From mboxrd@z Thu Jan 1 00:00:00 1970 From: Marcelo Ricardo Leitner Date: Tue, 17 Jan 2017 16:37:21 +0000 Subject: Re: Test-Suite for the Partial Reliability Extension Message-Id: <20170117163720.GJ3781@localhost.localdomain> List-Id: References: <20170117122328.GA4322@aomx-netbook> In-Reply-To: <20170117122328.GA4322@aomx-netbook> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable To: linux-sctp@vger.kernel.org On Tue, Jan 17, 2017 at 01:23:28PM +0100, Julian Cordes wrote: > Dear linux sctp mailing list, > I am a master student of the M=FCnster University of Applied Sciences > (Germany). > In the last months I have developed a test suite in packetdrill for the > partial reliability extension of sctp. > Many test cases can run the Linux and FreeBSD kernel and therefore I > have > found some potential problems / bugs with my test suite. > I want to share and discuss these potential issues with you on the > mailing list. Hi Julian, Thanks for coming forward with all this, much appreciated. >=20 > Because the testsuite is published on Github with a permissive license > (bsd-style) you can get all developed tests by cloning the following > repository with git: > https://github.com/nplab/PR_SCTP_Testsuite. > On Github I also have documented the test-results of the test-suite in > README-Files. > Please note that only the tests from the subfolder forward-tsn > are applicable for Linux, because the tests in the other directories > require > support for the NR-SACK and the User Message Interleaving extension. >=20 > Please also note that you need the nplab-version of packetdrill from > https://github.com/nplab/packetdrill > in order to run the tests on linux. The original version of packetdrill > from google misses SCTP support. >=20 > I will send a couple of mails, one for every potential issue that I have > found with the test suite on kernel 4.4.0-x under ubuntu 16.04. > All mails will contain a short description of the potential problem and > a link to the test-case or multiple test-cases that were used to detect > this behaviour. >=20 > Yours sincerely, > Julian Cordes