[Bug ada/27937] New: [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

[Bug ada/27937] New: [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

[gcc-bugzilla at gcc dot gnu dot org]

Bootstrapping mainline on Solaris 10/x86 (configured for
i686-pc-solaris2.10 to work around the still unfixed PR target/26146) fails
compiling several Ada files:

/vol/gccsrc/obj/gcc-4.2.0-20060606/10-gcc-gas-amd64/./gcc/xgcc
-B/vol/gccsrc/obj/gcc-4.2.0-20060606/10-gcc-gas-amd64/./gcc/
-B/vol/gcc/i686-pc-solaris2.10/bin/ -B/vol/gcc/i686-pc-solaris2.10/lib/
-isystem /vol/gcc/i686-pc-solaris2.10/include -isystem
/vol/gcc/i686-pc-solaris2.10/sys-include -c -g -O2 -fPIC      -W -Wall -gnatpg 
a-stwifi.adb -o a-stwifi.o
virtual memory exhausted: Not enough space

If one observes the compilation with top, one can see that VM usage grows
to 4 GB before gnat1 dies.  Compiling with -O instead works.

This is a regression from the 4.1 branch.

Environment:
System: SunOS afrika 5.10 Generic_Patch i86pc i386 i86pc
Architecture: i86pc


host: i686-pc-solaris2.10
build: i686-pc-solaris2.10
target: i686-pc-solaris2.10
configured with: /vol/gcc/src/gcc-dist/configure --prefix=/vol/gcc
--with-local-prefix=/vol/gcc --disable-nls --with-gnu-as
--with-as=/usr/sfw/bin/gas --with-gmp-dir=/vol/gnu/obj/gmp-4.1.3
--with-mpfr-dir=/vol/gnu/obj/gmp-4.1.3/mpfr
--enable-languages=c,c++,fortran,objc,ada --host i686-pc-solaris2.10 --build
i686-pc-solaris2.10 --target i686-pc-solaris2.10 --disable-libmudflap

How-To-Repeat:
Bootstrap mainline as described above.


-- 
           Summary: [4.2 Regression] Ada bootstrap failure on Solaris 10/x86
           Product: gcc
           Version: unknown
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: ada
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: ro at techfak dot uni-bielefeld dot de
 GCC build triplet: i686-pc-solaris2.10
  GCC host triplet: i686-pc-solaris2.10
GCC target triplet: i686-pc-solaris2.10


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27937

[Bug ada/27937] [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

[pinskia at gcc dot gnu dot org]

-- 

pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.2.0
            Version|unknown                     |4.2.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27937

[Bug ada/27937] [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

[ebotcazou at gcc dot gnu dot org]

------- Comment #1 from ebotcazou at gcc dot gnu dot org  2006-06-09 09:08 -------
Confirmed, this happens when the compiler is configured for SJLJ exceptions.


-- 

ebotcazou at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |ebotcazou at gcc dot gnu dot
                   |                            |org
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|0                           |1
   Last reconfirmed|0000-00-00 00:00:00         |2006-06-09 09:08:22
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27937

[Bug ada/27937] [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

[ro at techfak dot uni-bielefeld dot de]

------- Comment #2 from ro at techfak dot uni-bielefeld dot de  2006-06-09 15:36 -------
Subject: Re:  [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

ebotcazou at gcc dot gnu dot org writes:

> Confirmed, this happens when the compiler is configured for SJLJ exceptions.

But this seems not to be the case here:

% grep -i except gcc/auto-host.h 
/* Define 0/1 to force the choice for exception handling model. */
/* #undef CONFIG_SJLJ_EXCEPTIONS */

        Rainer


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27937

[Bug ada/27937] [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

[ebotcazou at gcc dot gnu dot org]

------- Comment #3 from ebotcazou at gcc dot gnu dot org  2006-06-11 15:43 -------
> But this seems not to be the case here:
> 
> % grep -i except gcc/auto-host.h 
> /* Define 0/1 to force the choice for exception handling model. */
> /* #undef CONFIG_SJLJ_EXCEPTIONS */

Right, it's another kind of SJLJ exceptions, purely front-end based.  Excerpt
from system-solaris-x86.ads:

   ZCX_By_Default            : constant Boolean := False;
   GCC_ZCX_Support           : constant Boolean := False;


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27937

[Bug ada/27937] [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

[ro at techfak dot uni-bielefeld dot de]

------- Comment #4 from ro at techfak dot uni-bielefeld dot de  2006-06-12 17:45 -------
Subject: Re:  [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

ebotcazou at gcc dot gnu dot org writes:

> Right, it's another kind of SJLJ exceptions, purely front-end based.  Excerpt
> from system-solaris-x86.ads:
> 
>    ZCX_By_Default            : constant Boolean := False;
>    GCC_ZCX_Support           : constant Boolean := False;

Ok, I see.  Thanks for the explanation.  Any reason this is done this way?
Solaris/SPARC doesn't use this, nor does Linux/x86, so it seems like both
the Solaris and x86 parts of the code are there.

        Rainer


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27937

[Bug ada/27937] [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

[ebotcazou at gcc dot gnu dot org]

------- Comment #5 from ebotcazou at gcc dot gnu dot org  2006-06-12 18:11 -------
> Ok, I see.  Thanks for the explanation.  Any reason this is done this way?
> Solaris/SPARC doesn't use this, nor does Linux/x86, so it seems like both
> the Solaris and x86 parts of the code are there.

Short answer: seamless support for synchronous signals.

Long answer: we want to be able to turn synchronous signals into regular Ada
exceptions.  This means unwinding through signal handler frames (easy) and
unwinding through kernel frames (tricky).  With SJLJ exceptions, the latter
turns out to be transparent; but with DWARF-2 EH exceptions, you need specific
support code under the form of MD_FALLBACK_FRAME_STATE_FOR macros, defined in
the *-unwind.h files.  The FSF tree contains some M_F_F_S_F macros, we have
a few more, in particular for SPARC/Solaris but not for x86/Solaris.  We should
already have contributed them, but there is a discrepancy between Java and Ada
in this domain that should be sorted out first.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27937

[Bug ada/27937] [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

[ro at techfak dot uni-bielefeld dot de]

------- Comment #6 from ro at techfak dot uni-bielefeld dot de  2006-06-21 17:35 -------
Subject: Re:  [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

ebotcazou at gcc dot gnu dot org writes:

> Long answer: we want to be able to turn synchronous signals into regular Ada
> exceptions.  This means unwinding through signal handler frames (easy) and
> unwinding through kernel frames (tricky).  With SJLJ exceptions, the latter
> turns out to be transparent; but with DWARF-2 EH exceptions, you need specific
> support code under the form of MD_FALLBACK_FRAME_STATE_FOR macros, defined in
> the *-unwind.h files.  The FSF tree contains some M_F_F_S_F macros, we have
> a few more, in particular for SPARC/Solaris but not for x86/Solaris.  We should
> already have contributed them, but there is a discrepancy between Java and Ada
> in this domain that should be sorted out first.

Ok, thanks for the explanation.  But right now, the FSF tree has M_F_F_S_F
definitions for neither Solaris/SPARC nor /x86, AFAIKT.  And Solaris/SPARC
and Solaris/SPARCv9 can turn on ZCX and are thus unaffected, while
Solaris/x86 has ZCX off, which breaks bootstrap.

Since this is a mainline regression from 3.x/4.[01], do you think there
could be a fix/workaround in place for 4.2.0?

Thanks.
        Rainer


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27937

[Bug ada/27937] [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

[ebotcazou at gcc dot gnu dot org]

------- Comment #7 from ebotcazou at gcc dot gnu dot org  2006-06-21 18:12 -------
> Ok, thanks for the explanation.  But right now, the FSF tree has M_F_F_S_F
> definitions for neither Solaris/SPARC nor /x86, AFAIKT.  And Solaris/SPARC
> and Solaris/SPARCv9 can turn on ZCX and are thus unaffected, while
> Solaris/x86 has ZCX off, which breaks bootstrap.

ZCX are already turned on for SPARC/Solaris.

> Since this is a mainline regression from 3.x/4.[01], do you think there
> could be a fix/workaround in place for 4.2.0?

Sure, someone has to write it. :-)


-- 

ebotcazou at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|ebotcazou at gcc dot gnu dot|
                   |org                         |
         AssignedTo|unassigned at gcc dot gnu   |ebotcazou at gcc dot gnu dot
                   |dot org                     |org
             Status|NEW                         |ASSIGNED
   Last reconfirmed|2006-06-09 09:08:22         |2006-06-21 18:12:29
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27937

[Bug ada/27937] [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

[ro at techfak dot uni-bielefeld dot de]

------- Comment #8 from ro at techfak dot uni-bielefeld dot de  2006-06-21 18:15 -------
Subject: Re:  [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

ebotcazou at gcc dot gnu dot org writes:

> > Ok, thanks for the explanation.  But right now, the FSF tree has M_F_F_S_F
> > definitions for neither Solaris/SPARC nor /x86, AFAIKT.  And Solaris/SPARC
> > and Solaris/SPARCv9 can turn on ZCX and are thus unaffected, while
> > Solaris/x86 has ZCX off, which breaks bootstrap.
> 
> ZCX are already turned on for SPARC/Solaris.

That's what I meant.  I wondered why it's turned on without M_F_F_S_F and
works while it's off on Solaris/x86.

> > Since this is a mainline regression from 3.x/4.[01], do you think there
> > could be a fix/workaround in place for 4.2.0?
> 
> Sure, someone has to write it. :-)

Which is true for most fixes :-)

Thanks.
        Rainer


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27937

[Bug tree-optimization/27937] [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

[rguenth at gcc dot gnu dot org]

------- Comment #9 from rguenth at gcc dot gnu dot org  2006-06-25 10:58 -------
Is there a testcase for this bug that can be examined on more common
host/target systems?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27937

[Bug tree-optimization/27937] [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

[mmitchel at gcc dot gnu dot org]

------- Comment #10 from mmitchel at gcc dot gnu dot org  2006-07-05 17:57 -------
Ada is not release-critical.


-- 

mmitchel at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P5


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27937

[Bug tree-optimization/27937] [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

[rguenth at gcc dot gnu dot org]

------- Comment #11 from rguenth at gcc dot gnu dot org  2006-07-24 08:33 -------
Is this now fixed on the mainline?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27937

[Bug tree-optimization/27937] [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

[ro at techfak dot uni-bielefeld dot de]

------- Comment #12 from ro at techfak dot uni-bielefeld dot de  2006-07-28 16:03 -------
Subject: Re:  [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

rguenth at gcc dot gnu dot org writes:

> Is this now fixed on the mainline?

A mainline bootstrap as of 20060724 was successful, so it seems so
(although I haven't seen a specific fix addressing this issue).

        Rainer


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27937

[Bug tree-optimization/27937] [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

[pinskia at gcc dot gnu dot org]

------- Comment #13 from pinskia at gcc dot gnu dot org  2006-07-28 16:21 -------
Fixed so closing as such.


-- 

pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27937

[Bug tree-optimization/27937] [4.2 Regression] Ada bootstrap failure on Solaris 10/x86

[ebotcazou at gcc dot gnu dot org]

------- Comment #14 from ebotcazou at gcc dot gnu dot org  2006-08-04 21:36 -------
> A mainline bootstrap as of 20060724 was successful, so it seems so
> (although I haven't seen a specific fix addressing this issue).

This is http://gcc.gnu.org/ml/gcc-patches/2006-07/msg00229.html and I asked
Zdenek to add the PR reference twice.

Need to learn a little bit of Czech someday :-)


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27937