| --- Log opened Mon Dec 21 00:00:11 2015 | ||
| -!- mafm3 is now known as mafm | 00:40 | |
| -!- Netsplit *.net <-> *.split quits: mboehnert | 09:49 | |
| -!- mboehnert1 is now known as mboehnert | 09:49 | |
| -!- Netsplit *.net <-> *.split quits: blueCmd, wallento, arand, andrzejr | 10:12 | |
| -!- Netsplit over, joins: wallento | 10:12 | |
| wallento | hey guys | 15:37 |
|---|---|---|
| wallento | playing around with travis, jenkins and docker again | 15:37 |
| wallento | stekern, olofk: Did you do something with it recently? | 15:38 |
| stekern | I just pointed olofk to something I had played with way back | 15:38 |
| wallento | okay, it seems we will rely on Munich university for toolchain jenkins | 15:51 |
| wallento | but I am thinking of some simple testing now | 15:51 |
| wallento | also found this platform.io for the first time | 15:51 |
| wallento | anyone used it? | 15:51 |
| olofk | wallento: Haven't seen | 17:07 |
| olofk | What the hell is it? :) | 17:08 |
| olofk | But I really feel the need for continous integration increasing. Been breaking stuff a bit too much lately | 17:08 |
| olofk | Being able to run automatic regression tests in orpsoc-cores would help a lot | 17:09 |
| GeneralStupid | olofk: i know regressiontests for embedded devices... | 17:21 |
| GeneralStupid | olofk: only from automotive industry | 17:21 |
| olofk | GeneralStupid: We have a proof of concept Jenkins system that does tests on one core in orpsoc-cores. I want to scale this to include most of the cores | 17:25 |
| GeneralStupid | olofk: sounds good, how do you do tests for a hardware design? | 17:30 |
| olofk | GeneralStupid: We don't :) | 17:43 |
| olofk | That would be too much work to set up and maintain. There are projects for many different boards in orpsoc-cores, no standardized tests, no standardized way of accessing the boards. It would probably require a full-time position to keep that going | 17:44 |
| GeneralStupid | i just thought about a test for a generic implementation | 17:59 |
| olofk | GeneralStupid: Do you mean on real hardware or in simulation? And what kind of test? | 18:03 |
| GeneralStupid | simulation. | 18:03 |
| GeneralStupid | yes, i dont know how to make a regression test for hardware. | 18:04 |
| wallento | philipp from TU Munich is working on CI with hardware | 18:11 |
| wallento | I also did it for the DE0 nano prototypically at TU Munich | 18:12 |
| wallento | like hooking up JTAG and UART, compile and run a program and check for the output on UART | 18:12 |
| GeneralStupid | special code or is it a "generic" application for regression tests on hardware? | 18:15 |
| -!- arand_ is now known as arand | 20:18 | |
| wallento | GeneralStupid: its generic code, like helloworld | 21:07 |
| wallento | we did not go far with this until now | 21:07 |
| wallento | currently consolidating CI | 21:07 |
| wallento | Finished travis-ci for newlib now as a first step, jenkins at TU Munich will still be there for building and deployment | 21:08 |
| --- Log closed Tue Dec 22 00:00:13 2015 | ||
Generated by irclog2html.py 2.15.2 by Marius Gedminas - find it at mg.pov.lt!