| Age | Commit message (Collapse) | Author |
|---|
|
ubiformat segfaults like this:
ubiformat: mtd0 (NAND), size 134217728 bytes (128.0 MiB), 16384 eraseblocks of 16384 bytes (16.0 KiB), min. I/O size 512 bytes
libscan: scanning eraseblock 8191 -- 100 % complete
ubiformat: 8190 eraseblocks have valid erase counter, mean value is 9
ubiformat: 2 eraseblocks are supposedly empty
ubiformat: warning!: VID header and data offsets on flash are 512 and 1024, which is different to calculated offsets 256 and 512
ubiformat: use new offsets 512 and 1024? (yes/no) no
Segmentation fault
The reason is that volume table size is calculated for 256/512
layout, and when user chooses 512/1024 - it is not re-calculated
which leads to segfault in 'ubigen_write_layout_vol()'.
This patch also fixes the message - in the above output new
offsets are 256/512, not 512/1024.
Also, this patch adds explicit printing of the selected
positions.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Remove the fd field from the mtd information data structure,
because libmtd does not really know the device node file name,
and serves only as a place to save the descriptor. The callers
should find a better place.
This patch improves code readability and prepares for further
changes.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
We know that sysfs is always at /sys, so we do not have to
store it in lib->sysfs.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
sysfs if guaranteed to be in /sys.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Just for the sake of being consistent with libubi, move the
comments for API functions to the header file.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Building out of tree currently fails with ubi-utils because make gets
confused and can't figure out the vpathed .c files. Using VPATH rather
than vpath seems to work around it.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
|
|
On nommu arch, local stack size is very limited and
can't be enlarged on demand. So, don't define large
array on local stack.
Signed-off-by: Sonic Zhang <sonic.zhang@analog.com>
|
|
Signed-off-by: Roger Quadros <ext-roger.quadros@nokia.com>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Few minor amendments by Artem. And increase the utility
version number.
Signed-off-by: Roger Quadros <ext-roger.quadros@nokia.com>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Move new-utils to ubi-utils and old ones to ubi-utils/old-utils.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Take the latest version from the kernel, where it was renamed
to ubi-media.h.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
The 2GiB limit is at erase block 8192. In mtd_is_bad() the computation
for the MEMGETBADBLOCK ioctl() looks like the following:
| seek = eb * mtd->eb_size;
with both eb and mtd->eb_size being a signed int results in seek being a
signed result.
The _FILE_OFFSET_BITS=64 define is required to switch off_t from 32bit
to 64bit an 32bit systems. This is required in order to keep using
lseek() as lseek64 on 32bit system. Without this change lseek() in
mtd_read() is called with a 32bit value with most significat bit set and
the kernel performs a sign extension for the 64bit value which is used
in the mtd layer.
The last change also changes the size of the parameter which is passed
to the MEMGETBADBLOCK ioctl() from 32 to 64bit. The counter part in
kernel is also defined as loff_t which is of type __kernel_loff_t and
this is "long long". So this must have been broken for a while unless I
missed something.
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
The execution of
|./ubiupdatevol /dev/ubi0 -t
will fail because 'argv[optind + 1]' is undefined and the later executed
'strcmp(args.img, "-")' will segfault.
So I can hack around and supply a dummy image or fix it that way.
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
In case of error, ubiformat prints error messages at the same
line as the previous messages. Fix this.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Add missing whitespace in ubimkvol help output
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
When building with gcc security warnings enabled, the ubimirror.c code
triggers this warning:
./src/ubimirror.c: In function 'main':
./src/ubimirror.c:206: error: format not a string literal and no format arguments
Since the buffer in question should be a straight string anyways, avoid
ugly printf exploits by outputting the string indirectly:
printf("%s", buffer)
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Rather than duplicating the same thing over and over in every Makefile,
move it all to common.mk.
Other things fixed here:
- doing subdirs in parallel
- fix src!=build compiling in subdirs
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
The previous change with CFLAGS handling was not entirely correct. So fix
CFLAGS/CPPFLAGS/LDFLAGS to work like standard build systems (i.e. autotools)
and change `ranlib` to $(RANLIB) to fix cross-compiling.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
A size_t should be printed using %zu (unsigned size_t) rather than %d. This
fixes the following warning on my system:
gcc -O2 -Werror -Wall -Iinclude -Isrc -I../../include src/ubiformat.c -c -o ubiformat.o
cc1: warnings being treated as errors
src/ubiformat.c: In function ‘read_all’:
src/ubiformat.c:345: error: format ‘%d’ expects type ‘int’, but argument 3 has type ‘size_t’
src/ubiformat.c:352: error: format ‘%d’ expects type ‘int’, but argument 3 has type ‘size_t’
make: *** [ubiformat.o] Error 1
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
gcc-4.3.x introduces a number of new warnings (and we use -Werror) for
things like not checking scanf return values and not using explicit
formatting. Finally, it caught that we were overflowing our own buffer
in unubi.c, so increate the variable size as we want to declare and
clear PATH_MAX + 1, then use only PATH_MAX of the variable.
Signed-off-by: Tom Rini <trini@embeddedalley.com>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Some 'install' versions (e.g. this of OpenEmbedded) require the mode
to be a separate argument and don't understand the '-m0755' syntax:
.../staging/x86_64-linux/usr/bin/install-sh: ./-m0755 does not exist.
Signed-off-by: Enrico Scholz <enrico.scholz@sigma-chemnitz.de>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Let's have the "new ubi-utils" makefile handle CFLAGS like the other
makefiles. This is important when cross compiling and CFLAGS is
externally provided. Without this, local include files are not found due
to the various -I arguments which are then lost.
Signed-off-by: Nicolas Pitre <nico@marvell.com>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
J. Scott Merritt wrote:
> Dear list,
>
> I believe that I have stumbled upon two problems with the latest version
> of ubinize.c in mtd-utils - one serious, and one not so serious.
>
> - First, the easy one: At line 457, I believe that we should be printing
> args.subpage_size rather than ui.min_io_size.
>
> - More seriously, I believe that the "flags" field in the ubigen_vol_info
> structures is not being properly initialized. Line 494 allocates memory
> for these structures with malloc, so they are not cleared to zero. The
> read_section function updates the flag field if the AUTO-RESIZE flag is
> specified, but does not otherwise initialize or clear it.
>
> I reckon the latter problem could be repaired either with calloc, or by
> modifying read_section to more directly set/clear the flag.
>
> Thanks, Scott.
Reported-by: J. Scott Merritt <merrij3@rpi.edu>
Signed-off-by: Adrian Hunter <ext-adrian.hunter@nokia.com>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Some recent changes to ubinize prevent the 'image' parameter from being
absent in the configuration file. The following patch fixes that.
Signed-off-by: Richard Titmuss <richard.titmuss@logitech.com>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Add -N option to ubirmvol to support deleting volumes
by name.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
We use term "volume ID", not "volume number". So change few
vol_num variables to vol_id to be consistent.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Check for free LEBs before trying to create a volume. This fixes
the following issue:
$ ubimkvol /dev/ubi0 -n 0 -N test_fs -m
ubimkvol: error!: cannot UBI create volume
error 22 (Invalid argument)
Error: cannot create UBI volume
"Invalid argument" ? Not understandable. But the thing is that
-m makes ubimkvol use available space, which is zero, then it
passes zero to the ioctl, which fails in request validation
code, because 0 invalid volume size.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Trivial patch to add a missing linefeed in static volume info:
| Data bytes: 113112 bytes (110.5 KiB)State: OK
Signed-off-by: Enrico Scholz <enrico.scholz@sigma-chemnitz.de>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
This patch allows to read the image from stdin and adds an
'--image-size' option to specify its size. 'upiupdatevol' has this
feature already and it is useful for 'ubiformat' too.
It adds a read_all() function which reads a block of data. Plain
'read(2)' can fail on
nc -l -p 1234 | ubiformat -f - -S 25427968
else.
Signed-off-by: Enrico Scholz <enrico.scholz@sigma-chemnitz.de>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
The '--flash-image' option of 'ubiformat' requires an argument and
'.has_arg' must not be 0 hence. The patch adds this option to the
commandline synopsis too.
Signed-off-by: Enrico Scholz <enrico.scholz@sigma-chemnitz.de>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
The UBI tools refuse to work with symlinks like
'/dev/ubi/bootloader -> ../ubi0_2' because they use 'lstat(2)' and get
information about the symlink but not about the device. This is
unwanted and fixed by this patch.
This patch renames 'struct stat stat' variables to 'st' to avoid
compilation errors.
Signed-off-by: Enrico Scholz <enrico.scholz@sigma-chemnitz.de>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
If you execute straight `make`, this prevents the active make env from
being passed on to sub children. This prevents parallel building as well
as build flag passing.
I also dropped the recursive make in ubi-utils/Makefile for the all target
as this causes problems when building in parallel. All of the targets
that we actually care about are handled by the $(NTARGETS) target anyways.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
-j 2 build was not possible without this patch.
Signed-off-by: Sebastian Siewior <bigeasy@linutronix.de>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
|
|
When static volumes is generated, the 'used_ebs' field of VID
header has to be calculated based on image file size, not on
the volume size. Fix this.
This patch also improves error reporting a little and re-arrages
the code so that all the validation is done in 'read_section()'
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Check that volume names and IDs are unique
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Undo my weird change which slipped accidentally to the commit.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
1. Adds -s/--size switch;
2. Allows the image file to be specified as '-', meaning stdin; if used,
the size must be specified.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Users can pass the -O option to ubiformat to override the VID header
offset on the command-line. If for some reason, the new offset does
not match what is on flash, ubiformat currently prompts the user if
they'd rather use the *old* offsets. This means that calling
ubiformat with --yes will not do what was requested on the
command-line.
This patch reverses the question so --yes is actually useful for
running in batch-mode.
Signed-off-by: Bernard Blackham <bernard@largestprime.net>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
|
|
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Josh Boyer <jwboyer@gmail.com>
|
