small patch

small patch

[Mike Stump]

Based on comment from Peter Leven
<leven@lpr.e-technik.tu-muenchen.de>, thanks Peter.

Fri Mar 27 17:46:16 1998  Mike Stump  <mrs@wrs.com>

    * inc/exception: Fix ordering of closing braces.

Doing diffs in gcc/cp/inc:
*** gcc/cp/inc/exception.~1~	Thu Jan 15 17:45:20 1998
--- gcc/cp/inc/exception	Fri Mar 27 17:44:59 1998
*************** void terminate (void) __attribute__ ((__
*** 33,42 ****
  unexpected_handler set_unexpected (unexpected_handler);
  void unexpected (void) __attribute__ ((__noreturn__));
  bool uncaught_exception ();
- } // extern "C++"
  
  #if 0
  } // namespace std
  #endif
  
  #endif
--- 33,43 ----
  unexpected_handler set_unexpected (unexpected_handler);
  void unexpected (void) __attribute__ ((__noreturn__));
  bool uncaught_exception ();
  
  #if 0
  } // namespace std
  #endif
+ 
+ } // extern "C++"
  
  #endif
--------------

Re: small patch

[Jeffrey A Law]

  In message < 199801162104.NAA24605@kankakee.wrs.com >you write:
  > Here is a small patch to correct these two test cases:
  > 
  > XPASS: g++.mike/dyncast1.C  Execution test
  > XPASS: g++.mike/dyncast2.C  Execution test
Thanks.  Installed.
jeff

small patch

[Mike Stump]

Here is a small patch to correct these two test cases:

XPASS: g++.mike/dyncast1.C  Execution test
XPASS: g++.mike/dyncast2.C  Execution test
 
                === g++ Summary ===
 
# of expected passes            3463
# of unexpected successes       2
# of expected failures          79
# of untested testcases         6

testsuite as off a little while ago, on a Sun solaris 2.5.1 sparc.

Doing diffs in testsuite:
*** testsuite/g++.old-deja/g++.mike/dyncast1.C.~1~	Wed Oct 29 16:07:54 1997
--- testsuite/g++.old-deja/g++.mike/dyncast1.C	Wed Nov  5 18:09:42 1997
***************
*** 1,6 ****
  // Special g++ Options: -fexceptions -w
  // excess errors test - XFAIL a29k-*-* sparc64-*-elf sh-*-* arm-*-pe
- // execution test fails - XFAIL powerpc-*-eabi
  
  #include <typeinfo>
  
--- 1,5 ----
*** testsuite/g++.old-deja/g++.mike/dyncast2.C.~1~	Wed Oct 29 16:07:55 1997
--- testsuite/g++.old-deja/g++.mike/dyncast2.C	Wed Nov  5 18:09:45 1997
***************
*** 1,6 ****
  // Special g++ Options: -fexceptions -w
  // excess errors test - XFAIL a29k-*-* sparc64-*-elf sh-*-* arm-*-pe
- // execution test fails - XFAIL powerpc-*-eabi
  
  // Ensure reference handling works.
  
--- 1,5 ----
--------------

Re: small patch

[Mike Stump]

> To: mrs@wrs.com (Mike Stump)
> cc: egcs@cygnus.com
> Date: Wed, 14 Jan 1998 00:24:05 -0700
> From: Jeffrey A Law <law@hurl.cygnus.com>

>   In message < 199801131847.KAA00320@kankakee.wrs.com >you write:
> And how does this relate to a *program* that you're compiling.

It doesn't relate to a program being compiled, only to programs that
are loaded into VxWorks.

The symbol table of the program that you load (the testcase), modifies
the command parser for the target shell.  This happens because there
is just one global symbol table on the target managed by the OS.  When
the object is loaded, along with the symbol table, the normal sp
command disappears, as the new sp value replaces the prior definition
of `sp'.  The prior definition was just a function to spawn a task,
with the testcase, it becomes a global variable.

> ie, you're talking about the stuff that dejagnu uses to spawn programs,
> not about the program that's actually being spawned.

I hope the above was more clear.

Re: small patch

[Jeffrey A Law]

  In message < 199801131847.KAA00320@kankakee.wrs.com >you write:
  > sp is command in the target shell, meaning spawn, that command is used
  > by dejagnu to run testcases.  One feature of VxWorks target shell is
  > that you can replace builtins with new definitions, though replacing
  > sp with a global variable called sp isn't very useful.
And how does this relate to a *program* that you're compiling.

ie, you're talking about the stuff that dejagnu uses to spawn programs,
not about the program that's actually being spawned.  Obviously I'm
missing something.


  > This patch is reasonably safe, as we still test what it is the
  > testcase is trying to test.
No doubt.  I'm just trying to figure out what's going on.  I'm probably
going to go ahead and install it, even though it seems to be just papering
over severe namespace pollution in VxWorks.

jeff

Re: small patch

[Mike Stump]

sp is command in the target shell, meaning spawn, that command is used
by dejagnu to run testcases.  One feature of VxWorks target shell is
that you can replace builtins with new definitions, though replacing
sp with a global variable called sp isn't very useful.

This patch is reasonably safe, as we still test what it is the
testcase is trying to test.

> To: mrs@wrs.com (Mike Stump)
> cc: egcs@cygnus.com
> Date: Tue, 13 Jan 1998 00:48:16 -0700
> From: Jeffrey A Law <law@hurl.cygnus.com>

>   In message < 199801122240.OAA16601@kankakee.wrs.com >you write:
>   > Here is a small patch to avoid a failure on VxWorks:
>   > 
>   > Doing diffs in gcc/testsuite/g++.old-deja/g++.bugs:
>   > *** gcc/testsuite/g++.old-deja/g++.bugs/900520_05.C.~1~	Mon Jan 12 13:5
>   > 3:27 1998
>   > --- gcc/testsuite/g++.old-deja/g++.bugs/900520_05.C	Mon Jan 12 13:54:48 199
>   > 8
>   > *************** struct struct_0 {
>   > *** 8,14 ****
>   >   };
>   >   
>   >   char *cp;
>   > ! struct_0 *sp;
>   >   
>   >   void test0 ()
>   >   {
>   > --- 8,14 ----
>   >   };
>   >   
>   >   char *cp;
>   > ! static struct_0 *sp;
>   >   
>   >   void test0 ()
>   >   {
>   > --------------
> Huh?  Something's rather bogus here.  Why does adding a static help
> VxWorks?

Re: small patch

[Jeffrey A Law]

  In message < 199801122240.OAA16601@kankakee.wrs.com >you write:
  > Here is a small patch to avoid a failure on VxWorks:
  > 
  > Doing diffs in gcc/testsuite/g++.old-deja/g++.bugs:
  > *** gcc/testsuite/g++.old-deja/g++.bugs/900520_05.C.~1~	Mon Jan 12 13:5
  > 3:27 1998
  > --- gcc/testsuite/g++.old-deja/g++.bugs/900520_05.C	Mon Jan 12 13:54:48 199
  > 8
  > *************** struct struct_0 {
  > *** 8,14 ****
  >   };
  >   
  >   char *cp;
  > ! struct_0 *sp;
  >   
  >   void test0 ()
  >   {
  > --- 8,14 ----
  >   };
  >   
  >   char *cp;
  > ! static struct_0 *sp;
  >   
  >   void test0 ()
  >   {
  > --------------
Huh?  Something's rather bogus here.  Why does adding a static help
VxWorks?

jeff

small patch

[Mike Stump]

Here is a small patch to avoid a failure on VxWorks:

Doing diffs in gcc/testsuite/g++.old-deja/g++.bugs:
*** gcc/testsuite/g++.old-deja/g++.bugs/900520_05.C.~1~	Mon Jan 12 13:53:27 1998
--- gcc/testsuite/g++.old-deja/g++.bugs/900520_05.C	Mon Jan 12 13:54:48 1998
*************** struct struct_0 {
*** 8,14 ****
  };
  
  char *cp;
! struct_0 *sp;
  
  void test0 ()
  {
--- 8,14 ----
  };
  
  char *cp;
! static struct_0 *sp;
  
  void test0 ()
  {
--------------