Re: [RFC/testsuite] Update "info float" output for i386

Re: [RFC/testsuite] Update "info float" output for i386

[Michael Elizabeth Chastain]

Side question: what is gcc 2.95.4?  I've heard that version number
from time to time, but I don't see it at ftp.gnu.org.  Just curious.

Michael C

Re: [RFC/testsuite] Update "info float" output for i386

[Mark Kettenis]

AFAIK this is the "end" of the gcc-2_95-branch with some
FreeBSD-specific patches.  There has never been an official 2.95.4
release even though things have been checked in on the branch after
2.95.3 has been released.

Mark

Re: [RFC/testsuite] Update "info float" output for i386

[Michael Elizabeth Chastain]

Proofread, but not tested.  Recommended for approval provided that you
have tested this at least once.

This is cool, this "info float" line shows up in my test results as a
regression from gdb 5.3 to gdb HEAD.  I looked it and came up with the
same (obvious) analysis that you did.  So now that's one fewer boring
paragraph I will have to write when it comes time for "gdb 5.3 versus
gdb HEAD" analysis.

> ... and I'll be right back where I started at dozens of failures.
> But I'll have a tremendous sense of accomplishment!

My metric is: "number of tests which have at least one non-PASS line
in at least one tested configuration."

  5.2 was better than 5.1.1
  5.2.1 was the same as 5.2
  5.3 was better than 5.2.1
  gdb HEAD is better than 5.3

The # of tests increases in every release so this is particularly good
progress.

Michael C

===

2003-01-13  Daniel Jacobowitz  <drow@mvista.com>

	* gdb.base/default.exp (info float): Update expected output for
	i386.

[RFC/testsuite] Update "info float" output for i386

[Daniel Jacobowitz]

I'm still in testsuite maintenance mode, for a little while longer.  I'm on
the home stretch for GCC 2.95.3 + stabs; as soon as I finish, I'll do my
monthly apt-get dist-upgrade, the default compiler on my Debian machine will
change to GCC 3.2 + DWARF-2, and I'll be right back where I started at
dozens of failures.  But I'll have a tremendous sense of accomplishment!

Some time ago Mark Kettenis updated i387-tdep.c to use frame_register_read
for "info float".  This changed the result of "info float" with no running
program; I think for the better, and so did he, if I recall rightly.  Can
anyone think of a reason not to update the testsuite correspondingly, as
with the attached patch?

For reference, the old output was a zeroed register dump; the new is "The
program has no registers now."  Which is quite entirely accurate.

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

2003-01-13  Daniel Jacobowitz  <drow@mvista.com>

	* gdb.base/default.exp (info float): Update expected output for
	i386.

Index: gdb.base/default.exp
===================================================================
RCS file: /cvs/src/src/gdb/testsuite/gdb.base/default.exp,v
retrieving revision 1.14
diff -u -p -r1.14 default.exp
--- gdb.base/default.exp	18 Oct 2002 18:54:55 -0000	1.14
+++ gdb.base/default.exp	14 Jan 2003 04:10:46 -0000
@@ -1,5 +1,5 @@
 #   Copyright 1988, 1990, 1991, 1992, 1994, 1995, 1996, 1997, 1998, 1999,
-#   2000, 2001, 2002
+#   2000, 2001, 2002, 2003
 #   Free Software Foundation, Inc.
 
 # This program is free software; you can redistribute it and/or modify
@@ -325,7 +325,7 @@ if { [istarget "arm*-*-*"] || \
 	[istarget "strongarm*-*-*"] } then {
     gdb_test "info float" "Software FPU type.*mask:.*flags:.*" "info float"
 } elseif [istarget "i\[3456\]86-*-*"] then {
-    gdb_test "info float" "R7:.*Status Word:.*Opcode:.*" "info float"
+    gdb_test "info float" "The program has no registers now." "info float"
 } else {
     gdb_test "info float" "No floating.point info available for this processor." "info float"
 }

Re: [RFC/testsuite] Update "info float" output for i386

[Andrew Cagney]

> Some time ago Mark Kettenis updated i387-tdep.c to use frame_register_read
> for "info float".  This changed the result of "info float" with no running
> program; I think for the better, and so did he, if I recall rightly.  Can
> anyone think of a reason not to update the testsuite correspondingly, as
> with the attached patch?

That message comes from print_float_info, so while possibly mark's 
changes, it wouldn't be frame_register_read.  Check with fernando, but i 
think that all targets should always output that message.

Andrew

Re: [RFC/testsuite] Update "info float" output for i386

[Daniel Jacobowitz]

On Tue, Jan 14, 2003 at 12:52:07AM -0500, Andrew Cagney wrote:
> 
> >Some time ago Mark Kettenis updated i387-tdep.c to use frame_register_read
> >for "info float".  This changed the result of "info float" with no running
> >program; I think for the better, and so did he, if I recall rightly.  Can
> >anyone think of a reason not to update the testsuite correspondingly, as
> >with the attached patch?
> 
> That message comes from print_float_info, so while possibly mark's 
> changes, it wouldn't be frame_register_read.  Check with fernando, but i 
> think that all targets should always output that message.

Thanks very much.  I completely missed that line at the beginning of
the function.  Since I don't see anything else overriding
print_float_info (i.e. changing the cfunc for the info command), you're
probably right.

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

Re: [RFC/testsuite] Update "info float" output for i386

[Andrew Cagney]

> 
> Thanks very much.  I completely missed that line at the beginning of
> the function.  Since I don't see anything else overriding
> print_float_info (i.e. changing the cfunc for the info command), you're
> probably right.

No code should be overriding a command's function.  So if that is 
happening, its a bug.

Andrew

Re: [RFC/testsuite] Update "info float" output for i386

[Mark Kettenis]

Daniel Jacobowitz <drow@mvista.com> writes:

> I'm still in testsuite maintenance mode, for a little while longer.  I'm on
> the home stretch for GCC 2.95.3 + stabs; as soon as I finish, I'll do my
> monthly apt-get dist-upgrade, the default compiler on my Debian machine will
> change to GCC 3.2 + DWARF-2, and I'll be right back where I started at
> dozens of failures.  But I'll have a tremendous sense of accomplishment!

For what it's worth, I will continue to use my i386-unknown-freebsd4.7
system, which has GCC 2.95.4 with stabs as my primary development
machine for the foreseeable future.  This GCC 2.95.4 is the default
compiler on FreeBSD 4.7-RELEASE which, to be honest, I've patched to
fix a debug info generation problem caused by the FreeBSD-specific
modifications.

> Some time ago Mark Kettenis updated i387-tdep.c to use frame_register_read
> for "info float".  This changed the result of "info float" with no running
> program; I think for the better, and so did he, if I recall rightly.  Can
> anyone think of a reason not to update the testsuite correspondingly, as
> with the attached patch?

At the time I checked in that patch, I did submit a patch to the testsuite:

   http://sources.redhat.com/ml/gdb-patches/2002-11/msg00278.html

Unfortunately it seems to have fallen between the cracks, and I
totally forgot about it; it has been happily sitting in my tree ever
since :-(.  I sort-of prefer my patch over yours.  Should I check it
in without Fernando's approval?

Mark

Re: [RFC/testsuite] Update "info float" output for i386

[Daniel Jacobowitz]

On Wed, Jan 15, 2003 at 01:48:27AM +0100, Mark Kettenis wrote:
> Daniel Jacobowitz <drow@mvista.com> writes:
> 
> > I'm still in testsuite maintenance mode, for a little while longer.  I'm on
> > the home stretch for GCC 2.95.3 + stabs; as soon as I finish, I'll do my
> > monthly apt-get dist-upgrade, the default compiler on my Debian machine will
> > change to GCC 3.2 + DWARF-2, and I'll be right back where I started at
> > dozens of failures.  But I'll have a tremendous sense of accomplishment!
> 
> For what it's worth, I will continue to use my i386-unknown-freebsd4.7
> system, which has GCC 2.95.4 with stabs as my primary development
> machine for the foreseeable future.  This GCC 2.95.4 is the default
> compiler on FreeBSD 4.7-RELEASE which, to be honest, I've patched to
> fix a debug info generation problem caused by the FreeBSD-specific
> modifications.
> 
> > Some time ago Mark Kettenis updated i387-tdep.c to use frame_register_read
> > for "info float".  This changed the result of "info float" with no running
> > program; I think for the better, and so did he, if I recall rightly.  Can
> > anyone think of a reason not to update the testsuite correspondingly, as
> > with the attached patch?
> 
> At the time I checked in that patch, I did submit a patch to the testsuite:
> 
>    http://sources.redhat.com/ml/gdb-patches/2002-11/msg00278.html
> 
> Unfortunately it seems to have fallen between the cracks, and I
> totally forgot about it; it has been happily sitting in my tree ever
> since :-(.  I sort-of prefer my patch over yours.  Should I check it
> in without Fernando's approval?

Aha!  I couldn't find it when I went looking.  Please ping Fernando
about this, now that he's back.

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

Re: [RFC/testsuite] Update "info float" output for i386

[Fernando Nasser]

Yes, I missed Mark's patch.  Sorry.

If Daniel does not object, I prefer Mark's patch because it preserves the nature 
of default.exp (tests with no inferior) and in doing so it can test for the 
specific message that we are expecting in that condition.

I was also surprised how few tests we have for floating point handling.  So 
naming the new file float.exp is a good idea -- it may encourage more tests for 
that.

We could create a catch all file for info, but I guess it is tested in all 
places where the object being "infoe'ed" is tested :-)

Thanks Mark.

Regards,
Fernando


Mark Kettenis wrote:> Daniel Jacobowitz <drow@mvista.com> writes:
> 
> 
>>I'm still in testsuite maintenance mode, for a little while longer.  I'm on
>>the home stretch for GCC 2.95.3 + stabs; as soon as I finish, I'll do my
>>monthly apt-get dist-upgrade, the default compiler on my Debian machine will
>>change to GCC 3.2 + DWARF-2, and I'll be right back where I started at
>>dozens of failures.  But I'll have a tremendous sense of accomplishment!
> 
> 
> For what it's worth, I will continue to use my i386-unknown-freebsd4.7
> system, which has GCC 2.95.4 with stabs as my primary development
> machine for the foreseeable future.  This GCC 2.95.4 is the default
> compiler on FreeBSD 4.7-RELEASE which, to be honest, I've patched to
> fix a debug info generation problem caused by the FreeBSD-specific
> modifications.
> 
> 
>>Some time ago Mark Kettenis updated i387-tdep.c to use frame_register_read
>>for "info float".  This changed the result of "info float" with no running
>>program; I think for the better, and so did he, if I recall rightly.  Can
>>anyone think of a reason not to update the testsuite correspondingly, as
>>with the attached patch?
> 
> 
> At the time I checked in that patch, I did submit a patch to the testsuite:
> 
>    http://sources.redhat.com/ml/gdb-patches/2002-11/msg00278.html
> 
> Unfortunately it seems to have fallen between the cracks, and I
> totally forgot about it; it has been happily sitting in my tree ever
> since :-(.  I sort-of prefer my patch over yours.  Should I check it
> in without Fernando's approval?
> 
> Mark
> 



-- 
Fernando Nasser
Red Hat - Toronto                       E-Mail:  fnasser@redhat.com
2323 Yonge Street, Suite #300
Toronto, Ontario   M4P 2C9

Re: [RFC/testsuite] Update "info float" output for i386

[Mark Kettenis]

   Date: Wed, 15 Jan 2003 09:48:41 -0500
   From: Fernando Nasser <fnasser@redhat.com>

   If Daniel does not object, I prefer Mark's patch because it
   preserves the nature of default.exp (tests with no inferior) and in
   doing so it can test for the specific message that we are expecting
   in that condition.

Here's what I actually checked in:

Index: ChangeLog
from  Mark Kettenis  <kettenis@gnu.org>

	* gdb.base/default.exp: Adapt "info float" test for recent changes
	to that command.  Add test for "info vector".
	* gdb.base/float.exp: New file.  Add test for "info float" that
	resembles the old test in gdb.base/default.exp.

Index: gdb.base/default.exp
===================================================================
RCS file: /cvs/src/src/gdb/testsuite/gdb.base/default.exp,v
retrieving revision 1.14
diff -u -p -r1.14 default.exp
--- gdb.base/default.exp 18 Oct 2002 18:54:55 -0000 1.14
+++ gdb.base/default.exp 15 Jan 2003 17:43:07 -0000
@@ -1,5 +1,5 @@
 #   Copyright 1988, 1990, 1991, 1992, 1994, 1995, 1996, 1997, 1998, 1999,
-#   2000, 2001, 2002
+#   2000, 2001, 2002, 2003
 #   Free Software Foundation, Inc.
 
 # This program is free software; you can redistribute it and/or modify
@@ -320,15 +320,7 @@ gdb_test "info frame" "No stack.|No sele
 #test info files
 gdb_test "info files" "" "info files"
 #test info float
-if { [istarget "arm*-*-*"] || \
-	[istarget "xscale*-*-*"] || \
-	[istarget "strongarm*-*-*"] } then {
-    gdb_test "info float" "Software FPU type.*mask:.*flags:.*" "info float"
-} elseif [istarget "i\[3456\]86-*-*"] then {
-    gdb_test "info float" "R7:.*Status Word:.*Opcode:.*" "info float"
-} else {
-    gdb_test "info float" "No floating.point info available for this processor." "info float"
-}
+gdb_test "info float" "The program has no registers now." "info float"
 #test info functions
 gdb_test "info functions" "All defined functions:" "info functions"
 #test info locals
@@ -359,6 +351,8 @@ gdb_test "info terminal" "No saved termi
 gdb_test "info types" "All defined types:" "info types"
 #test info variables
 gdb_test "info variables" "All defined variables:" "info variables"
+#test info vector
+gdb_test "info vector" "The program has no registers now." "info vector"
 #test info warranty
 gdb_test "info warranty" "NO WARRANTY(\[^\r\n\]*\[\r\n\])+  *11.  *BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY(\[^\r\n\]*\[\r\n\])+FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW.  EXCEPT WHEN(\[^\r\n\]*\[\r\n\])+OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES(\[^\r\n\]*\[\r\n\])+PROVIDE THE PROGRAM \"AS IS\" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED(\[^\r\n\]*\[\r\n\])+OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF(\[^\r\n\]*\[\r\n\])+MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.  THE ENTIRE RISK AS(\[^\r\n\]*\[\r\n\])+TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU.  SHOULD THE(\[^\r\n\]*\[\r\n\])+PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING,(\[^\r\n\]*\[\r\n\])+REPAIR OR CORRECTION.(\[^\r\n\]*\[\r\n\])+  *12.  *IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING(\[^\r\n\]*\[\r\n\])+WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PAR!
 TY WHO MAY MODIFY AND/OR(\[^\r\n\]*\[\r\n\])+REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES,(\[^\r\n\]*\[\r\n\])+INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING(\[^\r\n\]*\[\r\n\])+OUT OF THE USE OR INABILITY TO USE THE PROGRAM .INCLUDING BUT NOT LIMITED(\[^\r\n\]*\[\r\n\])+TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY(\[^\r\n\]*\[\r\n\])+YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER(\[^\r\n\]*\[\r\n\])+PROGRAMS., EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE(\[^\r\n\]*\[\r\n\])+POSSIBILITY OF SUCH DAMAGES.*" "info warranty"
 #test info watchpoints
Index: gdb.base/float.exp
===================================================================
RCS file: gdb.base/float.exp
diff -N gdb.base/float.exp
--- /dev/null	1 Jan 1970 00:00:00 -0000
+++ gdb.base/float.exp 15 Jan 2003 17:43:07 -0000
@@ -0,0 +1,62 @@
+# Copyright 2003 Free Software Foundation, Inc.
+
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  
+
+# Please email any bugs, comments, and/or additions to this file to:
+# bug-gdb@gnu.org
+
+# This file is part of the gdb testsuite.
+
+if $tracelevel {
+    strace $tracelevel
+}
+
+#
+# Test floating-point related functionality.
+#
+
+set prms_id 0
+set bug_id 0
+
+set testfile "run"
+set srcfile ${testfile}.c
+set binfile ${objdir}/${subdir}/${testfile}
+if { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable {debug}] != "" } {
+    gdb_suppress_entire_file "Testcase compile failed, so all tests in this file will automatically fail."
+}
+
+gdb_exit
+gdb_start
+gdb_reinitialize_dir $srcdir/$subdir
+gdb_load ${binfile}
+
+# Set it up at a breakpoint so we have its registers.
+
+if ![runto_main] then {
+    perror "couldn't run to breakpoint"
+    continue
+}
+
+# Test "info float".
+
+if { [istarget "arm*-*-*"] || \
+	[istarget "xscale*-*-*"] || \
+	[istarget "strongarm*-*-*"] } then {
+    gdb_test "info float" "Software FPU type.*mask:.*flags:.*" "info float"
+} elseif [istarget "i?86-*-*"] then {
+    gdb_test "info float" "R7:.*Status Word:.*Opcode:.*" "info float"
+} else {
+    gdb_test "info float" "No floating.point info available for this processor." "info float"
+}