public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug c/109528] New: Calling exit() at the end of function triggers infinite recursion warning
@ 2023-04-15 17:12 stavros9899 at gmail dot com
  2023-04-15 19:10 ` [Bug c/109528] " pinskia at gcc dot gnu.org
                   ` (3 more replies)
  0 siblings, 4 replies; 5+ messages in thread
From: stavros9899 at gmail dot com @ 2023-04-15 17:12 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109528

            Bug ID: 109528
           Summary: Calling exit() at the end of function triggers
                    infinite recursion warning
           Product: gcc
           Version: 12.2.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
          Assignee: unassigned at gcc dot gnu.org
          Reporter: stavros9899 at gmail dot com
  Target Milestone: ---

Created attachment 54868
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=54868&action=edit
code that triggers the warning

Hello, this codem falsy -imo- triggers infinite-recursion warning. The
interesting thing is that the error goes away when the last exit(), at the near
end of the function gets removed.

^ permalink raw reply	[flat|nested] 5+ messages in thread

* [Bug c/109528] Calling exit() at the end of function triggers infinite recursion warning
  2023-04-15 17:12 [Bug c/109528] New: Calling exit() at the end of function triggers infinite recursion warning stavros9899 at gmail dot com
@ 2023-04-15 19:10 ` pinskia at gcc dot gnu.org
  2023-04-15 19:15 ` [Bug c/109528] function with _Noreturn triggers infinite recursion warning while using the attribute does not pinskia at gcc dot gnu.org
                   ` (2 subsequent siblings)
  3 siblings, 0 replies; 5+ messages in thread
From: pinskia at gcc dot gnu.org @ 2023-04-15 19:10 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109528

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|analyzer                    |c
           Keywords|                            |diagnostic

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---

Actually it looks more like _Noreturn is not take effect on the declaration.
If I use __attribute__((noreturn)) or [[gnu::noreturn]] (or on the trunk
[[noreturn]] which is part of C23) instead, it just works.

^ permalink raw reply	[flat|nested] 5+ messages in thread

* [Bug c/109528] function with _Noreturn triggers infinite recursion warning while using the attribute does not
  2023-04-15 17:12 [Bug c/109528] New: Calling exit() at the end of function triggers infinite recursion warning stavros9899 at gmail dot com
  2023-04-15 19:10 ` [Bug c/109528] " pinskia at gcc dot gnu.org
@ 2023-04-15 19:15 ` pinskia at gcc dot gnu.org
  2023-04-15 19:19 ` [Bug middle-end/109528] " pinskia at gcc dot gnu.org
  2023-04-16  5:21 ` stavros9899 at gmail dot com
  3 siblings, 0 replies; 5+ messages in thread
From: pinskia at gcc dot gnu.org @ 2023-04-15 19:15 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109528

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
     Ever confirmed|0                           |1
            Summary|Calling exit() at the end   |function with _Noreturn
                   |of function triggers        |triggers infinite recursion
                   |infinite recursion warning  |warning while using the
                   |                            |attribute does not
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2023-04-15

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Reduced testcase:
```
#if 1
#define noreturn _Noreturn
#else
#define noreturn [[gnu::noreturn]]
#endif

struct tree_node {
        unsigned          nr_children;
        char              name[16];
        struct tree_node  *children;
};

noreturn void fork_my_tree(const struct tree_node* pRoot)
{
  for (unsigned i = 0u; i < pRoot->nr_children; i++)
    fork_my_tree(pRoot->children + i);

  __builtin_exit(0);
}
```
Changing `#if 1` to `#if 0` fixes the warning.

^ permalink raw reply	[flat|nested] 5+ messages in thread

* [Bug middle-end/109528] function with _Noreturn triggers infinite recursion warning while using the attribute does not
  2023-04-15 17:12 [Bug c/109528] New: Calling exit() at the end of function triggers infinite recursion warning stavros9899 at gmail dot com
  2023-04-15 19:10 ` [Bug c/109528] " pinskia at gcc dot gnu.org
  2023-04-15 19:15 ` [Bug c/109528] function with _Noreturn triggers infinite recursion warning while using the attribute does not pinskia at gcc dot gnu.org
@ 2023-04-15 19:19 ` pinskia at gcc dot gnu.org
  2023-04-16  5:21 ` stavros9899 at gmail dot com
  3 siblings, 0 replies; 5+ messages in thread
From: pinskia at gcc dot gnu.org @ 2023-04-15 19:19 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109528

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|c                           |middle-end

--- Comment #3 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
  /* Avoid diagnosing an apparently infinitely recursive function that
     doesn't return where the infinite recursion might be avoided by
     a call to another noreturn function.  */
  noreturn_p = lookup_attribute ("noreturn", DECL_ATTRIBUTES (m_func->decl));


This is the wrong check. It should use "flags_from_decl_or_type (m_func->decl)
& ECF_NOTHROW" instead or even just TREE_THIS_VOLATILE here.

^ permalink raw reply	[flat|nested] 5+ messages in thread

* [Bug middle-end/109528] function with _Noreturn triggers infinite recursion warning while using the attribute does not
  2023-04-15 17:12 [Bug c/109528] New: Calling exit() at the end of function triggers infinite recursion warning stavros9899 at gmail dot com
                   ` (2 preceding siblings ...)
  2023-04-15 19:19 ` [Bug middle-end/109528] " pinskia at gcc dot gnu.org
@ 2023-04-16  5:21 ` stavros9899 at gmail dot com
  3 siblings, 0 replies; 5+ messages in thread
From: stavros9899 at gmail dot com @ 2023-04-16  5:21 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109528

--- Comment #4 from Stavros Avramidis <stavros9899 at gmail dot com> ---
The warning makes a bit more sense now, I was thinking gcc was getting confused
about function ending after calling the internal exits().

Anyways, I can confirm that it works with `[[gnu::noreturn]]`

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2023-04-16  5:21 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-04-15 17:12 [Bug c/109528] New: Calling exit() at the end of function triggers infinite recursion warning stavros9899 at gmail dot com
2023-04-15 19:10 ` [Bug c/109528] " pinskia at gcc dot gnu.org
2023-04-15 19:15 ` [Bug c/109528] function with _Noreturn triggers infinite recursion warning while using the attribute does not pinskia at gcc dot gnu.org
2023-04-15 19:19 ` [Bug middle-end/109528] " pinskia at gcc dot gnu.org
2023-04-16  5:21 ` stavros9899 at gmail dot com

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).