public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug c/113232] New: wrong code at -fpack-struct on x86_64-pc-linux-gnu
@ 2024-01-04 12:01 jwzeng at nuaa dot edu.cn
  2024-01-04 15:01 ` [Bug c/113232] " pinskia at gcc dot gnu.org
  0 siblings, 1 reply; 2+ messages in thread
From: jwzeng at nuaa dot edu.cn @ 2024-01-04 12:01 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113232

            Bug ID: 113232
           Summary: wrong code at -fpack-struct on x86_64-pc-linux-gnu
           Product: gcc
           Version: 14.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
          Assignee: unassigned at gcc dot gnu.org
          Reporter: jwzeng at nuaa dot edu.cn
  Target Milestone: ---

I compiled the following code with gcc at -fpack-struct, and it produces the
wrong code. 

The correct output result should be 0, but 2048 was output under -fpack-struct. 

This bug seems to be a long-standing bug that exists on almost all gcc
versions.

Compiler explorer: https://godbolt.org/z/v4sq988P9

```c
$ cat test.c
int printf(const char *, ...);
struct a {
  char b;
  short c;
};
union {
  short b;
  struct a d;
} e;
int main() {
  e.b = 0;
  for (e.d.c = 0; e.d.c < 8; e.d.c++)
    ;  
  printf("%d\n", e.b);
}
$
$ gcc-tk test.c -O0; ./test.c
0
$ gcc-tk test.c -fpack-struct; ./test.c
2048
$ ccomp test.c -O0; ./a.out
0
$
$ gcc-tk --version
gcc (GCC) 13.1.0
Copyright (C) 2023 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
$
$ ccomp --version
The CompCert C verified compiler, version 3.12
```

^ permalink raw reply	[flat|nested] 2+ messages in thread

* [Bug c/113232] wrong code at -fpack-struct on x86_64-pc-linux-gnu
  2024-01-04 12:01 [Bug c/113232] New: wrong code at -fpack-struct on x86_64-pc-linux-gnu jwzeng at nuaa dot edu.cn
@ 2024-01-04 15:01 ` pinskia at gcc dot gnu.org
  0 siblings, 0 replies; 2+ messages in thread
From: pinskia at gcc dot gnu.org @ 2024-01-04 15:01 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113232

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

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

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
This is not a bug.

Without -fpacked-struct, e.b and e.d.c does not have any bits in common due to
a padding field in `struct a` because of the alignment of short is 2. But with
-fpacked-struct, the padding byte/field goes away and you get an overlap.

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2024-01-04 15:01 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2024-01-04 12:01 [Bug c/113232] New: wrong code at -fpack-struct on x86_64-pc-linux-gnu jwzeng at nuaa dot edu.cn
2024-01-04 15:01 ` [Bug c/113232] " pinskia at gcc dot gnu.org

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).