Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: sami wagiaalla <swagiaal@redhat.com>
To: gdb-patches@sourceware.org
Subject: Re: [patch] function eval cleanup
Date: Tue, 20 Jul 2010 19:05:00 -0000	[thread overview]
Message-ID: <4C45F37C.1080401@redhat.com> (raw)
In-Reply-To: <4C446F00.5060709@redhat.com>

[-- Attachment #1: Type: text/plain, Size: 470 bytes --]


>
> In this particular location and if the language is C++ we are dealing
> with an OP_VAR_VALUE since all the other constructs were eliminated by
> the if else chain. So it is always a symbol, but I could add a check for
> OP_VAR_VALUE before skipping evaluation.
>

Tom pointed out that the catch all is also responsible for evaluating 
complex expressions describing function pointers. So I added a check for 
OP_VAR_VALUE, and a test-case for the missed scenario.


[-- Attachment #2: overload-cleanup.patch --]
[-- Type: text/plain, Size: 4315 bytes --]

Corrected pre-overload-resolution function evaluation.

2010-07-20  Sami Wagiaalla  <swagiaal@redhat.com>

	* eval.c (evaluate_subexp_standard): Disabled evaluation of C++
	function symbols before overload resolution.

2010-07-20  Sami Wagiaalla  <swagiaal@redhat.com>

	* gdb.cp/fpointer.cc: New test.
	* gdb.cp/fpointer.exp: New test.

diff --git a/gdb/eval.c b/gdb/eval.c
index 9a60616..ba75952 100644
--- a/gdb/eval.c
+++ b/gdb/eval.c
@@ -1491,19 +1491,29 @@ evaluate_subexp_standard (struct type *expect_type,
 	{
 	  /* Non-method function call */
 	  save_pos1 = *pos;
-	  argvec[0] = evaluate_subexp_with_coercion (exp, pos, noside);
 	  tem = 1;
-	  type = value_type (argvec[0]);
-	  if (type && TYPE_CODE (type) == TYPE_CODE_PTR)
-	    type = TYPE_TARGET_TYPE (type);
-	  if (type && TYPE_CODE (type) == TYPE_CODE_FUNC)
+
+	  /* If this is a C++ function wait until overload resolution.  */
+	  if (op == OP_VAR_VALUE
+	      && overload_resolution
+	      && (exp->language_defn->la_language == language_cplus))
 	    {
-	      for (; tem <= nargs && tem <= TYPE_NFIELDS (type); tem++)
+	      (*pos) += 4; /* Skip the evaluation of the symbol.  */
+	      argvec[0] = NULL;
+	    }
+	  else
+	    {
+	      argvec[0] = evaluate_subexp_with_coercion (exp, pos, noside);
+	      type = value_type (argvec[0]);
+	      if (type && TYPE_CODE (type) == TYPE_CODE_PTR)
+		type = TYPE_TARGET_TYPE (type);
+	      if (type && TYPE_CODE (type) == TYPE_CODE_FUNC)
 		{
-		  /* pai: FIXME This seems to be coercing arguments before
-		   * overload resolution has been done! */
-		  argvec[tem] = evaluate_subexp (TYPE_FIELD_TYPE (type, tem - 1),
-						 exp, pos, noside);
+		  for (; tem <= nargs && tem <= TYPE_NFIELDS (type); tem++)
+		    {
+		      argvec[tem] = evaluate_subexp (TYPE_FIELD_TYPE (type, tem - 1),
+						     exp, pos, noside);
+		    }
 		}
 	    }
 	}
diff --git a/gdb/testsuite/gdb.cp/fpointer.cc b/gdb/testsuite/gdb.cp/fpointer.cc
new file mode 100644
index 0000000..007ea39
--- /dev/null
+++ b/gdb/testsuite/gdb.cp/fpointer.cc
@@ -0,0 +1,40 @@
+#include "stdio.h"
+
+int foo(){
+  return 11;
+}
+
+int bar(){
+  return 12;
+}
+
+int bar(int){
+  printf ("bar(int)\n");
+  return 13;
+}
+
+int bar(char){
+  printf ("bar(char)\n");
+  return 14;
+}
+
+int (*p1)() = &foo;
+int (*p2)() = &bar;
+int (*p[2])() = {p1,p2};
+
+int (*p3)(int) = &bar;
+int (*p4)(char) = &bar;
+
+int main ()
+{
+  p1 ();
+  p2 ();
+
+  p[0]();
+  p[1]();
+
+  p3 ('a');
+  p4 (1);
+
+  return 0;
+}
diff --git a/gdb/testsuite/gdb.cp/fpointer.exp b/gdb/testsuite/gdb.cp/fpointer.exp
new file mode 100644
index 0000000..6144bbe
--- /dev/null
+++ b/gdb/testsuite/gdb.cp/fpointer.exp
@@ -0,0 +1,54 @@
+# Copyright 2008, 2010 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 tests for evaluation of functions in the form of
+# complex expressions.
+
+if $tracelevel then {
+    strace $tracelevel
+}
+
+set testfile fpointer
+set srcfile ${testfile}.cc
+set binfile ${objdir}/${subdir}/${testfile}
+if  { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable \
+      {debug c++}] != "" } {
+    untested "Couldn't compile test program"
+    return -1
+}
+
+# Get things started.
+
+gdb_exit
+gdb_start
+gdb_reinitialize_dir $srcdir/$subdir
+gdb_load ${binfile}
+
+if ![runto_main] then {
+    perror "couldn't run to breakpoint main"
+    continue
+}
+
+# Test that function calls done through function pointers
+gdb_test "p p1()" "= 11"
+gdb_test "p p2()" "= 12"
+
+gdb_test {p p[0]()} "= 11"
+gdb_test {p p[1]()} "= 12"
+
+# Test that overload resolution is not performed for function
+# pointers.
+gdb_test "p p3('a')" "= 13"
+gdb_test "p p4(1)"   "= 14"

  reply	other threads:[~2010-07-20 19:05 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-07-12 18:51 sami wagiaalla
2010-07-14 23:00 ` Tom Tromey
2010-07-19 15:28   ` sami wagiaalla
2010-07-20 19:05     ` sami wagiaalla [this message]
2010-07-20 19:57       ` Tom Tromey

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=4C45F37C.1080401@redhat.com \
    --to=swagiaal@redhat.com \
    --cc=gdb-patches@sourceware.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox