Make ext2 use cache line aligned buffers for reading from the filesystem.
This is needed when caches are enabled because unaligned cache invalidates
are not safe.
Signed-off-by: Simon Glass <sjg@chromium.org>
The VFAT short alias checksum read from a long file name is only overwritten
when another long file name appears in a directory list. Until then it renders
short file names invisible that have the same checksum. Reset the checksum on
first match.
Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Martin Mueller <martin.mueller5@de.bosch.com>
* 'next' of ../next:
mkenvimage: Add version info switch (-V)
mkenvimage: Fix getopt() error handling
mkenvimage: Fix some typos
phy: add Micrel KS8721BL phy definition
net: introduce per device index
mvgbe: remove setting of ethaddr within the driver
x86: Add support for specifying an initrd with the zboot command
x86: Refactor the zboot innards so they can be reused with a vboot image
x86: Add infrastructure to extract an e820 table from the coreboot tables
x86: Add support for booting Linux using the 32 bit boot protocol
x86: Clean up the x86 zimage code in preparation to extend it
x86: Import code from coreboot's libpayload to parse the coreboot table
x86: Initial commit for running as a coreboot payload
CHECKPATCH: ./board/esd/hh405/logo_320_240_8bpp.c
CHECKPATCH: ./board/esd/hh405/logo_1024_768_8bpp.c
CHECKPATCH: ./board/esd/hh405/logo_320_240_4bpp.c
CHECKPATCH: ./board/esd/hh405/logo_640_480_24bpp.c
CHECKPATCH: ./board/esd/apc405/logo_640_480_24bpp.c
CHECKPATCH: ./board/esd/voh405/logo_320_240_4bpp.c
CHECKPATCH: ./board/esd/voh405/logo_640_480_24bpp.c
CHECKPATCH: ./board/esd/hh405/fpgadata.c
CHECKPATCH: ./board/esd/pci405/fpgadata.c
CHECKPATCH: ./board/esd/tasreg/fpgadata.c
CHECKPATCH: ./board/esd/apc405/fpgadata.c
CHECKPATCH: ./board/esd/voh405/fpgadata.c
CHECKPATCH: ./board/esd/ash405/fpgadata.c
CHECKPATCH: ./board/esd/dasa_sim/fpgadata.c
CHECKPATCH: ./board/esd/ar405/fpgadata_xl30.c
CHECKPATCH: ./board/esd/ar405/fpgadata.c
CHECKPATCH: ./board/esd/plu405/fpgadata.c
CHECKPATCH: ./board/esd/wuh405/fpgadata.c
CHECKPATCH: ./board/esd/cpci405/fpgadata_cpci405.c
CHECKPATCH: ./board/esd/cpci405/fpgadata_cpci405ab.c
CHECKPATCH: ./board/esd/cpci405/fpgadata_cpci4052.c
CHECKPATCH: ./board/esd/canbt/fpgadata.c
CHECKPATCH: ./board/esd/du405/fpgadata.c
CHECKPATCH: ./board/esd/cpciiser4/fpgadata.c
CHECKPATCH: ./board/dave/PPChameleonEVB/fpgadata.c
avr32:mmu.c: fix printf() length modifier
fat.c: fix printf() length modifier
cmd_sf.c: fix printf() length modifier
Make printf and vprintf safe from buffer overruns
vsprintf: Move function documentation into header file
Add safe vsnprintf and snprintf library functions
Move vsprintf functions into their own header
Conflicts:
tools/mkenvimage.c
Signed-off-by: Wolfgang Denk <wd@denx.de>
Writing a file to the FAT partition didn't work while a
test using a CF card. The test was done on mpc5200 based
board (powerpc). There is a number of problems in FAT
write code:
Compiler warning:
fat_write.c: In function 'file_fat_write':
fat_write.c:326: warning: 'counter' may be used uninitialized
in this function
fat_write.c:326: note: 'counter' was declared here
'l_filename' string is not terminated, so a file name
with garbage at the end is used as a file name as shown
by debug code.
Return value of set_contents() is not checked properly
so actually a file won't be written at all (as checked
using 'fatls' after a write attempt with 'fatwrite'
command).
do_fat_write() doesn't return the number of written bytes
if no error happened. However the return value of this
function is used to show the number of written bytes
in do_fat_fswrite().
The patch adds some debug code and fixes above mentioned
problems and also fixes a typo in error output.
NOTE: after a successful write to the FAT partition (under
U-Boot) the partition was checked under Linux using fsck.
The partition needed fixing FATs:
-bash-3.2# fsck -a /dev/sda1
fsck 1.39 (29-May-2006)
dosfsck 2.11, 12 Mar 2005, FAT32, LFN
FATs differ but appear to be intact. Using first FAT.
Performing changes.
Signed-off-by: Anatolij Gustschin <agust@denx.de>
Cc: Donggeun Kim <dg77.kim@samsung.com>
Cc: Aaron Williams <Aaron.Williams@cavium.com>
Acked-by: Donggeun Kim <dg77.kim@samsung.com>
The DIRENTSPERBLOCK utilizes sizeof() which will return a size_t which has no
fixed size. Therefor use correct length modifer for printf() statement to
prevent compiler warnings.
This patch fixes following warning:
---8<---
fat.c: In function 'do_fat_read':
fat.c:879: warning: format '%d' expects type 'int', but argument 4 has type 'long unsigned int'
--->8---
Signed-off-by: Andreas Bießmann <biessmann@corscience.de>
cc: Mike Frysinger <vapier@gentoo.org>
cc: Thomas Chou <thomas@wytron.com.tw>
cc: rjones@nexus-tech.net
cc: kharris@nexus-tech.net
Acked-by: Mike Frysinger <vapier@gentoo.org>
Fix:
yaffs_guts.c: In function 'yaffs_GarbageCollectBlock':
yaffs_guts.c:2761:6: warning: variable 'retVal' set but not used
[-Wunused-but-set-variable]
Here GCC actually detected a bug. The code was always returning OK
instead of the previously set retrun code. Fix that.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: William Juul <wiljuul@cisco.com>
Cc: Scott Wood <scottwood@freescale.com>
Acked-by: William Juul <wiljuul@cisco.com>
Sorry if this is already fixed somewhere - I could not find it.
This fixes the warnings show below.
yaffs_tagscompat.c: In function 'yaffs_TagsCompatabilityReadChunkWithTagsFromNAND':
yaffs_tagscompat.c:151: warning: dereferencing pointer 'tu' does break strict-aliasing rules
yaffs_tagscompat.c:150: warning: dereferencing pointer 'tu' does break strict-aliasing rules
yaffs_tagscompat.c:149: warning: dereferencing pointer 'tu' does break strict-aliasing rules
yaffs_tagscompat.c:148: warning: dereferencing pointer 'tu' does break strict-aliasing rules
yaffs_tagscompat.c:147: warning: dereferencing pointer 'tu' does break strict-aliasing rules
yaffs_tagscompat.c:146: warning: dereferencing pointer 'tu' does break strict-aliasing rules
yaffs_tagscompat.c:145: warning: dereferencing pointer 'tu' does break strict-aliasing rules
yaffs_tagscompat.c:144: warning: dereferencing pointer 'tu' does break strict-aliasing rules
yaffs_tagscompat.c:141: note: initialized from here
Signed-off-by: Simon Glass <sjg@chromium.org>
Fix:
yaffs_guts.c: In function 'yaffs_CheckChunkErased':
yaffs_guts.c:854:6: warning: variable 'result' set but not used
yaffs_guts.c: In function 'yaffs_UpdateObjectHeader':
yaffs_guts.c:3463:6: warning: variable 'result' set but not used
yaffs_guts.c: In function 'yaffs_GrabChunkCache':
yaffs_guts.c:3774:6: warning: variable 'pushout' set but not used
yaffs_guts.c: In function 'yaffs_Scan':
yaffs_guts.c:5237:6: warning: variable 'result' set but not used
yaffs_guts.c: In function 'yaffs_CheckObjectDetailsLoaded':
yaffs_guts.c:5748:6: warning: variable 'alloc_failed' set but not used
yaffs_guts.c:5747:6: warning: variable 'result' set but not used
yaffs_guts.c: In function 'yaffs_ScanBackwards':
yaffs_guts.c:5808:6: warning: variable 'deleted' set but not used
yaffs_guts.c:5806:6: warning: variable 'result' set but not used
yaffs_guts.c: In function 'yaffs_GetObjectName':
yaffs_guts.c:6657:7: warning: variable 'result' set but not used
[-Wunused-but-set-variable]
Signed-off-by: Anatolij Gustschin <agust@denx.de>
Fix:
fat.c: In function 'fat_register_device':
fat.c:74:19: warning: variable 'info' set but not used
[-Wunused-but-set-variable]
Signed-off-by: Wolfgang Denk <wd@denx.de>
yaffs_guts.c: In function 'yaffs_ReadDataFromFile':
yaffs_guts.c:4461:8: warning: 'chunk' may be used uninitialized in this function
yaffs_guts.c:4462:8: warning: 'start' may be used uninitialized in this function
yaffs_guts.c: In function 'yaffs_WriteDataToFile':
yaffs_guts.c:4581:8: warning: 'chunk' may be used uninitialized in this function
yaffs_guts.c:4582:8: warning: 'start' may be used uninitialized in this function
yaffs_guts.c: In function 'yaffs_ResizeFile':
yaffs_guts.c:4816:8: warning: 'newSizeOfPartialChunk' may be used uninitialized
in this function
yaffs_guts.c:4817:8: warning: 'newFullChunks' may be used uninitialized in this
function
Signed-off-by: Marek Vasut <marek.vasut@gmail.com>
Cc: Wolfgang Denk <wd@denx.de>
Cc: Simon Glass <sjg@chromium.org>
Cc: Mike Frysinger <vapier@gentoo.org>
Cc: William Juul <william.juul@tandberg.com>
Drop yaffs_DeleteWorker():
yaffs_guts.c:1556:12: warning: 'yaffs_DeleteWorker' defined but not used
Drop yaffs_VerifyTnodeWorker():
yaffs_guts.c:600:12: warning: 'yaffs_VerifyTnodeWorker' defined but not used
Signed-off-by: Marek Vasut <marek.vasut@gmail.com>
Cc: Wolfgang Denk <wd@denx.de>
Cc: Simon Glass <sjg@chromium.org>
Cc: Mike Frysinger <vapier@gentoo.org>
ATTR_VFAT condition requires multiple bits to be set but the present
condition checking in do_fat_read() & get_dentfromdir() ends up
passing on even a single bit being set.
Signed-off-by: J. Vijayanand <vijayanand.jayaraman@in.bosch.com>
Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Commit c30a15e "FAT: Add FAT write feature" introduced a compiler
warning. Fix this.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Donggeun Kim <dg77.kim@samsung.com>
Cc: Kyungmin Park <kyungmin.park@samsung.com>
In some cases, saving data in RAM as a file with FAT format is required.
This patch allows the file to be written in FAT formatted partition.
The usage is similar with reading a file.
First, fat_register_device function is called before file_fat_write function
in order to set target partition.
Then, file_fat_write function is invoked with desired file name,
start ram address for writing data, and file size.
Signed-off-by: Donggeun Kim <dg77.kim@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Currently, if a device read request is done that does not begin or end
on a sector boundary a stack allocated bounce buffer is used to perform
the read, and then just the part of the sector that is needed is copied
into the users buffer. This stack allocation can mean that the bounce
buffer will not be aligned to the dcache line size. This is a problem
when caches are enabled because unaligned cache invalidates are not
safe.
This patch uses ALLOC_CACHE_ALIGN_BUFFER to create a stack allocated
cache line size aligned bounce buffer.
Signed-off-by: Anton Staaf <robotboy@chromium.org>
Cc: Lukasz Majewski <l.majewski@samsung.com>
Cc: Mike Frysinger <vapier@gentoo.org>
Cc: Dave Liu <r63238@freescale.com>
Cc: Andy Fleming <afleming@gmail.com>
Cc: Albert ARIBAUD <albert.u.boot@aribaud.net>
Change-Id: I32e1594d90ef039137bb219b0f7ced55768744ff
Acked-by: Mike Frysinger <vapier@gentoo.org>
The top level Makefile does not do any recursion into subdirs when
cleaning, so these clean/distclean targets in random arch/board dirs
never get used. Punt them all.
MAKEALL didn't report any errors related to this that I could see.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
This patch fixes an issue when ubifs reads a bad superblock. Later it
tries to free memory, that was not allocated, which freezes u-boot.
This is fixed by looking for a non null pointer before free.
The message I got before u-boot freezes:
UBI: max/mean erase counter: 53/32
UBIFS: mounted UBI device 0, volume 1, name "rootfs"
UBIFS: mounted read-only
UBIFS: file system size: 49140 bytes (50319360 KiB, 0 MiB, 49140 LEBs)
UBIFS: journal size: 49 bytes (6838272 KiB, 0 MiB, 6678 LEBs)
UBIFS: media format: w4/r0 (latest is w4/r0)
UBIFS: default compressor: LZO
UBIFS: reserved for root: 0 bytes (0 KiB)
UBIFS error (pid 0): ubifs_read_node: bad node type (255 but expected 9)
UBIFS error (pid 0): ubifs_read_node: bad node at LEB 330:13104
UBIFS error (pid 0): ubifs_iget: failed to read inode 1, error -22
Error reading superblock on volume 'ubi:rootfs'!
Signed-off-by: Lars Poeschel <larsi@wh2.tu-dresden.de>
Cc: Kyungmin Park <kmpark@infradead.org>
Signed-off-by: Stefan Roese <sr@denx.de>
Fix:
jffs2_1pass.c: In function 'jffs2_1pass_read_inode':
jffs2_1pass.c:699:7: warning: variable 'ret' set but not used [-Wunused-but-set-variable]
jffs2_1pass.c: In function 'jffs2_1pass_build_lists':
jffs2_1pass.c:1578:14: warning: variable 'empty_start' set but not used [-Wunused-but-set-variable]
Signed-off-by: Wolfgang Denk <wd@denx.de>
Currently in do_fat_read() when reading FAT sectors, we have to divide down
LINEAR_PREFETCH_SIZE by the sector size, whereas it's defined as 2 sectors
worth of bytes. In order to avoid redundant multiplication/division, introduce
#define PREFETCH_BLOCKS instead of #define LINEAR_PREFETCH_SIZE.
Signed-off-by: Sergei Shtylyov <sshtylyov@mvista.com>
The root directory cluster field only exists in a FAT32 boot sector, so the
'root_cluster' variable in do_fat_read() contains garbage in case of FAT12/16.
Make it contain 0 instead as this is what is passed to get_vfatname() in that
case anyway.
Signed-off-by: Sergei Shtylyov <sshtylyov@mvista.com>
The code multiples the FAT size in sectors by the sector size and then tries to
compare that to the number of sectors in the 'getsize' variable. While fixing
this, also change the initial value of 'getsize' as the division of FATBUFSIZE
by the sector size gets us FATBUFBLOCKS.
Signed-off-by: Sergei Shtylyov <sshtylyov@mvista.com>
Apple iPod nanos have sector sizes of 2 or 4 KiB, which crashes U-Boot when it
tries to read the boot sector into 512-byte buffer situated on stack. Make the
FAT code indifferent to the sector size.
Signed-off-by: Sergei Shtylyov <sshtylyov@mvista.com>
Commit 46d7274 "UBIFS: Change ubifsload to set the filesize variable"
introduced the follwing compiler warning:
ubifs.c: In function 'ubifs_load':
ubifs.c:742: warning: format '%lX' expects type 'long unsigned int', but argument 3 has type 'u32'
Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Bastian Ruppert <Bastian.Ruppert@Sewerin.de>
This is the same behaviour like tftp or fatload command.
Signed-off-by: Bastian Ruppert <Bastian.Ruppert@Sewerin.de>
CC: kmpark@infradead.org
Acked-by: Detlev Zundel <dzu@denx.de>
Fix these:
yaffs_guts.c: In function 'yaffs_ReadDataFromFile':
yaffs_guts.c:4622: warning: pointer targets in passing argument 3 of 'yaffs_AddrToChunk' differ in signedness
yaffs_guts.c:4622: warning: pointer targets in passing argument 4 of 'yaffs_AddrToChunk' differ in signedness
yaffs_guts.c: In function 'yaffs_WriteDataToFile':
yaffs_guts.c:4745: warning: pointer targets in passing argument 3 of 'yaffs_AddrToChunk' differ in signedness
yaffs_guts.c:4745: warning: pointer targets in passing argument 4 of 'yaffs_AddrToChunk' differ in signedness
yaffs_guts.c: In function 'yaffs_ResizeFile':
yaffs_guts.c:4968: warning: pointer targets in passing argument 3 of 'yaffs_AddrToChunk' differ in signedness
yaffs_guts.c:4968: warning: pointer targets in passing argument 4 of 'yaffs_AddrToChunk' differ in signedness
yaffs_guts.c: In function 'yaffs_GutsInitialise':
yaffs_guts.c:7235: warning: assignment from incompatible pointer type
yaffs_guts.c: In function 'yaffs_CreateNewObject':
yaffs_guts.c:2143: warning: 'tn' may be used uninitialized in this function
yaffs_guts.c: In function 'yaffs_MknodObject':
yaffs_guts.c:2258: warning: 'str' may be used uninitialized in this function
Signed-off-by: Wolfgang Denk <wd@denx.de>
Fix these:
yaffs_guts.c: At top level:
yaffs_guts.c:400: warning: 'yaffs_SkipFullVerification' defined but not used
Testing shows no changes of the image sizes.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Fix these:
yaffs_guts.c: In function 'yaffs_Scan':
yaffs_guts.c:5436: warning: pointer targets in passing argument 4 of 'yaffs_QueryInitialBlockState' differ in signedness
yaffs_guts.c: In function 'yaffs_ScanBackwards':
yaffs_guts.c:6017: warning: pointer targets in passing argument 4 of 'yaffs_QueryInitialBlockState' differ in signedness
yaffs_nand.c: In function 'yaffs_QueryInitialBlockState':
yaffs_nand.c:109: warning: pointer targets in passing argument 4 of 'dev->queryNANDBlock' differ in signedness
yaffs_nand.c:113: warning: pointer targets in passing argument 4 of 'yaffs_TagsCompatabilityQueryNANDBlock' differ in signedness
Signed-off-by: Wolfgang Denk <wd@denx.de>
Drop the "-DNO_Y_INLINE" setting to fix these:
yaffs_guts.h:806: warning: 'yaffs_GetBlockInfo' defined but not used
Impact on image size is negligible - for the VCMA9 board the text
segment size grew from 496353 to 496357 bytes (i. e. 0.0008%);
total image size even remained constant.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Fix these:
yaffscfg.c: In function 'cmd_yaffs_mread_file':
yaffscfg.c:316: warning: format '%08x' expects type 'unsigned int', but argument 3 has type 'char *'
yaffscfg.c: In function 'cmd_yaffs_ls': yaffscfg.c:371: warning: format '%7d' expects type 'int', but argument 3 has type 'off_t'
Signed-off-by: Wolfgang Denk <wd@denx.de>
Free private_data member element before freeing file structure.
This was causing malloc to crash. Also remove unnecessary variable
assigments as file structure gets free'd as well.
Signed-off-by: Rod Boyce <uboot@teamboyce.co.uk>
Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Stefan Roese <sr@denx.de>
Signed-off-by: Stefan Roese <sr@denx.de>
There was a mix of UTF-8 and ISO-8859 files in the U-Boot source
tree, which could cause issues with the patchwork review system.
This commit converts all ISO-8859 files to UTF-8.
Signed-off-by: Albert ARIBAUD <albert.u.boot@aribaud.net>
Previously reading or writing zero full sectors (reading the end of
one sector and the beginning of the next for example) was special
cased and involved stack allocating a second sector buffer. This
change uses the same code path for this case as well as when there
are a non-zero number of full sectors to access. The result is
easier to read and reduces the maximum stack used.
Signed-off-by: Anton Staaf <robotboy@chromium.org>
Cc: Andy Fleming <afleming@freescale.com>
Acked-by: Detlev Zundel <dzu@denx.de>
Fix all checkpatch violations in the low level Ext2 block
device reading code. This is done in preparation for cleaning
up the partial sector access code.
Signed-off-by: Anton Staaf <robotboy@chromium.org>
Cc: Andy Fleming <afleming@freescale.com>
Acked-by: Detlev Zundel <dzu@denx.de>
Fat directory handling didn't check reaching the end of the root directory. It
relied on a stop condition based on a directory entry with a name starting with
a '\0' character. This check in itself is wrong ('\0' indicates free entry, not
end_of_directory) but outside the scope of this fix. For FAT32, the end of the
rootdir is reached when the end of the cluster chain is reached. The code didn't
check this condition and started to read an incorrect cluster. This caused a
subsequent read request of a sector outside the range of the usb stick in
use. On its turn, the usb stick protested with a stall handshake.
Both FAT32 and non-FAT32 (FAT16/FAT12) end or rootdir checks have been put in.
Signed-off-by: Erik Hansen <erik@makarta.com>
Fix compiler warning
In file included from ubifs.h:2137:0,
from ubifs.c:26:
misc.h: In function 'ubifs_idx_key':
misc.h:263:26: warning: dereferencing type-punned pointer will break strict-aliasing rules
seen with gcc version 4.5.1 (Sourcery G++ Lite 2010.09-50).
No functional change.
CC: Stefan Roese <sr@denx.de>
Signed-off-by: Dirk Behme <dirk.behme@googlemail.com>
Signed-off-by: Stefan Roese <sr@denx.de>
The link_name variable is declared inside the if block and it is used
outside it through the name pointer.
Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Signed-off-by: Stefan Roese <sr@denx.de>
Until now ubifsload pads the destination with 0 up to a multiple of
UBIFS_BLOCK_SIZE (4KiB) while reading a file to memory. This patch
changes this behaviour to only read to the requested length. This
is either the file length or the length/size provided as parameter
to the ubifsload command.
Signed-off-by: Stefan Roese <sr@denx.de>
Before this commit, weak symbols were not overridden by non-weak symbols
found in archive libraries when linking with recent versions of
binutils. As stated in the System V ABI, "the link editor does not
extract archive members to resolve undefined weak symbols".
This commit changes all Makefiles to use partial linking (ld -r) instead
of creating library archives, which forces all symbols to participate in
linking, allowing non-weak symbols to override weak symbols as intended.
This approach is also used by Linux, from which the gmake function
cmd_link_o_target (defined in config.mk and used in all Makefiles) is
inspired.
The name of each former library archive is preserved except for
extensions which change from ".a" to ".o". This commit updates
references accordingly where needed, in particular in some linker
scripts.
This commit reveals board configurations that exclude some features but
include source files that depend these disabled features in the build,
resulting in undefined symbols. Known such cases include:
- disabling CMD_NET but not CMD_NFS;
- enabling CONFIG_OF_LIBFDT but not CONFIG_QE.
Signed-off-by: Sebastien Carlier <sebastien.carlier@gmail.com>
By now, the majority of architectures have working relocation
support, so the few remaining architectures have become exceptions.
To make this more obvious, we make working relocation now the default
case, and flag the remaining cases with CONFIG_NEEDS_MANUAL_RELOC.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Tested-by: Heiko Schocher <hs@denx.de>
Tested-by: Reinhard Meyer <u-boot@emk-elektronik.de>
Last commit 3831530dcb7b71329c272ccd6181f8038b6a6dd0a was intended
"explicitly specify FAT12/16 root directory parsing buffer size, instead
of relying on cluster size". Howver, the underlying function requires
the size of the buffer in blocks, not in bytes, and instead of passing
a double sector size a request for 1024 blocks is sent. This generates
a buffer overflow with overwriting of other structure (in the case seen,
USB structures were overwritten).
Signed-off-by: Stefano Babic <sbabic@denx.de>
CC: Mikhail Zolotaryov <lebon@lebon.org.ua>
The U-Boot code has the following bugs related to the processing of Long File
Name (LFN) entries scattered across several clusters/sectors :
1) get_vfatname() function is designed to gather scattered LFN entries by
cluster chain processing - that doesn't work for FAT12/16 root directory.
In other words, the function expects the following input data:
1.1) FAT32 directory (which is cluster chain based);
OR
1.2) FAT12/16 non-root directory (which is also cluster chain based);
OR
1.3) FAT12/16 root directory (allocated as contiguous sectors area), but
all necessary information MUST be within the input buffer of filesystem cluster
size (thus cluster-chain jump is never initiated).
In order to accomplish the last condition, root directory parsing code in
do_fat_read() uses the following trick: read-out cluster-size block, process
only first sector (512 bytes), then shift 512 forward, read-out cluster-size
block and so on. This works great unless cluster size is equal to 512 bytes
(in a case you have a small partition), or long file name entries are scattered
across three sectors, see 4) for details.
2) Despite of the fact that get_vfatname() supports FAT32 root directory
browsing, do_fat_read() function doesn't send current cluster number correctly,
so root directory look-up doesn't work correctly.
3) get_vfatname() doesn't gather scattered entries correctly also is the case
when all LFN entries are located at the end of the source cluster, but real
directory entry (which must be returned) is at the only beginning of the
next one. No error detected, the resulting directory entry returned contains
a semi-random information (wrong size, wrong start cluster number and so on)
i.e. the entry is not accessible.
4) LFN (VFAT) allows up to 20 entries (slots) each containing 26 bytes (13
UTF-16 code units) to represent a single long file name i.e. up to 520 bytes.
U-Boot allocates 256 bytes buffer instead, i.e. 10 or more LFN slots record
may cause buffer overflow / memory corruption.
Also, it's worth to mention that 20+1 slots occupy 672 bytes space which may
take more than one cluster of 512 bytes (medium-size FAT32 or small FAT16
partition) - get_vfatname() function doesn't support such case as well.
The patch attached fixes these problems in the following way:
- keep using 256 bytes buffer for a long file name, but safely prevent a
possible buffer overflow (skip LFN processing, if it contains 10 or more
slots).
- explicitly specify FAT12/16 root directory parsing buffer size, instead
of relying on cluster size. The value used is a double sector size (to store
current sector and the next one). This fixes the first problem and increases
performance on big FAT12/16 partitions;
- send current cluster number (FAT32) to get_vfatname() during root
directory processing;
- use LFN counter to seek the real directory entry in get_vfatname() - fixes the
third problem;
- skip deleted entries in the root directory (to prevent bogus buffer
overflow detection and LFN counter steps).
Note: it's not advised to split up the patch, because a separate part may
operate incorrectly.
Signed-off-by: Mikhail Zolotaryov <lebon@lebon.org.ua>
Doubly-indirect block numbers are compared against the first-level
indirect block when checking for a cached copy. This is causing the
doubly-indirect block to be re-read each time it is accessed.
Repairing this reduces load time for a 70M file from 72 seconds
to 38 seconds.
Signed-off-by: Aaron Pace <Aaron.Pace@alcatel-lucent.com>