mirrors/scylladb
1
0
Fork 0
mirror of https://github.com/scylladb/scylladb.git synced 2026-05-31 12:06:44 +00:00
Code Issues Packages Projects Releases Wiki Activity
Files
17d68cb40817e36b4f9cf9e1da30fa63debbe76b
scylladb/utils
History
Yibo Cai (Arm Technology China) 6fadba56cc utils: optimize UTF-8 validation 
UTF-8 string is now validated by boost::locale::conv::utf_to_utf, it
actually does string conversions which is more than necessary.  As
observed on Arm server, UTF-8 validation can become bottleneck under
heavy loads.

This patch introduces a brand new SIMD implementation supporting both
NEON and SSE, as well as a naive approach to handle short strings.
The naive approach is 3x faster than boost utf_to_utf, whilst SIMD
method outperforms naive approach 3x ~ 5x on Arm and x86. Details at
https://github.com/cyb70289/utf8/.

UTF-8 unit test is added to check various corner cases.

Signed-off-by: Yibo Cai <yibo.cai@arm.com>
Message-Id: <1543978498-12123-1-git-send-email-yibo.cai@arm.com>
2018-12-05 21:51:01 +02:00
..
arch/powerpc/crc32-vpmsum
utils: crc32: mark power crc32 assembly as not requiring an executable stack
2018-10-02 18:48:23 +01:00
gz
utils/gz: Fix compilation on non-x86 archs
2018-12-04 18:17:27 +00:00
allocation_strategy.hh
lsa: provide migrator with the object size
2018-05-09 16:52:26 +01:00
anchorless_list.hh
mvcc: Introduce partition_version_list
2018-05-30 12:18:56 +02:00
big_decimal.cc
Update seastar submodule
2018-11-21 00:01:44 +02:00
big_decimal.hh
utils/big_decimal: Added necessary operators and methods for aggregate functions.
2017-11-12 15:51:29 +01:00
bloom_calculations.cc
Fix pre-ScyllaDB copyright statements
2016-04-08 08:12:47 +03:00
bloom_calculations.hh
utils: convert sprint() to format()
2018-11-01 13:16:17 +00:00
bloom_filter.cc
Update seastar submodule
2018-11-21 00:01:44 +02:00
bloom_filter.hh
utils: Use dedicated enum for Bloom filter format instead of a boolean.
2018-05-10 09:47:41 +03:00
bounded_stats_deque.hh
Fix pre-ScyllaDB copyright statements
2016-04-08 08:12:47 +03:00
buffer_input_stream.cc
buffer_input_stream: make it possible to specify chunk size
2018-04-16 21:11:13 +02:00
buffer_input_stream.hh
buffer_input_stream: make it possible to specify chunk size
2018-04-16 21:11:13 +02:00
chunked_vector.hh
utils: chunked_vector: Do not require T to be default-constructible for clear()
2018-07-11 16:55:20 +02:00
class_registrator.hh
class_registry: introduce a non-static variant of class_registry
2018-11-26 13:30:21 +00:00
clmul.hh
utils: Extract clmul() from crc.hh
2018-12-03 14:36:08 +01:00
config_file_impl.hh
config_file_impl: Remove ostream operators
2018-02-07 10:11:46 +00:00
config_file.cc
util::config_file: Add "add" config item overload
2018-03-19 12:24:04 +00:00
config_file.hh
conf: define named_value<log_level> externally
2018-04-02 19:23:06 +01:00
coroutine.hh
Introduce a coroutine wrapper
2018-05-30 14:41:40 +02:00
crc.hh
utils: Extract clmul() from crc.hh
2018-12-03 14:36:08 +01:00
data_input.hh
Update seastar submodule
2018-11-21 00:01:44 +02:00
date.h
date: use correct casts for years
2017-04-17 23:03:15 +03:00
div_ceil.hh
utils: introduce div_ceil()
2017-05-17 12:30:03 +03:00
dynamic_bitset.cc
dynamic_bitset: optimize for large sets
2018-04-07 14:52:58 +03:00
dynamic_bitset.hh
dynamic_bitset: optimize for large sets
2018-04-07 14:52:58 +03:00
estimated_histogram.hh
Update seastar submodule
2018-11-21 00:01:44 +02:00
exceptions.cc
utils/exceptions: Whitelist EEXIST and ENOENT in should_stop_on_system_error()
2016-06-09 10:03:04 +02:00
exceptions.hh
utils: Improve storage_io_exception error message
2016-06-09 09:58:00 +02:00
exponential_backoff_retry.hh
Update seastar submodule
2018-11-21 00:01:44 +02:00
extremum_tracking.hh
Add class for tracking both extremum values (min and max) on updates.
2018-05-03 17:05:06 -07:00
fb_utilities.hh
utils::fb_utilities: add is_me(addr) method
2017-12-14 15:05:48 -05:00
file_lock.cc
Update seastar submodule
2018-11-21 00:01:44 +02:00
file_lock.hh
Update seastar submodule
2018-11-21 00:01:44 +02:00
flush_queue.hh
utils: convert sprint() to format()
2018-11-01 13:16:17 +00:00
fragment_range.hh
util fragment_range: add general linearisation functions
2018-07-18 12:28:06 +01:00
fragmented_temporary_buffer.hh
utils: convert sprint() to format()
2018-11-01 13:16:17 +00:00
hash.hh
Update seastar submodule
2018-11-21 00:01:44 +02:00
histogram.hh
Update seastar submodule
2018-11-21 00:01:44 +02:00
i_filter.cc
utils: convert sprint() to format()
2018-11-01 13:16:17 +00:00
i_filter.hh
utils: Use dedicated enum for Bloom filter format instead of a boolean.
2018-05-10 09:47:41 +03:00
in.hh
utils::in: Add helper type for perfect forwarding initializer lists
2017-12-05 14:28:34 +00:00
input_stream.hh
Merge seastar upstream
2016-09-28 17:34:16 +03:00
int_range.hh
utils: convert sprint() to format()
2018-11-01 13:16:17 +00:00
joinpoint.hh
Fix pre-ScyllaDB copyright statements
2016-04-08 08:12:47 +03:00
large_bitset.cc
large_bitset: use a chunked_vector internally and simplify API
2018-04-10 10:25:06 +03:00
large_bitset.hh
large_bitset: be more accurate with memory usage
2018-05-15 11:22:21 -04:00
latency.hh
histograms: do not use latency_in_nano
2016-11-14 18:01:43 +02:00
limiting_data_source.cc
Introduce make_limiting_data_source
2018-04-16 20:56:30 +02:00
limiting_data_source.hh
Introduce make_limiting_data_source
2018-04-16 20:56:30 +02:00
loading_cache.hh
loading_cache: make iterator work on top of lru_list iterators instead of loading_shared_values'
2018-08-30 20:56:44 -04:00
loading_shared_values.hh
utils/loading_shared_values: Add missing stat update call in one of the cases
2018-10-25 15:15:05 +03:00
log_heap.hh
log_histogram: rename to log_heap
2017-09-18 12:44:05 +02:00
logalloc.cc
utils: convert sprint() to format()
2018-11-01 13:16:17 +00:00
logalloc.hh
database: Make soft-pressure memtable flusher not consider already flushed memtables
2018-08-26 11:02:34 +03:00
managed_bytes.cc
linearization_context: remove non-trivial operations from fast path
2018-01-30 18:33:25 +01:00
managed_bytes.hh
managed_bytes: Mark read_linearize() as an allocation point
2018-07-17 16:39:43 +02:00
managed_ref.hh
Fix pre-ScyllaDB copyright statements
2016-04-08 08:12:47 +03:00
managed_vector.hh
managed_vector: Make external_memory_usage() ignore reserved space
2018-09-03 17:09:54 +03:00
memory_data_sink.hh
utils: Introduce memory_data_sink
2018-11-21 14:04:27 +01:00
meta.hh
utils: add metaprogramming helper functions
2018-05-31 10:09:01 +01:00
murmur_hash.cc
Fix pre-ScyllaDB copyright statements
2016-04-08 08:12:47 +03:00
murmur_hash.hh
Fix pre-ScyllaDB copyright statements
2016-04-08 08:12:47 +03:00
mutable_view.hh
mutable_view: add default constructor and const_iterator
2018-05-09 16:52:26 +01:00
observable.hh
observable: allow an observable to disconnect() twice without penalty
2018-07-11 10:15:01 +01:00
overloaded_functor.hh
utils: Add overloaded_functor helper.
2018-07-20 13:50:17 -07:00
phased_barrier.hh
utils: phased_barrier: Make advance_and_await() have strong exception guarantees
2018-11-20 16:15:12 +00:00
preempt.hh
mutation_partition: Make merging preemtable
2018-06-27 12:48:30 +02:00
rate_limiter.cc
Fix pre-ScyllaDB copyright statements
2016-04-08 08:12:47 +03:00
rate_limiter.hh
Update seastar submodule
2018-11-21 00:01:44 +02:00
reusable_buffer.hh
utils/reusable_buffer: do not warn about large allocations
2018-09-30 11:12:23 +03:00
runtime.cc
Fix pre-ScyllaDB copyright statements
2016-04-08 08:12:47 +03:00
runtime.hh
Fix pre-ScyllaDB copyright statements
2016-04-08 08:12:47 +03:00
sequenced_set.hh
Fix pre-ScyllaDB copyright statements
2016-04-08 08:12:47 +03:00
serialization.hh
Update seastar submodule
2018-11-21 00:01:44 +02:00
serialized_action.hh
utils/serialized_action: Introduce trigger_later()
2017-10-18 08:49:52 +02:00
small_vector.hh
utils: Extract small_vector.hh
2018-05-30 14:41:41 +02:00
streaming_histogram.hh
streaming_histogram: fix coding style
2017-06-29 02:08:12 -03:00
top_k.hh
remove exec permission from top_k source files
2018-11-21 18:38:50 +02:00
utf8.cc
utils: optimize UTF-8 validation
2018-12-05 21:51:01 +02:00
utf8.hh
utils: optimize UTF-8 validation
2018-12-05 21:51:01 +02:00
UUID_gen.cc
Fix pre-ScyllaDB copyright statements
2016-04-08 08:12:47 +03:00
UUID_gen.hh
utils: add get_time_UUID(system_clock::time_point)
2016-07-19 18:21:58 +03:00
uuid.cc
Update seastar submodule
2018-11-21 00:01:44 +02:00
UUID.hh
Update seastar submodule
2018-11-21 00:01:44 +02:00
with_relational_operators.hh
tombstone: Extract out relational operators
2017-04-25 11:43:04 +02:00