> To be clear, are you saying "git bisect" told you that 76531d4166 was > the first bad commit?
Not exactly. I told you I got confused on how to use git-bisect, so I didn't use it (I didn't understand how is it supposed to check each commit, when another kernel is running). I simply used your first suggestion for debugging which was: git checkout 76531d4166... cp /boot/config... .config make -j4 deb-pkg The kernel produced here showed the problem, however another kernel compiled before with commit 7d339ae99758bc21033d4a19bcd4f7b55f96e24e (which is exactly the one prior) wasn't affected. > If so, that means that _both_ its parents (7d339ae and a353fbb) tested > fine and that the merge tested as broken. Which would mean that > either we have a heisenbug and all information so far is suspect, or > their interaction produced the bug. mmm... I believe I tested the "first" parent and it worked fine. On the git web link I sent you that lists all the kernel commits that occurred around this time, there's a clear series of commits. It follows that 76531d4166 is preceded by 7d339ae, so I tested both, the former showed the problem and the latter did not. Is this enough? > If you have been using "git bisect", could you attach the output of > "git bisect log"? Sorry I did not use it. If you point me in the right direction I can try to use it Thanks for all your help Andres -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/CAH=dYrEHw-m467atr1CbD+=+6x+ulkvrmcpp6xhovzvw6zy...@mail.gmail.com
