Jeff King <p...@peff.net> writes: > So it sounds like you did that and the result is not _too_ bad. But I'm > not sure about doing it automatically with sed. For example... > >> diff --git a/packfile.h b/packfile.h >> index d70c6d9afb..dab50405e0 100644 >> --- a/packfile.h >> +++ b/packfile.h >> [...] >> @@ -43,10 +43,10 @@ void for_each_file_in_pack_dir(const char *objdir, >> #define PACKDIR_FILE_PACK 1 >> #define PACKDIR_FILE_IDX 2 >> #define PACKDIR_FILE_GARBAGE 4 >> -extern void (*report_garbage)(unsigned seen_bits, const char *path); >> +void (*report_garbage)(unsigned seen_bits, const char *path); > > This one is a function pointer, and so the extern is actually changing > the visibility of the declared variable. It needs to stay. > > (I didn't read the whole patch carefully, but I knew to look for this > one in particular since I had to deal with it in my patch, too).
Yeah, after reviewing your recent packfile patch, the above hunk was what I went to hunt for directly in Denton's message ;-) As I care much more about correctness than "conflicts with in-flight topics too much", I'd prefer people to understand why we avoid the wholesale update is because of an unnecessary bug like this one, not the one-time conflict resolution load that can be subcontracted out to "rerere" once dealt with ;-) A wholesale rewrite using a tool more suited for the job (e.g. spatch) is a different story, and raises the confidence level of the end result a lot more than a "sed and then eyeball" rewrite, but any tool that allows transformation that is sophisticated enough allows bugs in the instructions we give to the tool, so...
