On Thu, Oct 11, 2018 at 06:04:11PM -0700, Andres Freund wrote: > culicidae tests EXEC_BACKEND, so there's an explanation as to why it > sometimes behaves differently. But here I don't immediately see how > that'd matter. Probably still worth verifying that it's not somehow > caused by that.
Thanks, that's the point of detail I needed about culicidae (you will need to explain me one day face-to-face how you pronounce it). I have been able to reproduce the problem, and that's a bug within pg_verify_checksums as it fails to consider that config_exec_params is not a file it should scan when using EXEC_BACKEND. The same can happen with config_exec_params.new. The attached, which fixes the issue for me, needs to be back-patched to v11. -- Michael
diff --git a/src/bin/pg_verify_checksums/pg_verify_checksums.c b/src/bin/pg_verify_checksums/pg_verify_checksums.c
index 1bc020ab6c..a6c884f149 100644
--- a/src/bin/pg_verify_checksums/pg_verify_checksums.c
+++ b/src/bin/pg_verify_checksums/pg_verify_checksums.c
@@ -54,6 +54,10 @@ static const char *const skip[] = {
"pg_filenode.map",
"pg_internal.init",
"PG_VERSION",
+#ifdef EXEC_BACKEND
+ "config_exec_params",
+ "config_exec_params.new",
+#endif
NULL,
};
signature.asc
Description: PGP signature
