[DNM] storage/concurrency: introduce concurrency control package, prototype SFU

c-deps/libroach/include/libroach.h

@@ -359,14 +359,15 @@ typedef struct {
  DBChunkedBuffer data;
  DBSlice intents;
  DBTimestamp uncertainty_timestamp;
+ DBTimestamp write_too_old;
  DBSlice resume_key;
 } DBScanResults;
 
 DBScanResults MVCCGet(DBIterator* iter, DBSlice key, DBTimestamp timestamp, DBTxn txn,
  bool inconsistent, bool tombstones);
 DBScanResults MVCCScan(DBIterator* iter, DBSlice start, DBSlice end, DBTimestamp timestamp,
  int64_t max_keys, DBTxn txn, bool inconsistent, bool reverse,
- bool tombstones);
+ bool tombstones, bool write_too_old);
 
 // DBStatsResult contains various runtime stats for RocksDB.
 typedef struct {

c-deps/libroach/mvcc.cc

@@ -276,19 +276,19 @@ DBScanResults MVCCGet(DBIterator* iter, DBSlice key, DBTimestamp timestamp, DBTx
  const DBSlice end = {0, 0};
  ScopedStats scoped_iter(iter);
  mvccForwardScanner scanner(iter, key, end, timestamp, 1 /* max_keys */, txn, inconsistent,
- tombstones);
+ tombstones, false /* write_too_old */);
  return scanner.get();
 }
 
 DBScanResults MVCCScan(DBIterator* iter, DBSlice start, DBSlice end, DBTimestamp timestamp,
  int64_t max_keys, DBTxn txn, bool inconsistent, bool reverse,
- bool tombstones) {
+ bool tombstones, bool write_too_old) {
  ScopedStats scoped_iter(iter);
  if (reverse) {
- mvccReverseScanner scanner(iter, end, start, timestamp, max_keys, txn, inconsistent, tombstones);
+ mvccReverseScanner scanner(iter, end, start, timestamp, max_keys, txn, inconsistent, tombstones, write_too_old);
  return scanner.scan();
  } else {
- mvccForwardScanner scanner(iter, start, end, timestamp, max_keys, txn, inconsistent, tombstones);
+ mvccForwardScanner scanner(iter, start, end, timestamp, max_keys, txn, inconsistent, tombstones, write_too_old);
  return scanner.scan();
  }
 }

c-deps/libroach/mvcc.h

@@ -48,7 +48,7 @@ static const int kMaxItersBeforeSeek = 10;
 template <bool reverse> class mvccScanner {
  public:
  mvccScanner(DBIterator* iter, DBSlice start, DBSlice end, DBTimestamp timestamp, int64_t max_keys,
- DBTxn txn, bool inconsistent, bool tombstones)
+ DBTxn txn, bool inconsistent, bool tombstones, bool write_too_old)
  : iter_(iter),
  iter_rep_(iter->rep.get()),
  start_key_(ToSlice(start)),
@@ -62,6 +62,7 @@ template <bool reverse> class mvccScanner {
  txn_ignored_seqnums_(txn.ignored_seqnums),
  inconsistent_(inconsistent),
  tombstones_(tombstones),
+ write_too_old_(write_too_old),
  check_uncertainty_(timestamp < txn.max_timestamp),
  kvs_(new chunkedBuffer),
  intents_(new rocksdb::WriteBatch),
@@ -261,6 +262,13 @@ template <bool reverse> class mvccScanner {
  return false;
  }
 
+ bool writeTooOld(DBTimestamp ts) {
+ results_.write_too_old = ts;
+ kvs_->Clear();
+ intents_->Clear();
+ return false;
+ }
+
  bool setStatus(const DBStatus& status) {
  results_.status = status;
  return false;
@@ -276,6 +284,10 @@ template <bool reverse> class mvccScanner {
  return addAndAdvance(cur_value_);
  }
 
+ if (write_too_old_) {
+ return writeTooOld(cur_timestamp_);
+ }
+
  if (check_uncertainty_) {
  // 2. Our txn's read timestamp is less than the max timestamp
  // seen by the txn. We need to check for clock uncertainty
@@ -326,7 +338,7 @@ template <bool reverse> class mvccScanner {
  // Intents for other transactions are visible at or below:
  // max(txn.max_timestamp, read_timestamp)
  const DBTimestamp max_visible_timestamp = check_uncertainty_ ? txn_max_timestamp_ : timestamp_;
- if (max_visible_timestamp < meta_timestamp && !own_intent) {
+ if (max_visible_timestamp < meta_timestamp && !own_intent && !write_too_old_) {
  // 5. The key contains an intent, but we're reading before the
  // intent. Seek to the desired version. Note that if we own the
  // intent (i.e. we're reading transactionally) we want to read
@@ -729,6 +741,7 @@ template <bool reverse> class mvccScanner {
  const DBIgnoredSeqNums txn_ignored_seqnums_;
  const bool inconsistent_;
  const bool tombstones_;
+ const bool write_too_old_;
  const bool check_uncertainty_;
  DBScanResults results_;
  std::unique_ptr<chunkedBuffer> kvs_;

pkg/ccl/changefeedccl/rowfetcher_cache.go

@@ -98,7 +98,11 @@ func (c *rowFetcherCache) RowFetcherForTableDesc(
 
  var rf row.Fetcher
  if err := rf.Init(
- false /* reverse */, false /* returnRangeInfo */, false /* isCheck */, &c.a,
+ false, /* reverse */
+ sqlbase.ScanLockingStrength_FOR_NONE,
+ false, /* returnRangeInfo */
+ false, /* isCheck */
+ &c.a,
  row.FetcherTableArgs{
  Spans: tableDesc.AllIndexSpans(),
  Desc: tableDesc,

pkg/kv/txn_interceptor_heartbeater.go

@@ -407,6 +407,9 @@ func firstWriteIndex(ba *roachpb.BatchRequest) (int, *roachpb.Error) {
  if roachpb.IsTransactionWrite(args) {
  return i, nil
  }
+ if t, ok := args.(*roachpb.ScanRequest); ok && t.SelectForUpdate {
+ return i, nil
+ }
  }
  return -1, nil
 }

pkg/kv/txn_interceptor_pipeliner.go

@@ -503,6 +503,10 @@ func (tp *txnPipeliner) updateWriteTracking(
  tp.footprint.insert(sp)
  }
  }
+ } else if scan, ok := req.(*roachpb.ScanRequest); ok && scan.SelectForUpdate {
+ if sp, ok := roachpb.ActualSpan(req, resp); ok {
+ tp.footprint.insert(sp)
+ }
  }
  }
 }