83 lines
3.1 KiB
Diff
83 lines
3.1 KiB
Diff
Based on 1930eed2a7855d2df06ccf51f9e394428bf547e2 Mon Sep 17 00:00:00 2001
|
|
From: Jon Severinsson <jon@severinsson.net>
|
|
Date: Tue, 8 Jul 2014 18:29:46 +0200
|
|
Subject: [PATCH] journal/compress: improve xz compression performance
|
|
|
|
The new lzma2 compression options at the top of compress_blob_xz are
|
|
equivalent to using preset "0", exept for using a 1 MiB dictionary
|
|
(the same as preset "1"). This makes the memory usage at most 7.5 MiB
|
|
in the compressor, and 1 MiB in the decompressor, instead of the
|
|
previous 92 MiB in the compressor and 8 MiB in the decompressor.
|
|
|
|
According to test-compress-benchmark this commit makes XZ compression
|
|
20 times faster, with no increase in compressed data size.
|
|
Using more realistic test data (an ELF binary rather than repeating
|
|
ASCII letters 'a' through 'z' in order) it only provides a factor 10
|
|
speedup, and at a cost if a 10% increase in compressed data size.
|
|
But that is still a worthwhile trade-off.
|
|
|
|
According to test-compress-benchmark XZ compression is still 25 times
|
|
slower than LZ4, but the compressed data is one eighth the size.
|
|
Using more realistic test data XZ compression is only 18 times slower
|
|
than LZ4, and the compressed data is only one quarter the size.
|
|
|
|
---
|
|
src/journal/compress.c | 33 ++++++++++++++-------------------
|
|
1 file changed, 14 insertions(+), 19 deletions(-)
|
|
|
|
--- src/journal/compress.c
|
|
+++ src/journal/compress.c 2014-07-09 12:09:45.814235274 +0000
|
|
@@ -28,8 +28,15 @@
|
|
#include "compress.h"
|
|
|
|
bool compress_blob(const void *src, uint64_t src_size, void *dst, uint64_t *dst_size) {
|
|
- lzma_stream s = LZMA_STREAM_INIT;
|
|
+ static const lzma_options_lzma opt = {
|
|
+ 1u << 20u, NULL, 0, LZMA_LC_DEFAULT, LZMA_LP_DEFAULT,
|
|
+ LZMA_PB_DEFAULT, LZMA_MODE_FAST, 128, LZMA_MF_HC3, 4};
|
|
+ static const lzma_filter filters[2] = {
|
|
+ {LZMA_FILTER_LZMA2, (lzma_options_lzma*) &opt},
|
|
+ {LZMA_VLI_UNKNOWN, NULL}
|
|
+ };
|
|
lzma_ret ret;
|
|
+ size_t out_pos = 0;
|
|
bool b = false;
|
|
|
|
assert(src);
|
|
@@ -40,29 +47,17 @@ bool compress_blob(const void *src, uint
|
|
/* Returns false if we couldn't compress the data or the
|
|
* compressed result is longer than the original */
|
|
|
|
- ret = lzma_easy_encoder(&s, LZMA_PRESET_DEFAULT, LZMA_CHECK_NONE);
|
|
- if (ret != LZMA_OK)
|
|
+ if (src_size < 80)
|
|
return false;
|
|
|
|
- s.next_in = src;
|
|
- s.avail_in = src_size;
|
|
- s.next_out = dst;
|
|
- s.avail_out = src_size;
|
|
-
|
|
- /* Does it fit? */
|
|
- if (lzma_code(&s, LZMA_FINISH) != LZMA_STREAM_END)
|
|
- goto fail;
|
|
-
|
|
- /* Is it actually shorter? */
|
|
- if (s.avail_out == 0)
|
|
- goto fail;
|
|
+ ret = lzma_stream_buffer_encode((lzma_filter*) filters, LZMA_CHECK_NONE, NULL,
|
|
+ src, src_size, dst, &out_pos, src_size - 1);
|
|
+ if (ret != LZMA_OK)
|
|
+ return false;
|
|
|
|
- *dst_size = src_size - s.avail_out;
|
|
+ *dst_size = out_pos;
|
|
b = true;
|
|
|
|
-fail:
|
|
- lzma_end(&s);
|
|
-
|
|
return b;
|
|
}
|
|
|