[Bug c/29030] New: gcc.dg/array-9.c produces internal compiler error on Darwin at -m64

[Bug c/29030] New: gcc.dg/array-9.c produces internal compiler error on Darwin at -m64

[howarth at nitro dot med dot uc dot edu]

The gcc.dg/array-9.c testcase results in an internal compiler error on
powerpc-apple-darwin8 when compiled at -m64.

gcc-4 -O3 -m64 array-9.c
array-9.c:7: error: declaration of 'x' as array of voids
array-9.c: In function 'foo':
array-9.c:10: internal compiler error: tree check: expected class 'type', have
'exceptional' (error_mark) in rs6000_darwin64_record_arg_advance_recurse, at
config/rs6000/rs6000.c:4610
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.

gcc-4 -v 
Using built-in specs.
Target: powerpc-apple-darwin8
Configured with: ../gcc-4.2-20060910/configure --prefix=/sw
--prefix=/sw/lib/gcc4 --enable-languages=c,c++,fortran
--infodir=/sw/lib/gcc4/share/info --with-gmp=/sw --with-included-gettext
--host=powerpc-apple-darwin8 --with-libiconv-prefix=/sw


-- 
           Summary: gcc.dg/array-9.c produces internal compiler error on
                    Darwin at -m64
           Product: gcc
           Version: 4.2.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: howarth at nitro dot med dot uc dot edu
 GCC build triplet: powerpc-apple-darwin8
  GCC host triplet: powerpc-apple-darwin8
GCC target triplet: powerpc-apple-darwin8


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29030

[Bug target/29030] gcc.dg/array-9.c produces internal compiler error on Darwin at -m64

[pinskia at gcc dot gnu dot org]

------- Comment #1 from pinskia at gcc dot gnu dot org  2006-09-12 04:16 -------
This is a target specific issue, I might fix this even though I don't have any
way of testing the fix (except maybe for changing the ABI on powerpc-linux-gnu
to use the darwin64 struct passing ABI).


-- 

pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |pinskia at gcc dot gnu dot
                   |                            |org
          Component|c                           |target


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29030

[Bug target/29030] gcc.dg/array-9.c produces internal compiler error on Darwin at -m64

[howarth at nitro dot med dot uc dot edu]

------- Comment #2 from howarth at nitro dot med dot uc dot edu  2006-09-12 04:17 -------
Created an attachment (id=12230)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=12230&action=view)
preprocessed file for array-9.s generated with -O3 -m64 on Darwin PPC


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29030

[Bug target/29030] gcc.dg/array-9.c produces internal compiler error on Darwin at -m64

[howarth at nitro dot med dot uc dot edu]

------- Comment #3 from howarth at nitro dot med dot uc dot edu  2006-09-12 04:20 -------
Andrew,
     I'd be happy to test any attempts to fix this.
                        Jack


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29030

[Bug target/29030] gcc.dg/array-9.c produces internal compiler error on Darwin at -m64

[pinskia at gcc dot gnu dot org]

------- Comment #4 from pinskia at gcc dot gnu dot org  2006-09-12 05:57 -------
Simple untested (even on the testcase) fix:
Index: rs6000.c
===================================================================
--- rs6000.c    (revision 116799)
+++ rs6000.c    (working copy)
@@ -4607,7 +4607,10 @@ rs6000_darwin64_record_arg_advance_recur
       {
        HOST_WIDE_INT bitpos = startbitpos;
        tree ftype = TREE_TYPE (f);
-       enum machine_mode mode = TYPE_MODE (ftype);
+       enum machine_mode mode;
+       if (ftype == error_mark_node)
+         continue;
+       mode = TYPE_MODE (ftype);

        if (DECL_SIZE (f) != 0
            && host_integerp (bit_position (f), 1))


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29030

[Bug target/29030] gcc.dg/array-9.c produces internal compiler error on Darwin at -m64

[howarth at nitro dot med dot uc dot edu]

------- Comment #5 from howarth at nitro dot med dot uc dot edu  2006-09-13 12:20 -------
Andrew,
      The proposed patch doesn't work. It converts the internal
compiler error into a compiler time error of...

gcc-4 -O3 -g -m64 array-9.c
array-9.c:7: error: declaration of 'x' as array of voids


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29030

[Bug target/29030] gcc.dg/array-9.c produces internal compiler error on Darwin at -m64

[pinskia at gcc dot gnu dot org]

------- Comment #6 from pinskia at gcc dot gnu dot org  2006-09-13 16:22 -------
(In reply to comment #5)
> Andrew,
>       The proposed patch doesn't work. It converts the internal
> compiler error into a compiler time error of...
Yes the patch does work as this is invalid code to begin with :).
Look at the testcase:
struct A
{
  int i;
  void x[1];  /* { dg-error "array of voids" } */
};

void foo(struct A a) {}


See how there is a dg-error there, that error is expected.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29030

[Bug target/29030] gcc.dg/array-9.c produces internal compiler error on Darwin at -m64

[howarth at nitro dot med dot uc dot edu]

------- Comment #7 from howarth at nitro dot med dot uc dot edu  2006-09-13 18:41 -------
Okay. I'll run a complete make check tonight to verify
that the patch introduces no regressions in at least
the c, c++ and fortran language build. Assuming no
regressions, do you want to submit the patch or
should I (assuming Geoff has no objections to it)?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29030

[Bug target/29030] gcc.dg/array-9.c produces internal compiler error on Darwin at -m64

[howarth at nitro dot med dot uc dot edu]

------- Comment #8 from howarth at nitro dot med dot uc dot edu  2006-09-14 02:18 -------
Andrew,
    I just noticed that the same code fragment exists in
rs6000_darwin64_record_arg_recurse
as well as in rs6000_darwin64_record_arg_advance_recurse. Shouldn't we then
make the patch...

Index: gcc/config/rs6000/rs6000.c
===================================================================
--- gcc/config/rs6000/rs6000.c  (revision 116941)
+++ gcc/config/rs6000/rs6000.c  (working copy)
@@ -4607,7 +4607,10 @@ rs6000_darwin64_record_arg_advance_recur
       {
        HOST_WIDE_INT bitpos = startbitpos;
        tree ftype = TREE_TYPE (f);
-       enum machine_mode mode = TYPE_MODE (ftype);
+       enum machine_mode mode;
+       if (ftype == error_mark_node)
+         continue;
+       mode = TYPE_MODE (ftype);

        if (DECL_SIZE (f) != 0
            && host_integerp (bit_position (f), 1))
@@ -4975,7 +4978,10 @@ rs6000_darwin64_record_arg_recurse (CUMU
       {
        HOST_WIDE_INT bitpos = startbitpos;
        tree ftype = TREE_TYPE (f);
-       enum machine_mode mode = TYPE_MODE (ftype);
+       enum machine_mode mode;
+       if (ftype == error_mark_node)
+         continue;
+       mode = TYPE_MODE (ftype);

        if (DECL_SIZE (f) != 0
            && host_integerp (bit_position (f), 1))

...for completeness?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29030

[Bug target/29030] gcc.dg/array-9.c produces internal compiler error on Darwin at -m64

[howarth at nitro dot med dot uc dot edu]

------- Comment #9 from howarth at nitro dot med dot uc dot edu  2006-09-15 01:32 -------
Andrew,
     To answer my own question, the following is from tonight's gcc make check
using the original
version of your patch which only impacted the
rs6000_darwin64_record_arg_advance_recurse...

Executing on host:
/sw/src/fink.build/gcc4-4.1.999-20060913/darwin_objdir/gcc/xgcc
-B/sw/src/fink.build/gcc4-4.1.999-20060913/darwin_objdir/gcc/
/sw/src/fink.build/gcc4-4.1.999-200
60913/gcc-4.2-20060913/gcc/testsuite/gcc.dg/array-9.c    -ansi -pedantic-errors
-fno-show-column -S  -m64 -o array-9.s    (timeout = 300)
/sw/src/fink.build/gcc4-4.1.999-20060913/gcc-4.2-20060913/gcc/testsuite/gcc.dg/array-9.c:7:
error: declaration of 'x' as array of voids
/sw/src/fink.build/gcc4-4.1.999-20060913/gcc-4.2-20060913/gcc/testsuite/gcc.dg/array-9.c:
In function 'foo':
/sw/src/fink.build/gcc4-4.1.999-20060913/gcc-4.2-20060913/gcc/testsuite/gcc.dg/array-9.c:10:
internal compiler error: tree check: expected class 'type', have 'exceptional'
(error_m
ark) in rs6000_darwin64_record_arg_recurse, at config/rs6000/rs6000.c:4981
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.
compiler exited with status 1
output is:
/sw/src/fink.build/gcc4-4.1.999-20060913/gcc-4.2-20060913/gcc/testsuite/gcc.dg/array-9.c:7:
error: declaration of 'x' as array of voids
/sw/src/fink.build/gcc4-4.1.999-20060913/gcc-4.2-20060913/gcc/testsuite/gcc.dg/array-9.c:
In function 'foo':
/sw/src/fink.build/gcc4-4.1.999-20060913/gcc-4.2-20060913/gcc/testsuite/gcc.dg/array-9.c:10:
internal compiler error: tree check: expected class 'type', have 'exceptional'
(error_m
ark) in rs6000_darwin64_record_arg_recurse, at config/rs6000/rs6000.c:4981
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.

FAIL: gcc.dg/array-9.c (internal compiler error)
PASS: gcc.dg/array-9.c  (test for errors, line 7)
FAIL: gcc.dg/array-9.c (test for excess errors)
Excess errors:
/sw/src/fink.build/gcc4-4.1.999-20060913/gcc-4.2-20060913/gcc/testsuite/gcc.dg/array-9.c:10:
internal compiler error: tree check: expected class 'type', have 'exceptional'
(error_m
ark) in rs6000_darwin64_record_arg_recurse, at config/rs6000/rs6000.c:4981

So I guess my expanded version of the patch is required. I'll build tonight's
svn pull with it and
do a make check.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29030

[Bug target/29030] gcc.dg/array-9.c produces internal compiler error on Darwin at -m64

[howarth at nitro dot med dot uc dot edu]

------- Comment #10 from howarth at nitro dot med dot uc dot edu  2006-09-15 02:40 -------
Created an attachment (id=12274)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=12274&action=view)
Patch to fix PR29030


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29030

[Bug target/29030] gcc.dg/array-9.c produces internal compiler error on Darwin at -m64

[howarth at nitro dot med dot uc dot edu]

------- Comment #11 from howarth at nitro dot med dot uc dot edu  2006-09-15 05:33 -------
Andrew,
    The PR29030.txt patch I posted eliminates the array-9 testsuite failures at
-m64...

Native configuration is powerpc-apple-darwin8

                === gcc tests ===

Schedule of variations:
    unix/-m32
    unix/-m64

Running target unix/-m32
Running
/sw/src/fink.build/gcc4-4.1.999-20060914/gcc-4.2-20060914/gcc/testsuite/gcc.dg/dg.exp
...
PASS: gcc.dg/array-9.c  (test for errors, line 7)
PASS: gcc.dg/array-9.c (test for excess errors)

                === gcc Summary for unix/-m32 ===

# of expected passes            2
Running target unix/-m64
Running
/sw/src/fink.build/gcc4-4.1.999-20060914/gcc-4.2-20060914/gcc/testsuite/gcc.dg/dg.exp
...
PASS: gcc.dg/array-9.c  (test for errors, line 7)
PASS: gcc.dg/array-9.c (test for excess errors)

                === gcc Summary for unix/-m64 ===

# of expected passes            2

                === gcc Summary ===

# of expected passes            4
/sw/src/fink.build/gcc4-4.1.999-20060914/darwin_objdir/gcc/xgcc  version 4.2.0
20060915 (experimental)

I'll run a complete make check but don't suspect any problems as the previous
partial patch
didn't cause any regressions.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29030

[Bug target/29030] gcc.dg/array-9.c produces internal compiler error on Darwin at -m64

[echristo at gcc dot gnu dot org]

------- Comment #12 from echristo at gcc dot gnu dot org  2006-09-16 09:01 -------
Subject: Bug 29030

Author: echristo
Date: Sat Sep 16 09:01:16 2006
New Revision: 116994

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=116994
Log:
2006-09-16   Andrew Pinski  <pinskia@physics.uc.edu>
             Jack Howarth  <howarth@bromo.med.uc.edu>

        PR target/29030
        * config/rs6000/rs6000.c (rs6000_darwin64_record_arg_advance_recurse)
        skip on error_mark_node.                                               
                   (rs6000_darwin64_record_arg_recurse): Likewise.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/config/rs6000/rs6000.c


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29030

[Bug target/29030] gcc.dg/array-9.c produces internal compiler error on Darwin at -m64

[pinskia at gcc dot gnu dot org]

------- Comment #13 from pinskia at gcc dot gnu dot org  2006-09-16 22:47 -------
Fixed.


-- 

pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |FIXED
   Target Milestone|---                         |4.2.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29030