[RFC/commmit,testsuite/Ada] stop using project files when building test programs

Message ID 20151223144303.GC18676@adacore.com
State New, archived
Headers

Commit Message

Joel Brobecker Dec. 23, 2015, 2:43 p.m. UTC
  > Humpf, that's a fair number of reasons showing that assuming 8.6
> may not be reasonable. Bouh...
> 
> > How hard would it be to avoid try/finally?  Wouldn't you just have to
> > use catch instead?
> 
> I don't think it would be very hard. I think catch will work, but
> will be a little more convoluted. I'll give it a try...

Still, for a complete dummy like myself, that took nearly 45 minutes,
just to make sure I understand TCL semantics better, and find a machine
that has TCL 8.5...

I would appreciate a review of the attached patch, because TCL and
I are not regular friends.

gdb/testsuite:

        * lib/ada.exp (target_compile_ada_from_dir): New function.
        (gdb_compile_ada): Reimplement avoiding the use of project files.
        * gdb.ada/gnat_ada.gpr: Delete.
        * gdb.ada/cond_lang.exp: Adjust test to make path before
        filename optional.
        * gdb.ada/small_reg_param.exp: Likewise.

Thanks!
  

Comments

Pedro Alves Dec. 23, 2015, 3:21 p.m. UTC | #1
On 12/23/2015 02:43 PM, Joel Brobecker wrote:
>> > Humpf, that's a fair number of reasons showing that assuming 8.6
>> > may not be reasonable. Bouh...
>> > 
>>> > > How hard would it be to avoid try/finally?  Wouldn't you just have to
>>> > > use catch instead?
>> > 
>> > I don't think it would be very hard. I think catch will work, but
>> > will be a little more convoluted. I'll give it a try...
> Still, for a complete dummy like myself, that took nearly 45 minutes,
> just to make sure I understand TCL semantics better, and find a machine
> that has TCL 8.5...
> 

Thanks for doing this.

> I would appreciate a review of the attached patch, because TCL and
> I are not regular friends.

:-)  Looks generally good to me.  A couple minor comments below.



> --- a/gdb/testsuite/gdb.ada/cond_lang.exp
> +++ b/gdb/testsuite/gdb.ada/cond_lang.exp
> @@ -39,7 +39,7 @@ gdb_test "show lang" \
>  # current language mode is auto, and the breakpoint is inside Ada code.
>  set bp_location [gdb_get_line_number "STOP" ${testdir}/mixed.adb]
>  gdb_test "break mixed.adb:${bp_location} if light = green" \
> -         "Breakpoint \[0-9\]* at .*: file .*/mixed.adb, line \[0-9\]*\\."
> +         "Breakpoint \[0-9\]* at .*: file (.*/)?mixed.adb, line \[0-9\]*\\."

Isn't that the same as just:

 -         "Breakpoint \[0-9\]* at .*: file .*/mixed.adb, line \[0-9\]*\\."
 +         "Breakpoint \[0-9\]* at .*: file .*mixed.adb, line \[0-9\]*\\."

?

>  gdb_test "continue" \
> -    "Breakpoint .*, pck\\.call_me \\(w=(w@entry=)?50\\) at .*/pck.adb:.*" \
> +    "Breakpoint .*, pck\\.call_me \\(w=(w@entry=)?50\\) at (.*)?/pck.adb:.*" \
>      "continue to call_me"

Likewise, I think this is a no-op.  Did you mean to put the / inside the
parens like in the other case?  If so I'd suggest:

 -    "Breakpoint .*, pck\\.call_me \\(w=(w@entry=)?50\\) at .*/pck.adb:.*" \
 +    "Breakpoint .*, pck\\.call_me \\(w=(w@entry=)?50\\) at .*pck.adb:.*" \

>  
>  # And just to make sure, we also verify that the parameter value
> diff --git a/gdb/testsuite/lib/ada.exp b/gdb/testsuite/lib/ada.exp
> index 6a1e192..28284ff 100644
> --- a/gdb/testsuite/lib/ada.exp
> +++ b/gdb/testsuite/lib/ada.exp
> @@ -13,6 +13,26 @@
>  # You should have received a copy of the GNU General Public License
>  # along with this program.  If not, see <http://www.gnu.org/licenses/>.
>  
> +# Call target_compile with SOURCE DEST TYPE and OPTIONS as argument,
> +# after having temporarily changed the current working directory to
> +# BUILDDIR.
> +#
> +# FIXME: brobecke/2015-12-23: We can get rid of this function entirely
> +# when we are able to migrate to TCL 8.6, which added support for
> +# try { ... } finally {...} constructs.  Despite being released on
> +# December 2012, that version is still far from being widely used
> +# (in particular in the more exotic systems of the GCC compile farm).

Not sure this FIXME comment here turns out useful going forward.  We have
many other instances of catch in the harness.  And, IMO, having a wrapper
function looks nicer than open coding try/finally at the caller.  Once we
rely on 8.6 and thus try/finally, I would think we'd keep the wrapper
function, but reimplement its body with try/finally instead.

Thanks,
Pedro Alves
  
Pedro Alves Dec. 23, 2015, 3:30 p.m. UTC | #2
On 12/23/2015 03:21 PM, Pedro Alves wrote:
>> > --- a/gdb/testsuite/gdb.ada/cond_lang.exp
>> > +++ b/gdb/testsuite/gdb.ada/cond_lang.exp
>> > @@ -39,7 +39,7 @@ gdb_test "show lang" \
>> >  # current language mode is auto, and the breakpoint is inside Ada code.
>> >  set bp_location [gdb_get_line_number "STOP" ${testdir}/mixed.adb]
>> >  gdb_test "break mixed.adb:${bp_location} if light = green" \
>> > -         "Breakpoint \[0-9\]* at .*: file .*/mixed.adb, line \[0-9\]*\\."
>> > +         "Breakpoint \[0-9\]* at .*: file (.*/)?mixed.adb, line \[0-9\]*\\."
> Isn't that the same as just:
> 
>  -         "Breakpoint \[0-9\]* at .*: file .*/mixed.adb, line \[0-9\]*\\."
>  +         "Breakpoint \[0-9\]* at .*: file .*mixed.adb, line \[0-9\]*\\."
> 
> ?
> 

Huh, sorry, brain malfunction.  My version would give out a spurious
PASS with foomixed.adb, yours wouldn't.

>> >  gdb_test "continue" \
>> > -    "Breakpoint .*, pck\\.call_me \\(w=(w@entry=)?50\\) at .*/pck.adb:.*" \
>> > +    "Breakpoint .*, pck\\.call_me \\(w=(w@entry=)?50\\) at (.*)?/pck.adb:.*" \
>> >      "continue to call_me"
> Likewise, I think this is a no-op.  Did you mean to put the / inside the
> parens like in the other case?  If so I'd suggest:

This one still looks odd to me, though.

Thanks,
Pedro Alves
  

Patch

From 672d8857a125959b51e5f760cfc7f1c5d5a851cc Mon Sep 17 00:00:00 2001
From: Joel Brobecker <brobecker@adacore.com>
Date: Tue, 22 Dec 2015 13:28:41 +0400
Subject: [PATCH] [testsuite/Ada] stop using project files when building test
 programs

The current approach when building Ada programs for testing is
based on the use of a project file (testsuite/gdb.ada/gnat_ada.gpr).
To do that, we pass a number of additional arguments to target_compile,
one of them being the project file (via "-P/path/to/gnat_ada.gpr").
This used to work well-enough, but AdaCore is currently working towards
removing project-file support from gnatmake (the prefered tool for
using project files is gprbuild). So, we need to either switch
the compilation to gprbuild, or stop using project files.

First, using gprbuild is not always what users will be using to
build their applications. So having the option of using gnatmake
provides more flexibility towards exactly reproducing past bugs.
If we ever need a testcase that requires the use of gprbuild, then
I believe support for a new target needs to be added to dejagnu's
target_compile.

Also, the only real reason behind using a project file in the first
place is that we wanted to make it easy to specify the directory
where all compilation artifacts get stored. This is a consequence
of the organization choice we made for gdb.ada to keep each testcase
well organized. It is very easy to achieve that goal without using
project files.

This is therefore what this patch does: It change gdb_compile_ada
to build any program using gnatmake without using a project file
(by temporarily changing the current working directory).

There is a small (beneficial) side-effect; in the situation where
GDB is built in-tree, gnatmake is called as...

        % gnatmake [...] unit.adb

... which means that the debugging info in unit.o will say contain
a filename whose name is 'unit.adb', rather than '/path/to/unit.adb'.
This also better matches what users might typically do. But the side-
effect is that the unit name in the GDB output is not always a full
path. This patch tweaks a couple of testcases to make the path part
optional.

gdb/testsuite:

        * lib/ada.exp (target_compile_ada_from_dir): New function.
        (gdb_compile_ada): Reimplement avoiding the use of project files.
        * gdb.ada/gnat_ada.gpr: Delete.
        * gdb.ada/cond_lang.exp: Adjust test to make path before
        filename optional.
        * gdb.ada/small_reg_param.exp: Likewise.

Tested on x86_64-linux, with both in-tree and out-of-tree builds.
---
 gdb/testsuite/gdb.ada/cond_lang.exp       |  2 +-
 gdb/testsuite/gdb.ada/gnat_ada.gpr        | 25 -----------------------
 gdb/testsuite/gdb.ada/small_reg_param.exp |  2 +-
 gdb/testsuite/lib/ada.exp                 | 33 +++++++++++++++++++++++++++----
 4 files changed, 31 insertions(+), 31 deletions(-)
 delete mode 100644 gdb/testsuite/gdb.ada/gnat_ada.gpr

diff --git a/gdb/testsuite/gdb.ada/cond_lang.exp b/gdb/testsuite/gdb.ada/cond_lang.exp
index 0dfb9e3..7c3ad6e 100644
--- a/gdb/testsuite/gdb.ada/cond_lang.exp
+++ b/gdb/testsuite/gdb.ada/cond_lang.exp
@@ -39,7 +39,7 @@  gdb_test "show lang" \
 # current language mode is auto, and the breakpoint is inside Ada code.
 set bp_location [gdb_get_line_number "STOP" ${testdir}/mixed.adb]
 gdb_test "break mixed.adb:${bp_location} if light = green" \
-         "Breakpoint \[0-9\]* at .*: file .*/mixed.adb, line \[0-9\]*\\."
+         "Breakpoint \[0-9\]* at .*: file (.*/)?mixed.adb, line \[0-9\]*\\."
 
 # Now, continue until we hit the breakpoint.  If the condition is
 # evaluated correctly, the first hit will be ignored, and the debugger
diff --git a/gdb/testsuite/gdb.ada/gnat_ada.gpr b/gdb/testsuite/gdb.ada/gnat_ada.gpr
deleted file mode 100644
index 2736206..0000000
--- a/gdb/testsuite/gdb.ada/gnat_ada.gpr
+++ /dev/null
@@ -1,25 +0,0 @@ 
---  Copyright 2004-2015 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 3 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, see <http://www.gnu.org/licenses/>.
-
---  This project file allows us to control the location where the
---  compilation artifacts produced when building the Ada examples
---  are stored.
-
-project Gnat_Ada is
-
-   for Source_Dirs use (external ("SRC"));
-   for Object_Dir use external ("OBJ");
-
-end Gnat_Ada;
diff --git a/gdb/testsuite/gdb.ada/small_reg_param.exp b/gdb/testsuite/gdb.ada/small_reg_param.exp
index bd5cfd6..0a3b972 100644
--- a/gdb/testsuite/gdb.ada/small_reg_param.exp
+++ b/gdb/testsuite/gdb.ada/small_reg_param.exp
@@ -33,7 +33,7 @@  gdb_breakpoint "call_me"
 # Continue until we hit the breakpoint inside `Call_Me'.  We verify
 # that the parameter value is correct.
 gdb_test "continue" \
-    "Breakpoint .*, pck\\.call_me \\(w=(w@entry=)?50\\) at .*/pck.adb:.*" \
+    "Breakpoint .*, pck\\.call_me \\(w=(w@entry=)?50\\) at (.*)?/pck.adb:.*" \
     "continue to call_me"
 
 # And just to make sure, we also verify that the parameter value
diff --git a/gdb/testsuite/lib/ada.exp b/gdb/testsuite/lib/ada.exp
index 6a1e192..28284ff 100644
--- a/gdb/testsuite/lib/ada.exp
+++ b/gdb/testsuite/lib/ada.exp
@@ -13,6 +13,26 @@ 
 # You should have received a copy of the GNU General Public License
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
+# Call target_compile with SOURCE DEST TYPE and OPTIONS as argument,
+# after having temporarily changed the current working directory to
+# BUILDDIR.
+#
+# FIXME: brobecke/2015-12-23: We can get rid of this function entirely
+# when we are able to migrate to TCL 8.6, which added support for
+# try { ... } finally {...} constructs.  Despite being released on
+# December 2012, that version is still far from being widely used
+# (in particular in the more exotic systems of the GCC compile farm).
+
+proc target_compile_ada_from_dir {builddir source dest type options} {
+    set saved_cwd [pwd]
+    catch {
+        cd $builddir
+        return [target_compile $source $dest $type $options]
+    } result options
+    cd $saved_cwd
+    return -options $options $result
+}
+
 # Compile some Ada code.
 
 proc gdb_compile_ada {source dest type options} {
@@ -21,12 +41,17 @@  proc gdb_compile_ada {source dest type options} {
     set gprdir [file dirname $srcdir]
     set objdir [file dirname $dest]
 
+    # Although strictly not necessary, we force the recompilation
+    # of all units (additional_flags=-f).  This is what is done
+    # when using GCC to build programs in the other languages,
+    # and it avoids using a stray objfile file from a long-past
+    # run, for instance.
     append options " ada"
-    append options " additional_flags=-P$gprdir/gnat_ada"
-    append options " additional_flags=-XSRC=[file tail $srcdir]"
-    append options " additional_flags=-XOBJ=$objdir"
+    append options " additional_flags=-f"
+    append options " additional_flags=-I$srcdir"
 
-    set result [target_compile [file tail $source] $dest $type $options]
+    set result [target_compile_ada_from_dir \
+                    $objdir [file tail $source] $dest $type $options]]
 
     # The Ada build always produces some output, even when the build
     # succeeds. Thus, we can not use the output the same way we do in
-- 
2.1.4