[CI crash] Hash table emplace operation in distinct transform

[robot-clickhouse]

Stack trace details

The sipHash64(st.trace_full) is 14145905254348956910
The trace is from the master or release branch: True

The query for CIDB to compare the trace with the known one:

WITH
    (
        SELECT groupArrayDistinct(cleanStackTrace(trace_full) AS trace) FROM default.stack_traces
        WHERE sipHash64(trace) IN (14145905254348956910, {ANOTHER_TRACE_HASH}) -- FIXME: replace with the known hash
    ) AS traces,
    1.97 AS alpha,
    stack_frame_weights AS (
        WITH
            (
                SELECT count()
                FROM default.stack_traces
                FINAL
            ) AS total,
            2.0 AS beta,
            3.7 AS gamma
        SELECT
            arrayJoin(cleanStackTrace(trace_full)) AS frame,
            countDistinct(trace_full) AS count,
            log(total / count) AS IDF,
            sigmoid(beta * (IDF - gamma)) AS weight
        FROM default.stack_traces
        FINAL
        GROUP BY frame
    ),
    (SELECT groupArray(weight) AS w, groupArray(frame) AS f FROM stack_frame_weights) AS weights,
    (trace -> arrayMap((_frame, pos) -> (pow(pos, -alpha) * arrayFirst(w, f -> (f = _frame), weights.w, weights.f)), trace, arrayEnumerate(trace))) AS get_trace_weights,
    (arr -> arrayStringConcat(arr, '\n')) AS joinArr

SELECT arraySimilarity(traces[1], traces[2], get_trace_weights(traces[1]) AS weights1, get_trace_weights(traces[2]) AS weights2) AS similarity,
    arrayLevenshteinDistanceWeighted(traces[1], traces[2], weights1, weights2),
    joinArr(traces[1]), joinArr(traces[2]), joinArr(weights1), joinArr(weights2)

The following new stack trace from CI Logs system.crash_log found:

   CRC32Hash::operator()(StringRef) const
   HashTable<StringRef, HashSetCellWithSavedHash<StringRef, DefaultHash<StringRef>, HashTableNoState>, DefaultHash<StringRef>, HashTableGrowerWithPrecalculation<8ul>, Allocator<true, true>>::hash(StringRef const&) const
   void HashTable<StringRef, HashSetCellWithSavedHash<StringRef, DefaultHash<StringRef>, HashTableNoState>, DefaultHash<StringRef>, HashTableGrowerWithPrecalculation<8ul>, Allocator<true, true>>::emplace<DB::ArenaKeyHolder&>(DB::ArenaKeyHolder&, HashSetCellWithSavedHash<StringRef, DefaultHash<StringRef>, HashTableNoState>*&, bool&)
   DB::ColumnsHashing::columns_hashing_impl::EmplaceResultImpl<void> DB::ColumnsHashing::columns_hashing_impl::HashMethodBase<DB::ColumnsHashing::HashMethodString<StringRef, void, true, false, false, false>, StringRef, void, false, false, false>::emplaceImpl<HashSetTable<StringRef, HashSetCellWithSavedHash<StringRef, DefaultHash<StringRef>, HashTableNoState>, DefaultHash<StringRef>, HashTableGrowerWithPrecalculation<8ul>, Allocator<true, true>>, DB::ArenaKeyHolder>(DB::ArenaKeyHolder&, HashSetTable<StringRef, HashSetCellWithSavedHash<StringRef, DefaultHash<StringRef>, HashTableNoState>, DefaultHash<StringRef>, HashTableGrowerWithPrecalculation<8ul>, Allocator<true, true>>&)
   DB::ColumnsHashing::columns_hashing_impl::EmplaceResultImpl<void> DB::ColumnsHashing::columns_hashing_impl::HashMethodBase<DB::ColumnsHashing::HashMethodString<StringRef, void, true, false, false, false>, StringRef, void, false, false, false>::emplaceKey<HashSetTable<StringRef, HashSetCellWithSavedHash<StringRef, DefaultHash<StringRef>, HashTableNoState>, DefaultHash<StringRef>, HashTableGrowerWithPrecalculation<8ul>, Allocator<true, true>>>(HashSetTable<StringRef, HashSetCellWithSavedHash<StringRef, DefaultHash<StringRef>, HashTableNoState>, DefaultHash<StringRef>, HashTableGrowerWithPrecalculation<8ul>, Allocator<true, true>>&, unsigned long, DB::Arena&)
   void DB::DistinctTransform::buildFilter<DB::SetMethodString<HashSetTable<StringRef, HashSetCellWithSavedHash<StringRef, DefaultHash<StringRef>, HashTableNoState>, DefaultHash<StringRef>, HashTableGrowerWithPrecalculation<8ul>, Allocator<true, true>>>>(DB::SetMethodString<HashSetTable<StringRef, HashSetCellWithSavedHash<StringRef, DefaultHash<StringRef>, HashTableNoState>, DefaultHash<StringRef>, HashTableGrowerWithPrecalculation<8ul>, Allocator<true, true>>>&, std::vector<DB::IColumn const*, std::allocator<DB::IColumn const*>> const&, DB::PODArray<char8_t, 4096ul, Allocator<false, false>, 63ul, 64ul>&, unsigned long, DB::SetVariantsTemplate<DB::NonClearableSet>&) const
   DB::DistinctTransform::transform(DB::Chunk&)
   DB::ISimpleTransform::transform(DB::Chunk&, DB::Chunk&)
   DB::ISimpleTransform::work()
   DB::executeJob(DB::ExecutingGraph::Node*, DB::ReadProgressCallback*)
   DB::ExecutionThreadContext::executeTask()
   DB::PipelineExecutor::executeStepImpl(unsigned long, DB::IAcquiredSlot*, std::atomic<bool>*)
   DB::PipelineExecutor::executeSingleThread(unsigned long, DB::IAcquiredSlot*)
   DB::PipelineExecutor::executeImpl(unsigned long, bool)
   DB::PipelineExecutor::execute(unsigned long, bool)
   DB::threadFunction(DB::PullingAsyncPipelineExecutor::Data&, std::shared_ptr<DB::ThreadGroup>, unsigned long, bool)
   operator()
   decltype(std::declval<DB::PullingAsyncPipelineExecutor::pull(DB::Chunk&, unsigned long)::$_0&>()()) std::__invoke[$ABI]<DB::PullingAsyncPipelineExecutor::pull(DB::Chunk&, unsigned long)::$_0&>(DB::PullingAsyncPipelineExecutor::pull(DB::Chunk&, unsigned long)::$_0&)
   decltype(auto) std::__apply_tuple_impl[$ABI]<DB::PullingAsyncPipelineExecutor::pull(DB::Chunk&, unsigned long)::$_0&, std::tuple<>&>(DB::PullingAsyncPipelineExecutor::pull(DB::Chunk&, unsigned long)::$_0&, std::tuple<>&, std::__tuple_indices<...>)
   decltype(auto) std::apply[$ABI]<DB::PullingAsyncPipelineExecutor::pull(DB::Chunk&, unsigned long)::$_0&, std::tuple<>&>(DB::PullingAsyncPipelineExecutor::pull(DB::Chunk&, unsigned long)::$_0&, std::tuple<>&)
   operator()
   decltype(std::declval<DB::PullingAsyncPipelineExecutor::pull(DB::Chunk&, unsigned long)::$_0>()()) std::__invoke[$ABI]<ThreadFromGlobalPoolImpl<true, true>::ThreadFromGlobalPoolImpl<DB::PullingAsyncPipelineExecutor::pull(DB::Chunk&, unsigned long)::$_0>(DB::PullingAsyncPipelineExecutor::pull(DB::Chunk&, unsigned long)::$_0&&)::'lambda'()&>(DB::PullingAsyncPipelineExecutor::pull(DB::Chunk&, unsigned long)::$_0&&)
   void std::__invoke_void_return_wrapper<void, true>::__call[$ABI]<ThreadFromGlobalPoolImpl<true, true>::ThreadFromGlobalPoolImpl<DB::PullingAsyncPipelineExecutor::pull(DB::Chunk&, unsigned long)::$_0>(DB::PullingAsyncPipelineExecutor::pull(DB::Chunk&, unsigned long)::$_0&&)::'lambda'()&>(ThreadFromGlobalPoolImpl<true, true>::ThreadFromGlobalPoolImpl<DB::PullingAsyncPipelineExecutor::pull(DB::Chunk&, unsigned long)::$_0>(DB::PullingAsyncPipelineExecutor::pull(DB::Chunk&, unsigned long)::$_0&&)::'lambda'()&)
   ThreadPoolImpl<std::thread>::ThreadFromThreadPool::worker()
   decltype(*std::declval<ThreadPoolImpl<std::thread>::ThreadFromThreadPool*>().*std::declval<void (ThreadPoolImpl<std::thread>::ThreadFromThreadPool::*)()>()()) std::__invoke[$ABI]<void (ThreadPoolImpl<std::thread>::ThreadFromThreadPool::*)(), ThreadPoolImpl<std::thread>::ThreadFromThreadPool*, void>(void (ThreadPoolImpl<std::thread>::ThreadFromThreadPool::*&&)(), ThreadPoolImpl<std::thread>::ThreadFromThreadPool*&&)
   void std::__thread_execute[$ABI]<std::unique_ptr<std::__thread_struct, std::default_delete<std::__thread_struct>>, void (ThreadPoolImpl<std::thread>::ThreadFromThreadPool::*)(), ThreadPoolImpl<std::thread>::ThreadFromThreadPool*, 2ul>(std::tuple<std::unique_ptr<std::__thread_struct, std::default_delete<std::__thread_struct>>, void (ThreadPoolImpl<std::thread>::ThreadFromThreadPool::*)(), ThreadPoolImpl<std::thread>::ThreadFromThreadPool*>&, std::__tuple_indices<2ul>)
   void* std::__thread_proxy[$ABI]<std::tuple<std::unique_ptr<std::__thread_struct, std::default_delete<std::__thread_struct>>, void (ThreadPoolImpl<std::thread>::ThreadFromThreadPool::*)(), ThreadPoolImpl<std::thread>::ThreadFromThreadPool*>>(void*)

Possible causes:

• Incorrect hash calculation leading to incorrect key placement
• Mismatch between key type and hash function
• Improper handling of string references in hash table
• Incorrect memory management in arena key holder
• Hash table overflow or incorrect resizing parameters

The stack trace appeared in the following checks:

• AST fuzzer (amd_debug)

[robot-ch-test-poll4]

Stack trace details

The sipHash64(st.trace_full) is 8045388553872222062
The trace is from the master or release branch: False
Similarity score: 0.80587

The query for CIDB to compare the trace with the known one:

WITH
    (
        SELECT groupArrayDistinct(cleanStackTrace(trace_full) AS trace) FROM default.stack_traces
        WHERE sipHash64(trace) IN (8045388553872222062, {ANOTHER_TRACE_HASH}) -- FIXME: replace with the known hash
    ) AS traces,
    1.97 AS alpha,
    stack_frame_weights AS (
        WITH
            (
                SELECT count()
                FROM default.stack_traces
                FINAL
            ) AS total,
            2.0 AS beta,
            3.7 AS gamma
        SELECT
            arrayJoin(cleanStackTrace(trace_full)) AS frame,
            countDistinct(trace_full) AS count,
            log(total / count) AS IDF,
            sigmoid(beta * (IDF - gamma)) AS weight
        FROM default.stack_traces
        FINAL
        GROUP BY frame
    ),
    (SELECT groupArray(weight) AS w, groupArray(frame) AS f FROM stack_frame_weights) AS weights,
    (trace -> arrayMap((_frame, pos) -> (pow(pos, -alpha) * arrayFirst(w, f -> (f = _frame), weights.w, weights.f)), trace, arrayEnumerate(trace))) AS get_trace_weights,
    (arr -> arrayStringConcat(arr, '\n')) AS joinArr

SELECT arraySimilarity(traces[1], traces[2], get_trace_weights(traces[1]) AS weights1, get_trace_weights(traces[2]) AS weights2) AS similarity,
    arrayLevenshteinDistanceWeighted(traces[1], traces[2], weights1, weights2),
    joinArr(traces[1]), joinArr(traces[2]), joinArr(weights1), joinArr(weights2)

The following new stack trace from CI Logs system.crash_log found:

   CRC32Hash::operator()(StringRef) const
   HashTable<StringRef, HashSetCellWithSavedHash<StringRef, DefaultHash<StringRef>, HashTableNoState>, DefaultHash<StringRef>, HashTableGrowerWithPrecalculation<8ul>, Allocator<true, true>>::hash(StringRef const&) const
   HashTable<StringRef, HashSetCellWithSavedHash<StringRef, DefaultHash<StringRef>, HashTableNoState>, DefaultHash<StringRef>, HashTableGrowerWithPrecalculation<8ul>, Allocator<true, true>>::find(StringRef const&)
   DB::ColumnsHashing::columns_hashing_impl::FindResultImpl<void, false> DB::ColumnsHashing::columns_hashing_impl::HashMethodBase<DB::ColumnsHashing::HashMethodString<StringRef, void, true, false, false, false>, StringRef, void, false, false, false>::findKeyImpl<HashSetTable<StringRef, HashSetCellWithSavedHash<StringRef, DefaultHash<StringRef>, HashTableNoState>, DefaultHash<StringRef>, HashTableGrowerWithPrecalculation<8ul>, Allocator<true, true>>, StringRef>(StringRef, HashSetTable<StringRef, HashSetCellWithSavedHash<StringRef, DefaultHash<StringRef>, HashTableNoState>, DefaultHash<StringRef>, HashTableGrowerWithPrecalculation<8ul>, Allocator<true, true>>&)
   DB::ColumnsHashing::columns_hashing_impl::FindResultImpl<void, false> DB::ColumnsHashing::columns_hashing_impl::HashMethodBase<DB::ColumnsHashing::HashMethodString<StringRef, void, true, false, false, false>, StringRef, void, false, false, false>::findKey<HashSetTable<StringRef, HashSetCellWithSavedHash<StringRef, DefaultHash<StringRef>, HashTableNoState>, DefaultHash<StringRef>, HashTableGrowerWithPrecalculation<8ul>, Allocator<true, true>>>(HashSetTable<StringRef, HashSetCellWithSavedHash<StringRef, DefaultHash<StringRef>, HashTableNoState>, DefaultHash<StringRef>, HashTableGrowerWithPrecalculation<8ul>, Allocator<true, true>>&, unsigned long, DB::Arena&)
   void DB::Set::executeImplCase<DB::SetMethodString<HashSetTable<StringRef, HashSetCellWithSavedHash<StringRef, DefaultHash<StringRef>, HashTableNoState>, DefaultHash<StringRef>, HashTableGrowerWithPrecalculation<8ul>, Allocator<true, true>>>, true>(DB::SetMethodString<HashSetTable<StringRef, HashSetCellWithSavedHash<StringRef, DefaultHash<StringRef>, HashTableNoState>, DefaultHash<StringRef>, HashTableGrowerWithPrecalculation<8ul>, Allocator<true, true>>>&, std::vector<DB::IColumn const*, std::allocator<DB::IColumn const*>> const&, DB::PODArray<char8_t, 4096ul, Allocator<false, false>, 63ul, 64ul>&, bool, unsigned long, DB::PODArray<char8_t, 4096ul, Allocator<false, false>, 63ul, 64ul> const*) const
   DB::Set::executeOrdinary(std::vector<DB::IColumn const*, std::allocator<DB::IColumn const*>> const&, DB::PODArray<char8_t, 4096ul, Allocator<false, false>, 63ul, 64ul>&, bool, DB::PODArray<char8_t, 4096ul, Allocator<false, false>, 63ul, 64ul> const*) const
   DB::Set::execute(std::vector<DB::ColumnWithTypeAndName, std::allocator<DB::ColumnWithTypeAndName>> const&, bool) const
   DB::FunctionToExecutableFunctionAdaptor::executeImpl(std::vector<DB::ColumnWithTypeAndName, std::allocator<DB::ColumnWithTypeAndName>> const&, std::shared_ptr<DB::IDataType const> const&, unsigned long) const
   DB::executeAction(DB::ExpressionActions::Action const&, DB::(anonymous namespace)::ExecutionContext&, bool, bool)
   DB::ExpressionActions::execute(DB::Block&, unsigned long&, bool, bool) const
   DB::FilterTransform::doTransform(DB::Chunk&)
   DB::FilterTransform::transform(DB::Chunk&)
   DB::ISimpleTransform::transform(DB::Chunk&, DB::Chunk&)
   DB::ISimpleTransform::work()
   DB::executeJob(DB::ExecutingGraph::Node*, DB::ReadProgressCallback*)
   DB::ExecutionThreadContext::executeTask()
   DB::PipelineExecutor::executeStepImpl(unsigned long, DB::IAcquiredSlot*, std::atomic<bool>*)
   DB::PipelineExecutor::executeSingleThread(unsigned long, DB::IAcquiredSlot*)
   operator()
   decltype(std::declval<DB::PipelineExecutor::spawnThreads(std::shared_ptr<DB::IAcquiredSlot>)::$_0&>()()) std::__invoke[$ABI]<DB::PipelineExecutor::spawnThreads(std::shared_ptr<DB::IAcquiredSlot>)::$_0&>(DB::PipelineExecutor::spawnThreads(std::shared_ptr<DB::IAcquiredSlot>)::$_0&)
   void std::__invoke_void_return_wrapper<void, true>::__call[$ABI]<DB::PipelineExecutor::spawnThreads(std::shared_ptr<DB::IAcquiredSlot>)::$_0&>(DB::PipelineExecutor::spawnThreads(std::shared_ptr<DB::IAcquiredSlot>)::$_0&)
   ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool::worker()
   decltype(*std::declval<ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool*&>().*std::declval<void (ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool::*&)()>()()) std::__invoke[$ABI]<void (ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool::*&)(), ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool*&, void>(void (ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool::*&)(), ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool*&)
   decltype(auto) std::__apply_tuple_impl[$ABI]<void (ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool::*&)(), std::tuple<ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool*>&, 0ul>(void (ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool::*&)(), std::tuple<ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool*>&, std::__tuple_indices<0ul>)
   decltype(auto) std::apply[$ABI]<void (ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool::*&)(), std::tuple<ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool*>&>(void (ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool::*&)(), std::tuple<ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool*>&)
   ThreadFromGlobalPoolImpl<false, true>::ThreadFromGlobalPoolImpl<void (ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool::*)(), ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool*>(void (ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool::*&&)(), ThreadPoolImpl<ThreadFromGlobalPoolImpl<false, true>>::ThreadFromThreadPool*&&)::'lambda'()::operator()()
   ThreadPoolImpl<std::thread>::ThreadFromThreadPool::worker()
   decltype(*std::declval<ThreadPoolImpl<std::thread>::ThreadFromThreadPool*>().*std::declval<void (ThreadPoolImpl<std::thread>::ThreadFromThreadPool::*)()>()()) std::__invoke[$ABI]<void (ThreadPoolImpl<std::thread>::ThreadFromThreadPool::*)(), ThreadPoolImpl<std::thread>::ThreadFromThreadPool*, void>(void (ThreadPoolImpl<std::thread>::ThreadFromThreadPool::*&&)(), ThreadPoolImpl<std::thread>::ThreadFromThreadPool*&&)
   void std::__thread_execute[$ABI]<std::unique_ptr<std::__thread_struct, std::default_delete<std::__thread_struct>>, void (ThreadPoolImpl<std::thread>::ThreadFromThreadPool::*)(), ThreadPoolImpl<std::thread>::ThreadFromThreadPool*, 2ul>(std::tuple<std::unique_ptr<std::__thread_struct, std::default_delete<std::__thread_struct>>, void (ThreadPoolImpl<std::thread>::ThreadFromThreadPool::*)(), ThreadPoolImpl<std::thread>::ThreadFromThreadPool*>&, std::__tuple_indices<2ul>)
   void* std::__thread_proxy[$ABI]<std::tuple<std::unique_ptr<std::__thread_struct, std::default_delete<std::__thread_struct>>, void (ThreadPoolImpl<std::thread>::ThreadFromThreadPool::*)(), ThreadPoolImpl<std::thread>::ThreadFromThreadPool*>>(void*)

Possible causes:

• Incorrect hash calculation leading to incorrect lookup in the hash table
• Hash collision or incorrect hash value for the key being looked up
• Misconfiguration or improper setup of the hash table parameters
• Data inconsistency or invalid input causing unexpected hash values

The stack trace appeared in the following checks:

• Stateless tests (amd_binary, ParallelReplicas, s3 storage, parallel)
• Stateless tests (amd_binary, old analyzer, s3 storage, DatabaseReplicated, parallel)

[PedroTadim]

This was fixed by #86270