"David Rowley" <dgrowley@gmail.com> writes:
> patching file src/include/catalog/pg_proc.h
> Hunk #4 FAILED at 106.
> 1 out of 4 hunks FAILED -- saving rejects to file
> src/include/catalog/pg_proc.h.rej
I imagine you'll find that "hunk #4" covers the entire DATA() body of
the file :-(. It can't possibly apply cleanly if anyone's added or
altered pg_proc entries since the patch was made.
What you'd need to do is manually insert proiswfunc = 'f' entries in
all the existing DATA lines (this is usually not too hard with sed or
an emacs macro), then add whatever new functions the patch defines.
Even figuring out the latter from the patch representation can be a
serious PITA, since they'll be a few lines out of a multi-thousand-line
failed diff hunk.
I'm not sure if Hitoshi is in a position to submit the pg_proc changes
as two separate diffs --- one to add the new column and a separate one
to add in the new functions --- but it'd be a lot easier to deal with
merge issues if he could.
(Now I'll sit back and wait for some fanboy to claim that
$his_favorite_scm could solve this automatically ...)
regards, tom lane