aboutsummaryrefslogtreecommitdiff
path: root/lib/sqfs/Makemodule.am
AgeCommit message (Collapse)Author
2020-06-09Cleanup: split libsquashfs xattr writer codeDavid Oberhollenzer
This commit moves the libsquashfs xattr related code into a sub directory and splits the xattr writer code up into several files. No actual code is changed. Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2020-05-24Cleanup: split the block processor common.c againDavid Oberhollenzer
This commit breaks the common code up again by moving the data submission code to a separate file, making both a little bit more readable. Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2020-05-19Cleanup: move hash table header to include directoryDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2020-04-22Import and use Mesa's hash tableMatt Turner
With `perf record`/`perf report` I saw that 30% of the time was spent in `sqfs_frag_table_find_tail_end` with tar2sqfs for a tarball containing the Gentoo ebuild repository (many thousands of small files). The reason was the bucketing hash table in frag_table.c: too many elements in too few buckets meant lots of walking over the linked lists. This patch replaces that hash table with the hash table implementation from Mesa. Its implementation is more complex (is is an open-addressing, linear-reprobing) hash table, but it is much better suited for the task. On my 4c/8t Skylake, the time to run tar2sqfs drops from 7.5s to less than 3s. CPU usage increases from ~207% to ~356%, presumably indicating an increase in available parallelism due to the removal of the hash table as a bottleneck. The `perf report` profile with this patch shows that the time spent in `sqfs_frag_table_find_tail_end` has dropped from ~30% to 0.01%. Output from ministat: x before + after N Min Max Median Avg Stddev x 20 7.476 7.685 7.5725 7.5615 0.051254268 + 20 2.79 2.901 2.846 2.84475 0.03543842 Difference at 95.0% confidence -4.71675 +/- 0.0282015 -62.3785% +/- 0.241477% (Student's t, pooled s = 0.0440618) I imported only the bits of the hash table implementation that were needed for frag_table.c. Among the changes I made after importing are - removed usage of ralloc, Mesa's recursive memory allocator - Replaced ralloc -> malloc ralloc_free -> free rzalloc_array -> calloc - Removed mem_ctx parameters - Added free()s to the appropriate places (valgrind confirms there are no leaks) - removed _mesa_-prefix from function names Fixes: #40 Signed-off-by: Matt Turner <mattst88@gmail.com>
2020-03-18Cleanup: Move xxhash32 code to libutilDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2020-02-21Cleanup: move utilities back out of libsquashfsDavid Oberhollenzer
This commit removes the allocation helpers and string table functions out of libsquashfs back into a "libutil.a". The problem of libsquashfs exporting stuff that it shouldn't is resolved by retaining the internal attributes and directly adding the source to libsquashfs instead of trying to somehow link against libutil.la. Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2020-02-16block processor: merge rest of fileapi.c into common.cDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2020-02-16Replace crc32 with xxhash32David Oberhollenzer
On the one hand, benchmarking and profiling determined xxhash32 to be faster than the zlib implementation of crc32, on the other hand profiling determined that crc32 computation contributed signifficantly to the overall runtime. Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2020-02-09block processor: merge left overs of block.c/fragment.c into common.cDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2020-01-31Split the block writing/deduplication away from the block processorDavid Oberhollenzer
This commit moves the entire block writing and deduplication of data blocks over to a different data type named "block writer". For simplicity, the interfaces of the block processor are left as is and are turned into warppers. Likewise, most of the code in the block writer is just verbatim from the block processor, to be cleaned up in subsequent commits. Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2020-01-29Rename sqfs_data_writer_t back to sqfs_block_processor_tDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2020-01-25Add missing headers to installed header listDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2020-01-24Add a fragment table primitive to libsquashfsDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-12-30Fix zlib paths in automake fileDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-12-29Add a small version of zlib that can be built in staticallyDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-12-28Add a small version of liblz4 that can be built in staticallyDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-12-13Fix name of libsquashfs pkg-config fileDavid Oberhollenzer
This fixes both the name inside the file, as well as the file name by adding the major version suffix. Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-12-13Merge windows and pthread thread pool implementationsDavid Oberhollenzer
Since they are both structured the same way using condition variables, they are only a few defines away from removing code duplication. Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-12-08Replace SetEvent synchronization with condition variablesDavid Oberhollenzer
It's cleaner, more stable and works pretty much the same way as the pthread version. The downside is that the minimum target for the library is now Windows Vista, or Server 2008. But both are over a decade old anyway, so this shouldn't be an issue. Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-12-08Add native Windows port of the multi-threaded data writerDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-11-25Cleanup: remove what is left of libutilDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-11-25Cleanup: move overflow safe alloc code into libsquashfsDavid Oberhollenzer
There were only a hand full of instances outside libsquashfs that used the alloc code. In most cases, the thing allocated hat its size derived from something already in memory anyway, so it is safe to assume its size fits into a size_t. At the same time, the opencoded Windows path conversion functions are all unified into a single helper function. Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-11-24Cleanup: completely move str_table into libsquashfs internalsDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-11-24Cleanup: split out sqfs_write_super similar to sqfs_read_superDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-11-24Fix: Move LZO compressor from libsquashfs to libcommonDavid Oberhollenzer
The liblzo2 library is licensed under GPLv2, so it is not possible to distribute binaries of libsquashfs that link against liblzo2 under LGPL. This commit moves the LZO compressor implementation to libcommon, where this isn't a problem, since the tools themselves are licensed under GPLv3. It removes the ability of libsquashfs to read or generate LZO compressed SquashFS images, but the tools still can. Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-10-23For Windows builds, remove the so-version, link libgcc staticallyDavid Oberhollenzer
The -static-libgcc flag has to be passed through the compiler with a "-Wc," prefix, because libtool tries to be clever about linker flags. If added directly to LDFLAGS, libtool removes it. Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-10-11Add the required libtool magic to build a Windows DLLDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-10-11Add Windows implementation for sqfs_file_tDavid Oberhollenzer
This commit moves the generic unix implementation into a "unix" subdirectory and adds a "win32" subdirectory with a winapi based implementation. Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-29Cleanup: rename "compress.h" to "compressor.h"David Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-28Add recoding implementation for the xattr writerDavid Oberhollenzer
This commit implements the part of the API responsible for recoding and deduplicating xattr key-value blocks. Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-28Add header with declarations for the sqfs_xattr_writer_tDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-28Move declarations for sqfs_xattr_reader_t to a seperate headerDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-27Add a pkg-config file for libsquashfsDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-27Cleanup: merge data.h into block.hDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-27Add xz-utils based lzma compressor implementationDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-26Remove the create I/O block from sqfs_file_t functionDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-26Add file API stub to sqfs data writerDavid Oberhollenzer
Basically move the state tracking from the old data writer over to the new one. Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-25Cleanup: move the stdin sqfs_file_t wrapper out of libsquashfsDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-25Move sqfs_block_t to its own headerDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-25Rename block processor to sqfs_data_writer_tDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-25More refactoring of the block processorDavid Oberhollenzer
Basically just moving functions around and renaming things. Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-25Refactor out common code of the block processorsDavid Oberhollenzer
Code that already is shared between the pthread and the serial processor as well as code that can be re-used by other threading API implementations. Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-24Breake some of the helper functions out of process_block.cDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-23Move the bulk of the work from the data writer to the block processorDavid Oberhollenzer
Instead of calling a callback, the block processor now takes care of writing the data blocks to the file in the correct order, keeping track of fragment blocks and deduplicating blocks. Some cleanup work remains to be done and the statistics have to be repaired (yet again). Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-22Add helper functions for working with inodesDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-21Merge some of serial & pthread block processor code paths/declarationsDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-20Integrate the data reader into libsquashfsDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-20Cleanup: split "dir.h" into "dir_reader.h" and "dir_writer.h"David Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-20Add fstree like tree deserialization function to dir readerDavid Oberhollenzer
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
2019-09-19Add directory reader data structureDavid Oberhollenzer
This moves a lot of the stuff that is done manually in the tree deserializer to a generic helper in libsquashfs. Due to how the fstree is implemented, as a work around, the inode needs to be temporarily stored in the tree node, but some of the directory details could be removed. Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>