Avoid integer overflow while sifting-up a heap in tuplesort.c.

If the number of tuples in the heap exceeds approximately INT_MAX/2,
this loop's calculation "2*i+1" could overflow, resulting in a crash.
Fix it by using unsigned int rather than int for the relevant local
variables; that shouldn't cost anything extra on any popular hardware.
Per bug #14722 from Sergey Koposov.

Original patch by Sergey Koposov, modified by me per a suggestion
from Heikki Linnakangas to use unsigned int not int64.

Back-patch to 9.4, where tuplesort.c grew the ability to sort as many
as INT_MAX tuples in-memory (commit 263865a48).

Discussion: https://postgr.es/m/20170629161637.1478.93109@wrigleys.postgresql.org

diff --git a/src/backend/utils/sort/tuplesort.c b/src/backend/utils/sort/tuplesort.c
index 18cb046f19876737c52d8269692a67cd013eed99..9b61a893ca1396447d3a4616604870f2441be726 100644 (file)
--- a/src/backend/utils/sort/tuplesort.c
+++ b/src/backend/utils/sort/tuplesort.c
@@ -2729,7 +2729,7 @@ tuplesort_heap_siftup(Tuplesortstate *state, bool checkIndex)
 {
    SortTuple  *memtuples = state->memtuples;
    SortTuple  *tuple;
-   int         i,
+   unsigned int i,
                n;
 
    if (--state->memtupcount <= 0)
@@ -2737,12 +2737,17 @@ tuplesort_heap_siftup(Tuplesortstate *state, bool checkIndex)
 
    CHECK_FOR_INTERRUPTS();
 
+   /*
+    * state->memtupcount is "int", but we use "unsigned int" for i, j, n.
+    * This prevents overflow in the "2 * i + 1" calculation, since at the top
+    * of the loop we must have i < n <= INT_MAX <= UINT_MAX/2.
+    */
    n = state->memtupcount;
    tuple = &memtuples[n];      /* tuple that must be reinserted */
    i = 0;                      /* i is where the "hole" is */
    for (;;)
    {
-       int         j = 2 * i + 1;
+       unsigned int j = 2 * i + 1;
 
        if (j >= n)
            break;