[5.3] What's the hold up?

[5.3] What's the hold up?

[Andrew Cagney]

If you search the bug data base for high GDB PR's you'll see:

PR Category State Priority Responsible Synopsis
378 gdb suspended high unassigned ``GNU/Linux'' ``Linux kernel''
394 gdb suspended high unassigned GDB 5.1.1 contains intl/ droppings
527 gdb suspended high unassigned GDB 5.2 incompatibilities with GNU 
textutils 2.0.21 in POSIX mode
676 gdb analyzed high unassigned Alpha OSF1, gcc, internal GDB error in 
mdebugread
708 gdb open high unassigned Can't build 5.3 branch (and probably trunk).
725 gdb open high unassigned Crash using debug target and regcaches (in 
5.3 branch?)
763 gdb open high unassigned gdb 5.2 removes the conditional breakpoints
840 gdb suspended high unassigned Expect on GDB 5.3 branch doesn't build 
on IRIX 6.5

It is these (actually the non-suspended ones) that I'm trying to get 
flushed.

Of most concern are 708 and 676.  They both indicate something failing the:

	./configure
	make
	./gdb/gdb ./gdb/gdb
	(gdb) b main
	(gdb) run

release criteria :-(

Andrew

Re: [5.3] What's the hold up?

[Joel Brobecker]

> 676 gdb analyzed high unassigned Alpha OSF1, gcc, internal GDB error in 
> mdebugread
> 708 gdb open high unassigned Can't build 5.3 branch (and probably trunk).
[...]
> Of most concern are 708 and 676.  They both indicate something failing the:
> 
> 	./configure
> 	make
> 	./gdb/gdb ./gdb/gdb
> 	(gdb) b main
> 	(gdb) run

I don't think that 676 is any worse than before, so I would not hold up
5.3 for this PR. I just tried your release criteria on Tru64 4.0f and
5.1a, and the test passed...

My 2 cents...
-- 
Joel

Re: [5.3] What's the hold up?

[Daniel Jacobowitz]

On Wed, Nov 20, 2002 at 12:50:46PM -0500, Andrew Cagney wrote:
> If you search the bug data base for high GDB PR's you'll see:
> 
> PR Category State Priority Responsible Synopsis
> 378 gdb suspended high unassigned ``GNU/Linux'' ``Linux kernel''
> 394 gdb suspended high unassigned GDB 5.1.1 contains intl/ droppings
> 527 gdb suspended high unassigned GDB 5.2 incompatibilities with GNU 
> textutils 2.0.21 in POSIX mode
> 676 gdb analyzed high unassigned Alpha OSF1, gcc, internal GDB error in 
> mdebugread
> 708 gdb open high unassigned Can't build 5.3 branch (and probably trunk).
> 725 gdb open high unassigned Crash using debug target and regcaches (in 
> 5.3 branch?)
> 763 gdb open high unassigned gdb 5.2 removes the conditional breakpoints
> 840 gdb suspended high unassigned Expect on GDB 5.3 branch doesn't build 
> on IRIX 6.5
> 
> It is these (actually the non-suspended ones) that I'm trying to get 
> flushed.
> 
> Of most concern are 708 and 676.  They both indicate something failing the:
> 
> 	./configure
> 	make
> 	./gdb/gdb ./gdb/gdb
> 	(gdb) b main
> 	(gdb) run
> 
> release criteria :-(

I hit gdb/708 on HP/UX... it's really not that complicated to fix, I
don't think...

Of course trying to reproduce it I can't even build the snapshot; it
tries to make distclean in itcl.

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

Re: [5.3] What's the hold up?

[Daniel Jacobowitz]

On Wed, Nov 20, 2002 at 01:52:27PM -0500, Daniel Jacobowitz wrote:
> On Wed, Nov 20, 2002 at 12:50:46PM -0500, Andrew Cagney wrote:
> > If you search the bug data base for high GDB PR's you'll see:
> > 
> > PR Category State Priority Responsible Synopsis
> > 378 gdb suspended high unassigned ``GNU/Linux'' ``Linux kernel''
> > 394 gdb suspended high unassigned GDB 5.1.1 contains intl/ droppings
> > 527 gdb suspended high unassigned GDB 5.2 incompatibilities with GNU 
> > textutils 2.0.21 in POSIX mode
> > 676 gdb analyzed high unassigned Alpha OSF1, gcc, internal GDB error in 
> > mdebugread
> > 708 gdb open high unassigned Can't build 5.3 branch (and probably trunk).
> > 725 gdb open high unassigned Crash using debug target and regcaches (in 
> > 5.3 branch?)
> > 763 gdb open high unassigned gdb 5.2 removes the conditional breakpoints
> > 840 gdb suspended high unassigned Expect on GDB 5.3 branch doesn't build 
> > on IRIX 6.5
> > 
> > It is these (actually the non-suspended ones) that I'm trying to get 
> > flushed.
> > 
> > Of most concern are 708 and 676.  They both indicate something failing the:
> > 
> > 	./configure
> > 	make
> > 	./gdb/gdb ./gdb/gdb
> > 	(gdb) b main
> > 	(gdb) run
> > 
> > release criteria :-(
> 
> I hit gdb/708 on HP/UX... it's really not that complicated to fix, I
> don't think...
> 
> Of course trying to reproduce it I can't even build the snapshot; it
> tries to make distclean in itcl.

However, my guess at a fix for gdb/708 is that dejagnu/Makefile.am
needs a distclean-local to parallel clean-local, which distcleans in
example/.

The failure in itcl/ appears to be a configure script bug with autoconf
2.5.  Keith fixed this on mainline on Sep. 5th:

        * configure.in (subdirs): New variable. Don't put AC_CONFIG_SUBDIRS
        macro into configure.in twice: it confuses autoconf. Instead use
        the variable to hold the subdirs to be configured.
        * configure: Regenerated.

This should probably go to the branch.

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

Re: [5.3] What's the hold up?

[Daniel Jacobowitz]

On Wed, Nov 20, 2002 at 02:07:05PM -0500, Daniel Jacobowitz wrote:
> On Wed, Nov 20, 2002 at 01:52:27PM -0500, Daniel Jacobowitz wrote:
> > On Wed, Nov 20, 2002 at 12:50:46PM -0500, Andrew Cagney wrote:
> > > If you search the bug data base for high GDB PR's you'll see:
> > > 
> > > PR Category State Priority Responsible Synopsis
> > > 378 gdb suspended high unassigned ``GNU/Linux'' ``Linux kernel''
> > > 394 gdb suspended high unassigned GDB 5.1.1 contains intl/ droppings
> > > 527 gdb suspended high unassigned GDB 5.2 incompatibilities with GNU 
> > > textutils 2.0.21 in POSIX mode
> > > 676 gdb analyzed high unassigned Alpha OSF1, gcc, internal GDB error in 
> > > mdebugread
> > > 708 gdb open high unassigned Can't build 5.3 branch (and probably trunk).
> > > 725 gdb open high unassigned Crash using debug target and regcaches (in 
> > > 5.3 branch?)
> > > 763 gdb open high unassigned gdb 5.2 removes the conditional breakpoints
> > > 840 gdb suspended high unassigned Expect on GDB 5.3 branch doesn't build 
> > > on IRIX 6.5
> > > 
> > > It is these (actually the non-suspended ones) that I'm trying to get 
> > > flushed.
> > > 
> > > Of most concern are 708 and 676.  They both indicate something failing the:
> > > 
> > > 	./configure
> > > 	make
> > > 	./gdb/gdb ./gdb/gdb
> > > 	(gdb) b main
> > > 	(gdb) run
> > > 
> > > release criteria :-(
> > 
> > I hit gdb/708 on HP/UX... it's really not that complicated to fix, I
> > don't think...
> > 
> > Of course trying to reproduce it I can't even build the snapshot; it
> > tries to make distclean in itcl.
> 
> However, my guess at a fix for gdb/708 is that dejagnu/Makefile.am
> needs a distclean-local to parallel clean-local, which distcleans in
> example/.
> 
> The failure in itcl/ appears to be a configure script bug with autoconf
> 2.5.  Keith fixed this on mainline on Sep. 5th:
> 
>         * configure.in (subdirs): New variable. Don't put AC_CONFIG_SUBDIRS
>         macro into configure.in twice: it confuses autoconf. Instead use
>         the variable to hold the subdirs to be configured.
>         * configure: Regenerated.
> 
> This should probably go to the branch.

It did, two months ago.  I haven't cvs up'd that tree in longer than I
had thought :)

Try this patch for the DejaGNU droppings.  And one in gdb/doc/.  That
wasn't so hard, was it?

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

--- src/gdb/doc/Makefile.in	2002-11-20 15:48:51.000000000 -0500
+++ tsrc/gdb/doc/Makefile.in	2002-11-20 15:57:21.000000000 -0500
@@ -423,7 +423,7 @@
 	rm -f gdb-cfg.texi
 
 distclean: clean
-	rm -f Makefile config.status
+	rm -f Makefile config.status config.log
 
 # GDBvn.texi, the dvi files, the info files, and the postscript files, 
 # are all part of the distribution, so it should not be removed by
--- src/dejagnu/Makefile.am	2002-04-21 04:46:47.000000000 -0400
+++ tsrc/dejagnu/Makefile.am	2002-11-20 16:35:17.000000000 -0500
@@ -34,10 +34,13 @@
 # We don't use SUBDIRS, so the excample or test cases don't get built
 # by default for a "make all".
 clean-local:
-	cd doc       ; $(MAKE) clean
 	cd testsuite ; $(MAKE) clean
 	cd example   ; $(MAKE) clean
 
+distclean-local:
+	cd testsuite ; $(MAKE) distclean
+	cd example   ; $(MAKE) distclean
+
 # Set ourselves up to build a RedHat package
 tarball: overview.html overview.ps overview.pdf
 	-rm -fr dejagnu-${VERSION}
--- src/dejagnu/Makefile.in	2002-09-02 07:53:24.000000000 -0400
+++ tsrc/dejagnu/Makefile.in	2002-11-20 16:35:31.000000000 -0500
@@ -1,6 +1,6 @@
-# Makefile.in generated automatically by automake 1.4-p5 from Makefile.am
+# Makefile.in generated automatically by automake 1.4 from Makefile.am
 
-# Copyright (C) 1994, 1995-8, 1999, 2001 Free Software Foundation, Inc.
+# Copyright (C) 1994, 1995-8, 1999 Free Software Foundation, Inc.
 # This Makefile.in is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
 # with or without modifications, as long as this notice is preserved.
@@ -123,7 +123,7 @@
 
 DISTFILES = $(DIST_COMMON) $(SOURCES) $(HEADERS) $(TEXINFOS) $(EXTRA_DIST)
 
-TAR = gtar
+TAR = tar
 GZIP_ENV = --best
 EXPECT = expect
 all: all-redirect
@@ -236,7 +236,7 @@
 	dot_seen=no; \
 	rev=''; list='$(SUBDIRS)'; for subdir in $$list; do \
 	  rev="$$subdir $$rev"; \
-	  test "$$subdir" != "." || dot_seen=yes; \
+	  test "$$subdir" = "." && dot_seen=yes; \
 	done; \
 	test "$$dot_seen" = "no" && rev=". $$rev"; \
 	target=`echo $@ | sed s/-recursive//`; \
@@ -406,7 +406,7 @@
 
 clean: clean-recursive
 
-distclean-am:  distclean-tags distclean-generic clean-am
+distclean-am:  distclean-tags distclean-generic clean-am distclean-local
 
 distclean: distclean-recursive
 	-rm -f config.status
@@ -443,10 +443,13 @@
 # We don't use SUBDIRS, so the excample or test cases don't get built
 # by default for a "make all".
 clean-local:
-	cd doc       ; $(MAKE) clean
 	cd testsuite ; $(MAKE) clean
 	cd example   ; $(MAKE) clean
 
+distclean-local:
+	cd testsuite ; $(MAKE) distclean
+	cd example   ; $(MAKE) distclean
+
 # Set ourselves up to build a RedHat package
 tarball: overview.html overview.ps overview.pdf
 	-rm -fr dejagnu-${VERSION}

Re: [5.3] What's the hold up?

[Andrew Cagney]

> On Wed, Nov 20, 2002 at 02:07:05PM -0500, Daniel Jacobowitz wrote:
> 
>> On Wed, Nov 20, 2002 at 01:52:27PM -0500, Daniel Jacobowitz wrote:
> 
>> > On Wed, Nov 20, 2002 at 12:50:46PM -0500, Andrew Cagney wrote:
> 
>> > > If you search the bug data base for high GDB PR's you'll see:
>> > > 
>> > > PR Category State Priority Responsible Synopsis
>> > > 378 gdb suspended high unassigned ``GNU/Linux'' ``Linux kernel''
>> > > 394 gdb suspended high unassigned GDB 5.1.1 contains intl/ droppings
>> > > 527 gdb suspended high unassigned GDB 5.2 incompatibilities with GNU 
>> > > textutils 2.0.21 in POSIX mode
>> > > 676 gdb analyzed high unassigned Alpha OSF1, gcc, internal GDB error in 
>> > > mdebugread
>> > > 708 gdb open high unassigned Can't build 5.3 branch (and probably trunk).
>> > > 725 gdb open high unassigned Crash using debug target and regcaches (in 
>> > > 5.3 branch?)
>> > > 763 gdb open high unassigned gdb 5.2 removes the conditional breakpoints
>> > > 840 gdb suspended high unassigned Expect on GDB 5.3 branch doesn't build 
>> > > on IRIX 6.5
>> > > 
>> > > It is these (actually the non-suspended ones) that I'm trying to get 
>> > > flushed.
>> > > 
>> > > Of most concern are 708 and 676.  They both indicate something failing the:
>> > > 
>> > > 	./configure
>> > > 	make
>> > > 	./gdb/gdb ./gdb/gdb
>> > > 	(gdb) b main
>> > > 	(gdb) run
>> > > 
>> > > release criteria :-(
> 
>> > 
>> > I hit gdb/708 on HP/UX... it's really not that complicated to fix, I
>> > don't think...
>> > 
>> > Of course trying to reproduce it I can't even build the snapshot; it
>> > tries to make distclean in itcl.
> 
>> 
>> However, my guess at a fix for gdb/708 is that dejagnu/Makefile.am
>> needs a distclean-local to parallel clean-local, which distcleans in
>> example/.
>> 
>> The failure in itcl/ appears to be a configure script bug with autoconf
>> 2.5.  Keith fixed this on mainline on Sep. 5th:
>> 
>>         * configure.in (subdirs): New variable. Don't put AC_CONFIG_SUBDIRS
>>         macro into configure.in twice: it confuses autoconf. Instead use
>>         the variable to hold the subdirs to be configured.
>>         * configure: Regenerated.
>> 
>> This should probably go to the branch.
> 
> 
> It did, two months ago.  I haven't cvs up'd that tree in longer than I
> had thought :)
> 
> Try this patch for the DejaGNU droppings.  And one in gdb/doc/.  That
> wasn't so hard, was it?

I think these can go in.

Andrew

Re: [5.3] What's the hold up?

[Andrew Cagney]

PS: Don't forget to send the deajgnu one up stream.

Andrew