Alexey Serbin has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/17598 )

Change subject: [tools] add --row_count_only and --report_scanner_stats flags
......................................................................


Patch Set 2:

(1 comment)

> (1 comment)
 >
 > Neat addition! Though it looks like the newly updated tests aren't
 > happy yet.

Thank you for the review!

I updated the tests, so no ASAN warnings are reported.

We talked about those warnings offline, and it seems it's still a mystery why 
with higher numbers for kudu perf loadgen the test reports ASAN warnings, 
although I ran the test with the following "instrumentation" and saw all 
allocated memory was deallocated successfully.  I guess we can take care of 
that separately.

diff --git a/src/kudu/common/partial_row.cc b/src/kudu/common/partial_row.cc
index 6951e6a6d..67a962aeb 100644
--- a/src/kudu/common/partial_row.cc
+++ b/src/kudu/common/partial_row.cc
@@ -253,6 +253,8 @@ void KuduPartialRow::DeallocateStringIfSet(int col_idx, 
const ColumnSchema& col)
         LOG(FATAL) << "Unexpected type " << col.type_info()->type();
         break;
     }
+    LOG(INFO) << Substitute("CALLING delete on $0",
+                            reinterpret_cast<uintptr_t>(dst->data()));
     delete [] dst->data();
     BitmapClear(owned_strings_bitmap_, col_idx);
   }
@@ -442,6 +444,8 @@ Status KuduPartialRow::SetSliceCopy(int col_idx, const 
Slice& val) {
     case STRING:
     case BINARY:
       auto relocated = new uint8_t[val.size()];
+      LOG(INFO) << Substitute("CALLING new: $0",
+                              reinterpret_cast<uintptr_t>(relocated));
       memcpy(relocated, val.data(), val.size());
       relocated_val = Slice(relocated, val.size());
       break;

http://gerrit.cloudera.org:8080/#/c/17598/2/src/kudu/tools/kudu-tool-test.cc
File src/kudu/tools/kudu-tool-test.cc:

http://gerrit.cloudera.org:8080/#/c/17598/2/src/kudu/tools/kudu-tool-test.cc@2848
PS2, Line 2848:     ASSERT_TRUE(s.ok()) << s.ToString() << ": " << err;
> nit: would it make sense to also check there are no row outputs? I guess th
Good point.  Done.



--
To view, visit http://gerrit.cloudera.org:8080/17598
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: kudu
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: Ia64e1a0b26996f0087d2473d15350d590581a69c
Gerrit-Change-Number: 17598
Gerrit-PatchSet: 2
Gerrit-Owner: Alexey Serbin <[email protected]>
Gerrit-Reviewer: Alexey Serbin <[email protected]>
Gerrit-Reviewer: Andrew Wong <[email protected]>
Gerrit-Reviewer: Grant Henke <[email protected]>
Gerrit-Reviewer: Kudu Jenkins (120)
Gerrit-Comment-Date: Thu, 17 Jun 2021 02:25:59 +0000
Gerrit-HasComments: Yes

Reply via email to