[Bug ada/103405] New: [12 Regression] c67005c FAILs

[Bug ada/103405] New: [12 Regression] c67005c FAILs

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103405

            Bug ID: 103405
           Summary: [12 Regression] c67005c FAILs
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: ada
          Assignee: unassigned at gcc dot gnu.org
          Reporter: rguenth at gcc dot gnu.org
  Target Milestone: ---

I see

                === acats tests ===
FAIL:   c67005c

                === acats Summary ===
# of expected passes            2327
# of unexpected failures        1
Native configuration is x86_64-pc-linux-gnu

it's reported to be failing since Nov 18th, also on powerpc64le

[Bug ada/103405] [12 Regression] c67005c FAILs

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103405

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Target|                            |x86_64-*-*,
                   |                            |powerpc64le-linux
   Target Milestone|---                         |12.0

[Bug ada/103405] [12 Regression] c67005c FAILs

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103405

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
           Keywords|                            |wrong-code
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2021-11-24

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
for me it started between r12-5291-g3200de91bc70bfd6d and
r12-5390-gd3152981f71eef1.

[Bug ada/103405] [12 Regression] c67005c FAILs

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103405

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jakub at gcc dot gnu.org

--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
For me it is r12-5347-g2c2148d8c144d7388abcb7c34b782be647fe81c9 known good and
r12-5385-g6f4ac4f81f89caac7e74127ed2e6db6bbb3d7426 known bad.

[Bug ipa/103405] [12 Regression] c67005c FAILs

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103405

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|ada                         |ipa
                 CC|                            |marxin at gcc dot gnu.org

--- Comment #3 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
-O2 fails but -O2 -fno-ipa-modref pass.

[Bug ipa/103405] [12 Regression] c67005c FAILs

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103405

--- Comment #4 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
DSE removes a few stores even though they are used in the check.

  _18 = .gnat_malloc (4);
  MEM[(integer *)_18] = 1; ;; this one
  FRAME.2.i = _18;
  _17 = .gnat_malloc (4);
  MEM[(integer *)_17] = 0; ;; this one
  FRAME.2.j = _17;
  c67005c.B_1.check (0); [static-chain: &FRAME.2]
  _16 = .gnat_malloc (4);
  MEM[(integer *)_16] = 0; ;; this one
  FRAME.2.i = _16;
  c67005c.B_1.check (1); [static-chain: &FRAME.2]
  report.result ();

[Bug ipa/103405] [12 Regression] c67005c FAILs

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103405

--- Comment #5 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
C testcase:
typedef __SIZE_TYPE__ size_t;
void *malloc(size_t);

static inline int *starter(int a)
{
    int *b = malloc(sizeof(int));
    *b = a;
    return b;
}

static inline _Bool equal(int *l, int *r)
{
    if (l == 0)
      __builtin_abort();
    if (r == 0)
      __builtin_abort();
    return *r == *l;
}

int main(void)
{
  int *i;
  int *j;
  void check(_Bool a)
  {
    _Bool t = equal(i, j);
    if (a && t) __builtin_abort ();
    _Bool t1 = equal(i, j);
    if (!a && !t1) __builtin_abort ();
  }
  i = starter(1);
  j = starter(0);
  check(1);
  i = starter(0);
  check(0);

}

[Bug ipa/103405] [12 Regression] c67005c FAILs

[hubicka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103405

Jan Hubicka <hubicka at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|unassigned at gcc dot gnu.org      |hubicka at gcc dot gnu.org
             Status|NEW                         |ASSIGNED

--- Comment #6 from Jan Hubicka <hubicka at gcc dot gnu.org> ---
mine.

[Bug ipa/103405] [12 Regression] c67005c FAILs with -fipa-modref

[ebotcazou at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103405

Eric Botcazou <ebotcazou at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |ebotcazou at gcc dot gnu.org
            Summary|[12 Regression] c67005c     |[12 Regression] c67005c
                   |FAILs                       |FAILs with -fipa-modref

--- Comment #7 from Eric Botcazou <ebotcazou at gcc dot gnu.org> ---
Yep, so far it's the only Ada issue I know of with -fipa-modref.

[Bug ipa/103405] [12 Regression] c67005c FAILs with -fipa-modref

[hubicka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103405

--- Comment #8 from Jan Hubicka <hubicka at gcc dot gnu.org> ---
Created attachment 51868
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=51868&action=edit
Patch in testing

Good to know that there are no more known modref & Ada problems ;)  I recently
extended modref to also handle static chains and some problem may pop out...

The problem is the IPA pass determining incorrect EAF flags for the static
chain (it get NO_INDIRECT_READ which is not true since equal reads indirectly).
This is missed propagation of EAF flags for static chain - I missed this spot
while extending retslot handling to static chain since retslots does not need
this.

[Bug ipa/103405] [12 Regression] c67005c FAILs with -fipa-modref

[hubicka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103405

--- Comment #9 from Jan Hubicka <hubicka at gcc dot gnu.org> ---
Fixed by g:16e85390507ea92331c9052393b591202007f5ab (forgot to add PR marker)

[Bug ipa/103405] [12 Regression] c67005c FAILs with -fipa-modref

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103405

Martin Liška <marxin at gcc dot gnu.org> changed:

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

--- Comment #10 from Martin Liška <marxin at gcc dot gnu.org> ---
Fixed.