linux/drivers/mtd/inftlmount.c

818 lines
23 KiB
C
Raw Normal View History

/*
* inftlmount.c -- INFTL mount code with extensive checks.
*
* Author: Greg Ungerer (gerg@snapgear.com)
* (C) Copyright 2002-2003, Greg Ungerer (gerg@snapgear.com)
*
* Based heavily on the nftlmount.c code which is:
* Author: Fabrice Bellard (fabrice.bellard@netgem.com)
* Copyright (C) 2000 Netgem S.A.
*
* $Id: inftlmount.c,v 1.18 2005/11/07 11:14:20 gleixner Exp $
*
* 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
*/
#include <linux/kernel.h>
#include <linux/module.h>
#include <asm/errno.h>
#include <asm/io.h>
#include <asm/uaccess.h>
#include <linux/miscdevice.h>
#include <linux/delay.h>
#include <linux/slab.h>
#include <linux/init.h>
#include <linux/mtd/mtd.h>
#include <linux/mtd/nftl.h>
#include <linux/mtd/inftl.h>
#include <linux/mtd/compatmac.h>
char inftlmountrev[]="$Revision: 1.18 $";
[MTD] Rework the out of band handling completely Hopefully the last iteration on this! The handling of out of band data on NAND was accompanied by tons of fruitless discussions and halfarsed patches to make it work for a particular problem. Sufficiently annoyed by I all those "I know it better" mails and the resonable amount of discarded "it solves my problem" patches, I finally decided to go for the big rework. After removing the _ecc variants of mtd read/write functions the solution to satisfy the various requirements was to refactor the read/write _oob functions in mtd. The major change is that read/write_oob now takes a pointer to an operation descriptor structure "struct mtd_oob_ops".instead of having a function with at least seven arguments. read/write_oob which should probably renamed to a more descriptive name, can do the following tasks: - read/write out of band data - read/write data content and out of band data - read/write raw data content and out of band data (ecc disabled) struct mtd_oob_ops has a mode field, which determines the oob handling mode. Aside of the MTD_OOB_RAW mode, which is intended to be especially for diagnostic purposes and some internal functions e.g. bad block table creation, the other two modes are for mtd clients: MTD_OOB_PLACE puts/gets the given oob data exactly to/from the place which is described by the ooboffs and ooblen fields of the mtd_oob_ops strcuture. It's up to the caller to make sure that the byte positions are not used by the ECC placement algorithms. MTD_OOB_AUTO puts/gets the given oob data automaticaly to/from the places in the out of band area which are described by the oobfree tuples in the ecclayout data structre which is associated to the devicee. The decision whether data plus oob or oob only handling is done depends on the setting of the datbuf member of the data structure. When datbuf == NULL then the internal read/write_oob functions are selected, otherwise the read/write data routines are invoked. Tested on a few platforms with all variants. Please be aware of possible regressions for your particular device / application scenario Disclaimer: Any whining will be ignored from those who just contributed "hot air blurb" and never sat down to tackle the underlying problem of the mess in the NAND driver grown over time and the big chunk of work to fix up the existing users. The problem was not the holiness of the existing MTD interfaces. The problems was the lack of time to go for the big overhaul. It's easy to add more mess to the existing one, but it takes alot of effort to go for a real solution. Improvements and bugfixes are welcome! Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2006-05-29 03:26:58 +02:00
extern int inftl_read_oob(struct mtd_info *mtd, loff_t offs, size_t len,
size_t *retlen, uint8_t *buf);
extern int inftl_write_oob(struct mtd_info *mtd, loff_t offs, size_t len,
size_t *retlen, uint8_t *buf);
/*
* find_boot_record: Find the INFTL Media Header and its Spare copy which
* contains the various device information of the INFTL partition and
* Bad Unit Table. Update the PUtable[] table according to the Bad
* Unit Table. PUtable[] is used for management of Erase Unit in
* other routines in inftlcore.c and inftlmount.c.
*/
static int find_boot_record(struct INFTLrecord *inftl)
{
struct inftl_unittail h1;
//struct inftl_oob oob;
unsigned int i, block;
u8 buf[SECTORSIZE];
struct INFTLMediaHeader *mh = &inftl->MediaHdr;
struct mtd_info *mtd = inftl->mbd.mtd;
struct INFTLPartition *ip;
size_t retlen;
DEBUG(MTD_DEBUG_LEVEL3, "INFTL: find_boot_record(inftl=%p)\n", inftl);
/*
* Assume logical EraseSize == physical erasesize for starting the
* scan. We'll sort it out later if we find a MediaHeader which says
* otherwise.
*/
inftl->EraseSize = inftl->mbd.mtd->erasesize;
inftl->nb_blocks = inftl->mbd.mtd->size / inftl->EraseSize;
inftl->MediaUnit = BLOCK_NIL;
/* Search for a valid boot record */
for (block = 0; block < inftl->nb_blocks; block++) {
int ret;
/*
* Check for BNAND header first. Then whinge if it's found
* but later checks fail.
*/
ret = mtd->read(mtd, block * inftl->EraseSize,
SECTORSIZE, &retlen, buf);
/* We ignore ret in case the ECC of the MediaHeader is invalid
(which is apparently acceptable) */
if (retlen != SECTORSIZE) {
static int warncount = 5;
if (warncount) {
printk(KERN_WARNING "INFTL: block read at 0x%x "
"of mtd%d failed: %d\n",
block * inftl->EraseSize,
inftl->mbd.mtd->index, ret);
if (!--warncount)
printk(KERN_WARNING "INFTL: further "
"failures for this block will "
"not be printed\n");
}
continue;
}
if (retlen < 6 || memcmp(buf, "BNAND", 6)) {
/* BNAND\0 not found. Continue */
continue;
}
/* To be safer with BIOS, also use erase mark as discriminant */
[MTD] Rework the out of band handling completely Hopefully the last iteration on this! The handling of out of band data on NAND was accompanied by tons of fruitless discussions and halfarsed patches to make it work for a particular problem. Sufficiently annoyed by I all those "I know it better" mails and the resonable amount of discarded "it solves my problem" patches, I finally decided to go for the big rework. After removing the _ecc variants of mtd read/write functions the solution to satisfy the various requirements was to refactor the read/write _oob functions in mtd. The major change is that read/write_oob now takes a pointer to an operation descriptor structure "struct mtd_oob_ops".instead of having a function with at least seven arguments. read/write_oob which should probably renamed to a more descriptive name, can do the following tasks: - read/write out of band data - read/write data content and out of band data - read/write raw data content and out of band data (ecc disabled) struct mtd_oob_ops has a mode field, which determines the oob handling mode. Aside of the MTD_OOB_RAW mode, which is intended to be especially for diagnostic purposes and some internal functions e.g. bad block table creation, the other two modes are for mtd clients: MTD_OOB_PLACE puts/gets the given oob data exactly to/from the place which is described by the ooboffs and ooblen fields of the mtd_oob_ops strcuture. It's up to the caller to make sure that the byte positions are not used by the ECC placement algorithms. MTD_OOB_AUTO puts/gets the given oob data automaticaly to/from the places in the out of band area which are described by the oobfree tuples in the ecclayout data structre which is associated to the devicee. The decision whether data plus oob or oob only handling is done depends on the setting of the datbuf member of the data structure. When datbuf == NULL then the internal read/write_oob functions are selected, otherwise the read/write data routines are invoked. Tested on a few platforms with all variants. Please be aware of possible regressions for your particular device / application scenario Disclaimer: Any whining will be ignored from those who just contributed "hot air blurb" and never sat down to tackle the underlying problem of the mess in the NAND driver grown over time and the big chunk of work to fix up the existing users. The problem was not the holiness of the existing MTD interfaces. The problems was the lack of time to go for the big overhaul. It's easy to add more mess to the existing one, but it takes alot of effort to go for a real solution. Improvements and bugfixes are welcome! Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2006-05-29 03:26:58 +02:00
if ((ret = inftl_read_oob(mtd, block * inftl->EraseSize +
SECTORSIZE + 8, 8, &retlen,
(char *)&h1) < 0)) {
printk(KERN_WARNING "INFTL: ANAND header found at "
"0x%x in mtd%d, but OOB data read failed "
"(err %d)\n", block * inftl->EraseSize,
inftl->mbd.mtd->index, ret);
continue;
}
/*
* This is the first we've seen.
* Copy the media header structure into place.
*/
memcpy(mh, buf, sizeof(struct INFTLMediaHeader));
/* Read the spare media header at offset 4096 */
mtd->read(mtd, block * inftl->EraseSize + 4096,
SECTORSIZE, &retlen, buf);
if (retlen != SECTORSIZE) {
printk(KERN_WARNING "INFTL: Unable to read spare "
"Media Header\n");
return -1;
}
/* Check if this one is the same as the first one we found. */
if (memcmp(mh, buf, sizeof(struct INFTLMediaHeader))) {
printk(KERN_WARNING "INFTL: Primary and spare Media "
"Headers disagree.\n");
return -1;
}
mh->NoOfBootImageBlocks = le32_to_cpu(mh->NoOfBootImageBlocks);
mh->NoOfBinaryPartitions = le32_to_cpu(mh->NoOfBinaryPartitions);
mh->NoOfBDTLPartitions = le32_to_cpu(mh->NoOfBDTLPartitions);
mh->BlockMultiplierBits = le32_to_cpu(mh->BlockMultiplierBits);
mh->FormatFlags = le32_to_cpu(mh->FormatFlags);
mh->PercentUsed = le32_to_cpu(mh->PercentUsed);
#ifdef CONFIG_MTD_DEBUG_VERBOSE
if (CONFIG_MTD_DEBUG_VERBOSE >= 2) {
printk("INFTL: Media Header ->\n"
" bootRecordID = %s\n"
" NoOfBootImageBlocks = %d\n"
" NoOfBinaryPartitions = %d\n"
" NoOfBDTLPartitions = %d\n"
" BlockMultiplerBits = %d\n"
" FormatFlgs = %d\n"
" OsakVersion = 0x%x\n"
" PercentUsed = %d\n",
mh->bootRecordID, mh->NoOfBootImageBlocks,
mh->NoOfBinaryPartitions,
mh->NoOfBDTLPartitions,
mh->BlockMultiplierBits, mh->FormatFlags,
mh->OsakVersion, mh->PercentUsed);
}
#endif
if (mh->NoOfBDTLPartitions == 0) {
printk(KERN_WARNING "INFTL: Media Header sanity check "
"failed: NoOfBDTLPartitions (%d) == 0, "
"must be at least 1\n", mh->NoOfBDTLPartitions);
return -1;
}
if ((mh->NoOfBDTLPartitions + mh->NoOfBinaryPartitions) > 4) {
printk(KERN_WARNING "INFTL: Media Header sanity check "
"failed: Total Partitions (%d) > 4, "
"BDTL=%d Binary=%d\n", mh->NoOfBDTLPartitions +
mh->NoOfBinaryPartitions,
mh->NoOfBDTLPartitions,
mh->NoOfBinaryPartitions);
return -1;
}
if (mh->BlockMultiplierBits > 1) {
printk(KERN_WARNING "INFTL: sorry, we don't support "
"UnitSizeFactor 0x%02x\n",
mh->BlockMultiplierBits);
return -1;
} else if (mh->BlockMultiplierBits == 1) {
printk(KERN_WARNING "INFTL: support for INFTL with "
"UnitSizeFactor 0x%02x is experimental\n",
mh->BlockMultiplierBits);
inftl->EraseSize = inftl->mbd.mtd->erasesize <<
mh->BlockMultiplierBits;
inftl->nb_blocks = inftl->mbd.mtd->size / inftl->EraseSize;
block >>= mh->BlockMultiplierBits;
}
/* Scan the partitions */
for (i = 0; (i < 4); i++) {
ip = &mh->Partitions[i];
ip->virtualUnits = le32_to_cpu(ip->virtualUnits);
ip->firstUnit = le32_to_cpu(ip->firstUnit);
ip->lastUnit = le32_to_cpu(ip->lastUnit);
ip->flags = le32_to_cpu(ip->flags);
ip->spareUnits = le32_to_cpu(ip->spareUnits);
ip->Reserved0 = le32_to_cpu(ip->Reserved0);
#ifdef CONFIG_MTD_DEBUG_VERBOSE
if (CONFIG_MTD_DEBUG_VERBOSE >= 2) {
printk(" PARTITION[%d] ->\n"
" virtualUnits = %d\n"
" firstUnit = %d\n"
" lastUnit = %d\n"
" flags = 0x%x\n"
" spareUnits = %d\n",
i, ip->virtualUnits, ip->firstUnit,
ip->lastUnit, ip->flags,
ip->spareUnits);
}
#endif
if (ip->Reserved0 != ip->firstUnit) {
struct erase_info *instr = &inftl->instr;
instr->mtd = inftl->mbd.mtd;
/*
* Most likely this is using the
* undocumented qiuck mount feature.
* We don't support that, we will need
* to erase the hidden block for full
* compatibility.
*/
instr->addr = ip->Reserved0 * inftl->EraseSize;
instr->len = inftl->EraseSize;
mtd->erase(mtd, instr);
}
if ((ip->lastUnit - ip->firstUnit + 1) < ip->virtualUnits) {
printk(KERN_WARNING "INFTL: Media Header "
"Partition %d sanity check failed\n"
" firstUnit %d : lastUnit %d > "
"virtualUnits %d\n", i, ip->lastUnit,
ip->firstUnit, ip->Reserved0);
return -1;
}
if (ip->Reserved1 != 0) {
printk(KERN_WARNING "INFTL: Media Header "
"Partition %d sanity check failed: "
"Reserved1 %d != 0\n",
i, ip->Reserved1);
return -1;
}
if (ip->flags & INFTL_BDTL)
break;
}
if (i >= 4) {
printk(KERN_WARNING "INFTL: Media Header Partition "
"sanity check failed:\n No partition "
"marked as Disk Partition\n");
return -1;
}
inftl->nb_boot_blocks = ip->firstUnit;
inftl->numvunits = ip->virtualUnits;
if (inftl->numvunits > (inftl->nb_blocks -
inftl->nb_boot_blocks - 2)) {
printk(KERN_WARNING "INFTL: Media Header sanity check "
"failed:\n numvunits (%d) > nb_blocks "
"(%d) - nb_boot_blocks(%d) - 2\n",
inftl->numvunits, inftl->nb_blocks,
inftl->nb_boot_blocks);
return -1;
}
inftl->mbd.size = inftl->numvunits *
(inftl->EraseSize / SECTORSIZE);
/*
* Block count is set to last used EUN (we won't need to keep
* any meta-data past that point).
*/
inftl->firstEUN = ip->firstUnit;
inftl->lastEUN = ip->lastUnit;
inftl->nb_blocks = ip->lastUnit + 1;
/* Memory alloc */
inftl->PUtable = kmalloc(inftl->nb_blocks * sizeof(u16), GFP_KERNEL);
if (!inftl->PUtable) {
printk(KERN_WARNING "INFTL: allocation of PUtable "
"failed (%zd bytes)\n",
inftl->nb_blocks * sizeof(u16));
return -ENOMEM;
}
inftl->VUtable = kmalloc(inftl->nb_blocks * sizeof(u16), GFP_KERNEL);
if (!inftl->VUtable) {
kfree(inftl->PUtable);
printk(KERN_WARNING "INFTL: allocation of VUtable "
"failed (%zd bytes)\n",
inftl->nb_blocks * sizeof(u16));
return -ENOMEM;
}
/* Mark the blocks before INFTL MediaHeader as reserved */
for (i = 0; i < inftl->nb_boot_blocks; i++)
inftl->PUtable[i] = BLOCK_RESERVED;
/* Mark all remaining blocks as potentially containing data */
for (; i < inftl->nb_blocks; i++)
inftl->PUtable[i] = BLOCK_NOTEXPLORED;
/* Mark this boot record (NFTL MediaHeader) block as reserved */
inftl->PUtable[block] = BLOCK_RESERVED;
/* Read Bad Erase Unit Table and modify PUtable[] accordingly */
for (i = 0; i < inftl->nb_blocks; i++) {
int physblock;
/* If any of the physical eraseblocks are bad, don't
use the unit. */
for (physblock = 0; physblock < inftl->EraseSize; physblock += inftl->mbd.mtd->erasesize) {
if (inftl->mbd.mtd->block_isbad(inftl->mbd.mtd, i * inftl->EraseSize + physblock))
inftl->PUtable[i] = BLOCK_RESERVED;
}
}
inftl->MediaUnit = block;
return 0;
}
/* Not found. */
return -1;
}
static int memcmpb(void *a, int c, int n)
{
int i;
for (i = 0; i < n; i++) {
if (c != ((unsigned char *)a)[i])
return 1;
}
return 0;
}
/*
* check_free_sector: check if a free sector is actually FREE,
* i.e. All 0xff in data and oob area.
*/
static int check_free_sectors(struct INFTLrecord *inftl, unsigned int address,
int len, int check_oob)
{
u8 buf[SECTORSIZE + inftl->mbd.mtd->oobsize];
struct mtd_info *mtd = inftl->mbd.mtd;
size_t retlen;
int i;
for (i = 0; i < len; i += SECTORSIZE) {
if (mtd->read(mtd, address, SECTORSIZE, &retlen, buf))
return -1;
if (memcmpb(buf, 0xff, SECTORSIZE) != 0)
return -1;
if (check_oob) {
[MTD] Rework the out of band handling completely Hopefully the last iteration on this! The handling of out of band data on NAND was accompanied by tons of fruitless discussions and halfarsed patches to make it work for a particular problem. Sufficiently annoyed by I all those "I know it better" mails and the resonable amount of discarded "it solves my problem" patches, I finally decided to go for the big rework. After removing the _ecc variants of mtd read/write functions the solution to satisfy the various requirements was to refactor the read/write _oob functions in mtd. The major change is that read/write_oob now takes a pointer to an operation descriptor structure "struct mtd_oob_ops".instead of having a function with at least seven arguments. read/write_oob which should probably renamed to a more descriptive name, can do the following tasks: - read/write out of band data - read/write data content and out of band data - read/write raw data content and out of band data (ecc disabled) struct mtd_oob_ops has a mode field, which determines the oob handling mode. Aside of the MTD_OOB_RAW mode, which is intended to be especially for diagnostic purposes and some internal functions e.g. bad block table creation, the other two modes are for mtd clients: MTD_OOB_PLACE puts/gets the given oob data exactly to/from the place which is described by the ooboffs and ooblen fields of the mtd_oob_ops strcuture. It's up to the caller to make sure that the byte positions are not used by the ECC placement algorithms. MTD_OOB_AUTO puts/gets the given oob data automaticaly to/from the places in the out of band area which are described by the oobfree tuples in the ecclayout data structre which is associated to the devicee. The decision whether data plus oob or oob only handling is done depends on the setting of the datbuf member of the data structure. When datbuf == NULL then the internal read/write_oob functions are selected, otherwise the read/write data routines are invoked. Tested on a few platforms with all variants. Please be aware of possible regressions for your particular device / application scenario Disclaimer: Any whining will be ignored from those who just contributed "hot air blurb" and never sat down to tackle the underlying problem of the mess in the NAND driver grown over time and the big chunk of work to fix up the existing users. The problem was not the holiness of the existing MTD interfaces. The problems was the lack of time to go for the big overhaul. It's easy to add more mess to the existing one, but it takes alot of effort to go for a real solution. Improvements and bugfixes are welcome! Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2006-05-29 03:26:58 +02:00
if(inftl_read_oob(mtd, address, mtd->oobsize,
&retlen, &buf[SECTORSIZE]) < 0)
return -1;
if (memcmpb(buf + SECTORSIZE, 0xff, mtd->oobsize) != 0)
return -1;
}
address += SECTORSIZE;
}
return 0;
}
/*
* INFTL_format: format a Erase Unit by erasing ALL Erase Zones in the Erase
* Unit and Update INFTL metadata. Each erase operation is
* checked with check_free_sectors.
*
* Return: 0 when succeed, -1 on error.
*
* ToDo: 1. Is it neceressary to check_free_sector after erasing ??
*/
int INFTL_formatblock(struct INFTLrecord *inftl, int block)
{
size_t retlen;
struct inftl_unittail uci;
struct erase_info *instr = &inftl->instr;
struct mtd_info *mtd = inftl->mbd.mtd;
int physblock;
DEBUG(MTD_DEBUG_LEVEL3, "INFTL: INFTL_formatblock(inftl=%p,"
"block=%d)\n", inftl, block);
memset(instr, 0, sizeof(struct erase_info));
/* FIXME: Shouldn't we be setting the 'discarded' flag to zero
_first_? */
/* Use async erase interface, test return code */
instr->mtd = inftl->mbd.mtd;
instr->addr = block * inftl->EraseSize;
instr->len = inftl->mbd.mtd->erasesize;
/* Erase one physical eraseblock at a time, even though the NAND api
allows us to group them. This way we if we have a failure, we can
mark only the failed block in the bbt. */
for (physblock = 0; physblock < inftl->EraseSize;
physblock += instr->len, instr->addr += instr->len) {
mtd->erase(inftl->mbd.mtd, instr);
if (instr->state == MTD_ERASE_FAILED) {
printk(KERN_WARNING "INFTL: error while formatting block %d\n",
block);
goto fail;
}
/*
* Check the "freeness" of Erase Unit before updating metadata.
* FixMe: is this check really necessary? Since we have check
* the return code after the erase operation.
*/
if (check_free_sectors(inftl, instr->addr, instr->len, 1) != 0)
goto fail;
}
uci.EraseMark = cpu_to_le16(ERASE_MARK);
uci.EraseMark1 = cpu_to_le16(ERASE_MARK);
uci.Reserved[0] = 0;
uci.Reserved[1] = 0;
uci.Reserved[2] = 0;
uci.Reserved[3] = 0;
instr->addr = block * inftl->EraseSize + SECTORSIZE * 2;
[MTD] Rework the out of band handling completely Hopefully the last iteration on this! The handling of out of band data on NAND was accompanied by tons of fruitless discussions and halfarsed patches to make it work for a particular problem. Sufficiently annoyed by I all those "I know it better" mails and the resonable amount of discarded "it solves my problem" patches, I finally decided to go for the big rework. After removing the _ecc variants of mtd read/write functions the solution to satisfy the various requirements was to refactor the read/write _oob functions in mtd. The major change is that read/write_oob now takes a pointer to an operation descriptor structure "struct mtd_oob_ops".instead of having a function with at least seven arguments. read/write_oob which should probably renamed to a more descriptive name, can do the following tasks: - read/write out of band data - read/write data content and out of band data - read/write raw data content and out of band data (ecc disabled) struct mtd_oob_ops has a mode field, which determines the oob handling mode. Aside of the MTD_OOB_RAW mode, which is intended to be especially for diagnostic purposes and some internal functions e.g. bad block table creation, the other two modes are for mtd clients: MTD_OOB_PLACE puts/gets the given oob data exactly to/from the place which is described by the ooboffs and ooblen fields of the mtd_oob_ops strcuture. It's up to the caller to make sure that the byte positions are not used by the ECC placement algorithms. MTD_OOB_AUTO puts/gets the given oob data automaticaly to/from the places in the out of band area which are described by the oobfree tuples in the ecclayout data structre which is associated to the devicee. The decision whether data plus oob or oob only handling is done depends on the setting of the datbuf member of the data structure. When datbuf == NULL then the internal read/write_oob functions are selected, otherwise the read/write data routines are invoked. Tested on a few platforms with all variants. Please be aware of possible regressions for your particular device / application scenario Disclaimer: Any whining will be ignored from those who just contributed "hot air blurb" and never sat down to tackle the underlying problem of the mess in the NAND driver grown over time and the big chunk of work to fix up the existing users. The problem was not the holiness of the existing MTD interfaces. The problems was the lack of time to go for the big overhaul. It's easy to add more mess to the existing one, but it takes alot of effort to go for a real solution. Improvements and bugfixes are welcome! Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2006-05-29 03:26:58 +02:00
if (inftl_write_oob(mtd, instr->addr + 8, 8, &retlen, (char *)&uci) < 0)
goto fail;
return 0;
fail:
/* could not format, update the bad block table (caller is responsible
for setting the PUtable to BLOCK_RESERVED on failure) */
inftl->mbd.mtd->block_markbad(inftl->mbd.mtd, instr->addr);
return -1;
}
/*
* format_chain: Format an invalid Virtual Unit chain. It frees all the Erase
* Units in a Virtual Unit Chain, i.e. all the units are disconnected.
*
* Since the chain is invalid then we will have to erase it from its
* head (normally for INFTL we go from the oldest). But if it has a
* loop then there is no oldest...
*/
static void format_chain(struct INFTLrecord *inftl, unsigned int first_block)
{
unsigned int block = first_block, block1;
printk(KERN_WARNING "INFTL: formatting chain at block %d\n",
first_block);
for (;;) {
block1 = inftl->PUtable[block];
printk(KERN_WARNING "INFTL: formatting block %d\n", block);
if (INFTL_formatblock(inftl, block) < 0) {
/*
* Cannot format !!!! Mark it as Bad Unit,
*/
inftl->PUtable[block] = BLOCK_RESERVED;
} else {
inftl->PUtable[block] = BLOCK_FREE;
}
/* Goto next block on the chain */
block = block1;
if (block == BLOCK_NIL || block >= inftl->lastEUN)
break;
}
}
void INFTL_dumptables(struct INFTLrecord *s)
{
int i;
printk("-------------------------------------------"
"----------------------------------\n");
printk("VUtable[%d] ->", s->nb_blocks);
for (i = 0; i < s->nb_blocks; i++) {
if ((i % 8) == 0)
printk("\n%04x: ", i);
printk("%04x ", s->VUtable[i]);
}
printk("\n-------------------------------------------"
"----------------------------------\n");
printk("PUtable[%d-%d=%d] ->", s->firstEUN, s->lastEUN, s->nb_blocks);
for (i = 0; i <= s->lastEUN; i++) {
if ((i % 8) == 0)
printk("\n%04x: ", i);
printk("%04x ", s->PUtable[i]);
}
printk("\n-------------------------------------------"
"----------------------------------\n");
printk("INFTL ->\n"
" EraseSize = %d\n"
" h/s/c = %d/%d/%d\n"
" numvunits = %d\n"
" firstEUN = %d\n"
" lastEUN = %d\n"
" numfreeEUNs = %d\n"
" LastFreeEUN = %d\n"
" nb_blocks = %d\n"
" nb_boot_blocks = %d",
s->EraseSize, s->heads, s->sectors, s->cylinders,
s->numvunits, s->firstEUN, s->lastEUN, s->numfreeEUNs,
s->LastFreeEUN, s->nb_blocks, s->nb_boot_blocks);
printk("\n-------------------------------------------"
"----------------------------------\n");
}
void INFTL_dumpVUchains(struct INFTLrecord *s)
{
int logical, block, i;
printk("-------------------------------------------"
"----------------------------------\n");
printk("INFTL Virtual Unit Chains:\n");
for (logical = 0; logical < s->nb_blocks; logical++) {
block = s->VUtable[logical];
if (block > s->nb_blocks)
continue;
printk(" LOGICAL %d --> %d ", logical, block);
for (i = 0; i < s->nb_blocks; i++) {
if (s->PUtable[block] == BLOCK_NIL)
break;
block = s->PUtable[block];
printk("%d ", block);
}
printk("\n");
}
printk("-------------------------------------------"
"----------------------------------\n");
}
int INFTL_mount(struct INFTLrecord *s)
{
struct mtd_info *mtd = s->mbd.mtd;
unsigned int block, first_block, prev_block, last_block;
unsigned int first_logical_block, logical_block, erase_mark;
int chain_length, do_format_chain;
struct inftl_unithead1 h0;
struct inftl_unittail h1;
size_t retlen;
int i;
u8 *ANACtable, ANAC;
DEBUG(MTD_DEBUG_LEVEL3, "INFTL: INFTL_mount(inftl=%p)\n", s);
/* Search for INFTL MediaHeader and Spare INFTL Media Header */
if (find_boot_record(s) < 0) {
printk(KERN_WARNING "INFTL: could not find valid boot record?\n");
return -ENXIO;
}
/* Init the logical to physical table */
for (i = 0; i < s->nb_blocks; i++)
s->VUtable[i] = BLOCK_NIL;
logical_block = block = BLOCK_NIL;
/* Temporary buffer to store ANAC numbers. */
ANACtable = kcalloc(s->nb_blocks, sizeof(u8), GFP_KERNEL);
if (!ANACtable) {
printk(KERN_WARNING "INFTL: allocation of ANACtable "
"failed (%zd bytes)\n",
s->nb_blocks * sizeof(u8));
return -ENOMEM;
}
/*
* First pass is to explore each physical unit, and construct the
* virtual chains that exist (newest physical unit goes into VUtable).
* Any block that is in any way invalid will be left in the
* NOTEXPLORED state. Then at the end we will try to format it and
* mark it as free.
*/
DEBUG(MTD_DEBUG_LEVEL3, "INFTL: pass 1, explore each unit\n");
for (first_block = s->firstEUN; first_block <= s->lastEUN; first_block++) {
if (s->PUtable[first_block] != BLOCK_NOTEXPLORED)
continue;
do_format_chain = 0;
first_logical_block = BLOCK_NIL;
last_block = BLOCK_NIL;
block = first_block;
for (chain_length = 0; ; chain_length++) {
if ((chain_length == 0) &&
(s->PUtable[block] != BLOCK_NOTEXPLORED)) {
/* Nothing to do here, onto next block */
break;
}
[MTD] Rework the out of band handling completely Hopefully the last iteration on this! The handling of out of band data on NAND was accompanied by tons of fruitless discussions and halfarsed patches to make it work for a particular problem. Sufficiently annoyed by I all those "I know it better" mails and the resonable amount of discarded "it solves my problem" patches, I finally decided to go for the big rework. After removing the _ecc variants of mtd read/write functions the solution to satisfy the various requirements was to refactor the read/write _oob functions in mtd. The major change is that read/write_oob now takes a pointer to an operation descriptor structure "struct mtd_oob_ops".instead of having a function with at least seven arguments. read/write_oob which should probably renamed to a more descriptive name, can do the following tasks: - read/write out of band data - read/write data content and out of band data - read/write raw data content and out of band data (ecc disabled) struct mtd_oob_ops has a mode field, which determines the oob handling mode. Aside of the MTD_OOB_RAW mode, which is intended to be especially for diagnostic purposes and some internal functions e.g. bad block table creation, the other two modes are for mtd clients: MTD_OOB_PLACE puts/gets the given oob data exactly to/from the place which is described by the ooboffs and ooblen fields of the mtd_oob_ops strcuture. It's up to the caller to make sure that the byte positions are not used by the ECC placement algorithms. MTD_OOB_AUTO puts/gets the given oob data automaticaly to/from the places in the out of band area which are described by the oobfree tuples in the ecclayout data structre which is associated to the devicee. The decision whether data plus oob or oob only handling is done depends on the setting of the datbuf member of the data structure. When datbuf == NULL then the internal read/write_oob functions are selected, otherwise the read/write data routines are invoked. Tested on a few platforms with all variants. Please be aware of possible regressions for your particular device / application scenario Disclaimer: Any whining will be ignored from those who just contributed "hot air blurb" and never sat down to tackle the underlying problem of the mess in the NAND driver grown over time and the big chunk of work to fix up the existing users. The problem was not the holiness of the existing MTD interfaces. The problems was the lack of time to go for the big overhaul. It's easy to add more mess to the existing one, but it takes alot of effort to go for a real solution. Improvements and bugfixes are welcome! Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2006-05-29 03:26:58 +02:00
if (inftl_read_oob(mtd, block * s->EraseSize + 8,
8, &retlen, (char *)&h0) < 0 ||
inftl_read_oob(mtd, block * s->EraseSize +
2 * SECTORSIZE + 8, 8, &retlen,
(char *)&h1) < 0) {
/* Should never happen? */
do_format_chain++;
break;
}
logical_block = le16_to_cpu(h0.virtualUnitNo);
prev_block = le16_to_cpu(h0.prevUnitNo);
erase_mark = le16_to_cpu((h1.EraseMark | h1.EraseMark1));
ANACtable[block] = h0.ANAC;
/* Previous block is relative to start of Partition */
if (prev_block < s->nb_blocks)
prev_block += s->firstEUN;
/* Already explored partial chain? */
if (s->PUtable[block] != BLOCK_NOTEXPLORED) {
/* Check if chain for this logical */
if (logical_block == first_logical_block) {
if (last_block != BLOCK_NIL)
s->PUtable[last_block] = block;
}
break;
}
/* Check for invalid block */
if (erase_mark != ERASE_MARK) {
printk(KERN_WARNING "INFTL: corrupt block %d "
"in chain %d, chain length %d, erase "
"mark 0x%x?\n", block, first_block,
chain_length, erase_mark);
/*
* Assume end of chain, probably incomplete
* fold/erase...
*/
if (chain_length == 0)
do_format_chain++;
break;
}
/* Check for it being free already then... */
if ((logical_block == BLOCK_FREE) ||
(logical_block == BLOCK_NIL)) {
s->PUtable[block] = BLOCK_FREE;
break;
}
/* Sanity checks on block numbers */
if ((logical_block >= s->nb_blocks) ||
((prev_block >= s->nb_blocks) &&
(prev_block != BLOCK_NIL))) {
if (chain_length > 0) {
printk(KERN_WARNING "INFTL: corrupt "
"block %d in chain %d?\n",
block, first_block);
do_format_chain++;
}
break;
}
if (first_logical_block == BLOCK_NIL) {
first_logical_block = logical_block;
} else {
if (first_logical_block != logical_block) {
/* Normal for folded chain... */
break;
}
}
/*
* Current block is valid, so if we followed a virtual
* chain to get here then we can set the previous
* block pointer in our PUtable now. Then move onto
* the previous block in the chain.
*/
s->PUtable[block] = BLOCK_NIL;
if (last_block != BLOCK_NIL)
s->PUtable[last_block] = block;
last_block = block;
block = prev_block;
/* Check for end of chain */
if (block == BLOCK_NIL)
break;
/* Validate next block before following it... */
if (block > s->lastEUN) {
printk(KERN_WARNING "INFTL: invalid previous "
"block %d in chain %d?\n", block,
first_block);
do_format_chain++;
break;
}
}
if (do_format_chain) {
format_chain(s, first_block);
continue;
}
/*
* Looks like a valid chain then. It may not really be the
* newest block in the chain, but it is the newest we have
* found so far. We might update it in later iterations of
* this loop if we find something newer.
*/
s->VUtable[first_logical_block] = first_block;
logical_block = BLOCK_NIL;
}
#ifdef CONFIG_MTD_DEBUG_VERBOSE
if (CONFIG_MTD_DEBUG_VERBOSE >= 2)
INFTL_dumptables(s);
#endif
/*
* Second pass, check for infinite loops in chains. These are
* possible because we don't update the previous pointers when
* we fold chains. No big deal, just fix them up in PUtable.
*/
DEBUG(MTD_DEBUG_LEVEL3, "INFTL: pass 2, validate virtual chains\n");
for (logical_block = 0; logical_block < s->numvunits; logical_block++) {
block = s->VUtable[logical_block];
last_block = BLOCK_NIL;
/* Check for free/reserved/nil */
if (block >= BLOCK_RESERVED)
continue;
ANAC = ANACtable[block];
for (i = 0; i < s->numvunits; i++) {
if (s->PUtable[block] == BLOCK_NIL)
break;
if (s->PUtable[block] > s->lastEUN) {
printk(KERN_WARNING "INFTL: invalid prev %d, "
"in virtual chain %d\n",
s->PUtable[block], logical_block);
s->PUtable[block] = BLOCK_NIL;
}
if (ANACtable[block] != ANAC) {
/*
* Chain must point back to itself. This is ok,
* but we will need adjust the tables with this
* newest block and oldest block.
*/
s->VUtable[logical_block] = block;
s->PUtable[last_block] = BLOCK_NIL;
break;
}
ANAC--;
last_block = block;
block = s->PUtable[block];
}
if (i >= s->nb_blocks) {
/*
* Uhoo, infinite chain with valid ANACS!
* Format whole chain...
*/
format_chain(s, first_block);
}
}
#ifdef CONFIG_MTD_DEBUG_VERBOSE
if (CONFIG_MTD_DEBUG_VERBOSE >= 2)
INFTL_dumptables(s);
if (CONFIG_MTD_DEBUG_VERBOSE >= 2)
INFTL_dumpVUchains(s);
#endif
/*
* Third pass, format unreferenced blocks and init free block count.
*/
s->numfreeEUNs = 0;
s->LastFreeEUN = BLOCK_NIL;
DEBUG(MTD_DEBUG_LEVEL3, "INFTL: pass 3, format unused blocks\n");
for (block = s->firstEUN; block <= s->lastEUN; block++) {
if (s->PUtable[block] == BLOCK_NOTEXPLORED) {
printk("INFTL: unreferenced block %d, formatting it\n",
block);
if (INFTL_formatblock(s, block) < 0)
s->PUtable[block] = BLOCK_RESERVED;
else
s->PUtable[block] = BLOCK_FREE;
}
if (s->PUtable[block] == BLOCK_FREE) {
s->numfreeEUNs++;
if (s->LastFreeEUN == BLOCK_NIL)
s->LastFreeEUN = block;
}
}
kfree(ANACtable);
return 0;
}