[Bug tree-optimization/56347] New: [4.8 Regression] FAIL: gfortran.dg/integer_exponentiation_2.f90 -O2 execution test

[Bug tree-optimization/56347] New: [4.8 Regression] FAIL: gfortran.dg/integer_exponentiation_2.f90 -O2 execution test

[danglin at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56347

             Bug #: 56347
           Summary: [4.8 Regression] FAIL:
                    gfortran.dg/integer_exponentiation_2.f90  -O2
                    execution test
    Classification: Unclassified
           Product: gcc
           Version: 4.8.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: danglin@gcc.gnu.org
              Host: hppa2.0w-hp-hpux11.11
            Target: hppa2.0w-hp-hpux11.11
             Build: hppa2.0w-hp-hpux11.11


Executing on host: /test/gnu/gcc/objdir/gcc/testsuite/gfortran/../../gfortran
-B
/test/gnu/gcc/objdir/gcc/testsuite/gfortran/../../
-B/test/gnu/gcc/objdir/hppa2.
0w-hp-hpux11.11/./libgfortran/
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integ
er_exponentiation_2.f90  -fno-diagnostics-show-caret   -O2   
-B/test/gnu/gcc/ob
jdir/hppa2.0w-hp-hpux11.11/./libgfortran/.libs
-L/test/gnu/gcc/objdir/hppa2.0w-h
p-hpux11.11/./libgfortran/.libs
-L/test/gnu/gcc/objdir/hppa2.0w-hp-hpux11.11/./l
ibgfortran/.libs
-B/test/gnu/gcc/objdir/hppa2.0w-hp-hpux11.11/./libquadmath/.lib
s -L/test/gnu/gcc/objdir/hppa2.0w-hp-hpux11.11/./libquadmath/.libs
-L/test/gnu/g
cc/objdir/hppa2.0w-hp-hpux11.11/./libquadmath/.libs  -lm   -o
./integer_exponent
iation_2.exe    (timeout = 300)
spawn /test/gnu/gcc/objdir/gcc/testsuite/gfortran/../../gfortran
-B/test/gnu/gcc
/objdir/gcc/testsuite/gfortran/../../
-B/test/gnu/gcc/objdir/hppa2.0w-hp-hpux11.
11/./libgfortran/
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentia
tion_2.f90 -fno-diagnostics-show-caret -O2
-B/test/gnu/gcc/objdir/hppa2.0w-hp-hp
ux11.11/./libgfortran/.libs
-L/test/gnu/gcc/objdir/hppa2.0w-hp-hpux11.11/./libgfortran/.libs
-L/test/gnu/gcc/objdir/hppa2.0w-hp-hpux11.11/./libgfortran/.libs
-B/test/gnu/gcc/objdir/hppa2.0w-hp-hpux11.11/./libquadmath/.libs
-L/test/gnu/gcc/objdir/hppa2.0w-hp-hpux11.11/./libquadmath/.libs
-L/test/gnu/gcc/objdir/hppa2.0w-hp-hpux11.11/./libquadmath/.libs -lm -o
./integer_exponentiation_2.exe
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:142.13:

  call gee_i(i**0_8) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:143.13:

  call gee_i(i**1_8) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:144.13:

  call gee_i(i**2_8) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:145.13:

  call gee_i(i**3_8) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:146.13:

  call gee_i(i**(-1_8)) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:147.13:

  call gee_i(i**(-2_8)) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:148.13:

  call gee_i(i**(-3_8)) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:149.13:

  call gee_i(i**huge(0_8)) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:150.13:

  call gee_i(i**(-huge(0_8))) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:151.13:

  call gee_i(i**(-huge(0_8)-1_8)) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
output is:
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:142.13:

  call gee_i(i**0_8) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:143.13:

  call gee_i(i**1_8) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:144.13:

  call gee_i(i**2_8) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:145.13:

  call gee_i(i**3_8) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:146.13:

  call gee_i(i**(-1_8)) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:147.13:

  call gee_i(i**(-2_8)) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:148.13:

  call gee_i(i**(-3_8)) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:149.13:

  call gee_i(i**huge(0_8)) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:150.13:

  call gee_i(i**(-huge(0_8))) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)
/test/gnu/gcc/gcc/gcc/testsuite/gfortran.dg/integer_exponentiation_2.f90:151.13:

  call gee_i(i**(-huge(0_8)-1_8)) ! { dg-warning "Type mismatch in argument" }
             1
Warning: Type mismatch in argument 'i' at (1); passed INTEGER(8) to INTEGER(4)

PASS: gfortran.dg/integer_exponentiation_2.f90  -O2   (test for warnings, line
142)
PASS: gfortran.dg/integer_exponentiation_2.f90  -O2   (test for warnings, line
143)
PASS: gfortran.dg/integer_exponentiation_2.f90  -O2   (test for warnings, line
144)
PASS: gfortran.dg/integer_exponentiation_2.f90  -O2   (test for warnings, line
145)
PASS: gfortran.dg/integer_exponentiation_2.f90  -O2   (test for warnings, line
146)
PASS: gfortran.dg/integer_exponentiation_2.f90  -O2   (test for warnings, line
147)
PASS: gfortran.dg/integer_exponentiation_2.f90  -O2   (test for warnings, line
148)
PASS: gfortran.dg/integer_exponentiation_2.f90  -O2   (test for warnings, line
149)
PASS: gfortran.dg/integer_exponentiation_2.f90  -O2   (test for warnings, line
150)
PASS: gfortran.dg/integer_exponentiation_2.f90  -O2   (test for warnings, line
151)
PASS: gfortran.dg/integer_exponentiation_2.f90  -O2  (test for excess errors)
Setting LD_LIBRARY_PATH to
.:/test/gnu/gcc/objdir/hppa2.0w-hp-hpux11.11/./libgfortran/.libs:/test/gnu/gcc/objdir/hppa2.0w-hp-hpux11.11/./libgfortran/.libs:/test/gnu/gcc/objdir/hppa2.0w-hp-hpux11.11/./libquadmath/.libs:/test/gnu/gcc/objdir/hppa2.0w-hp-hpux11.11/./libquadmath/.libs:/test/gnu/gcc/objdir/gcc:.:/test/gnu/gcc/objdir/hppa2.0w-hp-hpux11.11/./libgfortran/.libs:/test/gnu/gcc/objdir/hppa2.0w-hp-hpux11.11/./libgfortran/.libs:/test/gnu/gcc/objdir/hppa2.0w-hp-hpux11.11/./libquadmath/.libs:/test/gnu/gcc/objdir/hppa2.0w-hp-hpux11.11/./libquadmath/.libs:/test/gnu/gcc/objdir/gcc
spawn [open ...]

Program aborted. Backtrace:
#0  0xC1239FEF
FAIL: gfortran.dg/integer_exponentiation_2.f90  -O2  execution test

Component is just a guess.

Test appears to fail on this test:

36        call check_equal_r (a**(-2.0), (1/a)*(1/a))
(gdb)
85          if (abs(a - b) > 1.e-5 * abs(b)) call abort
(gdb)
95          if (a /= b) call abort
(gdb)
0x00002e08 in _gfortran_abort ()
    at ../../../gcc/libgfortran/intrinsics/abort.c:32
32      {

Introduced between 195125 and 195497.

[Bug tree-optimization/56347] [4.8 Regression] FAIL: gfortran.dg/integer_exponentiation_2.f90 -O2 execution test

[danglin at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56347

--- Comment #1 from John David Anglin <danglin at gcc dot gnu.org> 2013-02-15 18:35:58 UTC ---
Wrong assembly code is here:

   0x000030f4 <run_me_+396>:    fsub,sgl fr17,fr17,fr22
   0x000030f8 <run_me_+400>:    fabs,sgl fr22,fr22
   0x000030fc <run_me_+404>:    fcmp,sgl,!< fr13R,fr22
   0x00003100 <run_me_+408>:    ftest
   0x00003104 <run_me_+412>:    b,l,n 0x37f0 <run_me_+2184>,r0

This is with a slightly reduced testcase.  Bug disappears when I tried to
reduce run_me.

[Bug tree-optimization/56347] [4.8 Regression] FAIL: gfortran.dg/integer_exponentiation_2.f90 -O2 execution test

[danglin at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56347

--- Comment #2 from John David Anglin <danglin at gcc dot gnu.org> 2013-02-15 18:37:58 UTC ---
Created attachment 29471
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=29471
Slightly reduced testcase

[Bug tree-optimization/56347] [4.8 Regression] FAIL: gfortran.dg/integer_exponentiation_2.f90 -O2 execution test

[danglin at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56347

--- Comment #3 from John David Anglin <danglin at gcc dot gnu.org> 2013-02-15 19:07:29 UTC ---
In final:

(note 135 134 137 [bb 11] NOTE_INSN_BASIC_BLOCK)
(insn:TI 137 135 138 (set (reg:SF 68 %fr22 [orig:303 D.1185 ] [303])
        (minus:SF (reg:SF 58 %fr17 [orig:107 D.1185 ] [107])
            (reg:SF 58 %fr17 [orig:107 D.1185 ] [107])))
integer_exponentiation_2.f90:83 140 {subsf3}
     (nil))
(insn:TI 138 137 139 (set (reg:SF 68 %fr22 [orig:304 D.1185 ] [304])
        (abs:SF (reg:SF 68 %fr22 [orig:303 D.1185 ] [303])))
integer_exponentiation_2.f90:83 150 {abssf2}
     (nil))

In expand:

;; Generating RTL for gimple basic block 12

;; if (_131 < _143)

(insn 136 135 137 (set (reg:SF 302 [ D.1185 ])
        (mult:SF (reg:SF 104 [ D.1185 ])
            (reg:SF 200 [ D.1185 ]))) integer_exponentiation_2.f90:39 -1
     (nil))

(insn 137 136 138 (set (reg:SF 303 [ D.1185 ])
        (minus:SF (reg:SF 302 [ D.1185 ])
            (reg:SF 107 [ D.1185 ]))) integer_exponentiation_2.f90:83 -1
     (nil))

(insn 138 137 139 (set (reg:SF 304 [ D.1185 ])
        (abs:SF (reg:SF 303 [ D.1185 ]))) integer_exponentiation_2.f90:83 -1
     (nil))

The tree code is:

  _131 = _130 * 9.99999974737875163555145263671875e-6;
  ...

;;   basic block 12, loop depth 0
;;    pred:       11
  powmult_43 = _12 * powmult_177;
  _142 = powmult_43 - _17;
  _143 = ABS_EXPR <_142>;
  if (_131 < _143)
    goto <bb 3>;
  else
    goto <bb 13>;

[Bug tree-optimization/56347] [4.8 Regression] FAIL: gfortran.dg/integer_exponentiation_2.f90 -O2 execution test

[danglin at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56347

--- Comment #4 from John David Anglin <danglin at gcc dot gnu.org> 2013-02-15 20:07:47 UTC ---
> Wrong assembly code is here:
> 
>    0x000030f4 <run_me_+396>:    fsub,sgl fr17,fr17,fr22
>    0x000030f8 <run_me_+400>:    fabs,sgl fr22,fr22
>    0x000030fc <run_me_+404>:    fcmp,sgl,!< fr13R,fr22
>    0x00003100 <run_me_+408>:    ftest
>    0x00003104 <run_me_+412>:    b,l,n 0x37f0 <run_me_+2184

Actually test fails at next test:

(gdb) disass $pc-16,$pc+24
Dump of assembler code from 0x30ec to 0x3114:
   0x000030ec <run_me_+388>:    ftest
   0x000030f0 <run_me_+392>:    b,l,n 0x37f0 <run_me_+2184>,r0
   0x000030f4 <run_me_+396>:    fsub,sgl fr17,fr17,fr22
   0x000030f8 <run_me_+400>:    fabs,sgl fr22,fr22
=> 0x000030fc <run_me_+404>:    fcmp,sgl,!< fr13R,fr22
   0x00003100 <run_me_+408>:    ftest
   0x00003104 <run_me_+412>:    b,l,n 0x37f0 <run_me_+2184>,r0
   0x00003108 <run_me_+416>:    fcmp,sgl,!> fr12R,fr15
   0x0000310c <run_me_+420>:    ftest
   0x00003110 <run_me_+424>:    b,l,n 0x37f0 <run_me_+2184>,r0
End of assembler dump.
(gdb) p $fr12R
$3 = 2
(gdb) p $fr15
$4 = 9.09090886e-06

Ignore #3.

[Bug target/56347] [4.8 Regression] FAIL: gfortran.dg/integer_exponentiation_2.f90 -O2 execution test

[danglin at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56347

John David Anglin <danglin at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|tree-optimization           |target

--- Comment #5 from John David Anglin <danglin at gcc dot gnu.org> 2013-02-15 21:44:34 UTC ---
Floating register $fr12R is clobbered by call to powf.  Problem is similar
to PR fortran/31832.

[Bug target/56347] [4.8 Regression] FAIL: gfortran.dg/integer_exponentiation_2.f90 -O2 execution test

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56347

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.8.0

[Bug target/56347] [4.8 Regression] FAIL: gfortran.dg/integer_exponentiation_2.f90 -O2 execution test

[danglin at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56347

--- Comment #6 from John David Anglin <danglin at gcc dot gnu.org> 2013-02-18 16:29:56 UTC ---
Author: danglin
Date: Mon Feb 18 16:29:49 2013
New Revision: 196123

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=196123
Log:
    PR target/56347
    * config/pa/pa.c (pa_conditional_register_usage): On HP-UX, mark
    registers %fr12 and %fr12R as call used.


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

[Bug target/56347] [4.8 Regression] FAIL: gfortran.dg/integer_exponentiation_2.f90 -O2 execution test

[danglin at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56347

--- Comment #7 from John David Anglin <danglin at gcc dot gnu.org> 2013-02-18 18:04:23 UTC ---
Patch to remove old work around is here:
http://gcc.gnu.org/ml/gcc-patches/2013-02/msg00854.html

[Bug target/56347] [4.8 Regression] FAIL: gfortran.dg/integer_exponentiation_2.f90 -O2 execution test

[danglin at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56347

--- Comment #8 from John David Anglin <danglin at gcc dot gnu.org> 2013-02-19 14:01:51 UTC ---
Author: danglin
Date: Tue Feb 19 14:01:40 2013
New Revision: 196139

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=196139
Log:
    PR target/56347
    * acinclude.m4 (LIBGFOR_CHECK_FOR_BROKEN_POWF): Remove check for
    broken powf.
    * configure.ac (LIBGFOR_CHECK_FOR_BROKEN_POWF): Likewise.
    * intrinsics/c99_functions.c: Likewise.
    * configure: Rebuilt.
    * config.h.in: Rebuilt.


Modified:
    trunk/libgfortran/ChangeLog
    trunk/libgfortran/acinclude.m4
    trunk/libgfortran/config.h.in
    trunk/libgfortran/configure
    trunk/libgfortran/configure.ac
    trunk/libgfortran/intrinsics/c99_functions.c

[Bug target/56347] [4.8 Regression] FAIL: gfortran.dg/integer_exponentiation_2.f90 -O2 execution test

[danglin at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56347

John David Anglin <danglin at gcc dot gnu.org> changed:

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

--- Comment #9 from John David Anglin <danglin at gcc dot gnu.org> 2013-02-19 14:08:47 UTC ---
Fixed.

[Bug target/56347] [4.8 Regression] FAIL: gfortran.dg/integer_exponentiation_2.f90 -O2 execution test

[danglin at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56347

--- Comment #10 from John David Anglin <danglin at gcc dot gnu.org> 2013-03-11 00:44:33 UTC ---
Author: danglin
Date: Mon Mar 11 00:44:28 2013
New Revision: 196588

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=196588
Log:
    PR target/56347
    * config/pa/pa.md (call_value): Check for calls to powf and direct to
    new call patterns that clobber %fr12.
    (call_val_powf, call_val_powf_pic, call_val_powf_64bit): New insn,
    split and postreload patterns.
    * config/pa/pa.c (pa_conditional_register_usage): Revert marking
    registers %fr12 and %fr12R as call used.


Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/config/pa/pa.c
    trunk/gcc/config/pa/pa.md