AuroraMiddleware/zstd
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
3,747 Commits 1 Branch 0 Tags 27 MiB
2a62f48bf4
Commit Graph

473 Commits

Author SHA1 Message Date
Yann Collet
2a62f48bf4 release input buffers from inside worker thread 
buffers are released sooner, which makes them available faster for next job.
=> decreases total nb of buffers necessary
 2017-07-11 15:56:40 -07:00
Yann Collet
57236184af buffer pool : all buffers have same size 
to reduce memory fragmentation.
They can be used for in or out, interchangeably.
 2017-07-11 15:17:25 -07:00
Yann Collet
34b2b95631 zstdmt : intermediate outBuffer allocated from within worker 
reduces total amount of memory needed,
since jobs in queue do not have an outBuffer pre-reserved now
 2017-07-11 14:59:10 -07:00
Yann Collet
16261e6951 buffer pool can be invoked from multiple threads 2017-07-11 14:14:07 -07:00
Yann Collet
ef0ff7fe7f zstdmt: removed margin for improved memory usage 2017-07-11 08:54:29 -07:00
Yann Collet
4616fad18b improved ZSTDMT_compress() memory usage 
does not need the input buffer for streaming operations

also : reduced a few tests time length
 2017-07-10 17:16:41 -07:00
Yann Collet
670b1fc547 optimized memory usage for ZSTDMT_compress() 
Previously, each job would reserve a CCtx right before being posted.
The CCtx would be "part of the job description",
and only released when the job is completed (aka flushed).
For ZSTDMT_compress(), which creates all jobs first and only join at the end,
that meant one CCtx per job.
The nb of jobs used to be == nb of threads,
but since latest modification,
which reduces the size of jobs in order to spread the load of difficult areas,
it also increases the nb of jobs for large sources / small compression level.
This resulted in many more CCtx being created.

In this new version, CCtx are reserved within the worker thread.
It guaranteea there cannot be more CCtx reserved than workers (<= nb threads).

To do that, it required to make the CCtx Pool multi-threading-safe :
it can now be called from multiple threads in parallel.
 2017-07-10 16:30:55 -07:00
Yann Collet
3510efb02d fix : custom allocator correctly propagated to child contexts 2017-07-10 14:21:40 -07:00
Yann Collet
88da8f1816 fix : propagate custom allocator to ZSTDMT though ZSTD_CCtx_setParameter() 
also : compile fuzzer with MT enabled
 2017-07-10 14:02:33 -07:00
Yann Collet
e32fb0c1fe added ZSTD_sizeof_CCtx() test 2017-07-10 12:29:57 -07:00
Yann Collet
49af41820d clarified status of zstdmt_compress.h API 2017-07-05 17:21:13 -07:00
Yann Collet
27e883371d fixed wrong assert() condition 
A single job created by ZSTDMT_compress() can be < 256KB
if data to compress is < 256 KB
(in which case it is delegated to single thread mode)
 2017-07-04 19:33:16 -07:00
Yann Collet
2cb9774f5e more precise estimation of amount to flush at end of stream (single thread mode) 
also : can use DEBUGLEVEL variable in /tests
 2017-07-04 12:39:26 -07:00
Yann Collet
6383372dec fixed : 0-copy in NULL is UB 2017-07-04 10:36:41 -07:00
Yann Collet
2de2396a36 refactor ZSTDMT_compress() 2017-07-03 16:23:36 -07:00
Yann Collet
2084b041f4 fixed comments 2017-07-03 15:52:19 -07:00
Yann Collet
5a77361595 fixed wrong function name in comment 2017-07-03 15:21:24 -07:00
cyan4973
1bafe393e4 fix : ZSTDMT_compressStream_generic() can accept NULL input 
also : converge implementations towards new version of ZSTDMT_compressStream_generic()
 2017-07-01 06:59:24 -07:00
Yann Collet
58bd0e70fc fixed : dictionary compression with new advanced API in Multi-threading mode 2017-06-30 16:01:02 -07:00
Yann Collet
d8b33a598d Optimized ZSTDMT single-pass mode speed on large sources 
by ensuring job sizes remain "not too large"
 2017-06-30 15:44:57 -07:00
Yann Collet
d5c046c609 implemented shortcut for zstd_compress_generic() in MT mode 
added ZSTDMT_compress_advanced() API
 2017-06-30 14:51:01 -07:00
Yann Collet
a3d9926c40 compression optimization opportunity 
switch to single-pass mode directly into output buffer
when outputSize >= ZSTD_compressBound(inputSize).
Speed gains observed with fullbench (~+15% on level 1)
 2017-06-29 14:44:49 -07:00
Yann Collet
037466245f refactor ZSTD_check_compressionLevel_monotonicIncrease_memoryBudget() 
use less macro statements
the initial version was meant to work with STATIC_ASSERT
but since it doesn't work and needs assert()
it's possible to rewrite it using normally compiled code
which is better for compiler.

Downside : the error message is less precise.
There is a DEBUGLOG(3,) to compensate.
 2017-06-28 20:24:08 -07:00
Yann Collet
2bf428df45 Merge branch 'advancedAPI2' into refPrefix 2017-06-28 16:35:49 -07:00
Yann Collet
1ca76039af fixed -Wdeclaration-after-statement 2017-06-28 15:40:21 -07:00
Yann Collet
813535105b added function to control monotonic memory budget increase of ZSTD_defaultCParameters[0] 
It's a runtime test, based on assert(),
played once, on first ZSTD_getCParams() usage,
when ZSTD_DEBUG is enabled.
 2017-06-28 15:34:56 -07:00
Yann Collet
adbe74a8ac adjusted compression levels to guarantee a monotonically increasing memory budget 2017-06-28 13:22:37 -07:00
Yann Collet
33a6639039 fixed ZSTD_refPrefix with Multithread-enabled CCtx 2017-06-28 11:09:43 -07:00
Yann Collet
2e4274262d controlled dictMode 2017-06-27 17:09:12 -07:00
Yann Collet
b7372933b8 implemented ZSTD_refPrefix() 2017-06-27 15:49:12 -07:00
Yann Collet
7d3816183f exposed ZSTD_MAGIC_DICTIONARY in zstd.h 
makes it easier to explain ZSTD_dictMode
 2017-06-27 13:50:34 -07:00
Yann Collet
fecc721fd9 added parameter ZSTD_p_refDictContent 2017-06-27 11:46:39 -07:00
Yann Collet
dde10b23fe refactored ZSTD_estimateDStreamSize() 
now uses windowSize as argument.
Also : created ZSTD_estimateDStreamSize_fromFrame()
 2017-06-26 17:44:26 -07:00
Yann Collet
09ae03a570 ZSTD_estimateCDictSize_advanced() 
ZSTD_estimateCDictSize() now uses same arguments as ZSTD_createCDict()
ZSTD_estimateCDictSize_advanced() uses same arguments as ZSTD_createCDict_advanced()
 2017-06-26 16:47:32 -07:00
Yann Collet
0c9a915a28 ZSTD_estimateCStreamSize_advanced() 2017-06-26 16:02:25 -07:00
Yann Collet
31af8290d1 ZSTD_estimateCCtx_advanced() 
ZSTD_estimateCCtx() is now a "simple" function,
taking int compressionLevel as single argument.

ZSTD_estimateCCtx_advanced() takes a CParams argument,
which is both more complete and more complex to generate.
 2017-06-26 15:52:39 -07:00
Yann Collet
ef269c1b68 Merge pull request #725 from facebook/advancedAPI2 
New Advanced API
 2017-06-23 09:50:47 -07:00
Yann Collet
ecb0f46866 add controls over streaming buffers 2017-06-21 17:25:01 -07:00
Yann Collet
204b6b7ef6 fixed streaming buffered allocation with CDict compression 2017-06-21 15:13:00 -07:00
Yann Collet
83095970e6 free cdictLocal faster, suggested by @terrelln 2017-06-21 12:26:40 -07:00
Yann Collet
7bd1a2900e added ZSTD_dictMode_e to control dictionary loading mode 2017-06-21 11:50:33 -07:00
Yann Collet
e51d51bdf7 fixed memcpy() overlap 2017-06-20 17:44:55 -07:00
Yann Collet
466f92eaa6 removed one useless streaming compression stage, detected by @terrelln 2017-06-20 16:25:29 -07:00
Yann Collet
c3bce24ef4 fixed potential dangling pointer, detected by @terrelln 2017-06-20 16:09:11 -07:00
Yann Collet
b44ab82f7a ensure new ZSTD_strategy starts at value 1 2017-06-20 14:11:49 -07:00
Yann Collet
c08e649e95 first implementation of bench.c with new API ZSTD_compress_generic() 
Doesn't speed optimize this buffer-to-buffer scenario yet.
Still internally defers to streaming implementation.

Also : fixed a long standing bug in ZSTDMT streaming API.
 2017-06-19 18:25:35 -07:00
Nick Terrell
55f9cd4942 [libzstd] Fix UBSAN failure 2017-06-19 15:12:28 -07:00
Yann Collet
bf99150be3 update new api presentation in zstd.h and manual 2017-06-19 12:56:25 -07:00
Yann Collet
d7a3bffba9 new api : setting compression parameters is refused if a dictionary is already loaded 2017-06-19 11:53:01 -07:00
Yann Collet
381e66cfbd added ZSTD_clampCParams() 
now ZSTD_adjustCParams() is always successful,
it always produces a valid CParams
 2017-06-16 17:34:54 -07:00