On sparc64+ext4, suppress test failures from known WAL read failure.

nmisch · nmisch · commit ce6d79368efa · 2022-01-26T18:06:19.000-08:00
Buildfarm members kittiwake, tadarida and snapper began to fail frequently when commits 3cd9c3b and f47ed79 added tests of concurrency, but the problem was reachable before those commits. Back-patch to v10 (all supported versions). Discussion: https://postgr.es/m/20220116210241.GC756210@rfd.leadboat.com
diff --git a/contrib/amcheck/t/003_cic_2pc.pl b/contrib/amcheck/t/003_cic_2pc.pl
@@ -11,6 +11,8 @@
 
 use Test::More tests => 5;
 
+local $TODO = 'filesystem bug' if PostgreSQL::Test::Utils::has_wal_read_bug;
+
 my ($node, $result);
 
 #
diff --git a/src/test/perl/PostgreSQL/Test/Utils.pm b/src/test/perl/PostgreSQL/Test/Utils.pm
@@ -351,6 +351,29 @@ sub perl2host
 
 =pod
 
+=item has_wal_read_bug()
+
+Returns true if $tmp_check is subject to a sparc64+ext4 bug that causes WAL
+readers to see zeros if another process simultaneously wrote the same offsets.
+Consult this in tests that fail frequently on affected configurations.  The
+bug has made streaming standbys fail to advance, reporting corrupt WAL.  It
+has made COMMIT PREPARED fail with "could not read two-phase state from WAL".
+Non-WAL PostgreSQL reads haven't been affected, likely because those readers
+and writers have buffering systems in common.  See
+https://postgr.es/m/20220116210241.GC756210@rfd.leadboat.com for details.
+
+=cut
+
+sub has_wal_read_bug
+{
+	return
+	     $Config{osname} eq 'linux'
+	  && $Config{archname} =~ /^sparc/
+	  && !run_log([ qw(df -x ext4), $tmp_check ], '>', '/dev/null', '2>&1');
+}
+
+=pod
+
 =item system_log(@cmd)
 
 Run (via C<system()>) the command passed as argument; the return
diff --git a/src/test/recovery/t/027_stream_regress.pl b/src/test/recovery/t/027_stream_regress.pl
@@ -3,9 +3,20 @@
 use warnings;
 use PostgreSQL::Test::Cluster;
 use PostgreSQL::Test::Utils;
-use Test::More tests => 4;
+use Test::More;
 use File::Basename;
 
+if (PostgreSQL::Test::Utils::has_wal_read_bug)
+{
+	# We'd prefer to use "local $TODO", but the bug causes this test file to
+	# die(), not merely to fail.
+	plan skip_all => 'filesystem bug';
+}
+else
+{
+	plan tests => 4;
+}
+
 # Initialize primary node
 my $node_primary = PostgreSQL::Test::Cluster->new('primary');
 $node_primary->init(allows_streaming => 1);

Original file line number	Diff line number	Diff line change
`@@ -11,6 +11,8 @@`
`11`	`11`
`12`	`12`	`use Test::More tests => 5;`
`13`	`13`
	`14`	`+local $TODO = 'filesystem bug' if PostgreSQL::Test::Utils::has_wal_read_bug;`
	`15`	`+`
`14`	`16`	`my ($node, $result);`
`15`	`17`
`16`	`18`	`#`