Hi, On 2014-08-18 17:18, Wookey wrote: > We upgraded the kernel on the build machine to 3.16.0+ and tried a keyutils > build. It still failed:
Too bad... a simple solution such as this would have been nice. Two observations: > === /home/buildd/packages/keyutils-1.5.9/tests/keyctl/padd/bad-args/test.out > === > +++ CHECK EMPTY KEY TYPE > +++ CHECK UNSUPPORTED KEY TYPE > +++ CHECK INVALID KEY TYPE > FAILED The "bad-args" class of test cases are there to verify that bad arguments to the individual commands are caught. Here, adding a key with invalid type should have returned EPERM. I can't tell from this output but in the full test log, you should see either another error code (which wouldn't be all too bad, I guess) or no error at all (that would be very bad). > +++ CHECK MAXLEN KEY TYPE > +++ CHECK OVERLONG KEY TYPE > +++ CHECK ADD KEYRING WITH PAYLOAD > +++ CHECK MAXLEN DESC > +++ CHECK BAD KEY ID > Makefile:42: recipe for target 'run' failed > make[3]: *** [run] Error 1 > make[3]: Leaving directory '/home/buildd/packages/keyutils-1.5.9/tests' > Makefile:200: recipe for target 'test' failed > > I was hoping this would go away. This seems tobe a different place > from before. More investigation later. I should note that the order the tests are executed in is non-deterministic (tests are located using find(1)). So technically the same test could still be failing but the execution doesn't get that far. On the other hand, if the original problem did in fact go away, that would be a good sign for once... Christian -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
