Merge branch '3.0' of https://github.com/taosdata/TDengine into feat/vnode_compact

cmake/taosadapter_CMakeLists.txt.in

@@ -2,7 +2,7 @@
 # taosadapter
 ExternalProject_Add(taosadapter
         GIT_REPOSITORY https://github.com/taosdata/taosadapter.git
-        GIT_TAG f0c1753
+        GIT_TAG 5662a6d
         SOURCE_DIR "${TD_SOURCE_DIR}/tools/taosadapter"
         BINARY_DIR ""
         #BUILD_IN_SOURCE TRUE

docs/en/12-taos-sql/02-database.md

@@ -30,6 +30,8 @@ database_option: {
   | WAL_LEVEL {1 | 2}
   | VGROUPS value
   | SINGLE_STABLE {0 | 1}
+  | TABLE_PREFIX value
+  | TABLE_SUFFIX value
   | WAL_RETENTION_PERIOD value
   | WAL_ROLL_PERIOD value
   | WAL_RETENTION_SIZE value
@@ -67,6 +69,8 @@ database_option: {
 - SINGLE_STABLE: specifies whether the database can contain more than one supertable.
   - 0: The database can contain multiple supertables.
   - 1: The database can contain only one supertable.
+- TABLE_PREFIX：The prefix length in the table name that is ignored when distributing table to vnode based on table name.
+- TABLE_SUFFIX：The suffix length in the table name that is ignored when distributing table to vnode based on table name.
 - WAL_RETENTION_PERIOD: specifies the time after which WAL files are deleted. This parameter is used for data subscription. Enter a time in seconds. The default value of single copy is 0. A value of 0 indicates that each WAL file is deleted immediately after its contents are written to disk. -1: WAL files are never deleted. The default value of multiple copy is 4 days.
 - WAL_RETENTION_SIZE: specifies the size at which WAL files are deleted. This parameter is used for data subscription. Enter a size in KB. The default value of single copy is 0. A value of 0 indicates that each WAL file is deleted immediately after its contents are written to disk. -1: WAL files are never deleted. The default value of multiple copy is -1.
 - WAL_ROLL_PERIOD: specifies the time after which WAL files are rotated. After this period elapses, a new WAL file is created. The default value of single copy is 0. A value of 0 indicates that a new WAL file is created only after the previous WAL file was written to disk. The default values of multiple copy is 1 day.

docs/zh/12-taos-sql/02-database.md

@@ -30,6 +30,8 @@ database_option: {
   | WAL_LEVEL {1 | 2}
   | VGROUPS value
   | SINGLE_STABLE {0 | 1}
+  | TABLE_PREFIX value
+  | TABLE_SUFFIX value
   | WAL_RETENTION_PERIOD value
   | WAL_ROLL_PERIOD value
   | WAL_RETENTION_SIZE value
@@ -67,6 +69,8 @@ database_option: {
 - SINGLE_STABLE：表示此数据库中是否只可以创建一个超级表，用于超级表列非常多的情况。
   - 0：表示可以创建多张超级表。
   - 1：表示只可以创建一张超级表。
+- TABLE_PREFIX：内部存储引擎根据表名分配存储该表数据的 VNODE 时要忽略的前缀的长度。
+- TABLE_SUFFIX：内部存储引擎根据表名分配存储该表数据的 VNODE 时要忽略的后缀的长度。
 - WAL_RETENTION_PERIOD：wal 文件的额外保留策略，用于数据订阅。wal 的保存时长，单位为 s。单副本默认为 0，即落盘后立即删除。-1 表示不删除。多副本默认为 4 天。
 - WAL_RETENTION_SIZE：wal 文件的额外保留策略，用于数据订阅。wal 的保存的最大上限，单位为 KB。单副本默认为 0，即落盘后立即删除。多副本默认为-1，表示不删除。
 - WAL_ROLL_PERIOD：wal 文件切换时长，单位为 s。当 wal 文件创建并写入后，经过该时间，会自动创建一个新的 wal 文件。单副本默认为 0，即仅在落盘时创建新文件。多副本默认为 1 天。

include/libs/executor/executor.h

@@ -213,6 +213,7 @@ int32_t qStreamSourceRecoverStep1(qTaskInfo_t tinfo, int64_t ver);
 int32_t qStreamSourceRecoverStep2(qTaskInfo_t tinfo, int64_t ver);
 int32_t qStreamRecoverFinish(qTaskInfo_t tinfo);
 int32_t qStreamRestoreParam(qTaskInfo_t tinfo);
+bool    qStreamRecoverScanFinished(qTaskInfo_t tinfo);
 
 #ifdef __cplusplus
 }

source/dnode/vnode/src/meta/metaQuery.c

@@ -19,7 +19,7 @@ void metaReaderInit(SMetaReader *pReader, SMeta *pMeta, int32_t flags) {
   memset(pReader, 0, sizeof(*pReader));
   pReader->flags = flags;
   pReader->pMeta = pMeta;
-  if (!(flags & META_READER_NOLOCK)) {
+  if (pReader->pMeta && !(flags & META_READER_NOLOCK)) {
     metaRLock(pMeta);
   }
 }
@@ -152,7 +152,7 @@ bool metaIsTableExist(SMeta *pMeta, tb_uid_t uid) {
 }
 
 int metaGetTableEntryByUid(SMetaReader *pReader, tb_uid_t uid) {
-  SMeta *pMeta = pReader->pMeta;
+  SMeta  *pMeta = pReader->pMeta;
   int64_t version1;
 
   // query uid.idx
@@ -239,7 +239,6 @@ int metaGetTableSzNameByUid(void *meta, uint64_t uid, char *tbName) {
   return 0;
 }
 
-
 int metaGetTableUidByName(void *meta, char *tbName, uint64_t *uid) {
   int         code = 0;
   SMetaReader mr = {0};
@@ -756,9 +755,7 @@ int64_t metaGetTimeSeriesNum(SMeta *pMeta) {
   return pMeta->pVnode->config.vndStats.numOfTimeSeries + pMeta->pVnode->config.vndStats.numOfNTimeSeries;
 }
 
-int64_t metaGetNtbNum(SMeta *pMeta) {
-  return pMeta->pVnode->config.vndStats.numOfNTables;
-}
+int64_t metaGetNtbNum(SMeta *pMeta) { return pMeta->pVnode->config.vndStats.numOfNTables; }
 
 typedef struct {
   SMeta   *pMeta;

source/libs/executor/inc/executorimpl.h

@@ -136,6 +136,7 @@ typedef struct {
   SSchemaWrapper*     schema;
   char                tbName[TSDB_TABLE_NAME_LEN];
   int8_t              recoverStep;
+  int8_t              recoverScanFinished;
   SQueryTableDataCond tableCond;
   int64_t             fillHistoryVer1;
   int64_t             fillHistoryVer2;
@@ -182,7 +183,7 @@ struct SExecTaskInfo {
   SSubplan*             pSubplan;
   struct SOperatorInfo* pRoot;
   SLocalFetch           localFetch;
-  SArray*               pResultBlockList;// result block list
+  SArray*               pResultBlockList;  // result block list
   STaskStopInfo         stopInfo;
 };
 
@@ -199,7 +200,7 @@ typedef struct SOperatorFpSet {
   __optr_fn_t         getNextFn;
   __optr_fn_t         cleanupFn;  // call this function to release the allocated resources ASAP
   __optr_close_fn_t   closeFn;
-  __optr_reqBuf_fn_t  reqBufFn;   // total used buffer for blocking operator
+  __optr_reqBuf_fn_t  reqBufFn;  // total used buffer for blocking operator
   __optr_encode_fn_t  encodeResultRow;
   __optr_decode_fn_t  decodeResultRow;
   __optr_explain_fn_t getExplainFn;
@@ -255,22 +256,22 @@ typedef struct SLimitInfo {
 } SLimitInfo;
 
 typedef struct SExchangeInfo {
-  SArray*             pSources;
-  SArray*             pSourceDataInfo;
-  tsem_t              ready;
-  void*               pTransporter;
+  SArray* pSources;
+  SArray* pSourceDataInfo;
+  tsem_t  ready;
+  void*   pTransporter;
 
   // SArray<SSDataBlock*>, result block list, used to keep the multi-block that
   // passed by downstream operator
-  SArray*             pResultBlockList;
-  SArray*             pRecycledBlocks;// build a pool for small data block to avoid to repeatly create and then destroy.
-  SSDataBlock*        pDummyBlock;    // dummy block, not keep data
-  bool                seqLoadData;    // sequential load data or not, false by default
-  int32_t             current;
+  SArray*      pResultBlockList;
+  SArray*      pRecycledBlocks;  // build a pool for small data block to avoid to repeatly create and then destroy.
+  SSDataBlock* pDummyBlock;      // dummy block, not keep data
+  bool         seqLoadData;      // sequential load data or not, false by default
+  int32_t      current;
   SLoadRemoteDataInfo loadInfo;
   uint64_t            self;
   SLimitInfo          limitInfo;
-  int64_t             openedTs;       // start exec time stamp, todo: move to SLoadRemoteDataInfo
+  int64_t             openedTs;  // start exec time stamp, todo: move to SLoadRemoteDataInfo
 } SExchangeInfo;
 
 typedef struct SScanInfo {
@@ -305,9 +306,9 @@ typedef struct {
 } SAggOptrPushDownInfo;
 
 typedef struct STableMetaCacheInfo {
-  SLRUCache*             pTableMetaEntryCache; // 100 by default
-  uint64_t               metaFetch;
-  uint64_t               cacheHit;
+  SLRUCache* pTableMetaEntryCache;  // 100 by default
+  uint64_t   metaFetch;
+  uint64_t   cacheHit;
 } STableMetaCacheInfo;
 
 typedef struct STableScanBase {
@@ -325,16 +326,16 @@ typedef struct STableScanBase {
 } STableScanBase;
 
 typedef struct STableScanInfo {
-  STableScanBase         base;
-  SScanInfo              scanInfo;
-  int32_t                scanTimes;
-  SSDataBlock*           pResBlock;
-  SSampleExecInfo        sample;  // sample execution info
-  int32_t                currentGroupId;
-  int32_t                currentTable;
-  int8_t                 scanMode;
-  int8_t                 assignBlockUid;
-  bool                   hasGroupByTag;
+  STableScanBase  base;
+  SScanInfo       scanInfo;
+  int32_t         scanTimes;
+  SSDataBlock*    pResBlock;
+  SSampleExecInfo sample;  // sample execution info
+  int32_t         currentGroupId;
+  int32_t         currentTable;
+  int8_t          scanMode;
+  int8_t          assignBlockUid;
+  bool            hasGroupByTag;
 } STableScanInfo;
 
 typedef struct STableMergeScanInfo {
@@ -360,11 +361,11 @@ typedef struct STableMergeScanInfo {
 } STableMergeScanInfo;
 
 typedef struct STagScanInfo {
-  SColumnInfo*    pCols;
-  SSDataBlock*    pRes;
-  SColMatchInfo   matchInfo;
-  int32_t         curPos;
-  SReadHandle     readHandle;
+  SColumnInfo*  pCols;
+  SSDataBlock*  pRes;
+  SColMatchInfo matchInfo;
+  int32_t       curPos;
+  SReadHandle   readHandle;
 } STagScanInfo;
 
 typedef enum EStreamScanMode {
@@ -468,6 +469,11 @@ typedef struct SStreamScanInfo {
   SNodeList* pGroupTags;
   SNode*     pTagCond;
   SNode*     pTagIndexCond;
+
+  // recover
+  int32_t blockRecoverContiCnt;
+  int32_t blockRecoverTotCnt;
+
 } SStreamScanInfo;
 
 typedef struct {
@@ -499,8 +505,8 @@ typedef struct STableCountScanOperatorInfo {
 
   STableCountScanSupp supp;
 
-  int32_t   currGrpIdx;
-  SArray*   stbUidList; // when group by db_name and/or stable_name
+  int32_t currGrpIdx;
+  SArray* stbUidList;  // when group by db_name and/or stable_name
 } STableCountScanOperatorInfo;
 
 typedef struct SOptrBasicInfo {
@@ -678,19 +684,19 @@ void           setOperatorInfo(SOperatorInfo* pOperator, const char* name, int32
 void           destroyOperatorInfo(SOperatorInfo* pOperator);
 int32_t        optrDefaultBufFn(SOperatorInfo* pOperator);
 
-void    initBasicInfo(SOptrBasicInfo* pInfo, SSDataBlock* pBlock);
-void    cleanupBasicInfo(SOptrBasicInfo* pInfo);
+void initBasicInfo(SOptrBasicInfo* pInfo, SSDataBlock* pBlock);
+void cleanupBasicInfo(SOptrBasicInfo* pInfo);
 
 int32_t initExprSupp(SExprSupp* pSup, SExprInfo* pExprInfo, int32_t numOfExpr);
 void    cleanupExprSupp(SExprSupp* pSup);
 
-void    destroyExprInfo(SExprInfo* pExpr, int32_t numOfExprs);
+void destroyExprInfo(SExprInfo* pExpr, int32_t numOfExprs);
 
 int32_t initAggSup(SExprSupp* pSup, SAggSupporter* pAggSup, SExprInfo* pExprInfo, int32_t numOfCols, size_t keyBufSize,
                    const char* pkey);
 void    cleanupAggSup(SAggSupporter* pAggSup);
 
-void    initResultSizeInfo(SResultInfo* pResultInfo, int32_t numOfRows);
+void initResultSizeInfo(SResultInfo* pResultInfo, int32_t numOfRows);
 
 void doBuildStreamResBlock(SOperatorInfo* pOperator, SOptrBasicInfo* pbInfo, SGroupResInfo* pGroupResInfo,
                            SDiskbasedBuf* pBuf);
@@ -803,10 +809,10 @@ void setInputDataBlock(SExprSupp* pExprSupp, SSDataBlock* pBlock, int32_t order,
 
 int32_t checkForQueryBuf(size_t numOfTables);
 
-bool    isTaskKilled(SExecTaskInfo* pTaskInfo);
-void    setTaskKilled(SExecTaskInfo* pTaskInfo, int32_t rspCode);
-void    doDestroyTask(SExecTaskInfo* pTaskInfo);
-void    setTaskStatus(SExecTaskInfo* pTaskInfo, int8_t status);
+bool isTaskKilled(SExecTaskInfo* pTaskInfo);
+void setTaskKilled(SExecTaskInfo* pTaskInfo, int32_t rspCode);
+void doDestroyTask(SExecTaskInfo* pTaskInfo);
+void setTaskStatus(SExecTaskInfo* pTaskInfo, int8_t status);
 
 int32_t createExecTaskInfoImpl(SSubplan* pPlan, SExecTaskInfo** pTaskInfo, SReadHandle* pHandle, uint64_t taskId,
                                char* sql, EOPTR_EXEC_MODEL model);
@@ -828,8 +834,8 @@ bool isDeletedWindow(STimeWindow* pWin, uint64_t groupId, SAggSupporter* pSup);
 bool isDeletedStreamWindow(STimeWindow* pWin, uint64_t groupId, SStreamState* pState, STimeWindowAggSupp* pTwSup);
 void appendOneRowToStreamSpecialBlock(SSDataBlock* pBlock, TSKEY* pStartTs, TSKEY* pEndTs, uint64_t* pUid,
                                       uint64_t* pGp, void* pTbName);
-uint64_t    calGroupIdByData(SPartitionBySupporter* pParSup, SExprSupp* pExprSup, SSDataBlock* pBlock, int32_t rowId);
-void        calBlockTbName(SStreamScanInfo* pInfo, SSDataBlock* pBlock);
+uint64_t calGroupIdByData(SPartitionBySupporter* pParSup, SExprSupp* pExprSup, SSDataBlock* pBlock, int32_t rowId);
+void     calBlockTbName(SStreamScanInfo* pInfo, SSDataBlock* pBlock);
 
 int32_t finalizeResultRows(SDiskbasedBuf* pBuf, SResultRowPosition* resultRowPosition, SExprSupp* pSup,
                            SSDataBlock* pBlock, SExecTaskInfo* pTaskInfo);

source/libs/executor/src/executor.c

@@ -937,6 +937,11 @@ int32_t qStreamRestoreParam(qTaskInfo_t tinfo) {
   return 0;
 }
 
+bool qStreamRecoverScanFinished(qTaskInfo_t tinfo) {
+  SExecTaskInfo* pTaskInfo = (SExecTaskInfo*)tinfo;
+  return pTaskInfo->streamInfo.recoverScanFinished;
+}
+
 void* qExtractReaderFromStreamScanner(void* scanner) {
   SStreamScanInfo* pInfo = scanner;
   return (void*)pInfo->tqReader;

source/libs/executor/src/scanoperator.c

@@ -1785,11 +1785,18 @@ static SSDataBlock* doStreamScan(SOperatorInfo* pOperator) {
     pTSInfo->scanTimes = 0;
     pTSInfo->currentGroupId = -1;
     pTaskInfo->streamInfo.recoverStep = STREAM_RECOVER_STEP__SCAN;
+    pTaskInfo->streamInfo.recoverScanFinished = false;
   }
 
   if (pTaskInfo->streamInfo.recoverStep == STREAM_RECOVER_STEP__SCAN) {
+    if (pInfo->blockRecoverContiCnt > 100) {
+      pInfo->blockRecoverTotCnt += pInfo->blockRecoverContiCnt;
+      pInfo->blockRecoverContiCnt = 0;
+      return NULL;
+    }
     SSDataBlock* pBlock = doTableScan(pInfo->pTableScanOp);
     if (pBlock != NULL) {
+      pInfo->blockRecoverContiCnt++;
       calBlockTbName(pInfo, pBlock);
       if (pInfo->pUpdateInfo) {
         TSKEY maxTs = updateInfoFillBlockData(pInfo->pUpdateInfo, pBlock, pInfo->primaryTsIndex);
@@ -1807,6 +1814,7 @@ static SSDataBlock* doStreamScan(SOperatorInfo* pOperator) {
     pTSInfo->base.cond.startVersion = -1;
     pTSInfo->base.cond.endVersion = -1;
 
+    pTaskInfo->streamInfo.recoverScanFinished = true;
     return NULL;
   }
 

source/libs/parser/test/parExplainToSyncdbTest.cpp

@@ -52,8 +52,8 @@ TEST_F(ParserExplainToSyncdbTest, grant) {
     ASSERT_EQ(string(req.objname), string(expect.objname));
   });
 
-  setAlterUserReq(TSDB_ALTER_USER_ADD_ALL_DB, "wxy", "0.test");
-  run("GRANT ALL ON test.* TO wxy");
+  setAlterUserReq(TSDB_ALTER_USER_ADD_ALL_DB, "wxy", "0.*");
+  run("GRANT ALL ON *.* TO wxy");
 
   setAlterUserReq(TSDB_ALTER_USER_ADD_READ_DB, "wxy", "0.test");
   run("GRANT READ ON test.* TO wxy");
@@ -138,10 +138,38 @@ TEST_F(ParserExplainToSyncdbTest, redistributeVgroup) {
 TEST_F(ParserExplainToSyncdbTest, revoke) {
   useDb("root", "test");
 
-  run("REVOKE ALL ON test.* FROM wxy");
+  SAlterUserReq expect = {0};
+
+  auto setAlterUserReq = [&](int8_t alterType, const string& user, const string& obj) {
+    expect.alterType = alterType;
+    snprintf(expect.user, sizeof(expect.user), "%s", user.c_str());
+    snprintf(expect.objname, sizeof(expect.objname), "%s", obj.c_str());
+  };
+
+  setCheckDdlFunc([&](const SQuery* pQuery, ParserStage stage) {
+    ASSERT_EQ(nodeType(pQuery->pRoot), QUERY_NODE_REVOKE_STMT);
+    ASSERT_EQ(pQuery->pCmdMsg->msgType, TDMT_MND_ALTER_USER);
+    SAlterUserReq req = {0};
+    ASSERT_EQ(tDeserializeSAlterUserReq(pQuery->pCmdMsg->pMsg, pQuery->pCmdMsg->msgLen, &req), TSDB_CODE_SUCCESS);
+    ASSERT_EQ(req.alterType, expect.alterType);
+    ASSERT_EQ(string(req.user), string(expect.user));
+    ASSERT_EQ(string(req.objname), string(expect.objname));
+  });
+
+  setAlterUserReq(TSDB_ALTER_USER_REMOVE_ALL_DB, "wxy", "0.*");
+  run("REVOKE ALL ON *.* FROM wxy");
+
+  setAlterUserReq(TSDB_ALTER_USER_REMOVE_READ_DB, "wxy", "0.test");
   run("REVOKE READ ON test.* FROM wxy");
+
+  setAlterUserReq(TSDB_ALTER_USER_REMOVE_WRITE_DB, "wxy", "0.test");
   run("REVOKE WRITE ON test.* FROM wxy");
+
+  setAlterUserReq(TSDB_ALTER_USER_REMOVE_ALL_DB, "wxy", "0.test");
   run("REVOKE READ, WRITE ON test.* FROM wxy");
+
+  setAlterUserReq(TSDB_ALTER_USER_REMOVE_SUBSCRIBE_TOPIC, "wxy", "0.tp1");
+  run("REVOKE SUBSCRIBE ON tp1 FROM wxy");
 }
 
 // todo syncdb

source/libs/parser/test/parInitialDTest.cpp

@@ -99,7 +99,7 @@ TEST_F(ParserInitialDTest, dropDnode) {
     expect.force = force;
   };
 
-  auto setDropDnodeReqByEndpoint = [&](const char* pFqdn, int32_t port, bool force = false) {
+  auto setDropDnodeReqByEndpoint = [&](const char* pFqdn, int32_t port = tsServerPort, bool force = false) {
     strcpy(expect.fqdn, pFqdn);
     expect.port = port;
     expect.force = force;
@@ -131,6 +131,14 @@ TEST_F(ParserInitialDTest, dropDnode) {
   setDropDnodeReqByEndpoint("host2", 8030, true);
   run("DROP DNODE 'host2:8030' FORCE");
   clearDropDnodeReq();
+
+  setDropDnodeReqByEndpoint("host1");
+  run("DROP DNODE host1");
+  clearDropDnodeReq();
+
+  setDropDnodeReqByEndpoint("host2", tsServerPort, true);
+  run("DROP DNODE host2 FORCE");
+  clearDropDnodeReq();
 }
 
 // todo DROP function
@@ -174,7 +182,21 @@ TEST_F(ParserInitialDTest, dropMnode) {
 TEST_F(ParserInitialDTest, dropQnode) {
   useDb("root", "test");
 
-  run("DROP qnode on dnode 1");
+  SMDropQnodeReq expect = {0};
+
+  auto setDropQnodeReq = [&](int32_t dnodeId) { expect.dnodeId = dnodeId; };
+
+  setCheckDdlFunc([&](const SQuery* pQuery, ParserStage stage) {
+    ASSERT_EQ(nodeType(pQuery->pRoot), QUERY_NODE_DROP_QNODE_STMT);
+    SMDropQnodeReq req = {0};
+    ASSERT_TRUE(TSDB_CODE_SUCCESS ==
+                tDeserializeSCreateDropMQSNodeReq(pQuery->pCmdMsg->pMsg, pQuery->pCmdMsg->msgLen, &req));
+
+    ASSERT_EQ(req.dnodeId, expect.dnodeId);
+  });
+
+  setDropQnodeReq(1);
+  run("DROP QNODE ON DNODE 1");
 }
 
 TEST_F(ParserInitialDTest, dropSnode) {
@@ -237,7 +259,20 @@ TEST_F(ParserInitialDTest, dropUser) {
   login("root");
   useDb("root", "test");
 
-  run("DROP user wxy");
+  SDropUserReq expect = {0};
+
+  auto setDropUserReq = [&](const char* pUser) { sprintf(expect.user, "%s", pUser); };
+
+  setCheckDdlFunc([&](const SQuery* pQuery, ParserStage stage) {
+    ASSERT_EQ(nodeType(pQuery->pRoot), QUERY_NODE_DROP_USER_STMT);
+    SDropUserReq req = {0};
+    ASSERT_TRUE(TSDB_CODE_SUCCESS == tDeserializeSDropUserReq(pQuery->pCmdMsg->pMsg, pQuery->pCmdMsg->msgLen, &req));
+
+    ASSERT_EQ(std::string(req.user), std::string(expect.user));
+  });
+
+  setDropUserReq("wxy");
+  run("DROP USER wxy");
 }
 
 }  // namespace ParserTest

source/libs/stream/src/streamExec.c

@@ -112,7 +112,11 @@ int32_t streamScanExec(SStreamTask* pTask, int32_t batchSz) {
         ASSERT(0);
       }
       if (output == NULL) {
-        finished = true;
+        if (qStreamRecoverScanFinished(exec)) {
+          finished = true;
+        } else {
+          qSetStreamOpOpen(exec);
+        }
         break;
       }
 

source/os/src/osMath.c

@@ -16,6 +16,7 @@
 #define ALLOW_FORBID_FUNC
 #define _DEFAULT_SOURCE
 #include "os.h"
+#include <stdlib.h>
 
 #ifdef WINDOWS
 void swapStr(char* j, char* J, int width) {
@@ -33,16 +34,5 @@ void swapStr(char* j, char* J, int width) {
 
 // todo refactor: 1) move away; 2) use merge sort instead; 3) qsort is not a stable sort actually.
 void taosSort(void* arr, int64_t sz, int64_t width, __compar_fn_t compar) {
-#ifdef WINDOWS
-  int64_t i, j;
-  for (i = 0; i < sz - 1; i++) {
-    for (j = 0; j < sz - 1 - i; j++) {
-      if (compar((char*)arr + j * width, (char*)arr + (j + 1) * width) > 0.00) {
-        swapStr((char*)arr + j * width, (char*)arr + (j + 1) * width, width);
-      }
-    }
-  }
-#else
   qsort(arr, sz, width, compar);
-#endif
 }