I am using the same scripts on my test machine and if anyone would like I can dump the daily logs on a github for refference
Alin On Thu, Jan 2, 2020 at 6:50 PM Disruptive Solutions < disruptivesolution...@gmail.com> wrote: > Very interesting... I would like to use build scripts in the Jenkins > solution I have running to test my STM book configs... > > Verstuurd vanaf mijn iPhone > > > Op 2 jan. 2020 om 18:40 heeft Gregory Nutt <spudan...@gmail.com> het > volgende geschreven: > > > > > >> I have some scripts that I use to perform build testing. That is, just > build as many configurations. This only verifies that no existing builds > are broken. It does not prove that a change is correct or that a change > will not break a build that actually uses the change. It would take a much > smarter test to do that then such a brute force test (when, I don't > advocate such dumb build testing). > >> > >> But it would be nice to have it running as a 'cron' somewhere. It > currently builds 419 ARM configurations! plus several dozen sim, AVR, and > MIPS (PIC32 configurations). It would be nice if there were available > online some so that if someone wants to make a change to many files like > this in the future, we can at least be assured it won't break builds. > >> > >> The build test is based primarily on nuttx/tools/testbuild.sh which > will build a list of configurations. The configuration lists are elsewhere > at https://bitbucket.org/nuttx/tools/src/master/buildtest/ There are > also some bonehead scripts at the location that only work on my machine. > These basically just setup paths, and manage running > nuttx/tools/testbuild.sh. Those bonehead scripts are setup to work only in > my environment and would not work anywhere else without modification. > >> > >> But if we could make this runnable online (perhaps via Jenkins?), then > we could avoid such extensive code breakage in the future. > > > > I should also mention how I use this. > > > > * I keep the last "successful" build test output in, for example, > > armtest.log.last > > * I run the test generating a new armtest.log > > * I then 'diff -u armtest.log.last armtest.log | vim -' and I can see > > if any new warnings or errors occur. > > > > There is another bonehead script, monitor.sh, that just shows me the > running status of the test. I am sure that the people on this list with > more devops skills than I could make this all work smoother. > > > > > > >
