Pulls
-
PR #141: [IBD] measure all IBD optimizations together
-
PR #53: Alt Batch Verification implementation
-
PR #51: V1Transport::readData 256 MiB -> 512 MiB
-
PR #144: leveldb max file size regression
-
PR #120: build: remove need to test for endianness
-
PR #61: performance-unnecessary-value-param
-
PR #80: paranoid_checks = false [test]
-
PR #78: paranoid_checks = false
-
PR #77: block_size = 8 * 1024
-
PR #107: use a bitcoin core-specific flamegraph-rs
-
PR #137: Collect and sort unspent hashes from cursor
-
PR #88: Add parser & plotter for time, block height & cache size
-
PR #49: Empty commit for testing gh actions
-
PR #114: Group base/head into columns
-
PR #55: blockstorage: Optimize ReadBlockFromDisk with buffered reads
-
PR #74: NewLRUCache(options.max_open_files - 10)
-
PR #72: disable bloom filters
-
PR #108: fixup flame bin names
-
PR #91: [test bench]
-
PR #122: update to bitcoin/bitcoin
-
PR #102: TODO disable `run-assumeutxo` to speed up debugging
-
PR #66: refactor: Cache block[undo] serialized size for consecutive calls
-
PR #119: bench: measure `ProcessTransaction` performance for block413567 txs
-
PR #82: nMaxCoinsDBCache = 256
-
PR #41: Batch verification benchcoin
-
PR #90: Add parser & plotter for time, block height & cache size
-
PR #73: paranoid_checks = false
-
PR #44: Disable parallel script checking for testing purposes
-
PR #69: experiment: write_buffer_size -> nCacheSize * 2
-
PR #94: SortedBulkWrite
-
PR #138: Custom CoinEntry serialization
-
PR #64: refactor: Cache blockundo serialized size for consecutive calls
-
PR #68: experiment: write_buffer_size -> nCacheSize / 2
-
PR #65: Expriment
-
PR #89: optimization: assume validity happy-path in LogEscapeMessage
-
PR #83: cache CTransaction's serialized size (with/no witness)
-
PR #87: cache serialized block sizes
-
PR #50: hyperfine truthy
-
PR #71: max_open_files = 2000
-
PR #59: Optimize xor bench
-
PR #92: nDefaultDbBatchSize = 64 << 20
-
PR #19: Bench test
-
PR #118: Skip duplicate check for single inputs, skip set creation for 2 and optimize duplicate check for >=3
-
PR #113: Get total memory for every UpdateTipLog
-
PR #67: optimization: Buffer serialization reads/writes
-
PR #54: util/blockstorage: Add TRACE_RAII, slightly faster -reindex-chainstate with BufferedFile
-
PR #140: Benchmark PR26776
-
PR #109: validation: write chainstate to disk every ~hour~ 10 minutes
-
PR #86: test: block size
-
PR #135: Get IBD results with -disableschnorr param applied
-
PR #112: Add commit id to the plots to make sure they're not overwritten
-
PR #133: Test -O3 optimisations
-
PR #139: memusage: move MallocUsage into separate file
-
PR #57: Experiment: Reuse coin fetches in CheckTxInputs
-
PR #70: experiment: NewLRUCache(nCacheSize)
-
PR #136: WriteBatch::Reserve
-
PR #110: improve MallocUsage() accuracy
-
PR #129: 2502 ctx checks in connectblock
-
PR #131: dirty validation thread test v3
-
PR #116: reduce template bloat in primitive serialization
-
PR #81: Buffered Xor
-
PR #60: coins: Add move operations to CCoinsCacheEntry and Coin
-
PR #52: ser_writedata with ParamsStream of SizeComputer
-
PR #76: block_size = 2MiB
-
PR #84: assumeutxo benchmarking patch
-
PR #93: nCacheSize / 2
-
PR #117: Perf impr experiment
-
PR #43: util::Xor benchmark test
-
PR #123: optimization: buffer reads(23%)/writes(290%) in [undo]block [de]serialization, 6% faster IBD
-
PR #58: Show mainnet results first
-
PR #62: test: chainstate write test for periodic chainstate flush
-
PR #75: Dummy change to measure stability
-
PR #111: dirty validation thread
-
PR #45: max_file_size = 16 << 20
-
PR #79: try DCMAKE_BUILD_TYPE=Release
-
PR #121: dirty validation thread
-
PR #63: Xor with BufferedReadOnlyFile
-
PR #56: Experiment: Disable fsync during IBD