You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
One common mistake is to start preparing a patch series and before finishing it do a git fetch. When this is done the RESULT_BRANCH in the remote may not match the one the feature/fix started from. When we do a git pile format-patch it implicitly uses the remote branch to compare with git range-diff. That range-diff will not make sense if it contains changes outside what we really changed.
One idea to cope with that is: assuming pile_branch and result_branch from the remote are always in sync (i.e. nobody is pushing result_branch but rather generating from it), check if the merge-base of pile and origin/pile is origin/pile. If it is something else, it means the branches diverged and we should instruct the user to use the longer form of format-patch, passing a sensible result_branch for comparison.
The text was updated successfully, but these errors were encountered:
One common mistake is to start preparing a patch series and before finishing it do a
git fetch
. When this is done the RESULT_BRANCH in the remote may not match the one the feature/fix started from. When we do agit pile format-patch
it implicitly uses the remote branch to compare with git range-diff. That range-diff will not make sense if it contains changes outside what we really changed.One idea to cope with that is: assuming pile_branch and result_branch from the remote are always in sync (i.e. nobody is pushing result_branch but rather generating from it), check if the merge-base of pile and origin/pile is origin/pile. If it is something else, it means the branches diverged and we should instruct the user to use the longer form of format-patch, passing a sensible result_branch for comparison.
The text was updated successfully, but these errors were encountered: