From mboxrd@z Thu Jan 1 00:00:00 1970 From: Alexey Kodanev Date: Mon, 16 May 2016 16:57:31 +0300 Subject: [LTP] [RFC][PATCH 1/3] network: run network tests from network.sh In-Reply-To: <20160516124711.GD6454@rei.lan> References: <1462869965-25841-1-git-send-email-alexey.kodanev@oracle.com> <20160516123103.GC6454@rei.lan> <20160516124711.GD6454@rei.lan> Message-ID: <5739D1CB.6010500@oracle.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: ltp@lists.linux.it Hi, On 05/16/2016 03:47 PM, Cyril Hrubis wrote: > Hi! >> This does not say what is the reason for this change. >> >> How is this better than runing runtest files from runltp? >> >> You supposedly define a few variables in the script that starts the >> testcases rather than having a common script that is sourced at the >> start of the network testcases. Is there any other thing I'm missing? > Looking into the networkstress.sh and networktest.sh it starts ltp-pan > itself as well. So this more or less copies the functionality from there > to network.sh. Yeah, just making the one as I don't see the point of having two, almost equal, scripts. > > I would say that the defaults for variables should not be defined in the > script that runs selected (subset of) tests. That is because it should > be possible to run just the test script and expect it to have the same > defaults as when it's executed via the script. Why don't we stick the > defaults to test_net.sh and keep just the code that selects testcases > to run and start ltp-pan in the network.sh? > Agree, we can do that, there will be no difference when running them from network.sh, and it will be possible to run tests individually with already pre-defined default variables. Thanks, Alexey