Repository navigation
backport: v0.26 bitcoin#28385, bitcoin#28656 - #63
Conversation
|
This pull request has conflicts, please rebase. |
6e7dd1f to
7dc77e5
Compare
|
Branch rewritten at
🤖 backportsys, on behalf of the Dash backport pipeline. |
I have a fix ready, but re-gating the branch fails in bitcoin#25284 (verify: Removal of the legacy CHashVerifier class (upstream commit fa626af 'Remove unused legacy CHashVerifier'). Dash can't drop it yet: src/node/blockstorage.cpp UndoReadFromDisk still uses CHashVerifier, and upstream moved that to HashVerifier/AutoFile in bitcoin#26649 (eeee610). The Dash-only src/evo/snapshot.h also still uses it and will need its own adaptation.), a commit the fix does not touch. That backport is being dropped from the branch and I will apply the fix on the rewritten branch. 🤖 backportsys, on behalf of the Dash backport pipeline. |
I have a fix ready, but re-gating the branch fails in bitcoin#25284 (verify: Removal of the CHashVerifier class. Upstream can drop it because bitcoin#26649 had already moved every caller to HashVerifier/AutoFile. Dash never landed bitcoin#26649: src/node/blockstorage.cpp:558 still uses CHashVerifier in UndoReadFromDisk, and Dash-only src/evo/snapshot.h:170/434 uses it too. So the class had to stay, and the HashVerifier class that this PR's addrdb.cpp/addrman.cpp), a commit the fix does not touch. That backport is being dropped from the branch and I will apply the fix on the rewritten branch. 🤖 backportsys, on behalf of the Dash backport pipeline. |
I have a fix ready, but re-gating the branch fails in bitcoin#25284 (verify: Upstream commit fa626af ('Remove unused legacy CHashVerifier') deletes the CHashVerifier class. It is dead upstream but still used in Dash: node/blockstorage.cpp:558 (UndoReadFromDisk) and Dash-only evo/snapshot.h:170/434 (SnapshotBoundedInput relies on its GetType()/GetVersion() forwarding). Keeping it is correct, and the rest of bitcoin#25284 doesn't depend on its removal.; bitcoin#25284 does not add t), a commit the fix does not touch. That backport is being dropped from the branch and I will apply the fix on the rewritten branch. 🤖 backportsys, on behalf of the Dash backport pipeline. |
I have a fix ready, but re-gating the branch fails in bitcoin#25284 (verify: Removal of the legacy CHashVerifier 🤖 backportsys, on behalf of the Dash backport pipeline. |
I have a fix ready, but re-gating the branch fails in bitcoin#25284 (verify: Upstream commit fa626af ('Remove unused legacy CHashVerifier') deletes the CHashVerifier class. Dash still uses it: node/blockstorage.cpp:558 (UndoReadFromDisk), and Dash-only evo/snapshot.h:170 (SnapshotBoundedInput) and evo/snapshot.h:434, which depend on the GetType()/GetVersion() forwarding that HashVerifier does not have. Keeping the class is correct.; bitcoin#25284 does not add this class; its diff), a commit the fix does not touch. That backport is being dropped from the branch and I will apply the fix on the rewritten branch. 🤖 backportsys, on behalf of the Dash backport pipeline. |
I have a fix ready, but re-gating the branch fails in bitcoin#25284 (verify: Removal of the legacy CHashVerifier class (upstream commit fa626af, 'Remove unused legacy CHashVerifier'). Dash keeps it because node/blockstorage.cpp:558 still uses CHashVerifier. That call site is what bitcoin#26649 converts to AutoFile/HashVerifier. Dash-only evo/snapshot.h (SnapshotBoundedInput) will also need its own conversion before the class can go.), a commit the fix does not touch. That backport is being dropped from the branch and I will apply the fix on the rewritten branch. 🤖 backportsys, on behalf of the Dash backport pipeline. |
I have a fix ready, but re-gating the branch fails in bitcoin#25284 (verify: Upstream commit fa626af removes the legacy CHashVerifier class. Dash keeps it and the commit message calls this 'not for Dash'. But the class is only still needed because the change that converted its users was never backported: node/blockstorage.cpp:558 (UndoReadFromDisk) still uses CHashVerifier. evo/snapshot.h:170 and :434 also use it, and those are Dash-only. So this is a deferral, ), a commit the fix does not touch. That backport is being dropped from the branch and I will apply the fix on the rewritten branch. 🤖 backportsys, on behalf of the Dash backport pipeline. |
7dc77e5 to
20b8d98
Compare
The bench file this backport added passed temporary objects to emplace_back, which Dash's clang-tidy modernize-use-emplace check treats as an error. Lines 39-40 now pass the constructor arguments directly (tx.vin.emplace_back(COutPoint{prevout_hash, 0}); tx.vout.emplace_back(CENT, spk);), matching later upstream master. Not built locally because the tree isn't configured yet. (folded into the 🤖 backportsys, on behalf of the Dash backport pipeline. |
20b8d98 to
65a889c
Compare
|
Branch rewritten at
🤖 backportsys, on behalf of the Dash backport pipeline. |
|
This pull request has conflicts, please rebase. |
65a889c to
e4b82ec
Compare
e4b82ec to
40f3424
Compare
3325156 to
40f3424
Compare
… to not use boost 4313c77 make DisconnectedBlockTransactions responsible for its own memory management (glozow) cf5f1fa MOVEONLY: DisconnectedBlockTransactions to its own file (glozow) 2765d6f rewrite DisconnectedBlockTransactions as a list + map (glozow) 79ce9f0 add std::list to memusage (glozow) 59a35a7 [bench] DisconnectedBlockTransactions (glozow) 925bb72 [refactor] batch-add transactions to DisconnectedBlockTransactions (glozow) Pull request description: Motivation - I think it's preferable to use stdlib data structures instead of depending on boost if we can achieve the same thing. - Also see bitcoin#28335 for further context/motivation. This PR simplifies that one. Things done in this PR: - Add a bench for `DisconnectedBlockTransactions` where we reorg and the new chain has {100%, 90%, 10%} of the same transactions. AFAIU in practice, it's usually close to 100%. - Rewrite `DisconnectedBlockTransactions` as a `std::list` + `unordered_map` instead of a boost multi index container. - On my machine, the bench suggests the performance is very similar. - Move `DisconnectedBlockTransactions` from txmempool.h to its own kernel/disconnected_transactions.h. This struct isn't used by txmempool and doesn't have much to do with txmempool. My guess is that it's been living there for convenience since the boost includes are there. ACKs for top commit: ismaelsadeeq: Tested ACK 4313c77 stickies-v: ACK 4313c77 TheCharlatan: ACK 4313c77 Tree-SHA512: 273c80866bf3acd39b2a039dc082b7719d2d82e0940e1eb6c402f1c0992e997256722b85c7e310c9811238a770cfbdeb122ea4babbc23835d17128f214a1ef9e Dash adaptations: - src/Makefile.am: added only `kernel/disconnected_transactions.h`; upstream's neighbouring `kernel/cs_main.h` context line does not exist in Dash's BITCOIN_CORE_H list. - src/test/validation_chainstatemanager_tests.cpp: added `#include <kernel/disconnected_transactions.h>` in alphabetical position among Dash's evo/llmq/node includes; upstream's `node/kernel_notifications.h` context line is not present in Dash and was not introduced. - src/test/validation_chainstatemanager_tests.cpp: also converted the Dash-only second call site in `chainstatemanager_evodb_snapshot_only_flush_restart` (`DisconnectedBlockTransactions unused_pool{MAX_DISCONNECTED_TX_POOL_SIZE * 1000}`), which upstream has no equivalent of. - src/validation.cpp: added only `#include <kernel/disconnected_transactions.h>`; upstream's adjacent `kernel/chainparams.h` context line is not part of Dash's include block. - src/validation.cpp: in MaybeUpdateMempoolForReorg kept Dash's `m_mempool->exists((*it)->GetHash())` instead of upstream's `exists(GenTxid::Txid(...))` — Dash's CTxMemPool::exists has no GenTxid overload at this commit. - src/validation.cpp: updated the Dash-only ChainLocks call site `Chainstate::MarkConflictingBlock` to construct `DisconnectedBlockTransactions disconnectpool{MAX_DISCONNECTED_TX_POOL_SIZE * 1000}`, matching the new mandatory-argument constructor; upstream never touched this function. - src/validation.h: dropped upstream's duplicate `struct PrecomputedTransactionData;` line — Dash already forward-declares it a few lines above; only the `struct`→`class DisconnectedBlockTransactions` change was taken. Replayed onto a newer base. Dash adaptations: - src/Makefile.am: kept Dash develop's newer `kernel/cs_main.h` entry alongside the added `kernel/disconnected_transactions.h`; upstream's context had no cs_main.h line, which is what produced the conflict - src/Makefile.bench.include: Dash's bench target is `bench_bench_dash_SOURCES` (not `bench_bench_bitcoin_SOURCES`) and the neighbouring entries differ (Dash has `bench/ecdsa.cpp`); the new `bench/disconnected_transactions.cpp` entry is placed in the same alphabetical slot - src/validation.cpp: Dash-only `Chainstate::MarkConflictingBlock` also constructs a `DisconnectedBlockTransactions`; updated to the new `{MAX_DISCONNECTED_TX_POOL_SIZE * 1000}` constructor even though upstream never touched that call site - src/validation.cpp: `m_mempool->exists((*it)->GetHash())` kept as-is rather than upstream's `exists(GenTxid::Txid(...))` — Dash's CTxMemPool::exists still takes a bare uint256 - src/test/validation_chainstatemanager_tests.cpp: Dash-only test `chainstatemanager_evodb_snapshot_only_flush_restart` has a second `unused_pool`; both occurrences updated to the new constructor
faa5e06 fuzz: Allow multiple --m_dir args (MarcoFalke) Pull request description: This allows to merge the result from several servers (or just several folders) at the same time, instead of having to iterate over them. This should also allow the fuzz engine (libFuzzer) to optimize the final merge result more, because all fuzz inputs from all folders are available at the same time. ACKs for top commit: dergoegge: tACK faa5e06 Tree-SHA512: bf0da418b1f7b8a8af16bb7cc1e148b1ccd0f17062ce70758d1ca5b35c3eee77c0c30377d376befdd55480adfd1f1a1073cfc47118e7a710e6760e020abe24bb Dash adaptations: - test/fuzz/test_runner.py: merge_inputs() keeps Dash's `fuzz_bin=fuzz_bin` parameter (signature + call site) instead of upstream's `build_dir=config["environment"]["BUILDDIR"]`; Dash has not backported the build_dir/fuzz_bin refactor (bitcoin#28314), and that line is unchanged context in the upstream diff, so only the merge_dir->merge_dirs part of those lines was taken
40f3424 to
9dde9e4
Compare
|
merged as dashpay#7735 |
Automated Bitcoin Core v0.26 backports, batch
backport-0.26-b049-misc.9871dcc79040f3424a08Provenance
Each commit passed: cherry-pick (adapted by an Opus lane only where conflicts existed), build, touched tests, a mechanical diff-of-diffs check (every upstream hunk landed; no added line without an upstream counterpart), and an independent Opus verification lane where anything was adapted. Gate rows and lane artifacts are in the backportsys DB.