1MERGE STRATEGIES 2---------------- 3 4The merge mechanism (`git merge` and `git pull` commands) allows the 5backend 'merge strategies' to be chosen with `-s` option. Some strategies 6can also take their own options, which can be passed by giving `-X<option>` 7arguments to `git merge` and/or `git pull`. 8 9resolve:: 10 This can only resolve two heads (i.e. the current branch 11 and another branch you pulled from) using a 3-way merge 12 algorithm. It tries to carefully detect criss-cross 13 merge ambiguities and is considered generally safe and 14 fast. 15 16recursive:: 17 This can only resolve two heads using a 3-way merge 18 algorithm. When there is more than one common 19 ancestor that can be used for 3-way merge, it creates a 20 merged tree of the common ancestors and uses that as 21 the reference tree for the 3-way merge. This has been 22 reported to result in fewer merge conflicts without 23 causing mismerges by tests done on actual merge commits 24 taken from Linux 2.6 kernel development history. 25 Additionally this can detect and handle merges involving 26 renames. This is the default merge strategy when 27 pulling or merging one branch. 28+ 29The 'recursive' strategy can take the following options: 30 31ours;; 32 This option forces conflicting hunks to be auto-resolved cleanly by 33 favoring 'our' version. Changes from the other tree that do not 34 conflict with our side are reflected to the merge result. 35 For a binary file, the entire contents are taken from our side. 36+ 37This should not be confused with the 'ours' merge strategy, which does not 38even look at what the other tree contains at all. It discards everything 39the other tree did, declaring 'our' history contains all that happened in it. 40 41theirs;; 42 This is the opposite of 'ours'; note that, unlike 'ours', there is 43 no 'theirs' merge stragegy to confuse this merge option with. 44 45patience;; 46 With this option, 'merge-recursive' spends a little extra time 47 to avoid mismerges that sometimes occur due to unimportant 48 matching lines (e.g., braces from distinct functions). Use 49 this when the branches to be merged have diverged wildly. 50 See also linkgit:git-diff[1] `--patience`. 51 52diff-algorithm=[patience|minimal|histogram|myers];; 53 Tells 'merge-recursive' to use a different diff algorithm, which 54 can help avoid mismerges that occur due to unimportant matching 55 lines (such as braces from distinct functions). See also 56 linkgit:git-diff[1] `--diff-algorithm`. 57 58ignore-space-change;; 59ignore-all-space;; 60ignore-space-at-eol;; 61 Treats lines with the indicated type of whitespace change as 62 unchanged for the sake of a three-way merge. Whitespace 63 changes mixed with other changes to a line are not ignored. 64 See also linkgit:git-diff[1] `-b`, `-w`, and 65 `--ignore-space-at-eol`. 66+ 67* If 'their' version only introduces whitespace changes to a line, 68 'our' version is used; 69* If 'our' version introduces whitespace changes but 'their' 70 version includes a substantial change, 'their' version is used; 71* Otherwise, the merge proceeds in the usual way. 72 73renormalize;; 74 This runs a virtual check-out and check-in of all three stages 75 of a file when resolving a three-way merge. This option is 76 meant to be used when merging branches with different clean 77 filters or end-of-line normalization rules. See "Merging 78 branches with differing checkin/checkout attributes" in 79 linkgit:gitattributes[5] for details. 80 81no-renormalize;; 82 Disables the `renormalize` option. This overrides the 83 `merge.renormalize` configuration variable. 84 85no-renames;; 86 Turn off rename detection. 87 See also linkgit:git-diff[1] `--no-renames`. 88 89find-renames[=<n>];; 90 Turn on rename detection, optionally setting the similarity 91 threshold. This is the default. 92 See also linkgit:git-diff[1] `--find-renames`. 93 94rename-threshold=<n>;; 95 Deprecated synonym for `find-renames=<n>`. 96 97subtree[=<path>];; 98 This option is a more advanced form of 'subtree' strategy, where 99 the strategy makes a guess on how two trees must be shifted to 100 match with each other when merging. Instead, the specified path 101 is prefixed (or stripped from the beginning) to make the shape of 102 two trees to match. 103 104octopus:: 105 This resolves cases with more than two heads, but refuses to do 106 a complex merge that needs manual resolution. It is 107 primarily meant to be used for bundling topic branch 108 heads together. This is the default merge strategy when 109 pulling or merging more than one branch. 110 111ours:: 112 This resolves any number of heads, but the resulting tree of the 113 merge is always that of the current branch head, effectively 114 ignoring all changes from all other branches. It is meant to 115 be used to supersede old development history of side 116 branches. Note that this is different from the -Xours option to 117 the 'recursive' merge strategy. 118 119subtree:: 120 This is a modified recursive strategy. When merging trees A and 121 B, if B corresponds to a subtree of A, B is first adjusted to 122 match the tree structure of A, instead of reading the trees at 123 the same level. This adjustment is also done to the common 124 ancestor tree. 125 126With the strategies that use 3-way merge (including the default, 'recursive'), 127if a change is made on both branches, but later reverted on one of the 128branches, that change will be present in the merged result; some people find 129this behavior confusing. It occurs because only the heads and the merge base 130are considered when performing a merge, not the individual commits. The merge 131algorithm therefore considers the reverted change as no change at all, and 132substitutes the changed version instead.