From 46a38452ac14770053b3a3575ccb9a8a5b17a915 Mon Sep 17 00:00:00 2001
From: apostle <vladkakurin007@gmail.com>
Date: Wed, 4 May 2022 15:01:12 +0000
Subject: [PATCH 1/6] prototype of sampling code

---
 sql/handler.h                         |  4 +++
 sql/sql_statistics.cc                 | 13 ++++++++--
 storage/innobase/handler/ha_innodb.cc | 36 +++++++++++++++++++++++++++
 storage/innobase/handler/ha_innodb.h  |  2 ++
 4 files changed, 53 insertions(+), 2 deletions(-)

diff --git a/sql/handler.h b/sql/handler.h
index 8ad521e189a4f..305b2afdf6af2 100644
--- a/sql/handler.h
+++ b/sql/handler.h
@@ -4030,6 +4030,10 @@ class handler :public Sql_alloc
   virtual int ft_read(uchar *buf) { return HA_ERR_WRONG_COMMAND; }
   virtual int rnd_next(uchar *buf)=0;
   virtual int rnd_pos(uchar * buf, uchar *pos)=0;
+
+  virtual int sample_next(uchar *buf) { return 0;}
+
+
   /**
     This function only works for handlers having
     HA_PRIMARY_KEY_REQUIRED_FOR_POSITION set.
diff --git a/sql/sql_statistics.cc b/sql/sql_statistics.cc
index 1f034f490c8e2..ce856bc5204db 100644
--- a/sql/sql_statistics.cc
+++ b/sql/sql_statistics.cc
@@ -2634,6 +2634,11 @@ int collect_statistics_for_table(THD *thd, TABLE *table)
   table->collected_stats->cardinality_is_null= TRUE;
   table->collected_stats->cardinality= 0;
 
+  int b = file->sample_next(table->record[0]);
+
+  b += 1;
+
+
   if (thd->variables.sample_percentage == 0)
   {
     if (file->records() < MIN_THRESHOLD_FOR_SAMPLING)
@@ -2650,14 +2655,18 @@ int collect_statistics_for_table(THD *thd, TABLE *table)
 
   for (field_ptr= table->field; *field_ptr; field_ptr++)
   {
-    table_field= *field_ptr;   
+    table_field= *field_ptr;
     if (!table_field->collected_stats)
-      continue; 
+      continue;
     table_field->collected_stats->init(thd, table_field);
   }
 
   restore_record(table, s->default_values);
 
+
+
+
+
   /* Perform a full table scan to collect statistics on 'table's columns */
   if (!(rc= file->ha_rnd_init(TRUE)))
   {
diff --git a/storage/innobase/handler/ha_innodb.cc b/storage/innobase/handler/ha_innodb.cc
index 17a92a242fd41..9d9a445358882 100644
--- a/storage/innobase/handler/ha_innodb.cc
+++ b/storage/innobase/handler/ha_innodb.cc
@@ -9489,6 +9489,42 @@ ha_innobase::rnd_next(
 	DBUG_RETURN(error);
 }
 
+#include "../row/row0sel.cc"
+
+int
+ha_innobase::sample_next(
+/*=====================*/
+    uchar *buf)
+{
+        int rc= ha_rnd_init(TRUE);
+        rc = rc;
+        mtr_t		mtr;
+        btr_pcur_t*	pcur = m_prebuilt->pcur;
+        rec_t*          rec;
+
+
+        rec_offs	offsets_[REC_OFFS_NORMAL_SIZE];
+        rec_offs*	offsets				= offsets_;
+        rec_offs_init(offsets_);
+        mtr.start();
+        dict_index_t*	index = innobase_get_index(MAX_KEY);
+        bool res = btr_pcur_open_at_rnd_pos(index, BTR_SEARCH_LEAF, pcur, &mtr);
+
+        rec = btr_pcur_get_rec(pcur);
+
+        mem_heap_t*	heap	= NULL;
+        offsets = rec_get_offsets(rec, index, offsets, index->n_core_fields, ULINT_UNDEFINED, &heap);
+        res = row_sel_store_mysql_rec(
+            buf, m_prebuilt, rec, NULL, true,
+            index, offsets);
+
+        res = res;
+
+        mtr.commit();
+        ha_rnd_end();
+        return 0;
+}
+
 /**********************************************************************//**
 Fetches a row from the table based on a row reference.
 @return 0, HA_ERR_KEY_NOT_FOUND, or error code */
diff --git a/storage/innobase/handler/ha_innodb.h b/storage/innobase/handler/ha_innodb.h
index 08501859ec908..f82d61e66f665 100644
--- a/storage/innobase/handler/ha_innodb.h
+++ b/storage/innobase/handler/ha_innodb.h
@@ -155,6 +155,8 @@ class ha_innobase final : public handler
 
 	int rnd_next(uchar *buf) override;
 
+        int sample_next(uchar *buf) override;
+
 	int rnd_pos(uchar * buf, uchar *pos) override;
 
 	int ft_init() override;

From 06338602ed5a9f77f84d976276012599f137d624 Mon Sep 17 00:00:00 2001
From: apostle <vladkakurin007@gmail.com>
Date: Sat, 14 May 2022 14:17:57 +0000
Subject: [PATCH 2/6] add new HA_NATIVE_SAMPLING flag

---
 sql/handler.h                         |  5 ++-
 sql/sql_statistics.cc                 | 54 ++++++++++++++-------------
 storage/innobase/handler/ha_innodb.cc |  3 +-
 3 files changed, 35 insertions(+), 27 deletions(-)

diff --git a/sql/handler.h b/sql/handler.h
index 305b2afdf6af2..58ad863277a3c 100644
--- a/sql/handler.h
+++ b/sql/handler.h
@@ -366,7 +366,10 @@ enum chf_create_flags {
 /* Implements SELECT ... FOR UPDATE SKIP LOCKED */
 #define HA_CAN_SKIP_LOCKED  (1ULL << 61)
 
-#define HA_LAST_TABLE_FLAG HA_CAN_SKIP_LOCKED
+/* Implements native sampling (MDEV-19556) */
+#define HA_NATIVE_SAMPLING (1ULL << 62)
+
+#define HA_LAST_TABLE_FLAG HA_NATIVE_SAMPLING
 
 
 /* bits in index_flags(index_number) for what you can do with index */
diff --git a/sql/sql_statistics.cc b/sql/sql_statistics.cc
index ce856bc5204db..99c8bbd26b700 100644
--- a/sql/sql_statistics.cc
+++ b/sql/sql_statistics.cc
@@ -2634,11 +2634,6 @@ int collect_statistics_for_table(THD *thd, TABLE *table)
   table->collected_stats->cardinality_is_null= TRUE;
   table->collected_stats->cardinality= 0;
 
-  int b = file->sample_next(table->record[0]);
-
-  b += 1;
-
-
   if (thd->variables.sample_percentage == 0)
   {
     if (file->records() < MIN_THRESHOLD_FOR_SAMPLING)
@@ -2663,39 +2658,48 @@ int collect_statistics_for_table(THD *thd, TABLE *table)
 
   restore_record(table, s->default_values);
 
+  if(file->ha_table_flags() & HA_NATIVE_SAMPLING)
+  {
+    rc = -1;
 
+    int b = file->sample_next(table->record[0]);
 
+    b += 1;
 
-
-  /* Perform a full table scan to collect statistics on 'table's columns */
-  if (!(rc= file->ha_rnd_init(TRUE)))
+  }
+  else
   {
-    DEBUG_SYNC(table->in_use, "statistics_collection_start");
 
-    while ((rc= file->ha_rnd_next(table->record[0])) != HA_ERR_END_OF_FILE)
+    /* Perform a full table scan to collect statistics on 'table's columns */
+    if (!(rc= file->ha_rnd_init(TRUE)))
     {
-      if (thd->killed)
-        break;
-
-      if (rc)
-        break;
+      DEBUG_SYNC(table->in_use, "statistics_collection_start");
 
-      if (thd_rnd(thd) <= sample_fraction)
+      while ((rc= file->ha_rnd_next(table->record[0])) != HA_ERR_END_OF_FILE)
       {
-        for (field_ptr= table->field; *field_ptr; field_ptr++)
+        if (thd->killed)
+          break;
+
+        if (rc)
+          break;
+
+        if (thd_rnd(thd) <= sample_fraction)
         {
-          table_field= *field_ptr;
-          if (!table_field->collected_stats)
-            continue;
-          if ((rc= table_field->collected_stats->add()))
+          for (field_ptr= table->field; *field_ptr; field_ptr++)
+          {
+            table_field= *field_ptr;
+            if (!table_field->collected_stats)
+              continue;
+            if ((rc= table_field->collected_stats->add()))
+              break;
+          }
+          if (rc)
             break;
+          rows++;
         }
-        if (rc)
-          break;
-        rows++;
       }
+      file->ha_rnd_end();
     }
-    file->ha_rnd_end();
   }
   rc= (rc == HA_ERR_END_OF_FILE && !thd->killed) ? 0 : 1;
 
diff --git a/storage/innobase/handler/ha_innodb.cc b/storage/innobase/handler/ha_innodb.cc
index 9d9a445358882..b1d68485a967d 100644
--- a/storage/innobase/handler/ha_innodb.cc
+++ b/storage/innobase/handler/ha_innodb.cc
@@ -3020,7 +3020,8 @@ ha_innobase::ha_innobase(
                           | HA_CAN_ONLINE_BACKUPS
 			  | HA_CONCURRENT_OPTIMIZE
 			  | HA_CAN_SKIP_LOCKED
-			  |  (srv_force_primary_key ? HA_REQUIRE_PRIMARY_KEY : 0)
+			  |  (srv_force_primary_key ? HA_REQUIRE_PRIMARY_KEY : 0
+                          | HA_NATIVE_SAMPLING)
 		  ),
 	m_start_of_scan(),
         m_mysql_has_locked()

From 419379425de379413e1766f9f1cfbbdb90eda511 Mon Sep 17 00:00:00 2001
From: apostle <vladkakurin007@gmail.com>
Date: Sat, 14 May 2022 15:28:55 +0000
Subject: [PATCH 3/6] make ha_samlpe interface

---
 sql/handler.cc                        | 34 +++++++++++++++++++++++++
 sql/handler.h                         | 24 ++++++++++++++++--
 sql/sql_statistics.cc                 | 31 +++++++++++++++++++----
 storage/innobase/handler/ha_innodb.cc | 36 +++++++++++++++++++++------
 storage/innobase/handler/ha_innodb.h  |  4 +++
 5 files changed, 114 insertions(+), 15 deletions(-)

diff --git a/sql/handler.cc b/sql/handler.cc
index 269115bc0da86..c5997ecfef0e9 100644
--- a/sql/handler.cc
+++ b/sql/handler.cc
@@ -3379,6 +3379,40 @@ int handler::ha_close(void)
   DBUG_RETURN(close());
 }
 
+int handler::ha_sample_next(uchar *buf)
+{
+  int result;
+  DBUG_ENTER("handler::ha_sample_next");
+  DBUG_ASSERT(table_share->tmp_table != NO_TMP_TABLE ||
+              m_lock_type != F_UNLCK);
+  DBUG_ASSERT(inited == SAMPLING);
+
+  do
+  {
+    TABLE_IO_WAIT(tracker, PSI_TABLE_FETCH_ROW, MAX_KEY, result,
+                  { result= sample_next(buf); })
+    if (result != HA_ERR_RECORD_DELETED)
+      break;
+    //status_var_increment(table->in_use->status_var.ha_read_rnd_deleted_count);
+  } while (!table->in_use->check_killed(1));
+
+  if (result == HA_ERR_RECORD_DELETED)
+    result= HA_ERR_ABORTED_BY_USER;
+  else
+  {
+    if (!result)
+    {
+      update_rows_read();
+      if (table->vfield && buf == table->record[0])
+        table->update_virtual_fields(this, VCOL_UPDATE_FOR_READ);
+    }
+    //increment_statistics(&SSV::ha_read_rnd_next_count);
+  }
+
+  table->status=result ? STATUS_NOT_FOUND: 0;
+  DBUG_RETURN(result);
+}
+
 
 int handler::ha_rnd_next(uchar *buf)
 {
diff --git a/sql/handler.h b/sql/handler.h
index 58ad863277a3c..aeed16969786a 100644
--- a/sql/handler.h
+++ b/sql/handler.h
@@ -3248,7 +3248,7 @@ class handler :public Sql_alloc
   /** Length of ref (1-8 or the clustered key length) */
   uint ref_length;
   FT_INFO *ft_handler;
-  enum init_stat { NONE=0, INDEX, RND };
+  enum init_stat { NONE=0, INDEX, RND, SAMPLING };
   init_stat inited, pre_inited;
 
   const COND *pushed_cond;
@@ -3509,6 +3509,25 @@ class handler :public Sql_alloc
   }
   int ha_rnd_init_with_error(bool scan) __attribute__ ((warn_unused_result));
   int ha_reset();
+  int ha_sample_init()
+  {
+    DBUG_EXECUTE_IF("ha_sample_init_fail", return HA_ERR_TABLE_DEF_CHANGED;);
+    int result;
+    DBUG_ENTER("ha_sample_init");
+    DBUG_ASSERT(inited==NONE);
+    inited= (result= sample_init()) ? NONE: SAMPLING;
+    end_range= NULL;
+    DBUG_RETURN(result);
+  }
+  int ha_sample_next(uchar* buf);
+  int ha_sample_end()
+  {
+    DBUG_ENTER("ha_sample_end");
+    DBUG_ASSERT(inited==SAMPLING);
+    inited=NONE;
+    end_range= NULL;
+    DBUG_RETURN(sample_end());
+  }
   /* this is necessary in many places, e.g. in HANDLER command */
   int ha_index_or_rnd_end()
   {
@@ -4034,8 +4053,9 @@ class handler :public Sql_alloc
   virtual int rnd_next(uchar *buf)=0;
   virtual int rnd_pos(uchar * buf, uchar *pos)=0;
 
+  virtual int sample_init() {return 0;}
   virtual int sample_next(uchar *buf) { return 0;}
-
+  virtual int sample_end() {return 0;}
 
   /**
     This function only works for handlers having
diff --git a/sql/sql_statistics.cc b/sql/sql_statistics.cc
index 99c8bbd26b700..2bfa022e65abb 100644
--- a/sql/sql_statistics.cc
+++ b/sql/sql_statistics.cc
@@ -2660,12 +2660,32 @@ int collect_statistics_for_table(THD *thd, TABLE *table)
 
   if(file->ha_table_flags() & HA_NATIVE_SAMPLING)
   {
-    rc = -1;
-
-    int b = file->sample_next(table->record[0]);
+    if(!(rc= file->ha_sample_init())) {
+      DEBUG_SYNC(table->in_use, "statistics_collection_start");
+      rows = file->records() * (thd->variables.sample_percentage / 100);
 
-    b += 1;
+      for (ulonglong i= 0; i < rows; ++i)
+      {
+        rc = file->ha_sample_next(table->record[0]);
+        if (thd->killed)
+          break;
 
+        if (rc)
+          break;
+        for (field_ptr= table->field; *field_ptr; field_ptr++)
+        {
+          table_field= *field_ptr;
+          if (!table_field->collected_stats)
+            continue;
+          if ((rc= table_field->collected_stats->add()))
+            break;
+        }
+        if (rc)
+          break;
+      }
+      file->ha_sample_end();
+    }
+    rc= (rc == 0 && !thd->killed) ? 0 : 1;
   }
   else
   {
@@ -2700,8 +2720,9 @@ int collect_statistics_for_table(THD *thd, TABLE *table)
       }
       file->ha_rnd_end();
     }
+    rc= (rc == HA_ERR_END_OF_FILE && !thd->killed) ? 0 : 1;
   }
-  rc= (rc == HA_ERR_END_OF_FILE && !thd->killed) ? 0 : 1;
+
 
   /* 
     Calculate values for all statistical characteristics on columns and
diff --git a/storage/innobase/handler/ha_innodb.cc b/storage/innobase/handler/ha_innodb.cc
index b1d68485a967d..9db325c72789d 100644
--- a/storage/innobase/handler/ha_innodb.cc
+++ b/storage/innobase/handler/ha_innodb.cc
@@ -9492,24 +9492,47 @@ ha_innobase::rnd_next(
 
 #include "../row/row0sel.cc"
 
+int
+ha_innobase::sample_init()
+{
+  int		err;
+
+  /* Store the active index value so that we can restore the original
+  value after a scan */
+
+  if (m_prebuilt->clust_index_was_generated) {
+    err = change_active_index(MAX_KEY);
+  } else {
+    err = change_active_index(m_primary_key);
+  }
+
+  return(err);
+}
+
+int
+ha_innobase::sample_end()
+{
+  return(index_end());
+}
 int
 ha_innobase::sample_next(
 /*=====================*/
     uchar *buf)
 {
-        int rc= ha_rnd_init(TRUE);
-        rc = rc;
+
         mtr_t		mtr;
         btr_pcur_t*	pcur = m_prebuilt->pcur;
         rec_t*          rec;
-
-
         rec_offs	offsets_[REC_OFFS_NORMAL_SIZE];
         rec_offs*	offsets				= offsets_;
         rec_offs_init(offsets_);
         mtr.start();
         dict_index_t*	index = innobase_get_index(MAX_KEY);
         bool res = btr_pcur_open_at_rnd_pos(index, BTR_SEARCH_LEAF, pcur, &mtr);
+        if(!res) {
+          mtr.commit();
+          return -1;
+        }
 
         rec = btr_pcur_get_rec(pcur);
 
@@ -9519,11 +9542,8 @@ ha_innobase::sample_next(
             buf, m_prebuilt, rec, NULL, true,
             index, offsets);
 
-        res = res;
-
         mtr.commit();
-        ha_rnd_end();
-        return 0;
+        return res ? 0 : -1;
 }
 
 /**********************************************************************//**
diff --git a/storage/innobase/handler/ha_innodb.h b/storage/innobase/handler/ha_innodb.h
index f82d61e66f665..cd2df33b30ea1 100644
--- a/storage/innobase/handler/ha_innodb.h
+++ b/storage/innobase/handler/ha_innodb.h
@@ -155,6 +155,10 @@ class ha_innobase final : public handler
 
 	int rnd_next(uchar *buf) override;
 
+        int sample_init() override;
+
+        int sample_end() override;
+
         int sample_next(uchar *buf) override;
 
 	int rnd_pos(uchar * buf, uchar *pos) override;

From af4225c8420d321d739d9fddef9a3350d1d5c03b Mon Sep 17 00:00:00 2001
From: apostle <vladkakurin007@gmail.com>
Date: Mon, 16 May 2022 02:02:21 +0000
Subject: [PATCH 4/6] fix style mistakes

---
 sql/handler.h                         |  7 ++--
 storage/innobase/handler/ha_innodb.cc | 46 +++++++++++++--------------
 2 files changed, 27 insertions(+), 26 deletions(-)

diff --git a/sql/handler.h b/sql/handler.h
index aeed16969786a..d38c3da8da132 100644
--- a/sql/handler.h
+++ b/sql/handler.h
@@ -3514,8 +3514,9 @@ class handler :public Sql_alloc
     DBUG_EXECUTE_IF("ha_sample_init_fail", return HA_ERR_TABLE_DEF_CHANGED;);
     int result;
     DBUG_ENTER("ha_sample_init");
-    DBUG_ASSERT(inited==NONE);
-    inited= (result= sample_init()) ? NONE: SAMPLING;
+    DBUG_ASSERT(inited == NONE);
+    result= sample_init();
+    inited= result ? NONE: SAMPLING;
     end_range= NULL;
     DBUG_RETURN(result);
   }
@@ -3524,7 +3525,7 @@ class handler :public Sql_alloc
   {
     DBUG_ENTER("ha_sample_end");
     DBUG_ASSERT(inited==SAMPLING);
-    inited=NONE;
+    inited= NONE;
     end_range= NULL;
     DBUG_RETURN(sample_end());
   }
diff --git a/storage/innobase/handler/ha_innodb.cc b/storage/innobase/handler/ha_innodb.cc
index 9db325c72789d..2722c9edeac72 100644
--- a/storage/innobase/handler/ha_innodb.cc
+++ b/storage/innobase/handler/ha_innodb.cc
@@ -9519,31 +9519,31 @@ ha_innobase::sample_next(
 /*=====================*/
     uchar *buf)
 {
+  mtr_t		mtr;
+  btr_pcur_t*	pcur= m_prebuilt->pcur;
+  rec_t*          rec;
+  rec_offs	offsets_[REC_OFFS_NORMAL_SIZE];
+  rec_offs*	offsets= offsets_;
+  rec_offs_init(offsets_);
+  mtr.start();
+  dict_index_t*	index= innobase_get_index(MAX_KEY);
+  bool res= btr_pcur_open_at_rnd_pos(index, BTR_SEARCH_LEAF, pcur, &mtr);
+  if(!res) {
+    mtr.commit();
+    return HA_ERR_KEY_NOT_FOUND;
+  }
 
-        mtr_t		mtr;
-        btr_pcur_t*	pcur = m_prebuilt->pcur;
-        rec_t*          rec;
-        rec_offs	offsets_[REC_OFFS_NORMAL_SIZE];
-        rec_offs*	offsets				= offsets_;
-        rec_offs_init(offsets_);
-        mtr.start();
-        dict_index_t*	index = innobase_get_index(MAX_KEY);
-        bool res = btr_pcur_open_at_rnd_pos(index, BTR_SEARCH_LEAF, pcur, &mtr);
-        if(!res) {
-          mtr.commit();
-          return -1;
-        }
-
-        rec = btr_pcur_get_rec(pcur);
-
-        mem_heap_t*	heap	= NULL;
-        offsets = rec_get_offsets(rec, index, offsets, index->n_core_fields, ULINT_UNDEFINED, &heap);
-        res = row_sel_store_mysql_rec(
-            buf, m_prebuilt, rec, NULL, true,
-            index, offsets);
+  rec= btr_pcur_get_rec(pcur);
 
-        mtr.commit();
-        return res ? 0 : -1;
+  mem_heap_t*	heap= NULL;
+  offsets= rec_get_offsets(rec, index, offsets, index->n_core_fields, ULINT_UNDEFINED, &heap);
+  ut_ad(offsets != NULL);
+  ut_ad(heap == NULL);
+  res= row_sel_store_mysql_rec(
+      buf, m_prebuilt, rec, NULL, true,
+      index, offsets);
+  mtr.commit();
+  return res ? 0 : HA_ERR_INTERNAL_ERROR;
 }
 
 /**********************************************************************//**

From bd12c520d376c30d3b2aa10922dcd8e86bf5764c Mon Sep 17 00:00:00 2001
From: apostle <vladkakurin007@gmail.com>
Date: Sun, 22 May 2022 22:40:30 +0000
Subject: [PATCH 5/6] fix include and heap ptr, add offsets errors handling

---
 storage/innobase/handler/ha_innodb.cc | 21 ++++++++++++++++-----
 storage/innobase/include/row0sel.h    | 26 ++++++++++++++++++++++++++
 storage/innobase/row/row0sel.cc       |  2 +-
 3 files changed, 43 insertions(+), 6 deletions(-)

diff --git a/storage/innobase/handler/ha_innodb.cc b/storage/innobase/handler/ha_innodb.cc
index 2722c9edeac72..55939ab0af865 100644
--- a/storage/innobase/handler/ha_innodb.cc
+++ b/storage/innobase/handler/ha_innodb.cc
@@ -113,6 +113,7 @@ this program; if not, write to the Free Software Foundation, Inc.,
 #include "fil0pagecompress.h"
 #include "ut0mem.h"
 #include "row0ext.h"
+#include "row0sel.h"
 
 #include "lz4.h"
 #include "lzo/lzo1x.h"
@@ -9490,8 +9491,6 @@ ha_innobase::rnd_next(
 	DBUG_RETURN(error);
 }
 
-#include "../row/row0sel.cc"
-
 int
 ha_innobase::sample_init()
 {
@@ -9528,7 +9527,8 @@ ha_innobase::sample_next(
   mtr.start();
   dict_index_t*	index= innobase_get_index(MAX_KEY);
   bool res= btr_pcur_open_at_rnd_pos(index, BTR_SEARCH_LEAF, pcur, &mtr);
-  if(!res) {
+  if(!res)
+  {
     mtr.commit();
     return HA_ERR_KEY_NOT_FOUND;
   }
@@ -9537,12 +9537,23 @@ ha_innobase::sample_next(
 
   mem_heap_t*	heap= NULL;
   offsets= rec_get_offsets(rec, index, offsets, index->n_core_fields, ULINT_UNDEFINED, &heap);
-  ut_ad(offsets != NULL);
-  ut_ad(heap == NULL);
+  ut_ad(offsets);
+  if (!offsets)
+  {
+    mtr.commit();
+    if (heap)
+      mem_heap_free(heap);
+    return HA_ERR_INTERNAL_ERROR;
+  }
+
   res= row_sel_store_mysql_rec(
       buf, m_prebuilt, rec, NULL, true,
       index, offsets);
+
   mtr.commit();
+
+  if(heap)
+    mem_heap_free(heap);
   return res ? 0 : HA_ERR_INTERNAL_ERROR;
 }
 
diff --git a/storage/innobase/include/row0sel.h b/storage/innobase/include/row0sel.h
index eb83a4bcad656..ddff736a83fed 100644
--- a/storage/innobase/include/row0sel.h
+++ b/storage/innobase/include/row0sel.h
@@ -480,3 +480,29 @@ row_sel_field_store_in_mysql_format_func(
 #include "row0sel.inl"
 
 #endif
+
+/** Convert a row in the Innobase format to a row in the MySQL format.
+Note that the template in prebuilt may advise us to copy only a few
+columns to mysql_rec, other columns are left blank. All columns may not
+be needed in the query.
+@param[out]	mysql_rec	row in the MySQL format
+@param[in]	prebuilt	cursor
+@param[in]	rec		Innobase record in the index
+                                which was described in prebuilt's
+                                template, or in the clustered index;
+                                must be protected by a page latch
+@param[in]	vrow		virtual columns
+@param[in]	rec_clust	whether index must be the clustered index
+@param[in]	index		index of rec
+@param[in]	offsets		array returned by rec_get_offsets(rec)
+@retval true on success
+@retval false if not all columns could be retrieved */
+MY_ATTRIBUTE((warn_unused_result))
+bool row_sel_store_mysql_rec(
+    byte*		mysql_rec,
+    row_prebuilt_t*	prebuilt,
+    const rec_t*	rec,
+    const dtuple_t*	vrow,
+    bool		rec_clust,
+    const dict_index_t* index,
+    const rec_offs*	offsets);
diff --git a/storage/innobase/row/row0sel.cc b/storage/innobase/row/row0sel.cc
index f208ca51d69b3..19d30bd3f3bf0 100644
--- a/storage/innobase/row/row0sel.cc
+++ b/storage/innobase/row/row0sel.cc
@@ -3092,7 +3092,7 @@ be needed in the query.
 @retval true on success
 @retval false if not all columns could be retrieved */
 MY_ATTRIBUTE((warn_unused_result))
-static bool row_sel_store_mysql_rec(
+bool row_sel_store_mysql_rec(
 	byte*		mysql_rec,
 	row_prebuilt_t*	prebuilt,
 	const rec_t*	rec,

From c6dcf385da37e03d344e559a996996ea6690be1b Mon Sep 17 00:00:00 2001
From: apostle <vladkakurin007@gmail.com>
Date: Thu, 26 May 2022 23:07:36 +0000
Subject: [PATCH 6/6] change prototypes btr_[p]cur_open_at_rnd_pos, change
 random sampling method to (random + acceptance/rejection) sampling method

---
 sql/handler.cc                        |  2 +-
 sql/handler.h                         |  1 -
 storage/innobase/btr/btr0cur.cc       | 50 ++++++++++++++++++++-------
 storage/innobase/dict/dict0stats.cc   |  6 ++--
 storage/innobase/handler/ha_innodb.cc | 26 ++++++--------
 storage/innobase/ibuf/ibuf0ibuf.cc    |  8 ++---
 storage/innobase/include/btr0cur.h    | 18 ++++++----
 storage/innobase/include/btr0pcur.h   | 18 ++++++----
 storage/innobase/include/btr0pcur.inl | 26 ++++++++------
 9 files changed, 93 insertions(+), 62 deletions(-)

diff --git a/sql/handler.cc b/sql/handler.cc
index c5997ecfef0e9..05bd7459e7a27 100644
--- a/sql/handler.cc
+++ b/sql/handler.cc
@@ -3391,7 +3391,7 @@ int handler::ha_sample_next(uchar *buf)
   {
     TABLE_IO_WAIT(tracker, PSI_TABLE_FETCH_ROW, MAX_KEY, result,
                   { result= sample_next(buf); })
-    if (result != HA_ERR_RECORD_DELETED)
+    if (result != HA_ERR_RECORD_DELETED && result != HA_ERR_KEY_NOT_FOUND)
       break;
     //status_var_increment(table->in_use->status_var.ha_read_rnd_deleted_count);
   } while (!table->in_use->check_killed(1));
diff --git a/sql/handler.h b/sql/handler.h
index d38c3da8da132..c3f2908daf803 100644
--- a/sql/handler.h
+++ b/sql/handler.h
@@ -3511,7 +3511,6 @@ class handler :public Sql_alloc
   int ha_reset();
   int ha_sample_init()
   {
-    DBUG_EXECUTE_IF("ha_sample_init_fail", return HA_ERR_TABLE_DEF_CHANGED;);
     int result;
     DBUG_ENTER("ha_sample_init");
     DBUG_ASSERT(inited == NONE);
diff --git a/storage/innobase/btr/btr0cur.cc b/storage/innobase/btr/btr0cur.cc
index 03862caa62790..128afeb9ff709 100644
--- a/storage/innobase/btr/btr0cur.cc
+++ b/storage/innobase/btr/btr0cur.cc
@@ -2849,14 +2849,18 @@ btr_cur_open_at_index_side(
 
 /**********************************************************************//**
 Positions a cursor at a randomly chosen position within a B-tree.
-@return true if the index is available and we have put the cursor, false
-if the index is unavailable */
-bool
+@return DB_SUCCESS if the index is available and we have put the cursor,
+error code if the index is unavailable. If simulate_uniform=true, could be
+DB_RECORD_NOT_FOUND returned, which means that no record is chosen in the
+generated tree path. The caller should retry a call, that will
+try a new tree path */
+dberr_t
 btr_cur_open_at_rnd_pos(
-	dict_index_t*	index,		/*!< in: index */
-	ulint		latch_mode,	/*!< in: BTR_SEARCH_LEAF, ... */
-	btr_cur_t*	cursor,		/*!< in/out: B-tree cursor */
-	mtr_t*		mtr)		/*!< in: mtr */
+	dict_index_t*	index,	    /*!< in: index */
+	ulint		latch_mode, /*!< in: BTR_SEARCH_LEAF, ... */
+	btr_cur_t*	cursor,	    /*!< in/out: B-tree cursor */
+	mtr_t*		mtr,        /*!< in: mtr */
+        bool sim_uniform_dist)      /*!< in: uniform distribution simulation */
 {
 	page_cur_t*	page_cursor;
 	ulint		node_ptr_max_size = srv_page_size / 2;
@@ -2868,6 +2872,11 @@ btr_cur_open_at_rnd_pos(
 	ulint		n_blocks = 0;
 	ulint		n_releases = 0;
 	mem_heap_t*	heap		= NULL;
+	/* For uniform distribution simulation is used naive A/R sampling
+	    algorithm, check link below for details
+	   https://courses.cs.washington.edu/courses/cse590q/05au/papers/Olken-Sampling.pdf
+	*/
+        double          p = 1.0;
 	rec_offs	offsets_[REC_OFFS_NORMAL_SIZE];
 	rec_offs*	offsets		= offsets_;
 	rec_offs_init(offsets_);
@@ -2917,7 +2926,7 @@ btr_cur_open_at_rnd_pos(
 	}
 
 	DBUG_EXECUTE_IF("test_index_is_unavailable",
-			return(false););
+			return(DB_ERROR););
 
 	if (index->page == FIL_NULL) {
 		/* Since we don't hold index lock until just now, the index
@@ -2925,7 +2934,7 @@ btr_cur_open_at_rnd_pos(
 		statistics updater for referenced table, it could be marked
 		as unavailable by 'DROP TABLE' in the mean time, since
 		we don't hold lock for statistics updater */
-		return(false);
+		return(DB_ERROR);
 	}
 
 	const rw_lock_type_t root_leaf_rw_latch = btr_cur_latch_for_root_leaf(
@@ -3003,8 +3012,18 @@ btr_cur_open_at_rnd_pos(
 			/* We are in the root node */
 
 			height = btr_page_get_level(page);
+		} else {
+			if(sim_uniform_dist) {
+				ulint n_recs = page_get_n_recs(block->page.frame);
+				int extra_bytes = dict_table_is_comp(index->table) ?
+				REC_N_NEW_EXTRA_BYTES : REC_N_OLD_EXTRA_BYTES;
+				// Get max fanout from n_recs / (srv_page_size /
+				// extra_bytes +1), but exchange two divisons by
+				// one multiplication and one division
+				p *= (double)n_recs * (extra_bytes + 1) /
+					(double)srv_page_size;
+			}
 		}
-
 		if (height == 0) {
 			if (rw_latch == RW_NO_LATCH
 			    || srv_read_only_mode) {
@@ -3038,7 +3057,7 @@ btr_cur_open_at_rnd_pos(
 			}
 		}
 
-		page_cur_open_on_rnd_user_rec(block, page_cursor);
+                page_cur_open_on_rnd_user_rec(block, page_cursor);
 
 		if (height == 0) {
 
@@ -3133,7 +3152,14 @@ btr_cur_open_at_rnd_pos(
 		mem_heap_free(heap);
 	}
 
-	return err == DB_SUCCESS;
+	// We get max value of type by using ~(type)0 for
+	// getting 0..1 pseudo random number from ut_rnd_gen()
+	// and exchange division by multiplication like
+	// (b / c) < a <=> b < (a * c)
+	if(sim_uniform_dist && (ut_rnd_gen() < (p * ~(uint32_t)0)))
+		err = DB_RECORD_NOT_FOUND;
+
+	return err;
 }
 
 /*==================== B-TREE INSERT =========================*/
diff --git a/storage/innobase/dict/dict0stats.cc b/storage/innobase/dict/dict0stats.cc
index fd744e6e2f52f..b720d9a422b79 100644
--- a/storage/innobase/dict/dict0stats.cc
+++ b/storage/innobase/dict/dict0stats.cc
@@ -1224,12 +1224,12 @@ btr_estimate_number_of_different_key_vals(dict_index_t* index,
 	for (i = 0; i < n_sample_pages; i++) {
 		mtr.start();
 
-		bool	available;
+                dberr_t	available;
 
 		available = btr_cur_open_at_rnd_pos(index, BTR_SEARCH_LEAF,
-						    &cursor, &mtr);
+						    &cursor, &mtr, false);
 
-		if (!available || index->table->bulk_trx_id != bulk_trx_id) {
+		if (available != DB_SUCCESS || index->table->bulk_trx_id != bulk_trx_id) {
 			mtr.commit();
 			mem_heap_free(heap);
 
diff --git a/storage/innobase/handler/ha_innodb.cc b/storage/innobase/handler/ha_innodb.cc
index 55939ab0af865..aedb35dad7549 100644
--- a/storage/innobase/handler/ha_innodb.cc
+++ b/storage/innobase/handler/ha_innodb.cc
@@ -9520,40 +9520,34 @@ ha_innobase::sample_next(
 {
   mtr_t		mtr;
   btr_pcur_t*	pcur= m_prebuilt->pcur;
-  rec_t*          rec;
+  rec_t*        rec;
+  bool          res;
+  dberr_t       err;
   rec_offs	offsets_[REC_OFFS_NORMAL_SIZE];
   rec_offs*	offsets= offsets_;
   rec_offs_init(offsets_);
-  mtr.start();
+
   dict_index_t*	index= innobase_get_index(MAX_KEY);
-  bool res= btr_pcur_open_at_rnd_pos(index, BTR_SEARCH_LEAF, pcur, &mtr);
-  if(!res)
+  mtr.start();
+  err= btr_pcur_open_at_rnd_pos(index, BTR_SEARCH_LEAF, pcur, &mtr, true);
+  mtr.commit();
+  if(err != DB_SUCCESS)
   {
-    mtr.commit();
     return HA_ERR_KEY_NOT_FOUND;
   }
-
   rec= btr_pcur_get_rec(pcur);
-
   mem_heap_t*	heap= NULL;
+  auto _ = make_scope_exit([heap]() { if(heap) mem_heap_free(heap); });
+
   offsets= rec_get_offsets(rec, index, offsets, index->n_core_fields, ULINT_UNDEFINED, &heap);
   ut_ad(offsets);
   if (!offsets)
-  {
-    mtr.commit();
-    if (heap)
-      mem_heap_free(heap);
     return HA_ERR_INTERNAL_ERROR;
-  }
 
   res= row_sel_store_mysql_rec(
       buf, m_prebuilt, rec, NULL, true,
       index, offsets);
 
-  mtr.commit();
-
-  if(heap)
-    mem_heap_free(heap);
   return res ? 0 : HA_ERR_INTERNAL_ERROR;
 }
 
diff --git a/storage/innobase/ibuf/ibuf0ibuf.cc b/storage/innobase/ibuf/ibuf0ibuf.cc
index 2540d180fb02c..b08fc6a989307 100644
--- a/storage/innobase/ibuf/ibuf0ibuf.cc
+++ b/storage/innobase/ibuf/ibuf0ibuf.cc
@@ -2386,12 +2386,12 @@ ibuf_merge_pages(
 
 	/* Open a cursor to a randomly chosen leaf of the tree, at a random
 	position within the leaf */
-	bool available;
+        dberr_t err;
 
-	available = btr_pcur_open_at_rnd_pos(ibuf.index, BTR_SEARCH_LEAF,
-					     &pcur, &mtr);
+	err = btr_pcur_open_at_rnd_pos(ibuf.index, BTR_SEARCH_LEAF,
+					     &pcur, &mtr, false);
 	/* No one should make this index unavailable when server is running */
-	ut_a(available);
+	ut_a(err == DB_SUCCESS);
 
 	ut_ad(page_validate(btr_pcur_get_page(&pcur), ibuf.index));
 
diff --git a/storage/innobase/include/btr0cur.h b/storage/innobase/include/btr0cur.h
index f5f1c97295712..45875f7a6e5f2 100644
--- a/storage/innobase/include/btr0cur.h
+++ b/storage/innobase/include/btr0cur.h
@@ -207,14 +207,18 @@ btr_cur_open_at_index_side(
 
 /**********************************************************************//**
 Positions a cursor at a randomly chosen position within a B-tree.
-@return true if the index is available and we have put the cursor, false
-if the index is unavailable */
-bool
+@return DB_SUCCESS if the index is available and we have put the cursor,
+error code if the index is unavailable. If simulate_uniform=true, could be
+DB_RECORD_NOT_FOUND returned, which means that no record is chosen in the
+generated tree path. The caller should retry a call, that will
+try a new tree path */
+dberr_t
 btr_cur_open_at_rnd_pos(
-	dict_index_t*	index,		/*!< in: index */
-	ulint		latch_mode,	/*!< in: BTR_SEARCH_LEAF, ... */
-	btr_cur_t*	cursor,		/*!< in/out: B-tree cursor */
-	mtr_t*		mtr);		/*!< in: mtr */
+	dict_index_t*	index,	    /*!< in: index */
+	ulint		latch_mode, /*!< in: BTR_SEARCH_LEAF, ... */
+	btr_cur_t*	cursor,     /*!< in/out: B-tree cursor */
+	mtr_t*		mtr,        /*!< in: mtr */
+        bool sim_uniform_dist);     /*!< in: uniform distribution simulation */
 /*************************************************************//**
 Tries to perform an insert to a page in an index tree, next to cursor.
 It is assumed that mtr holds an x-latch on the page. The operation does
diff --git a/storage/innobase/include/btr0pcur.h b/storage/innobase/include/btr0pcur.h
index f6636816e7290..eed0cb86deceb 100644
--- a/storage/innobase/include/btr0pcur.h
+++ b/storage/innobase/include/btr0pcur.h
@@ -200,15 +200,19 @@ btr_pcur_open_on_user_rec(
 	mtr_t*		mtr);		/*!< in: mtr */
 /**********************************************************************//**
 Positions a cursor at a randomly chosen position within a B-tree.
-@return true if the index is available and we have put the cursor, false
-if the index is unavailable */
+@return DB_SUCCESS if the index is available and we have put the cursor,
+error code if the index is unavailable. If simulate_uniform=true, could be
+DB_RECORD_NOT_FOUND returned, which means that no record is chosen in the
+generated tree path. The caller should retry a call, that will
+try a new tree path */
 UNIV_INLINE
-bool
+dberr_t
 btr_pcur_open_at_rnd_pos(
-	dict_index_t*	index,		/*!< in: index */
-	ulint		latch_mode,	/*!< in: BTR_SEARCH_LEAF, ... */
-	btr_pcur_t*	cursor,		/*!< in/out: B-tree pcur */
-	mtr_t*		mtr);		/*!< in: mtr */
+	dict_index_t*	index,	    /*!< in: index */
+	ulint		latch_mode, /*!< in: BTR_SEARCH_LEAF, ... */
+	btr_pcur_t*	cursor,	    /*!< in/out: B-tree pcur */
+	mtr_t*		mtr,	    /*!< in: mtr */
+	bool sim_uniform_dist);     /*!< in: uniform distribution simulation */
 /**************************************************************//**
 Frees the possible memory heap of a persistent cursor and sets the latch
 mode of the persistent cursor to BTR_NO_LATCHES.
diff --git a/storage/innobase/include/btr0pcur.inl b/storage/innobase/include/btr0pcur.inl
index f5e59c7268e50..9e8ac32113763 100644
--- a/storage/innobase/include/btr0pcur.inl
+++ b/storage/innobase/include/btr0pcur.inl
@@ -479,15 +479,19 @@ btr_pcur_open_at_index_side(
 
 /**********************************************************************//**
 Positions a cursor at a randomly chosen position within a B-tree.
-@return true if the index is available and we have put the cursor, false
-if the index is unavailable */
+@return DB_SUCCESS if the index is available and we have put the cursor,
+error code if the index is unavailable. If simulate_uniform=true, could be
+DB_RECORD_NOT_FOUND returned, which means that no record is chosen in the
+generated tree path. The caller should retry a call, that will
+try a new tree path */
 UNIV_INLINE
-bool
+dberr_t
 btr_pcur_open_at_rnd_pos(
-	dict_index_t*	index,		/*!< in: index */
-	ulint		latch_mode,	/*!< in: BTR_SEARCH_LEAF, ... */
-	btr_pcur_t*	cursor,		/*!< in/out: B-tree pcur */
-	mtr_t*		mtr)		/*!< in: mtr */
+	dict_index_t*	index,	    /*!< in: index */
+	ulint		latch_mode, /*!< in: BTR_SEARCH_LEAF, ... */
+	btr_pcur_t*	cursor,	    /*!< in/out: B-tree pcur */
+	mtr_t*		mtr,	    /*!< in: mtr */
+	bool sim_uniform_dist)      /*!< in: uniform distribution simulation */
 {
 	/* Initialize the cursor */
 
@@ -496,17 +500,17 @@ btr_pcur_open_at_rnd_pos(
 
 	btr_pcur_init(cursor);
 
-	bool	available;
+	dberr_t	err;
 
-	available = btr_cur_open_at_rnd_pos(index, latch_mode,
+	err = btr_cur_open_at_rnd_pos(index, latch_mode,
 					    btr_pcur_get_btr_cur(cursor),
-					    mtr);
+					    mtr, sim_uniform_dist);
 	cursor->pos_state = BTR_PCUR_IS_POSITIONED;
 	cursor->old_stored = false;
 
 	cursor->trx_if_known = NULL;
 
-	return(available);
+	return err;
 }
 
 /**************************************************************//**