Thomas Lockhart <lockhart(at)alumni(dot)caltech(dot)edu> writes:
>> Uh ... what's your point?
> Any change to gram.y regenerates the local copy of parse.h and affects
> other files *in that local directory* (as well as elsewhere). The
> makefile *in that local directory* should be able to make the other
> files *in that same directory* at the same time.
Oh, right, the files in that directory are going to include parse.h
from the include dir now, instead of ".", aren't they? I see your
Probably the rule that installs parse.h into the include tree ought to
be pushed down from backend/Makefile to backend/parser/Makefile (but
backend/Makefile still needs to invoke it during its prebuildheaders
phase). Maybe likewise for fmgroids.h into backend/utils.
Peter, any thoughts here?
regards, tom lane
In response to
pgsql-hackers by date
|Next:||From: Tom Lane||Date: 2000-06-27 08:51:15|
|Subject: Re: SQL99 functions |
|Previous:||From: Giles Lean||Date: 2000-06-27 08:27:26|
|Subject: Re: Is *that* why debugging backend startup is so hard!? |