Johannes Schindelin <johannes.schinde...@gmx.de> writes: > The bug that bit me (hard!) and that triggered not only a long series of > curses but also my writing a patch and sending it to the list was that > `git rev-parse --git-path HEAD` would give *incorrect* output when run > in a subdirectory of a regular checkout, but *correct* output when run > in a subdirectory of an associated *worktree*. > > I had tested the script in question quite a bit, but in a worktree. And > in production, it quietly did exactly the wrong thing. > > Changes relative to v2: > > - the "iffy" test in t1700 was made "uniffy" > > - clarified in the commit message of 2/2 why we can get away with the > "reset then use" pattern
It is no longer relevant between "reset then use" and "use then reset", I think, because you did something much better, which is to move strbuf_release() up so that it comes before the possible early returns. Both patches look good. Let's queue this and move it to 'next' shortly. Personally, I think it is OK to fast-track this to 'master' before the final, but just like any other bugs, we've lived with the bug for some time, and it is not a big deal if we have to live with it for a bit longer. Thanks.
