| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
|
|
|
|
| |
feature request: #1051
Signed-off-by: Qi Wang <wangqi@linux.alibaba.com>
|
|
|
|
|
|
|
| |
ensure that `frametest` works fine with these values,
notably with low LZ4_MEMORY_USAGE (dict test notably)
following suggestions from @t-mat at #1016
|
|
|
|
|
| |
better name space isolation
suggested by @boris-kolpackov in #1053
|
| |
|
|
|
| |
The ARM64EC is a new Microsoft-designed ARM64 ABI that is compatible with AMD64 code. However, not all AMD64 intrinsic functions are supported. For, intrinsics that are lowered to AVX, AVX2 and AVX512 instructions are not supported, including the _tzcnt_u64. To make sure this file compiles for ARM64EC, the use of _tzcnt_u64 should be neutered.
|
|
|
|
|
|
|
|
|
|
|
| |
This makes decoding significantly faster on M1; measured on compressed source
code across 8 hardware threads, decompressing 294 MB to 1301 MB takes 513 ms
of cumulative work (2.53 GB/s) before, and 406 ms (3.2 GB/s) after this change
on M1 Pro.
There's no way to check if the target architecture is M1 specifically but the
gains are likely to be similar on recent iterations on Apple processors, and
the original performance issue was probably more specific to Qualcomm.
|
| |
|
| |
|
| |
|
|
|
|
| |
only do arithmetic if offset > 0
|
|
|
|
|
| |
ensure `dictBase` is only used
when there is an actual dictionary content.
|
|
|
|
| |
was blindly adding an offset (0) to `dictionary` which could be `NULL`.
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
only include <intrin.h> on vs2005+ (#947)
remove some useless #pragma
fix a few minor Visual warnings
|
|
|
|
| |
suggested by @aqrit in #947
|
|
|
|
|
|
| |
to remain similar to stdlib's calloc().
Updated test to use c++ compiler for stricter signature check.
|
|\ |
|
| |\
| | |
| | | |
fix #874
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
coverity reported a warning regarding a memcpy() overwrite.
This is a false positive (the memory area is large enough),
but it's true that it's not trivial to determine (encompassing struct),
and it's proper anyway to only memcpy() just the right amount of data.
|
| |/ |
|
|/
|
|
|
|
|
|
|
| |
makes it possible to replace at link time
malloc, calloc and free
by user-provided functions
which must be named LZ4_malloc(), LZ4_calloc() and LZ4_free().
answer #937
|
|
|
|
|
| |
- check alignment before casting a pointer
- saveDict : don't memmove() on NULL dst
|
|\
| |
| | |
Revert "Replace "static" to "LZ4_FORCE_INLINE" for small functions"
|
| |
| |
| |
| | |
This reverts commit 0e3933edd435c54cc2e21e38f5d4ba7bf644a24e.
|
|/
|
|
|
|
|
| |
minor: identical declaration and prototypes of `LZ4HC_compress_optimal()`
also :
very minor optimization of `LZ4_memcpy_using_offset()`
|
|\
| |
| | |
More alignment tests
|
| | |
|
| |
| |
| |
| | |
across lz4.c and lz4hc.c
|
| |
| |
| |
| | |
The "static" specifier does not guarantee that the function will be inlined.
|
| |
| |
| |
| | |
There is no reason to separate these two definitions!
|
|/
|
|
| |
This problem was reproduced on MSVC 2015 (32-bit). Both functions were called using the operator "call".
|
| |
|
|
|
|
|
| |
detected by scan-build, cppcheck and advanved compilation flags
fix #786
|
|\
| |
| | |
fix efficiency of LZ4_compress_HC_destSize()
|
| |
| |
| |
| |
| |
| | |
applying new more accurate formula from LZ4_compress_HC_destSize()
also : fix some minor display issue in tests/frametest
|
| |
| |
| |
| | |
must respect MFLIMIT distance from oend
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
LZ4_compress_HC_destSize() had a tendency
to discard its last match when this match overflowed specified dstBuffer limit.
The impact is generally moderate,
but occasionally huge,
typically when this last match is very large
(such as compressing a bunch of zeroes).
Issue #784 fixed for both Chain and Opt implementations.
Added a unit test suggested by @remittor checking this topic.
|
|\ \
| | |
| | | |
fix compressing into NULL
|
| |/
| |
| |
| |
| | |
fails properly
bug discovered by oss-fuzz
|
|/ |
|
|
|
|
| |
properly track history
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
LZ4_decompress_safe_partial()
now also supports a scenario where
nb_bytes_to_generate is <= block_decompressed_size
And
nb_bytes_to_read is >= block_compressed_size.
Previously, the only supported scenario was
nb_bytes_to_read == block_compress_size.
Pay attention that,
if nb_bytes_to_read is > block_compressed_size,
then, necessarily, it requires that
nb_bytes_to_generate is <= block_decompress_size.
If both are larger, it will generate corrupted data.
|
|\ |
|
| |
| |
| |
| |
| | |
Also : added memory-frugal software byte count for big endian 64-bit cpus.
Disabled by default.
|
| |\
| | |
| | | |
rejigger bit counting intrinsics
|
| | |
| | |
| | | |
MSVC debug mode complains
|