1 Date: Fri, 19 Dec 2008 00:45:19 -0800
2 From: Linus Torvalds <torvalds@linux-foundation.org>, Junio C Hamano <gitster@pobox.com>
3 Subject: Re: Odd merge behaviour involving reverts
4 Abstract: Sometimes a branch that was already merged to the mainline
5 is later found to be faulty. Linus and Junio give guidance on
6 recovering from such a premature merge and continuing development
7 after the offending branch is fixed.
8 Message-ID: <7vocz8a6zk.fsf@gitster.siamese.dyndns.org>
9 References: <alpine.LFD.2.00.0812181949450.14014@localhost.localdomain>
11 Alan <alan@clueserver.org> said:
13 I have a master branch. We have a branch off of that that some
14 developers are doing work on. They claim it is ready. We merge it
15 into the master branch. It breaks something so we revert the merge.
16 They make changes to the code. they get it to a point where they say
17 it is ok and we merge again.
19 When examined, we find that code changes made before the revert are
20 not in the master branch, but code changes after are in the master
21 branch.
23 and asked for help recovering from this situation.
25 The history immediately after the "revert of the merge" would look like
26 this:
28 ---o---o---o---M---x---x---W
29 /
30 ---A---B
32 where A and B are on the side development that was not so good, M is the
33 merge that brings these premature changes into the mainline, x are changes
34 unrelated to what the side branch did and already made on the mainline,
35 and W is the "revert of the merge M" (doesn't W look M upside down?).
36 IOW, "diff W^..W" is similar to "diff -R M^..M".
38 Such a "revert" of a merge can be made with:
40 $ git revert -m 1 M
42 After the develpers of the side branch fixes their mistakes, the history
43 may look like this:
45 ---o---o---o---M---x---x---W---x
46 /
47 ---A---B-------------------C---D
49 where C and D are to fix what was broken in A and B, and you may already
50 have some other changes on the mainline after W.
52 If you merge the updated side branch (with D at its tip), none of the
53 changes made in A nor B will be in the result, because they were reverted
54 by W. That is what Alan saw.
56 Linus explains the situation:
58 Reverting a regular commit just effectively undoes what that commit
59 did, and is fairly straightforward. But reverting a merge commit also
60 undoes the _data_ that the commit changed, but it does absolutely
61 nothing to the effects on _history_ that the merge had.
63 So the merge will still exist, and it will still be seen as joining
64 the two branches together, and future merges will see that merge as
65 the last shared state - and the revert that reverted the merge brought
66 in will not affect that at all.
68 So a "revert" undoes the data changes, but it's very much _not_ an
69 "undo" in the sense that it doesn't undo the effects of a commit on
70 the repository history.
72 So if you think of "revert" as "undo", then you're going to always
73 miss this part of reverts. Yes, it undoes the data, but no, it doesn't
74 undo history.
76 In such a situation, you would want to first revert the previous revert,
77 which would make the history look like this:
79 ---o---o---o---M---x---x---W---x---Y
80 /
81 ---A---B-------------------C---D
83 where Y is the revert of W. Such a "revert of the revert" can be done
84 with:
86 $ git revert W
88 This history would (ignoring possible conflicts between what W and W..Y
89 changed) be equivalent to not having W nor Y at all in the history:
91 ---o---o---o---M---x---x-------x----
92 /
93 ---A---B-------------------C---D
95 and merging the side branch again will not have conflict arising from an
96 earlier revert and revert of the revert.
98 ---o---o---o---M---x---x-------x-------*
99 / /
100 ---A---B-------------------C---D
102 Of course the changes made in C and D still can conflict with what was
103 done by any of the x, but that is just a normal merge conflict.
105 On the other hand, if the developers of the side branch discarded their
106 faulty A and B, and redone the changes on top of the updated mainline
107 after the revert, the history would have looked like this:
109 ---o---o---o---M---x---x---W---x---x
110 / \
111 ---A---B A'--B'--C'
113 If you reverted the revert in such a case as in the previous example:
115 ---o---o---o---M---x---x---W---x---x---Y---*
116 / \ /
117 ---A---B A'--B'--C'
119 where Y is the revert of W, A' and B'are rerolled A and B, and there may
120 also be a further fix-up C' on the side branch. "diff Y^..Y" is similar
121 to "diff -R W^..W" (which in turn means it is similar to "diff M^..M"),
122 and "diff A'^..C'" by definition would be similar but different from that,
123 because it is a rerolled series of the earlier change. There will be a
124 lot of overlapping changes that result in conflicts. So do not do "revert
125 of revert" blindly without thinking..
127 ---o---o---o---M---x---x---W---x---x
128 / \
129 ---A---B A'--B'--C'
131 In the history with rebased side branch, W (and M) are behind the merge
132 base of the updated branch and the tip of the mainline, and they should
133 merge without the past faulty merge and its revert getting in the way.
135 To recap, these are two very different scenarios, and they want two very
136 different resolution strategies:
138 - If the faulty side branch was fixed by adding corrections on top, then
139 doing a revert of the previous revert would be the right thing to do.
141 - If the faulty side branch whose effects were discarded by an earlier
142 revert of a merge was rebuilt from scratch (i.e. rebasing and fixing,
143 as you seem to have interpreted), then re-merging the result without
144 doing anything else fancy would be the right thing to do.
146 However, there are things to keep in mind when reverting a merge (and
147 reverting such a revert).
149 For example, think about what reverting a merge (and then reverting the
150 revert) does to bisectability. Ignore the fact that the revert of a revert
151 is undoing it - just think of it as a "single commit that does a lot".
152 Because that is what it does.
154 When you have a problem you are chasing down, and you hit a "revert this
155 merge", what you're hitting is essentially a single commit that contains
156 all the changes (but obviously in reverse) of all the commits that got
157 merged. So it's debugging hell, because now you don't have lots of small
158 changes that you can try to pinpoint which _part_ of it changes.
160 But does it all work? Sure it does. You can revert a merge, and from a
161 purely technical angle, git did it very naturally and had no real
162 troubles. It just considered it a change from "state before merge" to
163 "state after merge", and that was it. Nothing complicated, nothing odd,
164 nothing really dangerous. Git will do it without even thinking about it.
166 So from a technical angle, there's nothing wrong with reverting a merge,
167 but from a workflow angle it's something that you generally should try to
168 avoid.
170 If at all possible, for example, if you find a problem that got merged
171 into the main tree, rather than revert the merge, try _really_ hard to
172 bisect the problem down into the branch you merged, and just fix it, or
173 try to revert the individual commit that caused it.
175 Yes, it's more complex, and no, it's not always going to work (sometimes
176 the answer is: "oops, I really shouldn't have merged it, because it wasn't
177 ready yet, and I really need to undo _all_ of the merge"). So then you
178 really should revert the merge, but when you want to re-do the merge, you
179 now need to do it by reverting the revert.