| Age | Commit message (Collapse) | Author |
|---|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
The strategy is simple:
- The data writer function that write data/fragment blocks get
access to the list files.
- When writing a fragment, we look for an already written file that has
a fragment with the same size and checksum.
- If we find one, we throw away the fragment and reuse the existing one.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Instead of doing DFS on the fly in gensquashfs, churn out a linked list
of all files in an archive.
Future improvements in packing strategies can go into this file.
This can also be usefull for other purposes in the future, such as file
deduplication or as a work queue for the unpacker.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
If some library isn't present, the variable $(FOO_LIBS) simply evaluates
to empty string, so it can be simply be added to LDADD without any checks.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
The utility programs pretty much all have the same structure of including
one central header per C file that includes all required library headers,
so we can simply include config.h which we need for large file support
from there and remove it from the C files.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Requires that config.h be included before other headers, since the macro
_FILE_OFFSET_BITS changes the definitions of things like 'struct stat'.
I chose to simply include it at the top of every C file and at
immediately after the double-inclusion guards of every header.
Signed-off-by: Matt Turner <mattst88@gmail.com>
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
First of all, this commit adds a mod_time field to a tree node. When
creating the tree node, the field is set from the struct stat. When
scanning a directory, the time stamps from the input are used if set.
Second, the libsqfs code that reads inodes is modified to store the
mod_time from the inode in the fstree node and to write the tree node
into a generated inode.
Finally, tar2sqfs is modified to optionally keep the timestamps from
the tar archive instead of setting defaults. gensquashfs is similarly
modified to keep the input timestamps if specified.
The result is as follows:
- sqfs2tar will always carry the timestamps from the squashfs over
to the tar ball.
- tar2sqfs will set defaults, unless explicitly asked to preserve
the mtime from the tar ball.
- gensquashfs can optionally preserve the mtime from the input
hierarchy it processes if only --pack-dir is specified.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
This commit adds a mechanism to fstree_from_file to support filenames
with spaces in them by quoting the entire string. Quote marks can still
be used inside file names by escaping them with a backslash. Back slashes
(if that is your thing) can also be escaped.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
The added flags allow controlling the following on a per file level:
- forcing a file to be written uncompressed
- forcing a file to not have a fragment, i.e. the last truncated block
actually being written as a block
- padding a file to be alligned to device block size
The flags are not yet exposed to anything user controllable (such as
command line flags).
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Use a simple strtol for block size and device block size:
- We don't need to check the block size, sqfs_super_init does that.
- We only need to make sure the device block size is at least 1k.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Instead of decomposing a default string in gensquashfs option processing,
move that to fstree_init instead and pass the option string directly to
fstree_init.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
This commit removes handling of compressor names from gensquashfs. Instead,
functions are added to libcompress to obtain name from ID, ID from name
and to print out defaults.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
This commit removes the packdir/packfile based directory setup magic
from fstree_from_file and moves it to gensquashfs. Over there, the
common parts are deduplicated.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Make tree node list sort and recursive variant available and independend
of the fstree_t.
This is considered cleaner, since the fstree_t actually isn't needed for
any of this and we can just call the recusvie sort on the root instead,
and we can use the sort implementation directly for things like the
upcoming unit test.
Also this commit splits up the merge/sort implementation into a seperate
split and merge functions to make the code somewhat more readable.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
This commit moves the SELinux label code after the tree is sorted and
the inode table is generated. Sorting helps to make sure that the tree
will always be traversed in a defined, deterministic order and likewise
the creation of xattrs happens in a defined, deterministic order.
Second, we can now use the inode table instead of having to implement a
recursive tree traversal yet again.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Always try to gather all compressor libraries available, but only
complain about missing a one if it has been *explicitly* selected.
If a compressor has been explicityl disabled, we can still turn it
off after checking.
Also, rework gensquashfs to set the default compressor based on
what's available.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Add a generic one that reads and processes N bytes from an input
file descriptor.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
This commit tries to make the block and fragment processing code a
lot more readable by simplifying it as follows.
The following pattern is used repeatedly:
- Try to compress a block
- Write the compressed if smaller (the original otherwise) to a file
- Print errors on the way
- Set the magic (1 << 24) flag if we wrote the original
- Increment the super block byte counter
=> Move that to a seperate function called "write_compressed"
=> Move the code to grow the fragment table is moved to a seperate function.
=> Clearing the fragment buffer is pointless, so remove that
The add_fragment/add_block functions are now trivial
=> Merge them into process_file
=> Global block counter no longer needed, so remove that too
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Don't need that, we already know the index when we write add
the fragment.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
The idea is to move various higher level helper functions there.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Instead of allocating inode numbers as we go, generate and populat an
inode table from the fstree ahead of time. This makes processing nodes
a little bit simpler and we will need that table anyway for NFS export
support later on.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
This commit replaces the chdir to the input directory with pushd/popd
when building the fstree and again when packing files.
This simplifies handling of other file paths given on the command line
that have to be accessed and are relative to the original working
directories.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
The problem: SquashFS uses extended inode types for inodes with xattrs,
so for directories we need to generate an ldir inode. The ldir inode
type includes a directory index which cannot be empty (the counter is
off by 1, so 0 means 1 entry) and the entry is expected to contain a
file name with the same issue (must be at least 1 byte).
This has probably not really been thought through. What the mksquashfs
actually does is generating an ldir inode without an index. The kernel
does not bother to read the index, since the size of the directory is 0.
Its a hack that just so happens to work.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
If packdir and packfile are both specified, use packdir as
alternate root.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
For now, it is enough to reduce the comparison value, since the inode
numbers are increasing.
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
|
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
|
