From 85399b14f78e0917b82e42c390e7c825122c6f46 Mon Sep 17 00:00:00 2001
From: Levi Tamasi <ltamasi@fb.com>
Date: Tue, 11 Oct 2022 14:40:25 -0700
Subject: [PATCH] Consider wide columns when checksumming in the stress tests
 (#10788)

Summary:
There are two places in the stress test code where we compute the CRC
for a range of KVs for the purposes of checking consistency, namely in the
CF consistency test (to make sure CFs contain the same data), and when
performing `CompactRange` (to make sure the pre- and post-compaction
states are equivalent). The patch extends the logic so that wide columns
are also considered in both cases.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/10788

Test Plan: Tested using some simple blackbox crash test runs.

Reviewed By: riversand963

Differential Revision: D40191134

Pulled By: ltamasi

fbshipit-source-id: 542c21cac9077c6d225780deb210319bb5eee955
---
 db_stress_tool/cf_consistency_stress.cc |  8 ++++++++
 db_stress_tool/db_stress_test_base.cc   | 21 +++++++++++++++++----
 2 files changed, 25 insertions(+), 4 deletions(-)

diff --git a/db_stress_tool/cf_consistency_stress.cc b/db_stress_tool/cf_consistency_stress.cc
index 5fd4d7e5a..27c722284 100644
--- a/db_stress_tool/cf_consistency_stress.cc
+++ b/db_stress_tool/cf_consistency_stress.cc
@@ -510,11 +510,19 @@ class CfConsistencyStressTest : public StressTest {
     const auto checksum_column_family = [](Iterator* iter,
                                            uint32_t* checksum) -> Status {
       assert(nullptr != checksum);
+
       uint32_t ret = 0;
       for (iter->SeekToFirst(); iter->Valid(); iter->Next()) {
         ret = crc32c::Extend(ret, iter->key().data(), iter->key().size());
         ret = crc32c::Extend(ret, iter->value().data(), iter->value().size());
+
+        for (const auto& column : iter->columns()) {
+          ret = crc32c::Extend(ret, column.name().data(), column.name().size());
+          ret =
+              crc32c::Extend(ret, column.value().data(), column.value().size());
+        }
       }
+
       *checksum = ret;
       return iter->status();
     };
diff --git a/db_stress_tool/db_stress_test_base.cc b/db_stress_tool/db_stress_test_base.cc
index 4f2ce8b92..fdf6f8e42 100644
--- a/db_stress_tool/db_stress_test_base.cc
+++ b/db_stress_tool/db_stress_test_base.cc
@@ -2285,8 +2285,6 @@ uint32_t StressTest::GetRangeHash(ThreadState* thread, const Snapshot* snapshot,
                                   ColumnFamilyHandle* column_family,
                                   const Slice& start_key,
                                   const Slice& end_key) {
-  const std::string kCrcCalculatorSepearator = ";";
-  uint32_t crc = 0;
   // This `ReadOptions` is for validation purposes. Ignore
   // `FLAGS_rate_limit_user_ops` to avoid slowing any validation.
   ReadOptions ro;
@@ -2299,15 +2297,29 @@ uint32_t StressTest::GetRangeHash(ThreadState* thread, const Snapshot* snapshot,
     ts = ts_str;
     ro.timestamp = &ts;
   }
+
   std::unique_ptr<Iterator> it(db_->NewIterator(ro, column_family));
+
+  constexpr char kCrcCalculatorSepearator = ';';
+
+  uint32_t crc = 0;
+
   for (it->Seek(start_key);
        it->Valid() && options_.comparator->Compare(it->key(), end_key) <= 0;
        it->Next()) {
     crc = crc32c::Extend(crc, it->key().data(), it->key().size());
-    crc = crc32c::Extend(crc, kCrcCalculatorSepearator.data(), 1);
+    crc = crc32c::Extend(crc, &kCrcCalculatorSepearator, sizeof(char));
     crc = crc32c::Extend(crc, it->value().data(), it->value().size());
-    crc = crc32c::Extend(crc, kCrcCalculatorSepearator.data(), 1);
+    crc = crc32c::Extend(crc, &kCrcCalculatorSepearator, sizeof(char));
+
+    for (const auto& column : it->columns()) {
+      crc = crc32c::Extend(crc, column.name().data(), column.name().size());
+      crc = crc32c::Extend(crc, &kCrcCalculatorSepearator, sizeof(char));
+      crc = crc32c::Extend(crc, column.value().data(), column.value().size());
+      crc = crc32c::Extend(crc, &kCrcCalculatorSepearator, sizeof(char));
+    }
   }
+
   if (!it->status().ok()) {
     fprintf(stderr, "Iterator non-OK when calculating range CRC: %s\n",
             it->status().ToString().c_str());
@@ -2315,6 +2327,7 @@ uint32_t StressTest::GetRangeHash(ThreadState* thread, const Snapshot* snapshot,
     // Fail fast to preserve the DB state.
     thread->shared->SetVerificationFailure();
   }
+
   return crc;
 }