aboutsummaryrefslogtreecommitdiff
path: root/src/index
diff options
context:
space:
mode:
authorfanquake <fanquake@gmail.com>2022-07-07 17:55:15 +0100
committerfanquake <fanquake@gmail.com>2022-07-07 17:59:02 +0100
commit5abbc9afec1874efe139ccdc988b4ae122aa472c (patch)
tree04f904dd41e81721e30f8ab554c117bf38c7712b /src/index
parentd571cf2d2421c6f8efb2b61ca844034eaf230945 (diff)
parente734228d8585c0870c71ce8ba8c037f8cf8b249a (diff)
downloadbitcoin-5abbc9afec1874efe139ccdc988b4ae122aa472c.tar.xz
Merge bitcoin/bitcoin#24832: index: Verify the block filter hash when reading the filter from disk.
e734228d8585c0870c71ce8ba8c037f8cf8b249a Update GCSFilter benchmarks (Calvin Kim) aee9a8140b3a58b744766f9e89572f1d953a808b Add GCSFilterDecodeSkipCheck benchmark (Patrick Strateman) 299023c1d9962628d158fac0306f8531506a0123 Add GCSFilterDecode and GCSBlockFilterGetHash benchmarks. (Patrick Strateman) b0a53d50d9142bed51a8372eeb848816bfa94da8 Make sanity check in GCSFilter constructor optional (Patrick Strateman) Pull request description: This PR picks up the abandoned #19280 BlockFilterIndex was depending on `GolombRiceDecode()` during the filter decode to sanity check that the filter wasn't corrupt. However, we can check for corruption by ensuring that the encoded blockfilter's hash matches up with the one stored in the index database. Benchmarks that were added in #19280 showed that checking the hash is much faster. The benchmarks were changed to nanobench and the relevant benchmarks were like below, showing a clear win for the hash check method. ``` | ns/elem | elem/s | err% | ins/elem | bra/elem | miss% | total | benchmark |--------------------:|--------------------:|--------:|----------------:|---------------:|--------:|----------:|:---------- | 531.40 | 1,881,819.43 | 0.3% | 3,527.01 | 411.00 | 0.2% | 0.01 | `DecodeCheckedGCSFilter` | 258,220.50 | 3,872.66 | 0.1% | 2,990,092.00 | 586,706.00 | 1.7% | 0.01 | `DecodeGCSFilter` | 13,036.77 | 76,706.09 | 0.3% | 64,238.24 | 513.04 | 0.2% | 0.01 | `BlockFilterGetHash` ``` ACKs for top commit: mzumsande: Code Review ACK e734228d8585c0870c71ce8ba8c037f8cf8b249a theStack: Code-review ACK e734228d8585c0870c71ce8ba8c037f8cf8b249a stickies-v: ACK e734228d8585c0870c71ce8ba8c037f8cf8b249a ryanofsky: Code review ACK e734228d8585c0870c71ce8ba8c037f8cf8b249a, with caveat that I mostly paid attention to the main code, not the changes to the benchmark. Only changes since last review were changes to the benchmark code. Tree-SHA512: 02b86eab7b554e1a57a15b17a4d6d71faa91b556c637b0da29f0c9ee76597a110be8e3b4d0c158d4cab04af0623de18b764837be0ec2a72afcfe1ad9c78a83c6
Diffstat (limited to 'src/index')
-rw-r--r--src/index/blockfilterindex.cpp13
-rw-r--r--src/index/blockfilterindex.h2
2 files changed, 10 insertions, 5 deletions
diff --git a/src/index/blockfilterindex.cpp b/src/index/blockfilterindex.cpp
index c92b8c7e19..e7fad8eb64 100644
--- a/src/index/blockfilterindex.cpp
+++ b/src/index/blockfilterindex.cpp
@@ -5,6 +5,7 @@
#include <map>
#include <dbwrapper.h>
+#include <hash.h>
#include <index/blockfilterindex.h>
#include <node/blockstorage.h>
#include <util/system.h>
@@ -143,18 +144,22 @@ bool BlockFilterIndex::CommitInternal(CDBBatch& batch)
return BaseIndex::CommitInternal(batch);
}
-bool BlockFilterIndex::ReadFilterFromDisk(const FlatFilePos& pos, BlockFilter& filter) const
+bool BlockFilterIndex::ReadFilterFromDisk(const FlatFilePos& pos, const uint256& hash, BlockFilter& filter) const
{
CAutoFile filein(m_filter_fileseq->Open(pos, true), SER_DISK, CLIENT_VERSION);
if (filein.IsNull()) {
return false;
}
+ // Check that the hash of the encoded_filter matches the one stored in the db.
uint256 block_hash;
std::vector<uint8_t> encoded_filter;
try {
filein >> block_hash >> encoded_filter;
- filter = BlockFilter(GetFilterType(), block_hash, std::move(encoded_filter));
+ uint256 result;
+ CHash256().Write(encoded_filter).Finalize(result);
+ if (result != hash) return error("Checksum mismatch in filter decode.");
+ filter = BlockFilter(GetFilterType(), block_hash, std::move(encoded_filter), /*skip_decode_check=*/true);
}
catch (const std::exception& e) {
return error("%s: Failed to deserialize block filter from disk: %s", __func__, e.what());
@@ -381,7 +386,7 @@ bool BlockFilterIndex::LookupFilter(const CBlockIndex* block_index, BlockFilter&
return false;
}
- return ReadFilterFromDisk(entry.pos, filter_out);
+ return ReadFilterFromDisk(entry.pos, entry.hash, filter_out);
}
bool BlockFilterIndex::LookupFilterHeader(const CBlockIndex* block_index, uint256& header_out)
@@ -425,7 +430,7 @@ bool BlockFilterIndex::LookupFilterRange(int start_height, const CBlockIndex* st
filters_out.resize(entries.size());
auto filter_pos_it = filters_out.begin();
for (const auto& entry : entries) {
- if (!ReadFilterFromDisk(entry.pos, *filter_pos_it)) {
+ if (!ReadFilterFromDisk(entry.pos, entry.hash, *filter_pos_it)) {
return false;
}
++filter_pos_it;
diff --git a/src/index/blockfilterindex.h b/src/index/blockfilterindex.h
index 6deff59000..fef8b573e8 100644
--- a/src/index/blockfilterindex.h
+++ b/src/index/blockfilterindex.h
@@ -31,7 +31,7 @@ private:
FlatFilePos m_next_filter_pos;
std::unique_ptr<FlatFileSeq> m_filter_fileseq;
- bool ReadFilterFromDisk(const FlatFilePos& pos, BlockFilter& filter) const;
+ bool ReadFilterFromDisk(const FlatFilePos& pos, const uint256& hash, BlockFilter& filter) const;
size_t WriteFilterToDisk(FlatFilePos& pos, const BlockFilter& filter);
Mutex m_cs_headers_cache;