[Bug middle-end/94774] New: Uninitialized variable retval in function try_substitute_return_value

[Bug middle-end/94774] New: Uninitialized variable retval in function try_substitute_return_value

[stefansf at linux dot ibm.com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94774

            Bug ID: 94774
           Summary: Uninitialized variable retval in function
                    try_substitute_return_value
           Product: gcc
           Version: 10.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: middle-end
          Assignee: unassigned at gcc dot gnu.org
          Reporter: stefansf at linux dot ibm.com
  Target Milestone: ---

Created attachment 48377
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=48377&action=edit
If SAFE then retval was initialised

While bootstrapping GCC on S/390 the following warning/error gets thrown:

gcc/gimple-ssa-sprintf.c: In function 'bool
handle_printf_call(gimple_stmt_iterator*, const vr_values*)':                   
gcc/gimple-ssa-sprintf.c:4184:8: error: '*((void*)& retval +8)' may be used
uninitialized in this function [-Werror=maybe-uninitialized]
 4183 |    && (retval[0] < target_int_max ()
      |       ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 4184 |        && retval[1] < target_int_max ()))
      |        ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
gcc/gimple-ssa-sprintf.c:4123:26: note: '*((void*)& retval +8)' was declared
here
 4123 |   unsigned HOST_WIDE_INT retval[2];
      |                          ^~~~~~
gcc/gimple-ssa-sprintf.c:4214:14: error: 'retval' may be used uninitialized in
this function [-Werror=maybe-uninitialized]
 4214 |      fprintf (dump_file, "  %s %s-bounds return value "
      |      ~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 4215 |        HOST_WIDE_INT_PRINT_UNSIGNED ".\n",
      |        ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 4216 |        what, inbounds, retval[0]);
      |        ~~~~~~~~~~~~~~~~~~~~~~~~~~
gcc/gimple-ssa-sprintf.c:4123:26: note: 'retval' was declared here
 4123 |   unsigned HOST_WIDE_INT retval[2];
      |                          ^~~~~~

The attached patch fixes this for me. Otherwise, one could also just initialize
array retval which may be more appropriate?

[Bug tree-optimization/94774] Uninitialized variable retval in function try_substitute_return_value

[msebor at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94774

--- Comment #1 from Martin Sebor <msebor at gcc dot gnu.org> ---
It looks to me like only the second warning might be a true positive.  The
first one seems spurious since the uninitialized access is guarded by the test
for safe being true.

Moving the guard up would suppress the dump output in the "unsafe" case so I
don't think that's what we want.  OTOH, ether initializing the array, or
unconditionally assigning to it in is_call_safe, seems benign.

[Bug tree-optimization/94774] Uninitialized variable retval in function try_substitute_return_value

[stefansf at linux dot ibm.com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94774

--- Comment #2 from Stefan Schulze Frielinghaus <stefansf at linux dot ibm.com> ---
(In reply to Martin Sebor from comment #1)
> Moving the guard up would suppress the dump output in the "unsafe" case so I
> don't think that's what we want.  OTOH, ether initializing the array, or
> unconditionally assigning to it in is_call_safe, seems benign.

Since is_call_safe is called with a dummy argument in try_simplify_call I went
for initialising the array.  I will come up with a patch and post it to the
mailinglist.  Thanks for your help!

[Bug tree-optimization/94774] Uninitialized variable retval in function try_substitute_return_value

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94774

--- Comment #3 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Stefan Schulze Frielinghaus
<stefansf@gcc.gnu.org>:

https://gcc.gnu.org/g:1657178f59bf0c5b269a668bcdcc432fac3bbdd0

commit r10-8050-g1657178f59bf0c5b269a668bcdcc432fac3bbdd0
Author: Stefan Schulze Frielinghaus <stefansf@linux.ibm.com>
Date:   Mon Apr 27 18:09:07 2020 +0200

    tree-optimization: Fix use of uninitialized variable [PR94774]

    Array retval is not necessarily initialized by function is_call_safe and
    may be used afterwards.  Thus, initialize it explicitly.

    gcc/ChangeLog:

    2020-04-29  Stefan Schulze Frielinghaus  <stefansf@linux.ibm.com>

            PR tree-optimization/94774
            * gimple-ssa-sprintf.c (try_substitute_return_value): Initialize
            variable retval.

[Bug tree-optimization/94774] Uninitialized variable retval in function try_substitute_return_value

[manu at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94774

Manuel López-Ibáñez <manu at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|---                         |FIXED
                 CC|                            |manu at gcc dot gnu.org

--- Comment #4 from Manuel López-Ibáñez <manu at gcc dot gnu.org> ---
Seems fixed.