ddl: refactor V2 job args of dropping column DDL.

pkg/ddl/column.go

@@ -189,7 +189,12 @@ func onDropColumn(jobCtx *jobContext, t *meta.Meta, job *model.Job) (ver int64,
 		if err != nil {
 			return ver, errors.Trace(err)
 		}
-		job.Args = append(job.Args, indexInfosToIDList(idxInfos))
+		dropColumnArgs, err := model.GetDropColumnArgs(job)
+		if err != nil {
+			return ver, errors.Trace(err)
+		}
+		dropColumnArgs.IndexIDs = indexInfosToIDList(idxInfos)
+		job.FillArgs(dropColumnArgs)
 	case model.StateDeleteOnly:
 		// delete only -> reorganization
 		colInfo.State = model.StateDeleteReorganization
@@ -215,7 +220,12 @@ func onDropColumn(jobCtx *jobContext, t *meta.Meta, job *model.Job) (ver int64,
 		} else {
 			// We should set related index IDs for job
 			job.FinishTableJob(model.JobStateDone, model.StateNone, ver, tblInfo)
-			job.Args = append(job.Args, getPartitionIDs(tblInfo))
+			dropColumnArgs, err := model.GetDropColumnArgs(job)
+			if err != nil {
+				return ver, errors.Trace(err)
+			}
+			dropColumnArgs.PartitionIDs = getPartitionIDs(tblInfo)
+			job.FillArgs(dropColumnArgs)
 		}
 	default:
 		return ver, errors.Trace(dbterror.ErrInvalidDDLJob.GenWithStackByArgs("table", tblInfo.State))
@@ -231,16 +241,13 @@ func checkDropColumn(jobCtx *jobContext, t *meta.Meta, job *model.Job) (*model.T
 		return nil, nil, nil, false, errors.Trace(err)
 	}
 
-	var colName pmodel.CIStr
-	var ifExists bool
-	// indexIDs is used to make sure we don't truncate args when decoding the rawArgs.
-	var indexIDs []int64
-	err = job.DecodeArgs(&colName, &ifExists, &indexIDs)
+	dropColumnArgs, err := model.GetDropColumnArgs(job)
 	if err != nil {
 		job.State = model.JobStateCancelled
 		return nil, nil, nil, false, errors.Trace(err)
 	}
 
+	colName, ifExists := dropColumnArgs.ColName, dropColumnArgs.IfExists
 	colInfo := model.FindColumnInfo(tblInfo.Columns, colName.L)
 	if colInfo == nil || colInfo.Hidden {
 		job.State = model.JobStateCancelled

pkg/ddl/ddl_test.go

@@ -208,13 +208,13 @@ func TestBuildJobDependence(t *testing.T) {
 	}()
 	ctx := kv.WithInternalSourceType(context.Background(), kv.InternalTxnDDL)
 	// Add some non-add-index jobs.
-	job1 := &model.Job{ID: 1, TableID: 1, Type: model.ActionAddColumn}
-	job2 := &model.Job{ID: 2, TableID: 1, Type: model.ActionCreateTable}
-	job3 := &model.Job{ID: 3, TableID: 2, Type: model.ActionDropColumn}
-	job6 := &model.Job{ID: 6, TableID: 1, Type: model.ActionDropTable}
-	job7 := &model.Job{ID: 7, TableID: 2, Type: model.ActionModifyColumn}
-	job9 := &model.Job{ID: 9, SchemaID: 111, Type: model.ActionDropSchema}
-	job11 := &model.Job{Version: model.JobVersion1, ID: 11, TableID: 2, Type: model.ActionRenameTable, Args: []any{int64(111), "old db name"}}
+	job1 := &model.Job{ID: 1, TableID: 1, Version: model.JobVersion1, Type: model.ActionAddColumn}
+	job2 := &model.Job{ID: 2, TableID: 1, Version: model.JobVersion1, Type: model.ActionCreateTable}
+	job3 := &model.Job{ID: 3, TableID: 2, Version: model.JobVersion1, Type: model.ActionDropColumn}
+	job6 := &model.Job{ID: 6, TableID: 1, Version: model.JobVersion1, Type: model.ActionDropTable}
+	job7 := &model.Job{ID: 7, TableID: 2, Version: model.JobVersion1, Type: model.ActionModifyColumn}
+	job9 := &model.Job{ID: 9, SchemaID: 111, Version: model.JobVersion1, Type: model.ActionDropSchema}
+	job11 := &model.Job{ID: 11, TableID: 2, Version: model.JobVersion1, Type: model.ActionRenameTable, Args: []any{int64(111), "old db name"}}
 	err := kv.RunInNewTxn(ctx, store, false, func(ctx context.Context, txn kv.Transaction) error {
 		m := meta.NewMeta(txn)
 		require.NoError(t, m.EnQueueDDLJob(job1))

pkg/ddl/delete_range.go

@@ -29,7 +29,6 @@ import (
 	"github.com/pingcap/tidb/pkg/ddl/util"
 	"github.com/pingcap/tidb/pkg/kv"
 	"github.com/pingcap/tidb/pkg/meta/model"
-	pmodel "github.com/pingcap/tidb/pkg/parser/model"
 	"github.com/pingcap/tidb/pkg/parser/terror"
 	"github.com/pingcap/tidb/pkg/sessionctx"
 	"github.com/pingcap/tidb/pkg/tablecodec"
@@ -400,19 +399,17 @@ func insertJobIntoDeleteRangeTable(ctx context.Context, wrapper DelRangeExecWrap
 			}
 		}
 	case model.ActionDropColumn:
-		var colName pmodel.CIStr
-		var ifExists bool
-		var indexIDs []int64
-		var partitionIDs []int64
-		if err := job.DecodeArgs(&colName, &ifExists, &indexIDs, &partitionIDs); err != nil {
+		args, err := model.GetDropColumnArgs(job)
+		if err != nil {
 			return errors.Trace(err)
 		}
-		if len(indexIDs) > 0 {
-			if len(partitionIDs) == 0 {
-				return errors.Trace(doBatchDeleteIndiceRange(ctx, wrapper, job.ID, job.TableID, indexIDs, ea, "drop column: table ID"))
+
+		if len(args.IndexIDs) > 0 {
+			if len(args.PartitionIDs) == 0 {
+				return errors.Trace(doBatchDeleteIndiceRange(ctx, wrapper, job.ID, job.TableID, args.IndexIDs, ea, "drop column: table ID"))
 			}
-			for _, pid := range partitionIDs {
-				if err := doBatchDeleteIndiceRange(ctx, wrapper, job.ID, pid, indexIDs, ea, "drop column: partition table ID"); err != nil {
+			for _, pid := range args.PartitionIDs {
+				if err := doBatchDeleteIndiceRange(ctx, wrapper, job.ID, pid, args.IndexIDs, ea, "drop column: partition table ID"); err != nil {
 					return errors.Trace(err)
 				}
 			}

pkg/ddl/executor.go

@@ -3159,19 +3159,26 @@ func (e *executor) DropColumn(ctx sessionctx.Context, ti ast.Ident, spec *ast.Al
 	}
 
 	job := &model.Job{
+		// to do(joccau)
+		// we should  set Version = model.GetJobVerInUse() after refactor the actionMultiSchemaChange.
+		Version:        model.JobVersion1,
 		SchemaID:       schema.ID,
 		TableID:        t.Meta().ID,
 		SchemaName:     schema.Name.L,
 		SchemaState:    model.StatePublic,
 		TableName:      t.Meta().Name.L,
 		Type:           model.ActionDropColumn,
 		BinlogInfo:     &model.HistoryInfo{},
-		Args:           []any{colName, spec.IfExists},
 		CDCWriteSource: ctx.GetSessionVars().CDCWriteSource,
 		SQLMode:        ctx.GetSessionVars().SQLMode,
 	}
-
-	err = e.DoDDLJob(ctx, job)
+	args := &model.DropColumnArgs{
+		ColName:  colName,
+		IfExists: spec.IfExists,
+	}
+	// we need fill args here, because it will be added subjob which contains args and rawArgs from job.
+	job.FillArgs(args)
+	err = e.doDDLJob2(ctx, job, args)
 	return errors.Trace(err)
 }
 
@@ -6280,7 +6287,7 @@ func (e *executor) DoDDLJobWrapper(ctx sessionctx.Context, jobW *JobWrapper) (re
 	if mci := ctx.GetSessionVars().StmtCtx.MultiSchemaInfo; mci != nil {
 		// In multiple schema change, we don't run the job.
 		// Instead, we merge all the jobs into one pending job.
-		return appendToSubJobs(mci, job)
+		return appendToSubJobs(mci, jobW)
 	}
 	// Get a global job ID and put the DDL job in the queue.
 	setDDLJobQuery(ctx, job)

pkg/ddl/multi_schema_change.go

@@ -174,30 +174,30 @@ func handleRollbackException(runJobErr error, proxyJobErr *terror.Error) error {
 	return nil
 }
 
-func appendToSubJobs(m *model.MultiSchemaInfo, job *model.Job) error {
-	err := fillMultiSchemaInfo(m, job)
+func appendToSubJobs(m *model.MultiSchemaInfo, jobW *JobWrapper) error {
+	err := fillMultiSchemaInfo(m, jobW)
 	if err != nil {
 		return err
 	}
 	var reorgTp model.ReorgType
-	if job.ReorgMeta != nil {
-		reorgTp = job.ReorgMeta.ReorgTp
+	if jobW.ReorgMeta != nil {
+		reorgTp = jobW.ReorgMeta.ReorgTp
 	}
 	m.SubJobs = append(m.SubJobs, &model.SubJob{
-		Type:        job.Type,
-		Args:        job.Args,
-		RawArgs:     job.RawArgs,
-		SchemaState: job.SchemaState,
-		SnapshotVer: job.SnapshotVer,
+		Type:        jobW.Type,
+		Args:        jobW.Args,
+		RawArgs:     jobW.RawArgs,
+		SchemaState: jobW.SchemaState,
+		SnapshotVer: jobW.SnapshotVer,
 		Revertible:  true,
-		CtxVars:     job.CtxVars,
+		CtxVars:     jobW.CtxVars,
 		ReorgTp:     reorgTp,
 		UseCloud:    false,
 	})
 	return nil
 }
 
-func fillMultiSchemaInfo(info *model.MultiSchemaInfo, job *model.Job) (err error) {
+func fillMultiSchemaInfo(info *model.MultiSchemaInfo, job *JobWrapper) error {
 	switch job.Type {
 	case model.ActionAddColumn:
 		col := job.Args[0].(*table.Column)
@@ -210,7 +210,7 @@ func fillMultiSchemaInfo(info *model.MultiSchemaInfo, job *model.Job) (err error
 			info.PositionColumns = append(info.PositionColumns, pos.RelativeColumn.Name)
 		}
 	case model.ActionDropColumn:
-		colName := job.Args[0].(pmodel.CIStr)
+		colName := job.JobArgs.(*model.DropColumnArgs).ColName
 		info.DropColumns = append(info.DropColumns, colName)
 	case model.ActionDropIndex, model.ActionDropPrimaryKey:
 		indexName := job.Args[0].(pmodel.CIStr)

pkg/ddl/sanity_check.go

@@ -26,7 +26,6 @@ import (
 	"github.com/pingcap/tidb/pkg/meta/model"
 	"github.com/pingcap/tidb/pkg/parser"
 	"github.com/pingcap/tidb/pkg/parser/ast"
-	pmodel "github.com/pingcap/tidb/pkg/parser/model"
 	"github.com/pingcap/tidb/pkg/sessionctx"
 	"github.com/pingcap/tidb/pkg/util/intest"
 	"github.com/pingcap/tidb/pkg/util/mathutil"
@@ -152,15 +151,13 @@ func expectedDeleteRangeCnt(ctx delRangeCntCtx, job *model.Job) (int, error) {
 		}
 		return mathutil.Max(len(partitionIDs), 1), nil
 	case model.ActionDropColumn:
-		var colName pmodel.CIStr
-		var ifExists bool
-		var indexIDs []int64
-		var partitionIDs []int64
-		if err := job.DecodeArgs(&colName, &ifExists, &indexIDs, &partitionIDs); err != nil {
+		args, err := model.GetDropColumnArgs(job)
+		if err != nil {
 			return 0, errors.Trace(err)
 		}
-		physicalCnt := mathutil.Max(len(partitionIDs), 1)
-		return physicalCnt * len(indexIDs), nil
+
+		physicalCnt := mathutil.Max(len(args.PartitionIDs), 1)
+		return physicalCnt * len(args.IndexIDs), nil
 	case model.ActionModifyColumn:
 		var indexIDs []int64
 		var partitionIDs []int64

pkg/meta/model/BUILD.bazel

@@ -44,7 +44,7 @@ go_test(
     ],
     embed = [":model"],
     flaky = True,
-    shard_count = 34,
+    shard_count = 35,
     deps = [
         "//pkg/parser/charset",
         "//pkg/parser/model",

pkg/meta/model/job.go

@@ -490,39 +490,45 @@ func (job *Job) FillFinishedArgs(args FinishedJobArgs) {
 	args.fillFinishedJob(job)
 }
 
+func marshalArgs(jobVer JobVersion, args []any) (json.RawMessage, error) {
+	if jobVer <= JobVersion1 {
+		rawArgs, err := json.Marshal(args)
+		return rawArgs, errors.Trace(err)
+	}
+
+	intest.Assert(jobVer == JobVersion2, "job version is not v2")
+	var arg any
+	if len(args) > 0 {
+		intest.Assert(len(args) == 1, "Job.Args should have only one element")
+		arg = args[0]
+	}
+
+	rawArgs, err := json.Marshal(arg)
+	return rawArgs, errors.Trace(err)
+}
+
 // Encode encodes job with json format.
 // updateRawArgs is used to determine whether to update the raw args.
 func (job *Job) Encode(updateRawArgs bool) ([]byte, error) {
 	var err error
 	if updateRawArgs {
-		if job.Version == JobVersion1 {
-			job.RawArgs, err = json.Marshal(job.Args)
-			if err != nil {
-				return nil, errors.Trace(err)
-			}
-			if job.MultiSchemaInfo != nil {
-				for _, sub := range job.MultiSchemaInfo.SubJobs {
-					// Only update the args of executing sub-jobs.
-					if sub.Args == nil {
-						continue
-					}
-					sub.RawArgs, err = json.Marshal(sub.Args)
-					if err != nil {
-						return nil, errors.Trace(err)
-					}
+		job.RawArgs, err = marshalArgs(job.Version, job.Args)
+		if err != nil {
+			return nil, errors.Trace(err)
+		}
+
+		if job.MultiSchemaInfo != nil {
+			for _, sub := range job.MultiSchemaInfo.SubJobs {
+				// Only update the args of executing sub-jobs.
+				if sub.Args == nil {
+					continue
+				}
+
+				sub.RawArgs, err = marshalArgs(job.Version, sub.Args)
+				if err != nil {
+					return nil, errors.Trace(err)
 				}
 			}
-		} else {
-			var arg any
-			if len(job.Args) > 0 {
-				intest.Assert(len(job.Args) == 1, "Job.Args should have only one element")
-				arg = job.Args[0]
-			}
-			job.RawArgs, err = json.Marshal(arg)
-			if err != nil {
-				return nil, errors.Trace(err)
-			}
-			// TODO remember update sub-jobs' RawArgs when we do it.
 		}
 	}
 

pkg/meta/model/job_args.go

@@ -482,6 +482,49 @@ func GetResourceGroupArgs(job *Job) (*ResourceGroupArgs, error) {
 	return getOrDecodeArgsV2[*ResourceGroupArgs](job)
 }
 
+// DropColumnArgs is the arguments of dropping column job.
+type DropColumnArgs struct {
+	ColName  pmodel.CIStr `json:"column_name,omitempty"`
+	IfExists bool         `json:"if_exists,omitempty"`
+	// below 2 fields are filled during running.
+	IndexIDs     []int64 `json:"index_ids,omitempty"`
+	PartitionIDs []int64 `json:"partition_ids,omitempty"`
+}
+
+func (a *DropColumnArgs) fillJob(job *Job) {
+	if job.Version <= JobVersion1 {
+		job.Args = []any{a.ColName, a.IfExists, a.IndexIDs, a.PartitionIDs}
+	} else {
+		job.Args = []any{a}
+	}
+}
+
+// GetDropColumnArgs gets the args for drop column ddl.
+func GetDropColumnArgs(job *Job) (*DropColumnArgs, error) {
+	var (
+		colName      pmodel.CIStr
+		ifExists     bool
+		indexIDs     []int64
+		partitionIDs []int64
+	)
+
+	if job.Version <= JobVersion1 {
+		err := job.DecodeArgs(&colName, &ifExists, &indexIDs, &partitionIDs)
+		if err != nil {
+			return nil, errors.Trace(err)
+		}
+
+		return &DropColumnArgs{
+			ColName:      colName,
+			IfExists:     ifExists,
+			IndexIDs:     indexIDs,
+			PartitionIDs: partitionIDs,
+		}, nil
+	}
+
+	return getOrDecodeArgsV2[*DropColumnArgs](job)
+}
+
 // RenameTablesArgs is the arguments for rename tables job.
 type RenameTablesArgs struct {
 	RenameTableInfos []*RenameTableArgs `json:"rename_table_infos,omitempty"`

pkg/meta/model/job_args_test.go

@@ -313,3 +313,20 @@ func TestResourceGroupArgs(t *testing.T) {
 		}
 	}
 }
+
+func TestDropColumnArgs(t *testing.T) {
+	inArgs := &DropColumnArgs{
+		ColName:      model.NewCIStr("col_name"),
+		IfExists:     true,
+		IndexIDs:     []int64{1, 2, 3},
+		PartitionIDs: []int64{4, 5, 6},
+	}
+
+	for _, v := range []JobVersion{JobVersion1, JobVersion2} {
+		j2 := &Job{}
+		require.NoError(t, j2.Decode(getJobBytes(t, inArgs, v, ActionDropColumn)))
+		args, err := GetDropColumnArgs(j2)
+		require.NoError(t, err)
+		require.Equal(t, inArgs, args)
+	}
+}