[r11-6253 Regression] FAIL: gfortran.dg/unlimited_polymorphic_32.f03 -Os execution test on Linux/x86_64

[r11-6253 Regression] FAIL: gfortran.dg/unlimited_polymorphic_32.f03 -Os execution test on Linux/x86_64

[dhumieres.dominique]

Hi Paul,

I confirme the problem with -m32.
With the following changes:

% diff -up 
/opt/gcc/work/gcc/testsuite/gfortran.dg/unlimited_polymorphic_32.f03 
unlimited_polymorphic_32_db.f03
--- 
/opt/gcc/work/gcc/testsuite/gfortran.dg/unlimited_polymorphic_32.f03	2020-12-19 
12:39:07.000000000 +0100
+++ unlimited_polymorphic_32_db.f03	2020-12-20 13:52:36.000000000 +0100
@@ -17,8 +17,8 @@ contains
    subroutine driver
      class(*), allocatable :: obj
      type(Wrapper) w
-    integer(4) :: expected4(2) = [42_4, 43_4]
-    integer(8) :: expected8(3) = [42_8, 43_8, 44_8]
+    integer(4) :: expected4(3) = [42_4, 43_4, 44_4]
+    integer(8) :: expected8(5) = [42_8, 43_8, 44_8, 45_8, 46_8]

      w = new_wrapper (expected4)
      obj = w
@@ -42,18 +42,18 @@ contains
        type is (wrapper)
          select type (z => arg%elements)
            type is (integer(4))
-            if (any (z .ne. [42_4, 43_4])) stop 1 + idx
+            if (any (z .ne. [42_4, 43_4, 44_4])) stop 1 + idx
            type is (integer(8))
-            if (any (z .ne. [42_8, 43_8, 44_8])) stop 1 + idx
+            if (any (z .ne. [42_8, 43_8, 44_8, 45_8, 46_8])) print *, z
            type is (Mytype)
-            if (abs (z(1)%r - 99.0) .ge. 1e-6) stop 1 + idx
+            if (abs (z(1)%r - 99.0) .ge. 1e-6) stop 3 + idx
          class default
            stop 2 + idx
          end select
        type is (Mytype)
-        if (abs (arg%r - 42.0) .ge. 1e-6) stop 1 + idx
+        if (abs (arg%r - 42.0) .ge. 1e-6) stop 4 + idx
        class default
-        stop 3 + idx
+        stop 5 + idx
      end select
    end subroutine
  end

I get

% gfc unlimited_polymorphic_32_db.f03 -m32
% ./a.out
                    42                   43                   44          
          45                   42

Re: [r11-6253 Regression] FAIL: gfortran.dg/unlimited_polymorphic_32.f03 -Os execution test on Linux/x86_64

[Paul Richard Thomas]

Hi Dominique,

I find that with -m32, no matter how I modify it, that I get the error for
the integer(8) carrying wrapper of:
a.out: malloc.c:2394: sysmalloc: Assertion `(old_top == initial_top (av) &&
old_size == 0) || ((unsigned long) (old_size) >= MINSIZE && prev_inuse
(old_top) && ((unsigned long) old_end & (pagesize - 1)) == 0)' failed.

I'll probably set it to { dg-do compile } while I figure out what the
problem is.

Thanks

Paul


On Sun, 20 Dec 2020 at 13:00, <dhumieres.dominique@free.fr> wrote:

> Hi Paul,
>
> I confirme the problem with -m32.
> With the following changes:
>
> % diff -up
> /opt/gcc/work/gcc/testsuite/gfortran.dg/unlimited_polymorphic_32.f03
> unlimited_polymorphic_32_db.f03
> ---
> /opt/gcc/work/gcc/testsuite/gfortran.dg/unlimited_polymorphic_32.f03
> 2020-12-19
> 12:39:07.000000000 +0100
> +++ unlimited_polymorphic_32_db.f03     2020-12-20 13:52:36.000000000 +0100
> @@ -17,8 +17,8 @@ contains
>     subroutine driver
>       class(*), allocatable :: obj
>       type(Wrapper) w
> -    integer(4) :: expected4(2) = [42_4, 43_4]
> -    integer(8) :: expected8(3) = [42_8, 43_8, 44_8]
> +    integer(4) :: expected4(3) = [42_4, 43_4, 44_4]
> +    integer(8) :: expected8(5) = [42_8, 43_8, 44_8, 45_8, 46_8]
>
>       w = new_wrapper (expected4)
>       obj = w
> @@ -42,18 +42,18 @@ contains
>         type is (wrapper)
>           select type (z => arg%elements)
>             type is (integer(4))
> -            if (any (z .ne. [42_4, 43_4])) stop 1 + idx
> +            if (any (z .ne. [42_4, 43_4, 44_4])) stop 1 + idx
>             type is (integer(8))
> -            if (any (z .ne. [42_8, 43_8, 44_8])) stop 1 + idx
> +            if (any (z .ne. [42_8, 43_8, 44_8, 45_8, 46_8])) print *, z
>             type is (Mytype)
> -            if (abs (z(1)%r - 99.0) .ge. 1e-6) stop 1 + idx
> +            if (abs (z(1)%r - 99.0) .ge. 1e-6) stop 3 + idx
>           class default
>             stop 2 + idx
>           end select
>         type is (Mytype)
> -        if (abs (arg%r - 42.0) .ge. 1e-6) stop 1 + idx
> +        if (abs (arg%r - 42.0) .ge. 1e-6) stop 4 + idx
>         class default
> -        stop 3 + idx
> +        stop 5 + idx
>       end select
>     end subroutine
>   end
>
> I get
>
> % gfc unlimited_polymorphic_32_db.f03 -m32
> % ./a.out
>                     42                   43                   44
>           45                   42
>


-- 
"If you can't explain it simply, you don't understand it well enough" -
Albert Einstein

Re: [r11-6253 Regression] FAIL: gfortran.dg/unlimited_polymorphic_32.f03 -Os execution test on Linux/x86_64

[Paul Richard Thomas]

Hi Sunil,

Thanks for the heads up. I am just rebuilding with multilib enabled. I'll
investigate later on.

Regards

Paul


---------- Forwarded message ---------
From: sunil.k.pandey <skpandey@sc.intel.com>
Date: Fri, 18 Dec 2020 at 18:16
Subject: [r11-6253 Regression] FAIL:
gfortran.dg/unlimited_polymorphic_32.f03 -Os execution test on Linux/x86_64
To: <gcc-patches@gcc.gnu.org>, <gcc-regression@gcc.gnu.org>, <
pault@gcc.gnu.org>


On Linux/x86_64,

ce8dcc9105cbd4043d575d8b2c91309a423951a9 is the first bad commit
commit ce8dcc9105cbd4043d575d8b2c91309a423951a9
Author: Paul Thomas <pault@gcc.gnu.org>
Date:   Fri Dec 18 14:00:11 2020 +0000

    As well as the PR this patch fixes problems in handling class objects

caused

FAIL: gfortran.dg/unlimited_polymorphic_32.f03   -O0  execution test
FAIL: gfortran.dg/unlimited_polymorphic_32.f03   -O1  execution test
FAIL: gfortran.dg/unlimited_polymorphic_32.f03   -O2  execution test
FAIL: gfortran.dg/unlimited_polymorphic_32.f03   -O3 -fomit-frame-pointer
-funroll-loops -fpeel-loops -ftracer -finline-functions  execution test
FAIL: gfortran.dg/unlimited_polymorphic_32.f03   -O3 -g  execution test
FAIL: gfortran.dg/unlimited_polymorphic_32.f03   -Os  execution test

with GCC configured with

../../gcc/configure
--prefix=/local/skpandey/gccwork/toolwork/gcc-bisect-master/master/r11-6253/usr
--enable-clocale=gnu --with-system-zlib --with-demangler-in-ld
--with-fpmath=sse --enable-languages=c,c++,fortran --enable-cet
--without-isl --enable-libmpx x86_64-linux --disable-bootstrap

To reproduce:

$ cd {build_dir}/gcc && make check
RUNTESTFLAGS="dg.exp=gfortran.dg/unlimited_polymorphic_32.f03
--target_board='unix{-m32}'"
$ cd {build_dir}/gcc && make check
RUNTESTFLAGS="dg.exp=gfortran.dg/unlimited_polymorphic_32.f03
--target_board='unix{-m32\ -march=cascadelake}'"

(Please do not reply to this email, for question about this report, contact
me at skpgkp2 at gmail dot com)


-- 
"If you can't explain it simply, you don't understand it well enough" -
Albert Einstein