Re: file disappears after git rebase (missing one commit)
- From
Linus Torvalds <torvalds@linux-foundation.org>
- Date
- Aug 18, 2007, 20:01 UTC
- Message-ID
- <alpine.LFD.0.999.0708181247330.30176@woody.linux-foundation.org>
- In-Reply-To
- <e7bda7770708181237u34253bf1h7c3fe0987d13d3b3@mail.gmail.com>
On Sat, 18 Aug 2007, Torgil Svensson wrote:
Show 5 quoted lines
> > $ git rebase devel > First, rewinding head to replay your work on top of it... > HEAD is now at 57aa840... Add disk summarize tool (du.exe) > Nothing to do.
Ok. "git rebase" really does believe that there's nothing to do.
The reason, I think, is that I suspect that the newly added file is a binary file, no? That, in turn, will mean that the *patch* will have no patch ID (or rather, it will have an empty patch ID) - which in turn will make it invisible to "--ignore-if-in-upstream" if there are already some *other* patches that also just adds a binary file (which I think there is: I think upstream has "Add disk summarize tool (du.exe)" which I assume has exactly the same patch fingerprint).
In other words, "git rebase" really is just a series of cherry-picks, but it avoids patches that have the same patch ID as something that is already upstream. That helps *enormously*, but it so happens that the patch ID's don't work really well for binary diffs.
Try this patch - see if it helps. Totally untested! It will enable patch ID's on binary diffs too, which should avoid this issue.
Linus
---
diff --git a/patch-ids.c b/patch-ids.c index a288fac..4a3432e 100644 --- a/patch-ids.c +++ b/patch-ids.c @@ -122,6 +122,7 @@ int init_patch_ids(struct patch_ids *ids) memset(ids, 0, sizeof(*ids)); diff_setup(&ids->diffopts); ids->diffopts.recursive = 1; + ids->diffopts.binary = 1; if (diff_setup_done(&ids->diffopts) < 0) return error("diff_setup_done failed"); return 0;