On 2025-Jun-26, Fujii Masao wrote:
> However, with the patch applied, I encountered a segmentation fault in pg_dump
> as follows:
Ah, thanks for the test case. Yeah, I removed one 'if' condition too
many from Jian's patch. We just need to test the constraint name for
nullness and then things seem to work:
diff --git a/src/bin/pg_dump/pg_dump.c b/src/bin/pg_dump/pg_dump.c
index 5dd1f42631d..7b5024d4f71 100644
--- a/src/bin/pg_dump/pg_dump.c
+++ b/src/bin/pg_dump/pg_dump.c
@@ -17719,7 +17719,8 @@ dumpTableSchema(Archive *fout, const TableInfo *tbinfo)
for (j = 0; j < tbinfo->numatts; j++)
{
- if (tbinfo->notnull_comment[j] != NULL)
+ if (tbinfo->notnull_constrs[j] != NULL &&
+ tbinfo->notnull_comment[j] != NULL)
{
if (comment == NULL)
{
--
Álvaro Herrera 48°01'N 7°57'E — https://www.EnterpriseDB.com/