[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: ports/1902: ports: make on qt2 fails to apply patches
The following reply was made to PR ports/1902; it has been noted by GNATS.
From: "J.C. Roberts" <unknown@abac.com>
To: Peter Valchev <pvalchev@openbsd.org>
Cc: gnats@openbsd.org, bugs@openbsd.org
Subject: Re: ports/1902: ports: make on qt2 fails to apply patches
Date: Mon, 25 Jun 2001 20:30:34 -0700
On Mon, 25 Jun 2001 20:20:02 -0600 (MDT), you wrote:
>The following reply was made to PR ports/1902; it has been noted by GNATS.
>
>From: Peter Valchev <pvalchev@openbsd.org>
>To: "J.C. Roberts" <unknown@abac.com>
>Cc: gnats@openbsd.org
>Subject: Re: ports/1902: ports: make on qt2 fails to apply patches
>Date: Mon, 25 Jun 2001 20:12:41 -0600
>
> "J.C. Roberts" <unknown@abac.com> wrote:
>
> ..
> > ===> Patching for qt2-3.0
> > 1 out of 1 hunks failed--saving rejects to
> > examples/box/Makefile.in.rej
> > ***> patch-examples_box_Makefile_in did not apply cleanly
> > 1 out of 1 hunks failed--saving rejects to
> > examples/gear/Makefile.in.rej
> > ***> patch-examples_gear_Makefile_in did not apply cleanly
> > 1 out of 1 hunks failed--saving rejects to
> > examples/glpixmap/Makefile.in.rej
> > ***> patch-examples_glpixmap_Makefile_in did not apply cleanly
> > 1 out of 1 hunks failed--saving rejects to
> > examples/overlay/Makefile.in.rej
> > ***> patch-examples_overlay_Makefile_in did not apply cleanly
> > 1 out of 1 hunks failed--saving rejects to
> > examples/sharedbox/Makefile.in.rej
> > ***> patch-examples_sharedbix_Makefile_in did not apply cleanly
> > 1 out of 1 hunks failed--saving rejects to
> > tools/designer/designer/Makefile.in.rej
> > ***> patch-tools_designer_designer_Makefile.in did not apply cleanly
> > 1 out of 1 hunks failed--saving rejects to
> > tools/designer/uic/Makefile.in.rej
> > ***> patch-tools_designer_uic_Makefile.in did not apply correctly
> > *** Error code 1
>
> In fact, these do not exist at all. So just update your tree with the
> '-d' flag passed to cvs to go in all directories (specifically patches/)
> and it will work.
>
<second message>
> I cannot reproduce this, the patches are fine. Are you sure that you
> are up to date? How did you update your ports tree, have you missed
> some important flags? `cvs -q up -PAd` in that dir should clear the
> things -- can you try that and tell us what happened? Something with
> your tree is not right, possibly the way you updated it.
>
> Thanks,
> -Peter
My ports cvs tree was updated to current on 2001.06.23 (last Saturday)
just about as you stated but I used the CVS_ROOT. -It was either the
redwood city or moffit field (NASA) repository in California but I
can't remember which.
# cd /usr/ports
# cvs -d $CVS_ROOT up -PAd
I've been digging into this since I posted the bug. In a nutshell, it
seems these mystery patches I have in my tree all fail for basically
the same reason, namely, the source already includes the changes made
by the patches although the changes are not exactly the same exact
text (there's a different order on listed vars).
/usr/ports/x11/qt2/patches/
patch-examples_box_Makefile_in
patch-examples_gear_Makefile_in
patch-examples_glpixmap_Makefile_in
patch-examples_overlay_Makefile_in
patch-examples_sharedbox_Makefile_in
patch-tools_designer_designer_Makefile.in
patch-tools_designer_uic_Makefile.in
********************************************************************************
Error #1: /usr/ports/x11/qt2/examples/box/
patch-examples_box_Makefile_in
----------------------------------------------------------------------
From Makefile.in.orig
* CXXFLAGS= $(SYSCONF_CXXFLAGS_OPENGL) $(SYSCONF_CXXFLAGS_QT)
$(SYSCONF_CXXFLAGS)
From patch-examples_box_Makefile_in
- CXXFLAGS= $(SYSCONF_CXXFLAGS_QT) $(SYSCONF_CXXFLAGS)
+ CXXFLAGS= $(SYSCONF_CXXFLAGS_QT) $(SYSCONF_CXXFLAGS)
$(SYSCONF_CXXFLAGS_OPENGL)
----------------------------------------------------------------------
From Makefile.in.orig
* LFLAGS = $(SYSCONF_LFLAGS_OPENGL) $(SYSCONF_LFLAGS_QT)
$(SYSCONF_RPATH_QT) $(SYSCONF_LFLAGS) $(QT_LFLAGS_MT)
From patch-examples_box_Makefile_in
- LFLAGS = $(SYSCONF_LFLAGS_QT) $(SYSCONF_RPATH_QT)
$(SYSCONF_LFLAGS) $(QT_LFLAGS_MT)
+ LFLAGS = $(SYSCONF_LFLAGS_QT) $(SYSCONF_RPATH_QT)
$(SYSCONF_LFLAGS) $(QT_LFLAGS_MT) $(SYSCONF_LFLAGS_OPENGL)
----------------------------------------------------------------------
From Makefile.in.orig
* LIBS = $(SUBLIBS) $(SYSCONF_LIBS_OPENGL)
$(SYSCONF_LIBS_QTAPP) $(SYSCONF_LIBS_QT) $(SYSCONF_LIBS)
From patch-examples_box_Makefile_in
- LIBS = $(SUBLIBS) $(SYSCONF_LIBS_QTAPP) $(SYSCONF_LIBS_QT)
$(SYSCONF_LIBS)
+ LIBS = $(SUBLIBS) $(SYSCONF_LIBS_QTAPP) $(SYSCONF_LIBS_QT)
$(SYSCONF_LIBS) $(SYSCONF_LIBS_OPENGL)
********************************************************************************
You see what I mean... the $(*_OPENGL) vars are already in the
original Makefile.in files although the vars are in a different order
than what the patch would have done.
As you said, the patch files in my tree don't exist... weird. I should
have checked before posting the bug
http://www.openbsd.org/cgi-bin/cvsweb/ports/x11/qt2/patches/
It's not the most conventional way to do things but after updating my
source trees I back them up to tar files _before_ I do anything in
them. When I want to do a cvs up date, I rm the current /usr/ports and
untar the backup and then update the tree. -I'm a bit paranoid about
"cruft" problems so how this happened is a mystery. ;)
My standard routine for updating ports is:
# cd /usr
# rm -rf ports/
# tar -xzvf ports.tar.gz
# setenv CVS_RSH /usr/bin/ssh
# setenv CVS_ROOT anoncvs@anoncvs5.usa.openbsd.org:/cvs
# cd ports
# cvs -d $CVS_ROOT up -PAd
# cd ..
# rm ports.tar.gz
# tar -czvf ports.tar.gz ports/
Is there any chance the "-d $CVS_ROOT" option messes up the -PAd ?
I'll pull source again from Moffit Field (NASA), see how it goes and
get back to you when it's done. This seems to be either on my end or
in one of the California mirrors I normally use and I'm sorry for
wasting your time with it. I was actually just trying to help out in
the limited ways I can but you know what they say about good
intentions...
Best Regards,
JCR