From: Artem Bityutskiy <dedekind@infradead.org>
To: Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Cc: Christoph Hellwig <hch@infradead.org>,
Artem Bityutskiy <dedekind@infradead.org>,
Frank Haverkamp <haver@vnet.ibm.com>,
Josh Boyer <jwboyer@linux.vnet.ibm.com>,
Thomas Gleixner <tglx@linutronix.de>,
David Woodhouse <dwmw2@infradead.org>
Subject: [PATCH 19/44 take 2] [UBI] volume table unit header
Date: Sat, 17 Feb 2007 18:56:00 +0200 [thread overview]
Message-ID: <20070217165600.5845.37863.sendpatchset@localhost.localdomain> (raw)
In-Reply-To: <20070217165424.5845.4390.sendpatchset@localhost.localdomain>
diff -auNrp tmp-from/drivers/mtd/ubi/vtbl.h tmp-to/drivers/mtd/ubi/vtbl.h
--- tmp-from/drivers/mtd/ubi/vtbl.h 1970-01-01 02:00:00.000000000 +0200
+++ tmp-to/drivers/mtd/ubi/vtbl.h 2007-02-17 18:07:27.000000000 +0200
@@ -0,0 +1,302 @@
+/*
+ * 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 B. Bityutskiy
+ */
+
+/*
+ * The volume table unit.
+ *
+ * This unit is responsible for maintaining the volume table. The volume table
+ * is an on-flash table containing volume meta-data like volume name, number of
+ * reserved physical eraseblocks for this volume, volume type, etc. The volume
+ * table is stored in the so-called "layout volume".
+ *
+ * The layout volume is an internal volume where the volume table is stored.
+ * Actually, there are 2 equivalent copies of the volume table in the layout
+ * volume. The layout volume it organized as follows. It consists of two
+ * logical eraseblocks - LEB 0 and LEB 1. Each logical eraseblock stores a copy
+ * the volume table, i.e. LEB 0 and LEB 1 duplicate each other. This redundancy
+ * guarantees robustness and tolerance to unclean reboots. The volume table is
+ * a mere array of so-called "volume table records". Each record contains full
+ * information about the volume and is protected by a CRC checksum.
+ *
+ * The volume table is changed as follows. It is first changed in RAM. Then LEB
+ * 0 is erased, and the updated volume table is written back to LEB 0. The same
+ * is done with LEB 1. This scheme guarantees recoverability from unclean
+ * reboots.
+ *
+ * At this UBI implementation the on-flash volume table does not contain about
+ * how many data a static volume actually stores. This information may be found
+ * out while scanning (from the EB headers) so we do not store it in the
+ * on-flash volume table. So, as long as we have an unscalable UBI
+ * implementation which uses scanning, we may live without that. In case of a
+ * scalable implementation, this would be required.
+ *
+ * But it would be beneficial to store this information in the volume table.
+ * For example, suppose we have a static volume X, and all its physical
+ * eraseblocks have gone bad for some reasons. Suppose we are attaching the
+ * corresponding MTD device, the scanning unit finds no logical eraseblocks
+ * corresponding to the volume X. According to the volume table volume X does
+ * exist. So we don't know whether it is just empty or all its physical
+ * eraseblocks went bad. So we cannot alarm the user about this corruption.
+ *
+ * Note, although we don't store this information in the on-flash volume table,
+ * we keep it in the in-RAM copy of this table just because it is quite
+ * convenient.
+ *
+ * The volume table also stores so-called "update marker" which is used to
+ * implement the update operation. Before updating the volume, the update
+ * marker is set, after the update operation is finished, the update marker is
+ * cleared. So if the update operation was interrupted (e.g. by an unclean
+ * reboot) - the update marker is still there and we know that the volume's
+ * contents is damaged.
+ *
+ * Note, in this implementation we do not support concurrent updates and only
+ * one volume at a time may be updated.
+ */
+
+#ifndef __UBI_VTBL_H__
+#define __UBI_VTBL_H__
+
+#include <linux/mutex.h>
+#include <mtd/ubi-header.h>
+
+struct ubi_info;
+struct ubi_scan_info;
+struct ubi_vtbl_vtr;
+
+/**
+ * ubi_vtbl_mkvol - create volume table record for a new volume.
+ *
+ * @ubi: the UBI device description object
+ * @vol_id: ID of the new volume
+ * @vtr: volume table record of the new volume
+ *
+ * This function adds a volume described in @vtr to the volume table. This
+ * function uses only @vtr->reserved_pebs, @vtr->alignment, @vtr->data_pad,
+ * @vtr->vol_type, and @vtr->name_len fields of the @vtr object. The
+ * @vtr->usable_leb_size is calculated automatically. The data-related fields
+ * are set to zero for static volumes ant to volume size for dynamic ones.
+ *
+ * This function returns zero in case of success and a negative error code in
+ */
+int ubi_vtbl_mkvol(const struct ubi_info *ubi, int vol_id,
+ const struct ubi_vtbl_vtr *vtr);
+
+/**
+ * ubi_vtbl_rmvol - clear the volume table record of a volume.
+ *
+ * @ubi: the UBI device description object
+ * @vol_id: ID of the volume to remove
+ *
+ * This function returns zero in case of success and a negative error code in
+ * case of failure.
+ */
+int ubi_vtbl_rmvol(const struct ubi_info *ubi, int vol_id);
+
+/**
+ * ubi_vtbl_rsvol - change volume size in the volume table record.
+ *
+ * @ubi: the UBI device description object
+ * @vol_id: re-sized volume's ID
+ * @reserved_pebs: new size, i.e. new number of reserved eraseblocks.
+ *
+ * This function returns zero in case of success and a negative error code in
+ * case of failure.
+ */
+int ubi_vtbl_rsvol(const struct ubi_info *ubi, int vol_id, int reserved_pebs);
+
+/**
+ * ubi_vtbl_set_upd_marker - set the update marker flag.
+ *
+ * @ubi: the UBI device description object
+ * @vol_id: ID of the volume
+ *
+ * This function sets the update marker flag for volumr @vol_id. Returns zero
+ * in case of success and a negative error code in case of failure.
+ */
+int ubi_vtbl_set_upd_marker(const struct ubi_info *ubi, int vol_id);
+
+/**
+ * ubi_vtbl_clear_upd_marker - clear the update marker flag.
+ *
+ * @ubi: the UBI device description object
+ * @vol_id: ID of the volume
+ * @bytes: new data size in bytes
+ *
+ * This function clears the update marker for volume @vol_id, sets new volume
+ * data size and cleans the "corrupted" flag (static volume s only). This
+ * function returns zero in case of success and a negative error code in case
+ * of failure.
+ */
+int ubi_vtbl_clear_upd_marker(const struct ubi_info *ubi, int vol_id,
+ long long bytes);
+
+/**
+ * ubi_vtbl_set_corrupted - mark a volume as 'corrupted'.
+ *
+ * @ubi: the UBI device description object
+ * @vol_id: ID of the volume to mark
+ *
+ * This function marks volume @vol_id as corrupted. If the volume is not static
+ * it does nothing. Returns zero in case of success and a negative error code
+ * in case of failure.
+ */
+int ubi_vtbl_set_corrupted(const struct ubi_info *ubi, int vol_id);
+
+/**
+ * ubi_vtbl_get_vtr - retrieve a volume table record.
+ *
+ * @ubi: the UBI device description object
+ * @vol_id: the requested volume ID
+ *
+ * This function returns a pointer to the volume record or an error code.
+ * If the volume ID is incorrect, %-EINVAL is returned, if the volume does
+ * not exist, %-ENODEV is returned.
+ *
+ * This function does not access the flash media as retrieves the information
+ * from the in-RAM volume table copy. So it does not sleep.
+ */
+const struct ubi_vtbl_vtr *ubi_vtbl_get_vtr(const struct ubi_info *ubi,
+ int vol_id);
+
+/**
+ * ubi_vtbl_get_compat - get compatibility flags of a volume.
+ *
+ * @ubi: the UBI device description object
+ * @vol_id: volume ID
+ *
+ * This function returns compatibility flags of volumes. User volumes have no
+ * compatibility flags, so %0 is returned. The @vol_id must be correct.
+ */
+int ubi_vtbl_get_compat(const struct ubi_info *ubi, int vol_id);
+
+/**
+ * ubi_is_ivol - check if a volume is an internal volume.
+ *
+ * @vol_id: ID of the volume to test
+ *
+ * If the volume is internal volume, %1 is returned, otherwise %0 is returned.
+ */
+static inline int ubi_is_ivol(int vol_id)
+{
+ return vol_id >= UBI_INTERNAL_VOL_START &&
+ vol_id < UBI_INTERNAL_VOL_START + UBI_INT_VOL_COUNT;
+}
+
+/*
+ * ubi_ivol_is_known - check if this is a known internal volume.
+ *
+ * @vol_id: ID of the volume to check.
+ *
+ * This function returns non-zero if this is a known and supported internal
+ * volume and non-zero if not.
+ */
+static inline int ubi_ivol_is_known(int vol_id)
+{
+ return vol_id == UBI_LAYOUT_VOL_ID;
+}
+
+/**
+ * ubi_vtbl_init_scan - initialize the volume table unit using scanning
+ * information.
+ *
+ * @ubi: the UBI device description object
+ * @si: a pointer to the scanning information
+ *
+ * This function returns zero in case of success and a negative error code in
+ * case of failure.
+ */
+int ubi_vtbl_init_scan(struct ubi_info *ubi, struct ubi_scan_info *si);
+
+/**
+ * ubi_vtbl_close - close the volume table unit.
+ *
+ * @ubi: the UBI device description object
+ */
+void ubi_vtbl_close(const struct ubi_info *ubi);
+
+/**
+ * struct ubi_vtbl_vtr - in-memory representation of volume table records.
+ *
+ * @reserved_pebs: how many physical eraseblocks are reserved for this volume
+ * @alignment: volume alignment
+ * @data_pad: how many bytes are not used at the end of eraseblocks to
+ * satisfy the requested alignment
+ * @vol_type: volume type (%UBI_DYNAMIC_VOLUME or %UBI_STATIC_VOLUME)
+ * @name_len: volume name length
+ * @name: volume name
+ * @usable_leb_size: logical eraseblock size without padding
+ * @used_ebs: how many logical eraseblocks in this volume are contain data
+ * @last_eb_bytes: how many bytes are stored in the last logical eraseblock
+ * @used_bytes: how many bytes of data this volume contains
+ * @corrupted: non-zero if the data is corrupted (static volumes only)
+ * @upd_marker: non-zero if the update marker is set for this volume
+ *
+ * Note, the @usable_leb_size field is not stored on flash, as it is easily
+ * calculated with help of the @data_pad field. But it is just very handy, so
+ * we keep it in the in-RAM volume table record representation.
+ *
+ * Similar to the @used_ebs, @last_eb_bytes, @used_bytes and @corrupted. We do
+ * not store them in the on-flash volume table but keep handy in RAM.
+ *
+ * The @corrupted field indicates that the volume's contents is corrupted. And
+ * since UBI protects only the contents of static volumes, this field is only
+ * relevant to static volumes. In case of dynamic volumes it is user's
+ * responsibility to assure data integrity.
+ *
+ * The @upd_marker flag indicates that this volume is either being updated at
+ * the moment or is damaged because of an unclean reboot. Note, the @corrupted
+ * flag is always cleared if the @upd_marker flag is set.
+ */
+struct ubi_vtbl_vtr {
+ int reserved_pebs;
+ int alignment;
+ int data_pad;
+ int vol_type;
+ int name_len;
+ int usable_leb_size;
+ const char *name;
+ int used_ebs;
+ int last_eb_bytes;
+ long long used_bytes;
+ int corrupted;
+ int upd_marker;
+};
+
+/**
+ * struct ubi_vtbl_info - volume table unit description data structure.
+ *
+ * @vt_slots: how many volume table records are stored in the volume table
+ * @vt_size: size of the volume table in bytes
+ * @vt: the in-RAM copy of the volume table
+ * @mutex: serializes volume table changes
+ * @empty_rec: volume table record corresponding to an empty volume
+ * @ivol_vtrs: volume table records corresponding to internal volumes
+ */
+struct ubi_vtbl_info {
+ int vt_slots; /* public */
+ int vt_size; /* private */
+ struct ubi_vtbl_vtr *vt; /* private */
+ struct mutex mutex; /* private */
+ const struct ubi_vol_tbl_record empty_rec; /* private */
+ struct ubi_vtbl_vtr ivol_vtrs[UBI_INT_VOL_COUNT]; /* private */
+};
+
+#endif /* __UBI_VTBL_H__ */
next prev parent reply other threads:[~2007-02-17 16:57 UTC|newest]
Thread overview: 129+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-02-17 16:54 [PATCH 00/44 take 2] [UBI] Unsorted Block Images Artem Bityutskiy
2007-02-17 16:54 ` [PATCH 01/44 take 2] [UBI] Linux build integration Artem Bityutskiy
2007-02-17 16:54 ` [PATCH 02/44 take 2] [UBI] on-flash data structures header Artem Bityutskiy
2007-02-17 16:54 ` [PATCH 03/44 take 2] [UBI] user-space API header Artem Bityutskiy
2007-02-17 21:27 ` Arnd Bergmann
2007-02-20 13:07 ` Artem Bityutskiy
2007-02-20 13:17 ` Arnd Bergmann
2007-02-17 16:54 ` [PATCH 04/44 take 2] [UBI] kernel-spce " Artem Bityutskiy
2007-02-18 1:32 ` Greg KH
2007-02-18 2:08 ` Josh Boyer
2007-02-26 12:12 ` Artem Bityutskiy
2007-02-17 16:54 ` [PATCH 05/44 take 2] [UBI] internal common header Artem Bityutskiy
2007-02-17 21:05 ` Arnd Bergmann
2007-02-19 11:16 ` Artem Bityutskiy
2007-02-19 10:54 ` Christoph Hellwig
2007-02-19 12:38 ` Josh Boyer
2007-02-20 13:05 ` Artem Bityutskiy
2007-02-20 14:55 ` Theodore Tso
2007-02-20 15:15 ` David Woodhouse
2007-02-20 15:22 ` Theodore Tso
2007-02-20 15:33 ` David Woodhouse
2007-02-20 16:12 ` Theodore Tso
2007-02-20 16:47 ` David Woodhouse
2007-02-25 10:42 ` Pavel Machek
2007-02-20 15:24 ` Artem Bityutskiy
2007-02-25 5:45 ` Christoph Hellwig
2007-02-26 10:28 ` Artem Bityutskiy
2007-02-25 5:43 ` Christoph Hellwig
2007-02-25 6:04 ` David Woodhouse
2007-02-20 15:21 ` Artem Bityutskiy
2007-02-25 5:46 ` Christoph Hellwig
2007-02-20 15:25 ` Artem Bityutskiy
2007-02-25 5:50 ` Christoph Hellwig
2007-02-25 11:55 ` Theodore Tso
2007-02-26 10:09 ` Artem Bityutskiy
2007-02-17 16:54 ` [PATCH 06/44 take 2] [UBI] startup code Artem Bityutskiy
2007-02-19 10:59 ` Christoph Hellwig
2007-02-20 13:00 ` Artem Bityutskiy
2007-02-23 11:03 ` Artem Bityutskiy
2007-02-25 5:58 ` Christoph Hellwig
2007-02-25 22:03 ` Rusty Russell
2007-03-05 13:28 ` Frank Haverkamp
2007-02-26 11:54 ` Artem Bityutskiy
2007-05-17 14:44 ` Christoph Hellwig
2007-05-17 15:06 ` Artem Bityutskiy
2007-02-17 16:54 ` [PATCH 07/44 take 2] [UBI] misc unit header Artem Bityutskiy
2007-02-17 22:59 ` Theodore Tso
2007-02-19 11:00 ` Christoph Hellwig
2007-02-20 12:56 ` Artem Bityutskiy
2007-02-19 11:13 ` Artem Bityutskiy
2007-02-17 16:55 ` [PATCH 08/44 take 2] [UBI] misc unit implementation Artem Bityutskiy
2007-02-17 16:55 ` [PATCH 09/44 take 2] [UBI] debug unit header Artem Bityutskiy
2007-02-17 21:18 ` Arnd Bergmann
2007-02-19 11:00 ` Christoph Hellwig
2007-02-19 12:33 ` Artem Bityutskiy
2007-02-19 14:02 ` Josh Boyer
2007-02-19 14:04 ` Artem Bityutskiy
2007-02-17 16:55 ` [PATCH 10/44 take 2] [UBI] debug unit implementation Artem Bityutskiy
2007-02-17 21:00 ` Arnd Bergmann
2007-02-19 12:29 ` Artem Bityutskiy
2007-02-17 16:55 ` [PATCH 11/44 take 2] [UBI] allocation unit header Artem Bityutskiy
2007-02-17 16:55 ` [PATCH 12/44 take 2] [UBI] allocation unit implementation Artem Bityutskiy
2007-02-17 20:55 ` Arnd Bergmann
2007-02-19 11:05 ` Artem Bityutskiy
2007-02-19 11:13 ` Pekka Enberg
2007-02-20 11:30 ` Artem Bityutskiy
2007-02-17 16:55 ` [PATCH 13/44 take 2] [UBI] I/O unit header Artem Bityutskiy
2007-02-17 16:55 ` [PATCH 14/44 take 2] [UBI] I/O unit implementation Artem Bityutskiy
2007-02-17 16:55 ` [PATCH 15/44 take 2] [UBI] scanning unit header Artem Bityutskiy
2007-02-17 23:07 ` Theodore Tso
2007-02-18 2:17 ` Josh Boyer
2007-02-17 16:55 ` [PATCH 16/44 take 2] [UBI] scanning unit implementation Artem Bityutskiy
2007-02-19 11:05 ` Christoph Hellwig
2007-02-19 14:11 ` Artem Bityutskiy
2007-02-17 16:55 ` [PATCH 17/44 take 2] [UBI] build unit header Artem Bityutskiy
2007-02-17 16:55 ` [PATCH 18/44 take 2] [UBI] build unit implementation Artem Bityutskiy
2007-02-17 16:56 ` Artem Bityutskiy [this message]
2007-02-17 16:56 ` [PATCH 20/44 take 2] [UBI] volume table " Artem Bityutskiy
2007-02-17 16:56 ` [PATCH 21/44 take 2] [UBI] background thread unit header Artem Bityutskiy
2007-02-17 16:56 ` [PATCH 22/44 take 2] [UBI] background thread unit implementation Artem Bityutskiy
2007-02-19 11:09 ` Christoph Hellwig
2007-02-19 13:55 ` Artem Bityutskiy
2007-02-17 16:56 ` [PATCH 23/44 take 2] [UBI] wear-leveling unit header Artem Bityutskiy
2007-02-17 16:56 ` [PATCH 24/44 take 2] [UBI] wear-leveling unit implementation Artem Bityutskiy
2007-02-17 16:56 ` [PATCH 25/44 take 2] [UBI] EBA unit header Artem Bityutskiy
2007-02-17 16:56 ` [PATCH 26/44 take 2] [UBI] EBA unit implementation Artem Bityutskiy
2007-02-17 16:56 ` [PATCH 27/44 take 2] [UBI] bad block handling unit header Artem Bityutskiy
2007-02-17 16:56 ` [PATCH 28/44 take 2] [UBI] bad block handling unit implementation Artem Bityutskiy
2007-02-17 16:56 ` [PATCH 29/44 take 2] [UBI] update unit header Artem Bityutskiy
2007-02-17 16:56 ` [PATCH 30/44 take 2] [UBI] update unit implementation Artem Bityutskiy
2007-02-17 16:57 ` [PATCH 31/44 take 2] [UBI] accounting unit header Artem Bityutskiy
2007-02-17 16:57 ` [PATCH 32/44 take 2] [UBI] accounting unit implementation Artem Bityutskiy
2007-02-17 16:57 ` [PATCH 33/44 take 2] [UBI] volume management unit header Artem Bityutskiy
2007-02-17 16:57 ` [PATCH 34/44 take 2] [UBI] volume management unit implementation Artem Bityutskiy
2007-02-17 16:57 ` [PATCH 35/44 take 2] [UBI] user-interfaces unit header Artem Bityutskiy
2007-02-17 16:57 ` [PATCH 36/44 take 2] [UBI] user-interfaces unit implementation Artem Bityutskiy
2007-02-17 16:57 ` [PATCH 37/44 take 2] [UBI] sysfs handling unit header Artem Bityutskiy
2007-02-17 16:57 ` [PATCH 38/44 take 2] [UBI] sysfs handling unit implementation Artem Bityutskiy
2007-02-17 16:57 ` [PATCH 39/44 take 2] [UBI] character devices handling sub-unit header Artem Bityutskiy
2007-02-17 16:57 ` [PATCH 40/44 take 2] [UBI] character devices handling sub-unit implementation Artem Bityutskiy
2007-02-17 16:57 ` [PATCH 41/44 take 2] [UBI] gluebi unit header Artem Bityutskiy
2007-02-17 21:14 ` Arnd Bergmann
2007-02-18 2:04 ` Josh Boyer
2007-02-18 2:15 ` Arnd Bergmann
2007-02-18 3:02 ` Josh Boyer
2007-02-18 22:37 ` Arnd Bergmann
2007-02-19 13:52 ` Artem Bityutskiy
2007-02-19 14:01 ` Josh Boyer
2007-02-19 14:07 ` Jörn Engel
2007-02-19 12:29 ` Christoph Hellwig
2007-02-19 13:30 ` Artem Bityutskiy
2007-02-17 16:57 ` [PATCH 42/44 take 2] [UBI] gluebi unit implementation Artem Bityutskiy
2007-02-17 16:58 ` [PATCH 43/44 take 2] [UBI] JFFS2 UBI support Artem Bityutskiy
2007-02-17 16:58 ` [PATCH 44/44 take 2] [UBI] update MAINTAINERS Artem Bityutskiy
2007-02-17 22:49 ` [PATCH 00/44 take 2] [UBI] Unsorted Block Images Theodore Tso
2007-02-19 12:48 ` Artem Bityutskiy
2007-02-19 14:33 ` Theodore Tso
2007-02-19 17:07 ` Artem Bityutskiy
2007-02-19 23:34 ` Theodore Tso
2007-02-20 11:54 ` Artem Bityutskiy
2007-02-25 5:51 ` Christoph Hellwig
2007-02-26 10:11 ` Artem Bityutskiy
2007-02-19 10:50 ` Christoph Hellwig
2007-02-19 17:44 ` Artem Bityutskiy
2007-02-25 5:55 ` Christoph Hellwig
2007-02-20 14:52 ` John Stoffel
2007-02-20 17:41 ` Artem Bityutskiy
2007-02-20 17:44 ` Josh Boyer
2007-02-25 5:48 ` Christoph Hellwig
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20070217165600.5845.37863.sendpatchset@localhost.localdomain \
--to=dedekind@infradead.org \
--cc=dwmw2@infradead.org \
--cc=haver@vnet.ibm.com \
--cc=hch@infradead.org \
--cc=jwboyer@linux.vnet.ibm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=tglx@linutronix.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.