On 16/10/20 3:47 pm, Sebastian Huber wrote: > On 16/10/2020 06:38, Chris Johns wrote: >>> * has no tests, >> And all the other commands have tests? > > No, so this is admittedly not a very strong argument. > > It is a general problem that the shell code has little to no test coverage. > Fixing bugs and warnings in this area is always a bit risky. Most recent > warnings were in this area.
Yes. I have started to consider some sort of `expect` type support for rtems-test to handle input side but I have no idea how we could know the output is OK? A basic compare would make the tests fragile. Chris _______________________________________________ devel mailing list [email protected] http://lists.rtems.org/mailman/listinfo/devel
