Commit Graph

10 Commits

Author SHA1 Message Date
Avi Kivity
9d8507de09 Merge "Optimize checksum_combine() for CRC32" from Tomek
"
zlib's crc32_combine() is not very efficient. It is faster to re-combine
the buffer using crc32(). It's still substantial amount of work which
could be avoided.

This patch introduces a fast implementation of crc32_combine() which
uses a different algorithm than zlib. It also utilizes intrinsics for
carry-less multiplication instruction to perform the computation faster.
The details of the algorithm can be found in code comments.

Performance results using perf_checksum and second buffer of length 64 KiB:

zlib CRC32 combine:   38'851   ns
libdeflate CRC32:      4'797   ns
fast_crc32_combine():     11   ns

So the new implementation is 3500x faster than zlib's, and 417x faster than
re-checksumming the buffer using libdeflate.

Tested on i7-5960X CPU @ 3.00GHz

Performance was also evaluated using sstable writer benchmark:

  perf_fast_forward --populate --sstable-format=mc --data-directory /tmp/perf-mc \
     --value-size=10000 --rows 1000000 --datasets small-part

It yielded 9% improvement in median frag/s (129'055 vs 117'977).

Refs #3874
"

* tag 'fast-crc32-combine-v2' of github.com:tgrabiec/scylla:
  tests: perf_checksum: Test fast_crc32_combine()
  tests: Rename libdeflate_test to checksum_utils_test
  tests: libdeflate: Add more tests for checksum_combine()
  tests: libdeflate: Check both libdeflate and default checksummers
  sstables: Use fast_crc_combine() in the default checksummer
  utils/gz: Add fast implementation of crc32_combine()
  utils/gz: Add pre-computed polynomials
  utils/gz: Import Barett reduction implementation from libdeflate
  utils: Extract clmul() from crc.hh

(cherry picked from commit b098b5b987)
2018-12-08 13:42:43 +02:00
Tomasz Grabiec
07c980845d utils/crc: Add clmul_u32() implementation
Needed for backporting dependent changes.

Extracted from:

      commit 79136e895f
      Author: Yibo Cai (Arm Technology China) <Yibo.Cai@arm.com>
      Date:   Thu Nov 1 03:26:16 2018 +0000

        utils/crc: calculate crc in parallel
2018-12-08 13:42:43 +02:00
Paweł Dziepak
252cf0c681 utils: crc: accept FragmentRange 2018-09-18 11:29:36 +01:00
Vlad Zolotarov
0145ae2b4b utils::crc32: add power64 crc32 HW accelerated implementation
Based on the work of Anton Blanchard <anton@au.ibm.com>, IBM that may be found
here: https://github.com/antonblanchard/crc32-vpmsum

Signed-off-by: Vlad Zolotarov <vladz@scylladb.com>
2017-12-08 13:38:13 -05:00
Michael Munday
5158b3f484 utils::crc: introduce process_le/be(T) methods
Replace the oblique process(T) overloads for integer types with
explicit process_le/be(T) methods that would interpret the given integer
as a stream of bytes using the corresponding endiannes.

For instance

process_le(0x11223344) would treat this integer as the following array of bytes:
{0x44, 0x33, 0x22, 0x11}.

process_be(0x11223344) on the other hand would treat this integer as if it's
{0x11, 0x22, 0x33, 0x44}.

Signed-off-by: Vlad Zolotarov <vladz@scylladb.com>
2017-12-08 10:12:21 -05:00
Michael Munday
26b7c2622e utils/crc: use zlib for crc32 on non-x86 platforms
Ideally we should use the Castagnoli polynomial to match the SSE 4.2
crc32 instructions, but this works for now.
2017-12-08 09:47:50 -05:00
Asias He
00d7a35949 utils: Put crc32 under utils namespace
It conflicts with crc in zlib
Message-Id: <1480918984-4117-2-git-send-email-asias@scylladb.com>
2016-12-05 11:48:29 +02:00
Pekka Enberg
38a54df863 Fix pre-ScyllaDB copyright statements
People keep tripping over the old copyrights and copy-pasting them to
new files. Search and replace "Cloudius Systems" with "ScyllaDB".

Message-Id: <1460013664-25966-1-git-send-email-penberg@scylladb.com>
2016-04-08 08:12:47 +03:00
Avi Kivity
d5cf0fb2b1 Add license notices 2015-09-20 10:43:39 +03:00
Avi Kivity
d6351ecca7 utils: add crc32 class
C++ interface to the crc32 x86 instruction.
2015-08-09 00:05:33 +03:00