[m68k] gcc-4.0: internal consistency failure

[Debian Bug Importer]

Automatically imported from Debian bug report #344041 http://bugs.debian.org/344041

[Debian Bug Importer (debzilla)]

Message-ID: <email address hidden>
Date: Mon, 19 Dec 2005 09:34:47 -0600
From: Stephen R Marenka <email address hidden>
To: <email address hidden>
Subject: [m68k] gcc-4.0: internal consistency failure

--MAH+hnPXVZWQ5cD/
Content-Type: multipart/mixed; boundary="s9fJI615cBHmzTOP"
Content-Disposition: inline

--s9fJI615cBHmzTOP
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

Package: gcc-4.0
Version: 4.0.2-5
Severity: important

Recent versions of make fail with an error reproducible as follows.

$ gcc -O2 -c -o buggy.o buggy.c
buggy.c: In function 'pop_variable_scope':
buggy.c:62: fatal error: internal consistency failure
compilation terminated.

Workaround: change -O2 to -O1.

Hopefully, I've reduced this enough.

Thanks,

Stephen

--=20
Stephen R. Marenka If life's not fun, you're not doing it right!
<email address hidden>

--s9fJI615cBHmzTOP
Content-Type: text/x-csrc; charset=us-ascii
Content-Disposition: attachment; filename="buggy.c"

/* based on variable.c from make_3.80+3.81.b4-1 */

#define NULL 0
#include <assert.h>

struct variable_set
  {
    void * table; /* Hash table of variables. */
  };

struct variable_set_list
  {
    struct variable_set_list *next; /* Link in the chain. */
    struct variable_set *set; /* Variable set. */
  };

extern struct variable_set_list *current_variable_set_list;
static struct variable_set global_variable_set;
static struct variable_set_list global_setlist
  = { 0, &global_variable_set };

static void
free_variable_name_and_value (const void *item)
{
}

void
pop_variable_scope (void)
{
  struct variable_set_list *setlist;
  struct variable_set *set;

  /* Can't call this if there's no scope to pop! */
  assert(current_variable_set_list->next != NULL);

  if (current_variable_set_list != &global_setlist)
    {
      /* We're not pointing to the global setlist, so pop this one. */
      setlist = current_variable_set_list;
      set = setlist->set;
      current_variable_set_list = setlist->next;
    }
  else
    {
      /* This set is the one in the global_setlist, but there is another global
         set beyond that. We want to copy that set to global_setlist, then
         delete what used to be in global_setlist. */
      setlist = global_setlist.next;
      set = global_setlist.set;
      global_setlist.set = setlist->set;
      global_setlist.next = setlist->next;
    }

  /* Free the one we no longer need. */
  free ((char *) setlist);
  hash_map (&set->table, free_variable_name_and_value);
  hash_free (&set->table, 1);
  free ((char *) set);
}

--s9fJI615cBHmzTOP--

--MAH+hnPXVZWQ5cD/
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: Digital signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)

iD8DBQFDptMXVKM4J7YoSbMRAub8AJwJZd7OSUYy6l2/oKgECmN0+ShlJQCfVNAX
dDSkJA1ZgQE8E1crxq1+smM=
=4zgx
-----END PGP SIGNATURE-----

--MAH+hnPXVZWQ5cD/--

[Debian Bug Importer (debzilla)]

Message-ID: <email address hidden>
Date: Mon, 19 Dec 2005 13:52:54 -0600
From: Manoj Srivastava <email address hidden>
To: Stephen R Marenka <email address hidden>
Cc: <email address hidden>, <email address hidden>,
        <email address hidden>
Subject: Re: Bug#344046: FTBFS on m68k

reassign 344046 gcc-4.0
severity 344046 grave
severity 344041 grave
merge 344041 344046
thanks

Hi,

        If this is a bug in gcc-4.0, and it causes other packages to
 have FTBS bugs, then we should fix gcc.

        manoj
--
Newton's Little-Known Seventh Law: A bird in the hand is safer than
one overhead.
Manoj Srivastava <email address hidden> <http://www.debian.org/%7Esrivasta/>
1024D/BF24424C print 4966 F272 D093 B493 410B 924B 21BA DABB BF24 424C

[Debian Bug Importer (debzilla)]

Message-ID: <email address hidden>
Date: Mon, 19 Dec 2005 15:40:02 -0600
From: Stephen R Marenka <email address hidden>
To: <email address hidden>
Subject: fix severity

severity 344041 normal
thanks

--
Stephen R. Marenka If life's not fun, you're not doing it right!
<email address hidden>

[Debian Bug Importer (debzilla)]

Message-ID: <email address hidden>
Date: Mon, 19 Dec 2005 23:22:49 -0800
From: Steve Langasek <email address hidden>
To: Manoj Srivastava <email address hidden>, <email address hidden>
Cc: Stephen R Marenka <email address hidden>,
 <email address hidden>
Subject: Re: Bug#344046: FTBFS on m68k

--ZwgA9U+XZDXt4+m+
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

unmerge 344046
reassign 344046 make 3.80+3.81.b4-1
severity 344046 serious
block 344046 by 344041
thanks

On Mon, Dec 19, 2005 at 01:52:54PM -0600, Manoj Srivastava wrote:
> reassign 344046 gcc-4.0
> severity 344046 grave
> severity 344041 grave
> merge 344041 344046
> thanks

> If this is a bug in gcc-4.0, and it causes other packages to
> have FTBS bugs, then we should fix gcc.

By this standard, almost all toolchain bugs would be release-critical bugs.
This is not reasonable; there is an order of magnitude difference (at least)
between "fix an arbitrary bug in gcc which causes package foo to FTBFS" and
"fix package foo to use compile options that work", and one build failure
(or a few) doesn't exactly make gcc "unusable or mostly so" anyway.

When a package fails to build on a release architecture and the failure is
not caused by an RC bug in another package, that's a serious bug in the
package in question.

--=20
Steve Langasek Give me a lever long enough and a Free OS
Debian Developer to set it on, and I can move the world.
<email address hidden> http://www.debian.org/

--ZwgA9U+XZDXt4+m+
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: Digital signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)

iD8DBQFDp7FIKN6ufymYLloRAlIpAKCmurusQzQDn0xfHilks3t1gJdZDACgpjR/
/ovR4RnATmaR3NFZE2jaiB0=
=AQ2D
-----END PGP SIGNATURE-----

--ZwgA9U+XZDXt4+m+--

[Matthias Klose (doko)]

m68k only