[Bug middle-end/54327] [4.8 Regression] Segmentation fault in init_ggc

[dcb314 at hotmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54327

--- Comment #10 from David Binderman  ---
(In reply to Richard Biener from comment #9)
> Please open a new bugreport - this seems completely unrelated.

Done - # 58911

[Bug middle-end/54327] [4.8 Regression] Segmentation fault in init_ggc

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54327

Richard Biener  changed:

   What|Removed |Added

 Status|REOPENED|RESOLVED
 Resolution|--- |FIXED

--- Comment #9 from Richard Biener  ---
(In reply to David Binderman from comment #8)
> Seems to be newly broken. 20131023 ok, 20131027 not, as follows
> 
> htslib.c: In function ‘x_escape_http’:
> htslib.c:3804:7: internal compiler error: in operator[], at vec.h:827
> 0xb8fa27 vec::operator[](unsigned int)
> ../../src/trunk/gcc/vec.h:827
> 0xb91572 vec::operator[](unsigned int)
> ../../src/trunk/gcc/tree.h:2792
> 0xb91572 vec::operator[](unsigned int)
> ../../src/trunk/gcc/vec.h:1256
> 0xb91572 update_ops
> ../../src/trunk/gcc/tree-ssa-reassoc.c:2619
> 0xbf maybe_optimize_range_tests
> ../../src/trunk/gcc/tree-ssa-reassoc.c:2907
> 0xbf reassociate_bb
> ../../src/trunk/gcc/tree-ssa-reassoc.c:4325
> 0xb98eb7 reassociate_bb
> ../../src/trunk/gcc/tree-ssa-reassoc.c:4482
> 0xb9b7ab do_reassoc
> ../../src/trunk/gcc/tree-ssa-reassoc.c:4515
> 0xb9b7ab execute_reassoc
> ../../src/trunk/gcc/tree-ssa-reassoc.c:4597
> 0xb9b7ab execute
> ../../src/trunk/gcc/tree-ssa-reassoc.c:4639
> Please submit a full bug report,
> with preprocessed source if appropriate.
> Please include the complete backtrace with any bug report.
> See  for instructions.

Please open a new bugreport - this seems completely unrelated.

[Bug middle-end/54327] [4.8 Regression] Segmentation fault in init_ggc

[dcb314 at hotmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54327

David Binderman  changed:

   What|Removed |Added

 Status|RESOLVED|REOPENED
 Resolution|FIXED   |---

--- Comment #8 from David Binderman  ---
Seems to be newly broken. 20131023 ok, 20131027 not, as follows

htslib.c: In function ‘x_escape_http’:
htslib.c:3804:7: internal compiler error: in operator[], at vec.h:827
0xb8fa27 vec::operator[](unsigned int)
../../src/trunk/gcc/vec.h:827
0xb91572 vec::operator[](unsigned int)
../../src/trunk/gcc/tree.h:2792
0xb91572 vec::operator[](unsigned int)
../../src/trunk/gcc/vec.h:1256
0xb91572 update_ops
../../src/trunk/gcc/tree-ssa-reassoc.c:2619
0xbf maybe_optimize_range_tests
../../src/trunk/gcc/tree-ssa-reassoc.c:2907
0xbf reassociate_bb
../../src/trunk/gcc/tree-ssa-reassoc.c:4325
0xb98eb7 reassociate_bb
../../src/trunk/gcc/tree-ssa-reassoc.c:4482
0xb9b7ab do_reassoc
../../src/trunk/gcc/tree-ssa-reassoc.c:4515
0xb9b7ab execute_reassoc
../../src/trunk/gcc/tree-ssa-reassoc.c:4597
0xb9b7ab execute
../../src/trunk/gcc/tree-ssa-reassoc.c:4639
Please submit a full bug report,
with preprocessed source if appropriate.
Please include the complete backtrace with any bug report.
See  for instructions.

[Bug middle-end/54327] [4.8 Regression] Segmentation fault in init_ggc

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54327

Richard Guenther  changed:

   What|Removed |Added

 Status|ASSIGNED|RESOLVED
 Resolution||FIXED

--- Comment #7 from Richard Guenther  2012-08-20 
11:10:57 UTC ---
Fixed.

[Bug middle-end/54327] [4.8 Regression] Segmentation fault in init_ggc

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54327

--- Comment #6 from Richard Guenther  2012-08-20 
10:26:52 UTC ---
Author: rguenth
Date: Mon Aug 20 10:26:45 2012
New Revision: 190528

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=190528
Log:
2012-08-20  Richard Guenther  

PR tree-optimization/54327
* gimple-fold.c (get_maxval_strlen): Do not walk use-def chains
if the use is registered for SSA update.

* gcc.dg/torture/pr54327.c: New testcase.

Added:
trunk/gcc/testsuite/gcc.dg/torture/pr54327.c
Modified:
trunk/gcc/ChangeLog
trunk/gcc/gimple-fold.c
trunk/gcc/testsuite/ChangeLog

[Bug middle-end/54327] [4.8 Regression] Segmentation fault in init_ggc

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54327

--- Comment #5 from Richard Guenther  2012-08-20 
08:53:13 UTC ---
The issue is that we now merge two blocks we didn't merge before which
introduces SSA names into a STMT that are marked for update (thus, in
old_ssa_names):

  :
  # a_12 = PHI <0B(3)>
  _13 = *a_1;
  _14 = (int) _4;
  _15 = _4 == 0;
  _16 = (int) _6;
  goto ;

  :
  # a_11 = PHI 
  _9 = strlen (a_11);
  if (_9 <= 1)
goto ;
  else
goto ;

a_1 is marked to be replaced by a_12, block merging from CFG cleanup
(called via TODO, which calls CFG cleanup before updating SSA form ...)
now replaces a_11 with a_1 in _9 = strlen (a_11) via replace_uses_by
which ends up folding the stmt (but a_1 has a dead DEF stmt).

update_ssa only needs unreachable blocks removed to make DOM info
computable.  So this might be solved by re-ordering cfgcleanup/SSA update
in the TODO.  OTOH passes registering things for manual SSA update should
eventually call update_ssa themselves, not relying on TODO (in this case
it is jump threading from DOM).  OTOH calling into SSA def stmt walking
folders from this situation is bad, the folding in question is
gimple_fold_builtin calling get_maxval_strlen (which may even result in
wrong-code with a non-up-to-date SSA form).

So the bug IMHO is fold_stmt walking SSA def stmts.

[Bug middle-end/54327] [4.8 Regression] Segmentation fault in init_ggc

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54327

Richard Guenther  changed:

   What|Removed |Added

 Status|NEW |ASSIGNED
 AssignedTo|unassigned at gcc dot   |rguenth at gcc dot gnu.org
   |gnu.org |

--- Comment #4 from Richard Guenther  2012-08-20 
07:17:02 UTC ---
Mine.

[Bug middle-end/54327] [4.8 Regression] Segmentation fault in init_ggc

[markus at trippelsdorf dot de]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54327

Markus Trippelsdorf  changed:

   What|Removed |Added

 CC||markus at trippelsdorf dot
   ||de

--- Comment #3 from Markus Trippelsdorf  
2012-08-20 03:40:24 UTC ---
Small testcase:

#include 
#include 
void treathead ()
{
char *a = ';' == '\0' ? : 0;
if (*a == '=')
{
while (*a == (*a == 0) || *a == '\'')
a++;
if (strlen (a) < 2)
abort ();
}
}

[Bug middle-end/54327] [4.8 Regression] Segmentation fault in init_ggc

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54327

H.J. Lu  changed:

   What|Removed |Added

 CC||rguenth at gcc dot gnu.org
  Component|c   |middle-end

--- Comment #2 from H.J. Lu  2012-08-19 23:22:32 
UTC ---
It is caused by revision 189915:

http://gcc.gnu.org/ml/gcc-cvs/2012-07/msg00820.html