From: Hugo van der Sanden, Pirum Date: Mon, 29 Jan 2018 15:36:11 GMT Subject: rebase oddity Message-ID: Hi, I'm encountering a rebase oddity, and would welcome advice on how I could debug it far enough to characterize it so I can (one or both of) report it as a bug, or detect and avoid the situation. I can reproduce this on my primary development host, running git-2.11.0 on a debian (jessie) VM hosted on a Windows-10 machine via VirtualBox 5.2.2, on a filesystem native to a real debian machine mounted over Samba and imported by VirtualBox into the VM. I have not been able to reproduce it on similar checkouts on the real debian machine, nor in a Cygwin shell over a native Windows partition (running git-2.11.0 in each case). I have not yet attempted to update git version (which would involve getting permission to do so). In short: rebase stops on the second commit with the claim "Your local changes to the following files would be overwritten by merge: lib/NL/Control/Enrich.pm"; this file did not have local changes, and was not modified by any commit other than this one beyond the merge-base in either branch. I would expect it to apply cleanly during rebase, and indeed cherry-picking it at the same point manages to apply it with no problem. GIT_TRACE=1 does not show enough detail for me to guess what's going wrong, are there other diagnostic options that would show more detail, short of strace(1)? I've done numerous similar rebases in this environment, and not had anything like this happen before. The long version follows, any clues gratefully received. Hugo van der Sanden <