From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from ganesha.gnumonks.org (ganesha.gnumonks.org [IPv6:2001:780:45:1d:225:90ff:fe52:c662]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id E83EFD47 for ; Wed, 6 Dec 2023 09:18:07 -0800 (PST) Received: from [78.30.43.141] (port=50702 helo=gnumonks.org) by ganesha.gnumonks.org with esmtpsa (TLS1.3) tls TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384 (Exim 4.94.2) (envelope-from ) id 1rAvXA-0042D2-0Q; Wed, 06 Dec 2023 18:18:06 +0100 Date: Wed, 6 Dec 2023 18:18:03 +0100 From: Pablo Neira Ayuso To: Thomas Haller Cc: NetFilter Subject: Re: [PATCH nft v2 0/5] add infrastructure for unit tests Message-ID: References: <20231105150955.349966-1-thaller@redhat.com> <792015e09808892af632b8297195dd6bc449b1ae.camel@redhat.com> <66b156090ff987567645b9e84aa2d1469823fc2b.camel@redhat.com> Precedence: bulk X-Mailing-List: netfilter-devel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <66b156090ff987567645b9e84aa2d1469823fc2b.camel@redhat.com> X-Spam-Score: -1.8 (-) On Wed, Dec 06, 2023 at 08:53:23AM +0100, Thomas Haller wrote: > On Tue, 2023-11-21 at 13:37 +0100, Pablo Neira Ayuso wrote: > > On Tue, Nov 21, 2023 at 01:34:54PM +0100, Thomas Haller wrote: > > > Hi Pablo, > > > > > > any concerns about this? Could it be merged? > > > > Sorry. JSON support is not working, I had to locally revert those > > patches to run tests on -stable 5.4 here. > > > > Let's agree on some basic rule from now on: One series at a time > > only, > > anything else coming after will be marked as deferred in patchwork. > > > > Thanks. > > > > > Hi, > > Could this be considered? > > This provides the basis for unit tests (and the possibility to even add > any such tests). We are still discussing the json integration into tests/shell. I suggest, let dust settle on each front before making more changes. > It also hooks up tests to `make check`. Which would be desirable to > build upon. `make check` currently does nothing. For example, Florian's > afl++ patches could hook into `make check` (or `make check-more`), if > this basis was there. I still doubt `make check` provides any benefit to the release process, which will exercise this path because of `make distcheck' which I might have to relax it to `make dist' to skip this to ensure release process is reliable. I think all these tests should continously and provide reports to us, but not necessarily integrate them into `make check'.