1997-10-31 egcs snapshot available

1997-10-31 egcs snapshot available

[Jeffrey A Law]

egcs-ss-971105 is now available in ftp.cygnus.com:/pub/egcs/snapshots/1997-11-05
(and on various mirrors shortly, see the home page for mirror sites).

Since we're close to a release I'm going to point out what's actually
changed.

	* Compiler hang/core dumps due to alias analysis bug(s) have
	been fixed.

	* Fix problem with "hypot" in NeXT include files.

	* Fix a couple alpha bugs.

	* Fix combine problems exposed by alpha port.

	* Fix sco5 tar vs cpio problem.

	* Fix some filenames to work on systems with 14char name limits.

	* Fix a couple minor arm bugs.

	* Enable stabs for sco5.

	* Fix most of the flow control problems with exception handling
	(yippie).

	* Fix problem with dwarf2 unwind change from 1031 snapshot.

	* Fix some 32x64 (or 64x32) cross compile problems in the g++
	front end

	* Nested exceptions should work.

	* Minor stl updates.


We still need to fix:

	* ppro internal error relating to conditional move patterns.

	* -funroll-loops problems.

	* review some ppc patches that may be necessary for linux.


I think those are the only two critical technical issues on the agenda!


egcs-ss-971105 is now available in ftp.cygnus.com:/pub/egcs/snapshots/1997-11-05
(and on various mirrors shortly, see the home page for mirror sites).

You'll find:

  egcs-971105.tar.gz                    The full egcs snapshot, including all
                                        languages runtime libraries and
                                        testsuites.

  egcs-core-971105.tar.gz               Just the C and objc language front ends.

  egcs-g++-971105.tar.gz                The g++ language and runtime.

  egcs-g77-971105.tar.gz                The g77 language and runtime.
   
  egcs-objc-971105.tar.gz               The objc runtime.


  egcs-testsuite-971105.tar.gz          The c & g++ testusites.
  egcs-g++-testsuite-971105.tar.gz      The g++ testsuite.
  egcs-g77-testsuite-971105.tar.gz      The g77 testsuite.


Diffs to bring the 971031 to 971105 snapshot are available.


Jeff Law (law@cygnus.com)
Cygnus Solutions		EGCS GNU Compiler System
http://www.cygnus.com		http://www.cygnus.com/egcs

Re: 1997-10-31 egcs snapshot available

[H.J. Lu]

> 
> We still need to fix:
> 
> 	* ppro internal error relating to conditional move patterns.
> 
> 	* -funroll-loops problems.
> 
> 	* review some ppc patches that may be necessary for linux.
> 

How about the vtable thunks bug on alpha? -fvtable-thunks is broken
on alpha, which is the default for linux. I have sent in a patch:

Sat Oct 11 17:24:53 1997  H.J. Lu  (hjl@gnu.ai.mit.edu)
 
    * toplev.c (rest_of_compilation): Don't defer thunks.

What happened to it?

Thanks.

H.J.

Re: 1997-10-31 egcs snapshot available

[Jason Merrill]

>>>>> H J Lu <hjl@lucon.org> writes:

> How about the vtable thunks bug on alpha? -fvtable-thunks is broken
> on alpha, which is the default for linux. I have sent in a patch:

> Sat Oct 11 17:24:53 1997  H.J. Lu  (hjl@gnu.ai.mit.edu)
 
>     * toplev.c (rest_of_compilation): Don't defer thunks.

> What happened to it?

Hadn't gotten to it, sorry.  I think I prefer this fix:

Thu Nov  6 11:28:14 1997  Jason Merrill  <jason@yorick.cygnus.com>

	* method.c (emit_thunk): Don't let the backend defer generic thunks.

Index: method.c
===================================================================
RCS file: /cvs/cvsfiles/devo/gcc/cp/method.c,v
retrieving revision 1.166
diff -c -r1.166 method.c
*** method.c	1997/10/22 23:59:59	1.166
--- method.c	1997/11/06 19:39:12
***************
*** 1927,1932 ****
--- 1927,1939 ----
      c_expand_return (t);
  
      finish_function (lineno, 0, 0);
+ 
+     /* Don't let the backend defer this function.  */
+     if (DECL_DEFER_OUTPUT (thunk_fndecl))
+       {
+ 	output_inline_function (thunk_fndecl);
+ 	permanent_allocation (1);
+       }
  #endif /* ASM_OUTPUT_MI_THUNK */
    }
  

Re: 1997-11-05 egcs snapshot available

[Oleg Krivosheev]

hi,

please find enclosed patch for new teml.firends
declaration in iomanip.h.

regards

OK

ps

SGI STL oct 31 release is out. Will we upgrade
now or wait after release ?



*** iomanip.h.orig	Thu Nov  6 11:46:36 1997
--- iomanip.h	Thu Nov  6 11:52:07 1997
***************
*** 102,107 ****
--- 102,110 ----
         { return imanip<TP>(_f, a); }
  };
  
+ template <class TP>
+ inline istream& operator>>(istream&, const imanip<TP>&);
+ 
  template <class TP> class imanip {
      istream& (*_f)(istream&, TP);
      TP _a;
***************
*** 109,115 ****
      imanip(istream& (*f)(istream&, TP), TP a) : _f(f), _a(a) {}
      //
      friend
!       istream& operator>>(istream& i, const imanip<TP>& m);
  };
  
  template <class TP>
--- 112,118 ----
      imanip(istream& (*f)(istream&, TP), TP a) : _f(f), _a(a) {}
      //
      friend
!       istream& operator>> <>(istream& i, const imanip<TP>& m);
  };
  
  template <class TP>
***************
*** 131,136 ****
--- 134,142 ----
        { return omanip<TP>(_f, a); }
  };
  
+ template <class TP>
+ inline ostream& operator<<(ostream&, const omanip<TP>&);
+ 
  template <class TP> class omanip {
      ostream& (*_f)(ostream&, TP);
      TP _a;
***************
*** 138,144 ****
      omanip(ostream& (*f)(ostream&, TP), TP a) : _f(f), _a(a) {}
      //
      friend
!       ostream& operator<<(ostream& o, const omanip<TP>& m);
  };
  
  template <class TP>
--- 144,150 ----
      omanip(ostream& (*f)(ostream&, TP), TP a) : _f(f), _a(a) {}
      //
      friend
!       ostream& operator<< <>(ostream& o, const omanip<TP>& m);
  };
  
  template <class TP>

Re: 1997-11-05 egcs snapshot available

[Jason Merrill]

>>>>> Oleg Krivosheev <kriol@fnal.gov> writes:

> please find enclosed patch for new teml.firends
> declaration in iomanip.h.

Thanks.  Next time please include a ChangeLog entry.

> SGI STL oct 31 release is out. Will we upgrade
> now or wait after release ?

We already upgraded, because it fixes similar issues.

Jason

1997-10-31 egcs snapshot available

[Jeffrey A Law]

egcs-ss-971031 is now available in ftp.cygnus.com:/pub/egcs/snapshots/1997-10-31
(and on various mirrors shortly, see the home page for mirror sites).

You'll find:

  egcs-971031.tar.gz                    The full egcs snapshot, including all
                                        languages runtime libraries and
                                        testsuites.

  egcs-core-971031.tar.gz               Just the C and objc language front ends.

  egcs-g++-971031.tar.gz                The g++ language and runtime.

  egcs-g77-971031.tar.gz                The g77 language and runtime.
   
  egcs-objc-971031.tar.gz               The objc runtime.


  egcs-testsuite-971031.tar.gz          The c & g++ testusites.
  egcs-g++-testsuite-971031.tar.gz      The g++ testsuite.
  egcs-g77-testsuite-971031.tar.gz      The g77 testsuite.


Diffs to bring the 971023 to 971031 snapshot are available.





Jeff Law (law@cygnus.com)
Cygnus Solutions		EGCS GNU Compiler System
http://www.cygnus.com		http://www.cygnus.com/egcs

Re: 1997-10-31 egcs snapshot available

[H.J. Lu]

> 
> 
> egcs-ss-971031 is now available in ftp.cygnus.com:/pub/egcs/snapshots/1997-10-31
> (and on various mirrors shortly, see the home page for mirror sites).
> 

I got this on linux/x86. It seems that dwarf EH is broken again.


H.J.
---
FAIL: g++.benjamin/warn01.C (test for excess errors)
FAIL: g++.jason/template31.C (test for excess errors)
FAIL: g++.mike/eh10.C  Execution test
FAIL: g++.mike/eh12.C  Execution test
FAIL: g++.mike/eh14.C  Execution test
FAIL: g++.mike/eh16.C  Execution test
FAIL: g++.mike/eh17.C  Execution test
FAIL: g++.mike/eh18.C  Execution test
FAIL: g++.mike/eh2.C  Execution test
FAIL: g++.mike/eh21.C  Execution test
FAIL: g++.mike/eh24.C  Execution test
FAIL: g++.mike/eh25.C  Execution test
FAIL: g++.mike/eh26.C  Execution test
FAIL: g++.mike/eh27.C  Execution test
FAIL: g++.mike/eh28.C  Execution test
FAIL: g++.mike/eh29.C  Execution test
FAIL: g++.mike/eh3.C  Execution test
FAIL: g++.mike/eh31.C  Execution test
FAIL: g++.mike/eh33.C  Execution test
FAIL: g++.mike/eh34.C  Execution test
FAIL: g++.mike/eh35.C  Execution test
FAIL: g++.mike/eh36.C  Execution test
FAIL: g++.mike/eh37.C  Execution test
FAIL: g++.mike/eh38.C  Execution test
FAIL: g++.mike/eh39.C  Execution test
FAIL: g++.mike/eh40.C  Execution test
FAIL: g++.mike/eh41.C  Execution test
FAIL: g++.mike/eh42.C  Execution test
FAIL: g++.mike/eh44.C  Execution test
FAIL: g++.mike/eh47.C  Execution test
FAIL: g++.mike/eh48.C  Execution test
FAIL: g++.mike/eh49.C  Execution test
FAIL: g++.mike/eh5.C  Execution test
FAIL: g++.mike/eh50.C  Execution test
FAIL: g++.mike/eh51.C  Execution test
FAIL: g++.mike/eh6.C  Execution test
FAIL: g++.mike/eh8.C  Execution test
FAIL: g++.mike/p7912.C  Execution test
FAIL: g++.mike/p9706.C  Execution test

Re: 1997-10-31 egcs snapshot available

[Jeffrey A Law]

  In message <m0xRUeu-0004ecC@ocean.lucon.org>you write:
  > I got this on linux/x86. It seems that dwarf EH is broken again.
Just a note, after fixing the NULL termination problem in dwarf2out.c
I get the following:

Running /hurl/puke/law/egcs/gcc/testsuite/g++.old-deja/old-deja.exp ...
FAIL: g++.benjamin/warn01.C (test for excess errors)
XPASS: g++.jason/destruct3.C - (test for bogus messages, line 38)
FAIL: g++.jason/template31.C (test for excess errors)
FAIL: g++.mike/eh23.C  Execution test
FAIL: g++.mike/eh3.C  Execution test
FAIL: g++.mike/eh33.C  Execution test
FAIL: g++.mike/eh35.C  Execution test
FAIL: g++.mike/eh44.C  Execution test
FAIL: g++.mike/eh5.C  Execution test
FAIL: g++.mike/eh50.C  Execution test
FAIL: g++.mike/eh51.C  Execution test

                === g++ Summary ===

# of expected passes            3371
# of unexpected failures        10
# of unexpected successes       1
# of expected failures          81
# of untested testcases         6


I haven't looked through my testresults archives to see how that
compares to other x86 testruns.

jeff

Re: 1997-10-31 egcs snapshot available

[H.J. Lu]

> 
> 
>   In message <m0xRUeu-0004ecC@ocean.lucon.org>you write:
>   > I got this on linux/x86. It seems that dwarf EH is broken again.
> Just a note, after fixing the NULL termination problem in dwarf2out.c
> I get the following:
> 
> Running /hurl/puke/law/egcs/gcc/testsuite/g++.old-deja/old-deja.exp ...
> FAIL: g++.benjamin/warn01.C (test for excess errors)
> XPASS: g++.jason/destruct3.C - (test for bogus messages, line 38)
> FAIL: g++.jason/template31.C (test for excess errors)
> FAIL: g++.mike/eh23.C  Execution test
> FAIL: g++.mike/eh3.C  Execution test
> FAIL: g++.mike/eh33.C  Execution test
> FAIL: g++.mike/eh35.C  Execution test
> FAIL: g++.mike/eh44.C  Execution test
> FAIL: g++.mike/eh5.C  Execution test
> FAIL: g++.mike/eh50.C  Execution test
> FAIL: g++.mike/eh51.C  Execution test
> 

After applying your dwarf2out.c patch, these are the only failures I
got on linux/libc5/x86:

FAIL: g++.benjamin/warn01.C (test for excess errors)
FAIL: g++.jason/template31.C (test for excess errors)
FAIL: g++.mike/eh35.C  Execution test

It looks quite good.

BTW, after applying the patch, I have to start from scratch to get
this result.


H.J.

Re: 1997-10-31 egcs snapshot available

[Jeffrey A Law]

  In message < m0xRmP5-0004ecC@ocean.lucon.org >you write:
  > FAIL: g++.benjamin/warn01.C (test for excess errors)
  > FAIL: g++.jason/template31.C (test for excess errors)
  > FAIL: g++.mike/eh35.C  Execution test
  > 
  > It looks quite good.
  > 
  > BTW, after applying the patch, I have to start from scratch to get
  > this result.
Well, I still have the old binutils, so that probably accounts
for the difference.

jeff