NUM]  =
+	{
+		&bootenv_pdd_keep,
+		&bootenv_pdd_merge,
+		&bootenv_pdd_overwrite
+	};
+/**< An array of PDD function pointers indexed by the algorithm. */
+
+
+typedef enum ubi_update_process_t {
+	UBI_REMOVE = 0,
+	UBI_WRITE,
+} ubi_update_process_t;
+
+static int
+skip_raw_sections(FILE* pfi, list_t pfi_raws,
+		  char* err_buf, size_t err_buf_size)
+{
+	int rc = 0;
+
+	void *i;
+	list_t ptr;
+	size_t j, skip_size;
+
+	if (is_empty(pfi_raws))
+		return 0;
+
+	foreach(i, ptr, pfi_raws) {
+		skip_size = ((pfi_raw_t)i)->data_size;
+		for(j = 0;  j < skip_size; j++) {
+			fgetc(pfi);
+			if (ferror(pfi)) {
+				EBUF("Cannot skip raw section in PFI.");
+				rc = -EIO;
+				goto err;
+			}
+		}
+	}
+ err:
+	return rc;
+}
+
+/**
+ * @brief Wraps the ubi_mkvol functions and implements a hook for the bootenv
+ *	  update.
+ * @param devno UBI device number.
+ * @param s Current seqnum.
+ * @param u Information about the UBI volume from the PFI.
+ * @param err_buf An error buffer.
+ * @param err_buf_size The size of the error buffer.
+ * @return 0 On Sucess.
+ * @return else Error.
+ */
+static int
+my_ubi_mkvol(int devno, int s, pfi_ubi_t u, char *err_buf, size_t err_buf_size)
+{
+	int rc = 0;
+	int type;
+	ubi_lib_t ulib = NULL;
+
+	log_msg("%s(vol_id=%d, size=%d, data_size=%d, type=%d, "
+		"alig=%d, nlen=%d, name=%s)", __func__,
+		u->ids[s], u->size, u->data_size, u->type, u->alignment,
+		strnlen(u->names[s], PFI_UBI_VOL_NAME_LEN), u->names[s]);
+
+	rc = ubi_open(&ulib);
+	if (rc != 0) {
+		goto err;
+	}
+
+	switch (u->type) {
+	case pfi_ubi_static:
+		type = UBI_STATIC_VOLUME; break;
+	case pfi_ubi_dynamic:
+		type = UBI_DYNAMIC_VOLUME; break;
+	default:
+		type = UBI_DYNAMIC_VOLUME;
+	}
+
+	rc = ubi_mkvol(ulib, devno, u->ids[s], type, u->size, u->alignment,
+		       u->names[s]);
+	if (rc != 0) {
+		EBUF("Cannot create volume: %d", u->ids[s]);
+		goto err;
+	}
+
+ err:
+	if (ulib != NULL)
+		ubi_close(&ulib);
+	return rc;
+}
+
+/**
+ * @brief A wrapper around the UBI library function ubi_rmvol.
+ * @param devno UBI device number.
+ * @param s Current seqnum.
+ * @param u Information about the UBI volume from the PFI.
+ * @param err_buf An error buffer.
+ * @param err_buf_size The size of the error buffer.
+ *
+ * If the volume does not exist, the function will return success.
+ */
+static int
+my_ubi_rmvol(int devno, uint32_t id, char *err_buf, size_t err_buf_size)
+{
+	int rc = 0;
+	ubi_lib_t ulib = NULL;
+	int fd;
+
+	log_msg("%s(id=%d)", __func__, id);
+
+	rc = ubi_open(&ulib);
+	if (rc != 0)
+		goto err;
+
+	/**
+	 * Truncate if it exist or not.
+	 */
+	fd = ubi_vol_open(ulib, devno, id, O_RDWR);
+	if (fd == -1)
+		return 0;	/* not existent, return */
+
+	rc = ubi_vol_update(fd, 0);
+	if (rc < 0) {
+		fprintf(stderr, "update failed rc=%d errno=%d\n", rc, errno);
+		ubi_vol_close(fd);
+		goto err;	/* if EBUSY than empty device, continue */
+	}
+	ubi_vol_close(fd);
+
+	rc = ubi_rmvol(ulib, devno, id);
+	if (rc != 0) {
+		/* @TODO Define a ubi_rmvol return value which says
+		 * sth like EUBI_NOSUCHDEV. In this case, a failed
+		 * operation is acceptable. Everything else has to be
+		 * classified as real error. But talk to Andreas Arnez
+		 * before defining something odd...
+		 */
+		/* if ((errno == EINVAL) || (errno == ENODEV))
+		   return 0; */ /* currently it is EINVAL or ENODEV */
+
+		dbg_msg("Remove UBI volume %d returned with error: %d "
+			"errno=%d", id, rc, errno);
+		goto err;
+	}
+ err:
+	if (ulib != NULL)
+		ubi_close(&ulib);
+	return rc;
+}
+
+static int
+read_bootenv_volume(int devno, uint32_t id, bootenv_t bootenv_old,
+		    char *err_buf, size_t err_buf_size)
+{
+	int rc = 0;
+	ubi_lib_t ulib = NULL;
+	FILE* fp_in = NULL;
+
+	rc = ubi_open(&ulib);
+	if (rc)
+		return rc;
+
+	fp_in = ubi_vol_fopen_read(ulib, devno, id);
+	if (!fp_in) {
+		EBUF("Cannot open bootenv volume");
+		rc = -EIO;
+		goto err;
+	}
+
+	log_msg("%s reading old bootenvs", __func__);
+
+	/* Save old bootenvs for reference */
+	rc = bootenv_read(fp_in, bootenv_old, BOOTENV_MAXSIZE);
+	if (rc)
+		EBUF("Cannot read bootenv_old");
+ err:
+	if (fp_in)
+		fclose(fp_in);
+	if (ulib)
+		ubi_close(&ulib);
+	return rc;
+}
+
+static int
+write_bootenv_volume(int devno, uint32_t id, bootenv_t bootenv_old,
+		     pdd_func_t pdd_f,
+		     FILE* fp_in, /* new pdd data contained in pfi */
+		     size_t fp_in_size,	/* data size of new pdd data in pfi */
+		     char *err_buf, size_t err_buf_size)
+{
+	int rc = 0;
+	int warnings = 0;
+	ubi_lib_t ulib = NULL;
+	bootenv_t bootenv_new = NULL;
+	bootenv_t bootenv_res = NULL;
+	size_t update_size = 0;
+	FILE *fp_out = NULL;
+
+	log_msg("%s(id=%d, fp_in=%p)", __func__, id, fp_in);
+
+	/* Workflow:
+	 * 1. Apply PDD operation and get the size of the returning
+	 *    bootenv_res section. Without the correct size it wouldn't
+	 *    be possible to call UBI update vol.
+	 * 2. Call UBI update vol
+	 * 3. Get FILE* to vol dev
+	 * 4. Write to FILE*
+	 */
+
+	rc = ubi_open(&ulib);
+	if (rc != 0) {
+		goto err;
+	}
+
+	rc = bootenv_create(&bootenv_new);
+	if (rc != 0)
+		goto err;
+	rc = bootenv_create(&bootenv_res);
+	if (rc != 0)
+		goto err;
+
+	rc = bootenv_read(fp_in, bootenv_new, fp_in_size);
+	if (rc != 0)
+		goto err;
+
+	rc = pdd_f(bootenv_old, bootenv_new, &bootenv_res, &warnings,
+		   err_buf, err_buf_size);
+	if (rc != 0)
+		goto err;
+	if (warnings) {
+		/* @TODO Do sth with the warning */
+		dbg_msg("A warning in the PDD operation occured: %d",
+			warnings);
+	}
+	log_msg("... (2)");
+
+	rc = bootenv_size(bootenv_res, &update_size);
+	if (rc != 0)
+		goto err;
+
+	fp_out = ubi_vol_fopen_update(ulib, devno, id, update_size);
+	if (fp_out == NULL)
+		goto err;
+
+	rc = bootenv_write(fp_out, bootenv_res);
+	if (rc != 0) {
+		EBUF("Write operation on ubi%d_%d failed.", devno, id);
+		rc = -EIO;
+		goto err;
+	}
+
+ err:
+	if (ulib != NULL)
+		ubi_close(&ulib);
+	if (bootenv_new != NULL)
+		bootenv_destroy(&bootenv_new);
+	if (bootenv_res != NULL)
+		bootenv_destroy(&bootenv_res);
+	if (fp_out)
+		fclose(fp_out);
+	return rc;
+}
+
+static int
+write_normal_volume(int devno, uint32_t id, size_t update_size, FILE* fp_in,
+		    char *err_buf, size_t err_buf_size)
+{
+	int rc = 0;
+	ubi_lib_t ulib = NULL;
+	FILE* fp_out = NULL;
+	int c;
+	size_t i;
+
+	log_msg("%s(id=%d, update_size=%d fp_in=%p)",
+		__func__, id, update_size, fp_in);
+
+	rc = ubi_open(&ulib);
+	if (rc)
+		return rc;
+
+	fp_out = ubi_vol_fopen_update(ulib, devno, id, update_size);
+	if (fp_out == NULL) {
+		rc = -1;
+		goto err;
+	}
+
+	log_msg("starting the update ... "); /* FIXME DBG */
+	for (i = 0; i < update_size; i++) {
+		c = getc(fp_in);
+		if (c == EOF && ferror(fp_in)) {
+			rc = -EIO;
+			goto err;
+		}
+		if (putc(c, fp_out) == EOF) {
+			rc = -EIO;
+			goto err;
+		}
+		/*  FIXME DBG */
+		/* if ((i & 0xFFF) == 0xFFF) log_msg("."); */
+	}
+	/* log_msg("\n"); */		/*  FIXME DBG */
+ err:
+	if (fp_out)
+		fclose(fp_out);
+	if (ulib)
+		ubi_close(&ulib);
+	return rc;
+}
+
+
+/**
+ * @brief ...
+ * @precondition	The PFI file contains at least one ubi_id entry.
+ *			This is assured by the PFI read process.
+ * @postcondition	The used seqnum number is set in the UBI PFI
+ *			header list.
+ *			The UBI volumes specified by seqnum are processed.
+ */
+static int
+process_ubi_volumes(FILE* pfi, int seqnum, list_t pfi_ubis,
+		    bootenv_t bootenv_old, pdd_func_t pdd_f,
+		    ubi_update_process_t ubi_update_process,
+		    char *err_buf, size_t err_buf_size)
+{
+	int rc = 0;
+	pfi_ubi_t u;
+	list_t ptr;
+
+	foreach(u, ptr, pfi_ubis) {
+		int s = seqnum;
+		if (seqnum > (u->ids_size - 1)) {
+			s = 0; /* per default use the first */
+		}
+		u->curr_seqnum = s;
+
+		switch (ubi_update_process) {
+		case UBI_REMOVE:
+			if ((u->ids[s] == EXAMPLE_BOOTENV_VOL_ID_1) ||
+			    (u->ids[s] == EXAMPLE_BOOTENV_VOL_ID_2)) {
+				rc =read_bootenv_volume(EXAMPLE_UBI_DEVICE,
+							u->ids[s],
+							bootenv_old, err_buf,
+							err_buf_size);
+				if (rc != 0)
+					goto err;
+				}
+			rc = my_ubi_rmvol(EXAMPLE_UBI_DEVICE,  u->ids[s],
+					  err_buf, err_buf_size);
+			if (rc != 0)
+				goto err;
+			break;
+		case UBI_WRITE:
+			rc = my_ubi_mkvol(EXAMPLE_UBI_DEVICE, s, u,
+					  err_buf, err_buf_size);
+			if (rc != 0)
+				goto err;
+			if ((u->ids[s] == EXAMPLE_BOOTENV_VOL_ID_1) ||
+			    (u->ids[s] == EXAMPLE_BOOTENV_VOL_ID_2)) {
+				rc = write_bootenv_volume(EXAMPLE_UBI_DEVICE,
+							  u->ids[s],
+							  bootenv_old, pdd_f,
+							  pfi,
+							  u->data_size,
+							  err_buf,
+							  err_buf_size);
+			}
+			else {
+				rc = write_normal_volume(EXAMPLE_UBI_DEVICE,
+							 u->ids[s],
+							 u->data_size, pfi,
+							 err_buf,
+							 err_buf_size);
+			}
+			if (rc != 0)
+				goto err;
+			break;
+		default:
+			EBUF("Invoked unknown UBI operation.");
+			rc = -1;
+			goto err;
+
+		}
+		if (rc != 0) {
+			goto err;
+		}
+	}
+ err:
+	return rc;
+
+}
+
+static int
+erase_unmapped_ubi_volumes(int devno, list_t pfi_ubis,
+			   char *err_buf, size_t err_buf_size)
+{
+	int rc = 0;
+	list_t ptr;
+	pfi_ubi_t u;
+	size_t i;
+	uint8_t ubi_volumes[PFI_UBI_MAX_VOLUMES];
+
+	for (i = 0; i < PFI_UBI_MAX_VOLUMES; i++) {
+		ubi_volumes[i] = 1;
+	}
+
+	foreach(u, ptr, pfi_ubis) {
+		/* iterate over each vol_id */
+		for(i = 0; i < u->ids_size; i++) {
+			if (u->ids[i] > PFI_UBI_MAX_VOLUMES) {
+				EBUF("PFI file contains an invalid "
+				     "volume id: %d", u->ids[i]);
+				goto err;
+			}
+			/* remove from removal list */
+			ubi_volumes[u->ids[i]] = 0;
+		}
+	}
+
+	for (i = 0; i < PFI_UBI_MAX_VOLUMES; i++) {
+		if (ubi_volumes[i]) {
+			rc = my_ubi_rmvol(devno, i, err_buf, err_buf_size);
+			if (rc != 0)
+				goto err;
+		}
+	}
+ err:
+	return rc;
+}
+
+static int
+mirror_ubi_volumes(uint32_t devno, list_t pfi_ubis,
+		   char *err_buf, size_t err_buf_size)
+{
+	int rc = 0;
+	list_t ptr;
+	uint32_t j;
+	pfi_ubi_t i;
+	ubi_lib_t  ulib = NULL;
+
+	log_msg("%s(...)", __func__);
+
+	rc = ubi_open(&ulib);
+	if (rc != 0)
+		goto err;
+
+	/**
+	 * Execute all mirror operations on redundant groups.
+	 * Create a volume within a redundant group if it does
+	 * not exist already (this is a precondition of
+	 * ubimirror).
+	 */
+	foreach(i, ptr, pfi_ubis) {
+		for(j = 0; j < i->ids_size; j++) {
+			/* skip self-match */
+			if (i->ids[j] == i->ids[i->curr_seqnum])
+				continue;
+
+			rc = my_ubi_rmvol(devno, i->ids[j], err_buf,
+					  err_buf_size);
+			if (rc != 0)
+				goto err;
+
+			rc = my_ubi_mkvol(devno, j, i, err_buf, err_buf_size);
+			if (rc != 0)
+				goto err;
+		}
+	}
+
+	foreach(i, ptr, pfi_ubis) {
+		rc = ubimirror(devno, i->curr_seqnum, i->ids,
+			       i->ids_size, err_buf, err_buf_size);
+		if (rc != 0)
+			goto err;
+	}
+
+
+ err:
+	if (ulib != NULL)
+		ubi_close(&ulib);
+	return rc;
+}
+
+int
+pfiflash(FILE* pfi, int complete, int seqnum, pdd_handling_t pdd_handling,
+	 char *err_buf, size_t err_buf_size)
+{
+	int rc = 0;
+	pdd_func_t pdd_f = NULL;
+
+	if (pfi == NULL)
+		return -EINVAL;
+
+	/**
+	 * If the user didnt specify a seqnum we start per default
+	 * with the index 0
+	 */
+	int curr_seqnum = seqnum < 0 ? 0 : seqnum;
+
+	list_t pfi_raws   = mk_empty(); /* list of raw sections from a pfi */
+	list_t pfi_ubis   = mk_empty(); /* list of ubi sections from a pfi */
+
+	bootenv_t bootenv;
+	rc = bootenv_create(&bootenv);
+	if (rc != 0) {
+		EBUF("Cannot create bootenv variable");
+	}
+
+	rc = read_pfi_headers(&pfi_raws, &pfi_ubis, pfi,
+			      err_buf, err_buf_size);
+	if (rc != 0) {
+		EBUF("Cannot read PFI headers.");
+		goto err;
+	}
+
+	/* @TODO: If you want to implement an IPL update - start here. */
+	rc = skip_raw_sections(pfi, pfi_raws, err_buf, err_buf_size);
+	if (rc != 0) {
+		goto err;
+	}
+
+	if (complete) {
+		rc = erase_unmapped_ubi_volumes(EXAMPLE_UBI_DEVICE, pfi_ubis,
+						err_buf, err_buf_size);
+		if (rc != 0) {
+			EBUF("Cannot delete unmapped UBI volumes.");
+			goto err;
+		}
+	}
+
+	if ((pdd_handling >= 0) && (pdd_handling < PDD_HANDLING_NUM)) {
+		pdd_f = pdd_funcs[pdd_handling];
+	}
+	else {
+		EBUF("Used unknown PDD handling algorithm (pdd_handling)");
+	}
+
+	rc = process_ubi_volumes(pfi, curr_seqnum, pfi_ubis, bootenv, pdd_f,
+				 UBI_REMOVE, err_buf, err_buf_size);
+	if  (rc != 0) {
+		goto err;
+	}
+	rc = process_ubi_volumes(pfi, curr_seqnum, pfi_ubis, bootenv, pdd_f,
+				 UBI_WRITE, err_buf, err_buf_size);
+	if  (rc != 0) {
+		goto err;
+	}
+	if (seqnum < 0) { /* mirror redundant pairs */
+		rc = mirror_ubi_volumes(EXAMPLE_UBI_DEVICE, pfi_ubis,
+					err_buf, err_buf_size);
+		if (rc != 0)
+			goto err;
+	}
+
+ err:
+	pfi_raws = remove_all((free_func_t)&free_pfi_raw, pfi_raws);
+	pfi_ubis = remove_all((free_func_t)&free_pfi_ubi, pfi_ubis);
+	bootenv_destroy(&bootenv);
+	return rc;
+}