protect pg_stat_statements_info() for being used without the library loaded
Hi,
Attached is a small patch for ${subject}
--
Jaime Casanova
Director de Servicios Profesionales
SystemGuards - Consultores de PostgreSQL
Attachments:
fix_pg_stat_statements_info.patchtext/x-patch; charset=US-ASCII; name=fix_pg_stat_statements_info.patchDownload
diff --git a/contrib/pg_stat_statements/pg_stat_statements.c b/contrib/pg_stat_statements/pg_stat_statements.c
index 72a117fc19..62cccbfa44 100644
--- a/contrib/pg_stat_statements/pg_stat_statements.c
+++ b/contrib/pg_stat_statements/pg_stat_statements.c
@@ -1898,6 +1898,11 @@ pg_stat_statements_info(PG_FUNCTION_ARGS)
Datum values[PG_STAT_STATEMENTS_INFO_COLS];
bool nulls[PG_STAT_STATEMENTS_INFO_COLS];
+ if (!pgss || !pgss_hash)
+ ereport(ERROR,
+ (errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
+ errmsg("pg_stat_statements must be loaded via shared_preload_libraries")));
+
/* Build a tuple descriptor for our result type */
if (get_call_result_type(fcinfo, NULL, &tupdesc) != TYPEFUNC_COMPOSITE)
elog(ERROR, "return type must be a row type");
On Thu, Jan 28, 2021 at 3:53 AM Jaime Casanova
<jcasanov@systemguards.com.ec> wrote:
Hi,
Attached is a small patch for ${subject}
Good catch, and patch looks good to me.
On Thu, Jan 28, 2021 at 08:49:54AM +0800, Julien Rouhaud wrote:
Good catch, and patch looks good to me.
This crashes the server, cash. Looking at all the other modules in
the tree, I am not seeing any other hole. This is new as of 9fbc3f3,
and I will apply it on HEAD.
--
Michael
On 2021/01/28 15:42, Michael Paquier wrote:
On Thu, Jan 28, 2021 at 08:49:54AM +0800, Julien Rouhaud wrote:
Good catch, and patch looks good to me.
This crashes the server, cash. Looking at all the other modules in
the tree, I am not seeing any other hole. This is new as of 9fbc3f3,
and I will apply it on HEAD.
Thanks!
Regards,
--
Fujii Masao
Advanced Computing Technology Center
Research and Development Headquarters
NTT DATA CORPORATION