From: Eric Whitney Subject: latest xfstests-bld and nojournal test failures Date: Tue, 16 Aug 2016 17:45:19 -0400 Message-ID: <20160816214519.GA18888@localhost.localdomain> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: tytso@mit.edu To: linux-ext4@vger.kernel.org Return-path: Received: from mail-qk0-f195.google.com ([209.85.220.195]:35866 "EHLO mail-qk0-f195.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751266AbcHPVlp (ORCPT ); Tue, 16 Aug 2016 17:41:45 -0400 Received: by mail-qk0-f195.google.com with SMTP id v123so7514913qkh.3 for ; Tue, 16 Aug 2016 14:41:44 -0700 (PDT) Content-Disposition: inline Sender: linux-ext4-owner@vger.kernel.org List-ID: FYI for those running the latest version of the test appliance: Commit 5e574181f3bf ("test appliance - automate selection of test/scratch devices") deleted the ^has_journal option from the MKFS_OPTIONS string in the configuration file for the nojournal test case. This leads to failures for generic/034 and 18 other tests in the nojournal test case; all of them were prevented from running in the previous version of the test appliance by the xfstest function _require_metadata_journaling(). That function determines whether a test file system supports metadata journaling by using dumpe2fs to see whether the test or scratch file system contains a journal. Use of the noload mount option as specified by EXT_MOUNT_OPTIONS doesn't affect that determination. It looks like we'll want to add code to _require_metadata_journaling to check for use of the noload mount option as well as presence/absence of a journal. Eric