Обсуждение: Re: [QUESTIONS] Snapshot has not compiled in a couple of days

>         Is anyone aware that the last 2 or 3 days of snapshots will not
>         compile?
>
>         The problem is in the ecpq stuff. There is a duplicate def of S_SIGNED.
>         Cab we fix this, so I can test the compile on HPUX?

My cvs tree from 980223 03:00 UTC (about 24 hours ago) compiled to completion, but
I do see a duplicate definition of S_SIGNED in preproc.y. Remove that symbol from
line 234 of that file and try again.

                                                - Tom

> >         The problem is in the ecpq stuff. There is a duplicate def of S_SIGNED.
> >         Cab we fix this, so I can test the compile on HPUX?
>
> My cvs tree from 980223 03:00 UTC (about 24 hours ago) compiled to completion, but
> I do see a duplicate definition of S_SIGNED in preproc.y. Remove that symbol from
> line 234 of that file and try again.

/usr/bin/bison -y -d preproc.y
conflicts:  5 shift/reduce
gcc -I../include -O2 -g -Wall -DMAJOR_VERSION=1 -DMINOR_VERSION=0 -DPATCHLEVEL=0   -c
y.tab.c -o y.tab.o
...
preproc.y:289: warning: passing arg 1 of `free' discards `const' from pointer target
type
gcc -g -O2 -Wall -o ecpg y.tab.o pgc.o type.o ecpg.o ../lib/typename.o

I removed the duplicate S_SIGNED from my source code and noticed these remaining
messages. Can we get rid of the shift/reduce conflicts? I haven't looked at the code,
but there is a debugging flag on yacc/bison ("-v") which will write a log file
identifying the conflicting statements. I have some experience debugging this on the
main parser, but need to work on docs so shouldn't do this at the moment :(

                                            - Tom

On Wed, 25 Feb 1998, Thomas G. Lockhart wrote:

> > >         The problem is in the ecpq stuff. There is a duplicate def of S_SIGNED.
> > >         Cab we fix this, so I can test the compile on HPUX?
> >
> > My cvs tree from 980223 03:00 UTC (about 24 hours ago) compiled to completion, but
> > I do see a duplicate definition of S_SIGNED in preproc.y. Remove that symbol from
> > line 234 of that file and try again.
>
> /usr/bin/bison -y -d preproc.y
> conflicts:  5 shift/reduce
> gcc -I../include -O2 -g -Wall -DMAJOR_VERSION=1 -DMINOR_VERSION=0 -DPATCHLEVEL=0   -c
> y.tab.c -o y.tab.o
> ...
> preproc.y:289: warning: passing arg 1 of `free' discards `const' from pointer target
> type
> gcc -g -O2 -Wall -o ecpg y.tab.o pgc.o type.o ecpg.o ../lib/typename.o
>
> I removed the duplicate S_SIGNED from my source code and noticed these
> remaining messages. Can we get rid of the shift/reduce conflicts? I
> haven't looked at the code, but there is a debugging flag on yacc/bison
> ("-v") which will write a log file identifying the conflicting
> statements. I have some experience debugging this on the main parser,
> but need to work on docs so shouldn't do this at the moment :(

    If I can figure out bison, I'll try and work these out...one way
to learn, I guess :)

Marc G. Fournier
Systems Administrator @ hub.org
primary: scrappy@hub.org           secondary: scrappy@{freebsd|postgresql}.org

> > remaining messages. Can we get rid of the shift/reduce conflicts? I
> > haven't looked at the code, but there is a debugging flag on yacc/bison
> > ("-v") which will write a log file identifying the conflicting
> > statements. I have some experience debugging this on the main parser,
> > but need to work on docs so shouldn't do this at the moment :(
>
>     If I can figure out bison, I'll try and work these out...one way
> to learn, I guess :)

This ecps is hot, but it seems it is still 'in process' as we approach
release date.  Can we do what we did with pgaccess, and get a web site
that has the most recent version, and just ship a current version with
6.3, and people can go to the web site to get the newest copy?  We put a
nice README in the pgaccess directory pointing people to a web site
having the most recent version.  Seemed to work well.

--
Bruce Momjian                          |  830 Blythe Avenue
maillist@candle.pha.pa.us              |  Drexel Hill, Pennsylvania 19026
  +  If your life is a hard drive,     |  (610) 353-9879(w)
  +  Christ can be your backup.        |  (610) 853-3000(h)

On Tue, 24 Feb 1998, Bruce Momjian wrote:

> > > remaining messages. Can we get rid of the shift/reduce conflicts? I
> > > haven't looked at the code, but there is a debugging flag on yacc/bison
> > > ("-v") which will write a log file identifying the conflicting
> > > statements. I have some experience debugging this on the main parser,
> > > but need to work on docs so shouldn't do this at the moment :(
> >
> >     If I can figure out bison, I'll try and work these out...one way
> > to learn, I guess :)
>
> This ecps is hot, but it seems it is still 'in process' as we approach
> release date.  Can we do what we did with pgaccess, and get a web site
> that has the most recent version, and just ship a current version with
> 6.3, and people can go to the web site to get the newest copy?  We put a
> nice README in the pgaccess directory pointing people to a web site
> having the most recent version.  Seemed to work well.

    ecpg isn't a seperate package, so this isn't really an alternative
in this case :(

    The above problem is more an annoyance then anything, and, quite
frankly, I don't understand enough about yacc/bison to work out where.
Well, I understand more now then when I started, but still have a long way
to go.  Next Nutshell book, I fear :)

    I'm going to do a run through of the code again tomorrow at work,
on the two Solaris boxes, and check for compiler problems...

    Jan, how goes the pg_shadow stuff?  I believe we still have
breakage in initdb, but Bruce is waiting for that patch before
investigating further?

Marc G. Fournier
Systems Administrator @ hub.org
primary: scrappy@hub.org           secondary: scrappy@{freebsd|postgresql}.org

Thomas G. Lockhart writes:
> /usr/bin/bison -y -d preproc.y
> conflicts:  5 shift/reduce
> gcc -I../include -O2 -g -Wall -DMAJOR_VERSION=1 -DMINOR_VERSION=0 -DPATCHLEVEL=0   -c
> y.tab.c -o y.tab.o
> ...
> preproc.y:289: warning: passing arg 1 of `free' discards `const' from pointer target
> type
> gcc -g -O2 -Wall -o ecpg y.tab.o pgc.o type.o ecpg.o ../lib/typename.o
>
> I removed the duplicate S_SIGNED from my source code and noticed these remaining
> messages. Can we get rid of the shift/reduce conflicts? I haven't looked at the code,
> but there is a debugging flag on yacc/bison ("-v") which will write a log file
> identifying the conflicting statements. I have some experience debugging this on the
> main parser, but need to work on docs so shouldn't do this at the moment :(

I see what I can do. Unfortunately I'm not really a bison expert either.

Michael

--
Dr. Michael Meskes, Project-Manager    | topsystem Systemhaus GmbH
meskes@topsystem.de                    | Europark A2, Adenauerstr. 20
meskes@debian.org                      | 52146 Wuerselen
Go SF49ers! Go Rhein Fire!             | Tel: (+49) 2405/4670-44
Use Debian GNU/Linux!                  | Fax: (+49) 2405/4670-10