MDEV-6255 DUPLICATE KEY Errors on SELECT .. GROUP BY that uses temporary and filesort.

The problem was that my_hash_sort didn't properly delete end-space characters properly, so strings that should compare
identically was seen as different strings.  (Space was handled correctly, but not NBSP)
This caused duplicate key errors when a heap table was converted to Aria as part of overflow in group by.

Fixed by removing all characters that compares as end space when creating a hash.

Other things:
- Fixed that --sorted_results also works for errors in mysqltest.
- Speed up hash by not comparing strings that has different hash.
- Speed up many my_hash_sort functions by using registers to calculate hash instead of pointers.
  This was previously done for some functions, but not for all.
- Made a macro of the hash function, to simplify code and to be able to experiment with new hash functions.







client/mysqltest.cc:
  Fixed that --sorted_results also works for error messages.
mysql-test/r/ctype_partitions.result:
  New test to ensure that partitions on hash works
mysql-test/suite/multi_source/gtid.result:
  Updated result
mysql-test/suite/multi_source/gtid.test:
  Test that --sorted_result works for error messages
mysql-test/suite/multi_source/gtid_ignore_duplicates.result:
  Updated result
mysql-test/suite/multi_source/gtid_ignore_duplicates.test:
  Updated result
mysql-test/suite/multi_source/load_data.result:
  Updated result
mysql-test/suite/multi_source/load_data.test:
  Updated result
mysql-test/t/ctype_partitions.test:
  New test to ensure that partitions on hash works
storage/heap/hp_write.c:
  Speed up hash by not comparing strings that has different hash.
storage/maria/ma_check.c:
  Extra debug
strings/ctype-bin.c:
  Use macro for hash function
strings/ctype-latin1.c:
  Use macro for hash function
  Use registers to calculate hash (speedup)
strings/ctype-mb.c:
  Use macro for hash function
  Use registers to calculate hash (speedup)
strings/ctype-simple.c:
  Use macro for hash function
  Use same variable names as in other my_hash_sort functions.
  Update my_hash_sort_simple() to properly remove end space (patch by Bar)
strings/ctype-uca.c:
  Ignore duplicated space inside strings and end space in my_hash_sort_uca(). This fixed MDEV-6255
  Use macro for hash function
  Use registers to calculate hash (speedup)
strings/ctype-ucs2.c:
  Use macro for hash function
  Use registers to calculate hash (speedup)
strings/ctype-utf8.c:
  Use macro for hash function
  Use registers to calculate hash (speedup)
strings/strings_def.h:
  Made a macro of the hash function, to simplify code and to be able to experiment with new hash functions.

client/mysqltest.cc

@@ -7719,6 +7719,7 @@ int append_warnings(DYNAMIC_STRING *ds, MYSQL* mysql)
 {
   uint count;
   MYSQL_RES *warn_res;
+  DYNAMIC_STRING res;
   DBUG_ENTER("append_warnings");
 
   if (!(count= mysql_warning_count(mysql)))
@@ -7738,11 +7739,18 @@ int append_warnings(DYNAMIC_STRING *ds, MYSQL* mysql)
     die("Warning count is %u but didn't get any warnings",
 	count);
 
-  append_result(ds, warn_res);
+  init_dynamic_string(&res, "", 1024, 1024);
+
+  append_result(&res, warn_res);
   mysql_free_result(warn_res);
 
-  DBUG_PRINT("warnings", ("%s", ds->str));
+  DBUG_PRINT("warnings", ("%s", res.str));
 
+  if (display_result_sorted)
+    dynstr_append_sorted(ds, &res, 0);
+  else
+    dynstr_append_mem(ds, res.str, res.length);
+  dynstr_free(&res);
   DBUG_RETURN(count);
 }
 

mysql-test/r/ctype_partitions.result

+#
+# MDEV-6255 DUPLICATE KEY Errors on SELECT .. GROUP BY that uses temporary and filesort
+#
+CREATE TABLE t1 (a VARCHAR(10) CHARACTER SET cp1251 COLLATE cp1251_ukrainian_ci);
+INSERT INTO t1 VALUES (0x20),(0x60),(0x6060),(0x606060);
+SELECT HEX(a) FROM t1 WHERE a=0x60;
+HEX(a)
+20
+60
+6060
+606060
+ALTER TABLE t1  PARTITION BY KEY(a) PARTITIONS 3;
+SELECT HEX(a) FROM t1 WHERE a=0x60;
+HEX(a)
+20
+60
+6060
+606060
+DROP TABLE t1;
+CREATE TABLE t1 (a VARCHAR(10) CHARACTER SET koi8u COLLATE koi8u_general_ci);
+INSERT INTO t1 VALUES (0x20),(0x60),(0x6060),(0x606060);
+SELECT HEX(a) FROM t1 WHERE a=0x60;
+HEX(a)
+20
+60
+6060
+606060
+ALTER TABLE t1  PARTITION BY KEY(a) PARTITIONS 3;
+SELECT HEX(a) FROM t1 WHERE a=0x60;
+HEX(a)
+20
+60
+6060
+606060
+DROP TABLE t1;
+CREATE TABLE t1 (a VARCHAR(10) CHARACTER SET cp1250 COLLATE cp1250_general_ci);
+INSERT INTO t1 VALUES (0x20),(0xA0),(0xA0A0),(0xA0A0A0);
+SELECT HEX(a) FROM t1 WHERE a=0xA0;
+HEX(a)
+20
+A0
+A0A0
+A0A0A0
+ALTER TABLE t1  PARTITION BY KEY(a) PARTITIONS 3;
+SELECT HEX(a) FROM t1 WHERE a=0xA0;
+HEX(a)
+20
+A0
+A0A0
+A0A0A0
+DROP TABLE t1;

mysql-test/suite/multi_source/gtid.result

@@ -141,8 +141,8 @@ include/reset_master_slave.inc
 SET GLOBAL gtid_domain_id=0;
 STOP ALL SLAVES;
 Warnings:
-Note	1938	SLAVE 'slave2' stopped
 Note	1938	SLAVE 'slave1' stopped
+Note	1938	SLAVE 'slave2' stopped
 include/reset_master_slave.inc
 SET GLOBAL gtid_domain_id=0;
 include/reset_master_slave.inc

mysql-test/suite/multi_source/gtid.test

@@ -141,12 +141,14 @@ DROP TABLE t3;
 SET GLOBAL gtid_domain_id=0;
 --let $wait_condition= SELECT COUNT(*)=0 FROM information_schema.tables WHERE table_name IN ("t1", "t2", "t3") AND table_schema = "test"
 --source include/wait_condition.inc
+--sorted_result
 STOP ALL SLAVES;
 --source reset_master_slave.inc
 --disconnect slave1
 
 --connection slave2
 SET GLOBAL gtid_domain_id=0;
+--sorted_result
 STOP ALL SLAVES;
 --source reset_master_slave.inc
 --disconnect slave2

mysql-test/suite/multi_source/gtid_ignore_duplicates.result

@@ -245,8 +245,8 @@ a
 SET GLOBAL gtid_domain_id=0;
 STOP ALL SLAVES;
 Warnings:
-Note	1938	SLAVE 'c2a' stopped
 Note	1938	SLAVE 'b2a' stopped
+Note	1938	SLAVE 'c2a' stopped
 SET GLOBAL slave_parallel_threads= @old_parallel;
 SET GLOBAL gtid_ignore_duplicates= @old_ignore_duplicates;
 SET GLOBAL gtid_domain_id=0;

mysql-test/suite/multi_source/gtid_ignore_duplicates.test

@@ -261,24 +261,28 @@ SELECT * FROM t1 WHERE a >= 20 ORDER BY a;
 # Clean up.
 --connection server_1
 SET GLOBAL gtid_domain_id=0;
+--sorted_result
 STOP ALL SLAVES;
 SET GLOBAL slave_parallel_threads= @old_parallel;
 SET GLOBAL gtid_ignore_duplicates= @old_ignore_duplicates;
 
 --connection server_2
 SET GLOBAL gtid_domain_id=0;
+--sorted_result
 STOP ALL SLAVES;
 SET GLOBAL slave_parallel_threads= @old_parallel;
 SET GLOBAL gtid_ignore_duplicates= @old_ignore_duplicates;
 
 --connection server_3
 SET GLOBAL gtid_domain_id=0;
+--sorted_result
 STOP ALL SLAVES;
 SET GLOBAL slave_parallel_threads= @old_parallel;
 SET GLOBAL gtid_ignore_duplicates= @old_ignore_duplicates;
 
 --connection server_4
 SET GLOBAL gtid_domain_id=0;
+--sorted_result
 STOP ALL SLAVES;
 SET GLOBAL slave_parallel_threads= @old_parallel;
 SET GLOBAL gtid_ignore_duplicates= @old_ignore_duplicates;

mysql-test/suite/multi_source/load_data.result

@@ -23,8 +23,8 @@ drop table t1;
 drop table t2;
 stop all slaves;
 Warnings:
-Note	1938	SLAVE 'master2' stopped
 Note	1938	SLAVE '' stopped
+Note	1938	SLAVE 'master2' stopped
 include/reset_master_slave.inc
 include/reset_master_slave.inc
 include/reset_master_slave.inc

mysql-test/suite/multi_source/load_data.test

@@ -58,6 +58,7 @@ drop table t2;
 --sync_with_master 0,'master2'
 
 --connection slave
+--sorted_result
 stop all slaves;
 
 --source reset_master_slave.inc

mysql-test/t/ctype_partitions.test

+--source include/have_partition.inc
+
+--echo #
+--echo # MDEV-6255 DUPLICATE KEY Errors on SELECT .. GROUP BY that uses temporary and filesort
+--echo #
+
+# cp1251_ukrainian_ci: 0x20 SPACE is equal to 0x60 GRAVE ACCENT 
+CREATE TABLE t1 (a VARCHAR(10) CHARACTER SET cp1251 COLLATE cp1251_ukrainian_ci);
+INSERT INTO t1 VALUES (0x20),(0x60),(0x6060),(0x606060);
+SELECT HEX(a) FROM t1 WHERE a=0x60;
+ALTER TABLE t1  PARTITION BY KEY(a) PARTITIONS 3;
+SELECT HEX(a) FROM t1 WHERE a=0x60;
+DROP TABLE t1;
+
+# koi8u_general_ci: 0x20 SPACE is equal to 0x60 GRAVE ACCENT 
+CREATE TABLE t1 (a VARCHAR(10) CHARACTER SET koi8u COLLATE koi8u_general_ci);
+INSERT INTO t1 VALUES (0x20),(0x60),(0x6060),(0x606060);
+SELECT HEX(a) FROM t1 WHERE a=0x60;
+ALTER TABLE t1  PARTITION BY KEY(a) PARTITIONS 3;
+SELECT HEX(a) FROM t1 WHERE a=0x60;
+DROP TABLE t1;
+
+# cp1250_general_ci: 0x20 SPACE is equal to 0xA0 NO-BREAK SPACE
+CREATE TABLE t1 (a VARCHAR(10) CHARACTER SET cp1250 COLLATE cp1250_general_ci);
+INSERT INTO t1 VALUES (0x20),(0xA0),(0xA0A0),(0xA0A0A0);
+SELECT HEX(a) FROM t1 WHERE a=0xA0;
+ALTER TABLE t1  PARTITION BY KEY(a) PARTITIONS 3;
+SELECT HEX(a) FROM t1 WHERE a=0xA0;
+DROP TABLE t1;

storage/heap/hp_write.c

@@ -385,7 +385,8 @@ int hp_write_key(HP_INFO *info, HP_KEYDEF *keyinfo,
       pos=empty;
       do
       {
-	if (! hp_rec_key_cmp(keyinfo, record, pos->ptr_to_rec, 1))
+	if (pos->hash_of_key == hash_of_key &&
+            ! hp_rec_key_cmp(keyinfo, record, pos->ptr_to_rec, 1))
 	{
 	  DBUG_RETURN(my_errno=HA_ERR_FOUND_DUPP_KEY);
 	}

storage/maria/ma_check.c

@@ -5515,6 +5515,8 @@ static int sort_key_write(MARIA_SORT_PARAM *sort_param, const uchar *a)
   }
   if ((sort_param->keyinfo->flag & HA_NOSAME) && cmp == 0)
   {
+    DBUG_EXECUTE("key", _ma_print_keydata(DBUG_FILE, sort_param->seg, a,
+                                          USE_WHOLE_KEY););
     sort_info->dupp++;
     sort_info->info->cur_row.lastpos= get_record_for_key(sort_param->keyinfo,
                                                          a);

strings/ctype-bin.c

@@ -288,9 +288,7 @@ void my_hash_sort_8bit_bin(CHARSET_INFO *cs __attribute__((unused)),
 
   for (; key < end ; key++)
   {
-    tmp1^= (ulong) ((((uint) tmp1 & 63) + tmp2) *
-                    ((uint) *key)) + (tmp1 << 8);
-    tmp2+= 3;
+    MY_HASH_ADD(tmp1, tmp2, (uint) *key);
   }
 
   *nr1= tmp1;
@@ -307,9 +305,7 @@ void my_hash_sort_bin(CHARSET_INFO *cs __attribute__((unused)),
 
   for (; key < end ; key++)
   {
-    tmp1^= (ulong) ((((uint) tmp1 & 63) + tmp2) *
-                    ((uint) *key)) + (tmp1 << 8);
-    tmp2+= 3;
+    MY_HASH_ADD(tmp1, tmp2, (uint) *key);
   }
 
   *nr1= tmp1;

strings/ctype-latin1.c

@@ -691,6 +691,8 @@ void my_hash_sort_latin1_de(CHARSET_INFO *cs __attribute__((unused)),
 			    ulong *nr1, ulong *nr2)
 {
   const uchar *end;
+  register ulong m1= *nr1, m2= *nr2;
+    
   /*
     Remove end space. We have to do this to be able to compare
     'AE' and '' as identical
@@ -700,14 +702,14 @@ void my_hash_sort_latin1_de(CHARSET_INFO *cs __attribute__((unused)),
   for (; key < end ; key++)
   {
     uint X= (uint) combo1map[(uint) *key];
-    nr1[0]^=(ulong) ((((uint) nr1[0] & 63)+nr2[0]) * X) + (nr1[0] << 8);
-    nr2[0]+=3;
+    MY_HASH_ADD(m1, m2, X);
     if ((X= combo2map[*key]))
     {
-      nr1[0]^=(ulong) ((((uint) nr1[0] & 63)+nr2[0]) * X) + (nr1[0] << 8);
-      nr2[0]+=3;
+      MY_HASH_ADD(m1, m2, X);
     }
   }
+  *nr1= m1;
+  *nr2= m2;
 }
 
 

strings/ctype-mb.c

@@ -680,6 +680,8 @@ void
 my_hash_sort_mb_bin(CHARSET_INFO *cs __attribute__((unused)),
                     const uchar *key, size_t len,ulong *nr1, ulong *nr2)
 {
+  register ulong m1= *nr1, m2= *nr2;
+
   /*
      Remove trailing spaces. We have to do this to be able to compare
     'A ' and 'A' as identical
@@ -688,10 +690,10 @@ my_hash_sort_mb_bin(CHARSET_INFO *cs __attribute__((unused)),
   
   for (; key < end ; key++)
   {
-    nr1[0]^=(ulong) ((((uint) nr1[0] & 63)+nr2[0]) * 
-	     ((uint)*key)) + (nr1[0] << 8);
-    nr2[0]+=3;
+    MY_HASH_ADD(m1, m2, (uint)*key);
   }
+  *nr1= m1;
+  *nr2= m2;
 }
 
 

strings/ctype-simple.c

@@ -306,24 +306,48 @@ void my_hash_sort_simple(CHARSET_INFO *cs,
 {
   register const uchar *sort_order=cs->sort_order;
   const uchar *end;
-  ulong n1, n2;
+  register ulong m1= *nr1, m2= *nr2;
+  uint16 space_weight= sort_order[' '];
 
   /*
-    Remove end space. We have to do this to be able to compare
-    'A ' and 'A' as identical
+    Remove all trailing characters that are equal to space.
+    We have to do this to be able to compare 'A ' and 'A' as identical.
+
+    If the key is long enough, cut the trailing spaces (0x20) using an
+    optimized function implemented in skip_trailing_spaces().
+
+    "len > 16" is just some heuristic here.
+    Calling skip_triling_space() for short values is not desirable,
+    because its initialization block may be more expensive than the
+    performance gained.
+  */
+
+  end= len > 16 ? skip_trailing_space(key, len) : key + len;
+
+  /*
+    We removed all trailing characters that are binary equal to space 0x20.
+    Now remove all trailing characters that have weights equal to space.
+    Some 8bit simple collations may have such characters:
+    - cp1250_general_ci    0xA0 NO-BREAK SPACE == 0x20 SPACE
+    - cp1251_ukrainian_ci  0x60 GRAVE ACCENT   == 0x20 SPACE
+    - koi8u_general_ci     0x60 GRAVE ACCENT   == 0x20 SPACE
   */
-  end= skip_trailing_space(key, len);
 
-  n1= *nr1;
-  n2= *nr2;
+  for ( ; key < end ; )
+  {
+    if (sort_order[*--end] != space_weight)
+    {
+      end++;
+      break;
+    }
+  }
+
   for (; key < (uchar*) end ; key++)
   {
-    n1^=(ulong) ((((uint) n1 & 63)+n2) *
-	     ((uint) sort_order[(uint) *key])) + (n1 << 8);
-    n2+=3;
+    MY_HASH_ADD(m1, m2, (uint) sort_order[(uint) *key]);
   }
-  *nr1= n1;
-  *nr2= n2;
+  *nr1= m1;
+  *nr2= m2;
 }
 
 

strings/ctype-uca.c

@@ -20873,21 +20873,45 @@ static int my_strnncollsp_uca(CHARSET_INFO *cs,
 static void my_hash_sort_uca(CHARSET_INFO *cs,
                              my_uca_scanner_handler *scanner_handler,
 			     const uchar *s, size_t slen,
-			     ulong *n1, ulong *n2)
+			     ulong *nr1, ulong *nr2)
 {
   int   s_res;
   my_uca_scanner scanner;
-  
-  slen= cs->cset->lengthsp(cs, (char*) s, slen);
+  int space_weight= my_space_weight(cs);
+  register ulong m1= *nr1, m2= *nr2;
+
   scanner_handler->init(&scanner, cs, &cs->uca->level[0], s, slen);
   
   while ((s_res= scanner_handler->next(&scanner)) >0)
   {
-    n1[0]^= (((n1[0] & 63)+n2[0])*(s_res >> 8))+ (n1[0] << 8);
-    n2[0]+=3;
-    n1[0]^= (((n1[0] & 63)+n2[0])*(s_res & 0xFF))+ (n1[0] << 8);
-    n2[0]+=3;
+    if (s_res == space_weight)
+    {
+      /* Combine all spaces to be able to skip end spaces */
+      uint count= 0;
+      do
+      {
+        count++;
+        if ((s_res= scanner_handler->next(&scanner)) <= 0)
+        {
+          /* Skip strings at end of string */
+          goto end;
+        }
+      }
+      while (s_res == space_weight);
+
+      /* Add back that has for the space characters */
+      do
+      {
+        MY_HASH_ADD_16(m1, m2, space_weight);
+      }
+      while (--count != 0);
+
+    }
+    MY_HASH_ADD_16(m1, m2, s_res);
   }
+end:
+  *nr1= m1;
+  *nr2= m2;
 }
 
 

strings/ctype-ucs2.c

@@ -1222,23 +1222,23 @@ my_caseup_utf16(CHARSET_INFO *cs, char *src, size_t srclen,
 
 static void
 my_hash_sort_utf16(CHARSET_INFO *cs, const uchar *s, size_t slen,
-                   ulong *n1, ulong *n2)
+                   ulong *nr1, ulong *nr2)
 {
   my_wc_t wc;
   my_charset_conv_mb_wc mb_wc= cs->cset->mb_wc;
   int res;
   const uchar *e= s + cs->cset->lengthsp(cs, (const char *) s, slen);
   MY_UNICASE_INFO *uni_plane= cs->caseinfo;
+  register ulong m1= *nr1, m2= *nr2;
 
   while ((s < e) && (res= mb_wc(cs, &wc, (uchar *) s, (uchar *) e)) > 0)
   {
     my_tosort_utf16(uni_plane, &wc);
-    n1[0]^= (((n1[0] & 63) + n2[0]) * (wc & 0xFF)) + (n1[0] << 8);
-    n2[0]+= 3;
-    n1[0]^= (((n1[0] & 63) + n2[0]) * (wc >> 8)) + (n1[0] << 8);
-    n2[0]+= 3;
+    MY_HASH_ADD_16(m1, m2, wc);
     s+= res;
   }
+  *nr1= m1;
+  *nr2= m2;
 }
 
 
@@ -1611,12 +1611,14 @@ my_hash_sort_utf16_bin(CHARSET_INFO *cs,
                        const uchar *pos, size_t len, ulong *nr1, ulong *nr2)
 {
   const uchar *end= pos + cs->cset->lengthsp(cs, (const char *) pos, len);
+  register ulong m1= *nr1, m2= *nr2;
+
   for ( ; pos < end ; pos++)
   {
-    nr1[0]^= (ulong) ((((uint) nr1[0] & 63) + nr2[0]) * 
-              ((uint)*pos)) + (nr1[0] << 8);
-    nr2[0]+= 3;
+    MY_HASH_ADD(m1, m2, (uint)*pos);
   }
+  *nr1= m1;
+  *nr2= m2;
 }
 
 
@@ -2007,22 +2009,15 @@ my_caseup_utf32(CHARSET_INFO *cs, char *src, size_t srclen,
 }
 
 
-static inline void
-my_hash_add(ulong *n1, ulong *n2, uint ch)
-{
-  n1[0]^= (((n1[0] & 63) + n2[0]) * (ch)) + (n1[0] << 8);
-  n2[0]+= 3;
-}
-
-
 static void
 my_hash_sort_utf32(CHARSET_INFO *cs, const uchar *s, size_t slen,
-                   ulong *n1, ulong *n2)
+                   ulong *nr1, ulong *nr2)
 {
   my_wc_t wc;
   int res;
   const uchar *e= s + slen;
   MY_UNICASE_INFO *uni_plane= cs->caseinfo;
+  register ulong m1= *nr1, m2= *nr2;
 
   /* Skip trailing spaces */
   while (e > s + 3 && e[-1] == ' ' && !e[-2] && !e[-3] && !e[-4])
@@ -2031,12 +2026,14 @@ my_hash_sort_utf32(CHARSET_INFO *cs, const uchar *s, size_t slen,
   while ((res= my_utf32_uni(cs, &wc, (uchar*) s, (uchar*) e)) > 0)
   {
     my_tosort_utf32(uni_plane, &wc);
-    my_hash_add(n1, n2, (uint) (wc >> 24));
-    my_hash_add(n1, n2, (uint) (wc >> 16) & 0xFF);
-    my_hash_add(n1, n2, (uint) (wc >> 8)  & 0xFF);
-    my_hash_add(n1, n2, (uint) (wc & 0xFF));
+    MY_HASH_ADD(m1, m2, (uint) (wc >> 24));
+    MY_HASH_ADD(m1, m2, (uint) (wc >> 16) & 0xFF);
+    MY_HASH_ADD(m1, m2, (uint) (wc >> 8)  & 0xFF);
+    MY_HASH_ADD(m1, m2, (uint) (wc & 0xFF));
     s+= res;
   }
+  *nr1= m1;
+  *nr2= m2;
 }
 
 
@@ -2976,12 +2973,13 @@ static size_t my_caseup_ucs2(CHARSET_INFO *cs, char *src, size_t srclen,
 
 
 static void my_hash_sort_ucs2(CHARSET_INFO *cs, const uchar *s, size_t slen,
-			      ulong *n1, ulong *n2)
+			      ulong *nr1, ulong *nr2)
 {
   my_wc_t wc;
   int res;
   const uchar *e=s+slen;
   MY_UNICASE_INFO *uni_plane= cs->caseinfo;
+  register ulong m1= *nr1, m2= *nr2;
 
   while (e > s+1 && e[-1] == ' ' && e[-2] == '\0')
     e-= 2;
@@ -2989,12 +2987,11 @@ static void my_hash_sort_ucs2(CHARSET_INFO *cs, const uchar *s, size_t slen,
   while ((s < e) && (res=my_ucs2_uni(cs,&wc, (uchar *)s, (uchar*)e)) >0)
   {
     my_tosort_ucs2(uni_plane, &wc);
-    n1[0]^= (((n1[0] & 63)+n2[0])*(wc & 0xFF))+ (n1[0] << 8);
-    n2[0]+=3;
-    n1[0]^= (((n1[0] & 63)+n2[0])*(wc >> 8))+ (n1[0] << 8);
-    n2[0]+=3;
+    MY_HASH_ADD_16(m1, m2, wc);
     s+=res;
   }
+  *nr1= m1;
+  *nr2= m2;
 }
 
 
@@ -3312,16 +3309,17 @@ void my_hash_sort_ucs2_bin(CHARSET_INFO *cs __attribute__((unused)),
 			   const uchar *key, size_t len,ulong *nr1, ulong *nr2)
 {
   const uchar *end = key + len;
+  register ulong m1= *nr1, m2= *nr2;
 
   while (end > key+1 && end[-1] == ' ' && end[-2] == '\0')
     end-= 2;
 
   for (; key < (uchar*) end ; key++)
   {
-    nr1[0]^=(ulong) ((((uint) nr1[0] & 63)+nr2[0]) * 
-	     ((uint)*key)) + (nr1[0] << 8);
-    nr2[0]+=3;
+    MY_HASH_ADD(m1, m2, (uint)*key);
   }
+  *nr1= m1;
+  *nr2= m2;
 }
 
 

strings/ctype-utf8.c

@@ -5087,12 +5087,13 @@ static size_t my_caseup_utf8(CHARSET_INFO *cs, char *src, size_t srclen,
 
 
 static void my_hash_sort_utf8(CHARSET_INFO *cs, const uchar *s, size_t slen,
-                              ulong *n1, ulong *n2)
+                              ulong *nr1, ulong *nr2)
 {
   my_wc_t wc;
   int res;
   const uchar *e=s+slen;
   MY_UNICASE_INFO *uni_plane= cs->caseinfo;
+  register ulong m1= *nr1, m2= *nr2;
 
   /*
     Remove end space. We have to do this to be able to compare
@@ -5104,12 +5105,11 @@ static void my_hash_sort_utf8(CHARSET_INFO *cs, const uchar *s, size_t slen,
   while ((s < e) && (res=my_utf8_uni(cs,&wc, (uchar *)s, (uchar*)e))>0 )
   {
     my_tosort_unicode(uni_plane, &wc, cs->state);
-    n1[0]^= (((n1[0] & 63)+n2[0])*(wc & 0xFF))+ (n1[0] << 8);
-    n2[0]+=3;
-    n1[0]^= (((n1[0] & 63)+n2[0])*(wc >> 8))+ (n1[0] << 8);
-    n2[0]+=3;
+    MY_HASH_ADD_16(m1, m2, wc);
     s+=res;
   }
+  *nr1= m1;
+  *nr2= m2;
 }
 
 
@@ -7597,22 +7597,15 @@ my_caseup_utf8mb4(CHARSET_INFO *cs, char *src, size_t srclen,
 }
 
 
-static inline void
-my_hash_add(ulong *n1, ulong *n2, uint ch)
-{
-  n1[0]^= (((n1[0] & 63) + n2[0]) * (ch)) + (n1[0] << 8);
-  n2[0]+= 3;
-}
-
-
 static void
 my_hash_sort_utf8mb4(CHARSET_INFO *cs, const uchar *s, size_t slen,
-                     ulong *n1, ulong *n2)
+                     ulong *nr1, ulong *nr2)
 {
   my_wc_t wc;
   int res;
   const uchar *e= s + slen;
   MY_UNICASE_INFO *uni_plane= cs->caseinfo;
+  register ulong m1= *nr1, m2= *nr2;
 
   /*
     Remove end space. We do this to be able to compare
@@ -7624,8 +7617,7 @@ my_hash_sort_utf8mb4(CHARSET_INFO *cs, const uchar *s, size_t slen,
   while ((res= my_mb_wc_utf8mb4(cs, &wc, (uchar*) s, (uchar*) e)) > 0)
   {
     my_tosort_unicode(uni_plane, &wc, cs->state);
-    my_hash_add(n1, n2, (uint) (wc & 0xFF));
-    my_hash_add(n1, n2, (uint) (wc >> 8)  & 0xFF);
+    MY_HASH_ADD_16(m1, m2, (uint) (wc & 0xFFFF));
     if (wc > 0xFFFF)
     {
        /*
@@ -7635,10 +7627,12 @@ my_hash_sort_utf8mb4(CHARSET_INFO *cs, const uchar *s, size_t slen,
         This is useful to keep order of records in
         test results, e.g. for "SHOW GRANTS".
       */
-      my_hash_add(n1, n2, (uint) (wc >> 16) & 0xFF);
+      MY_HASH_ADD(m1, m2, (uint) ((wc >> 16) & 0xFF));
     }
     s+= res;
   }
+  *nr1= m1;
+  *nr2= m2;
 }
 
 

strings/strings_def.h

@@ -100,4 +100,20 @@ static inline const uchar *skip_trailing_space(const uchar *ptr,size_t len)
     end--;
   return (end);
 }
+
+/* Macros for hashing characters */
+
+#define MY_HASH_ADD(A, B, value) \
+  do { A^= (((A & 63)+B)*((value)))+ (A << 8); B+=3; } while(0)
+
+#define MY_HASH_ADD_16(A, B, value) \
+  do { MY_HASH_ADD(A, B, ((value) & 0xFF)) ; MY_HASH_ADD(A, B, ((value >>8 ))); } while(0) 
+
+/*
+  This one is needed to ensure we get the exact same hash as MariaDB 5.1
+  This is needed to ensure that old partitioned tables still work as before.
+*/
+#define MY_HASH_ADD_16_INV(A, B, value) \
+  do { MY_HASH_ADD(A, B, ((value >> 8))) ; MY_HASH_ADD(A, B, ((value & 0xFF ))); } while(0) 
+
 #endif