> It does make sense to move this to run-command.c from submodule.c > and the function name is already suitable for being global. I > however cannot help wondering if we should also pay attention to the > GIT_CONFIG_KEY_$n and GIT_CONFIG_VALUE_$n pairs (which is not a new > problem in this patch). Note that I changed the function name (the previous one was too submodule-specific). As for the config pairs, they are currently being passed through - do you have a situation in mind in which they should not be passed through? > This helper may sit better next to prep_childenv(), instead of just > saying "the location does not matter, just append randomly at the > end", though. > > Otherwise looking good. > > Thanks. Thanks - I'll move it.
