[Bug c/49801] New: df_live_verify_transfer_functions fails with to use of CC_REGNUM and checking enabled in rx backend

[Bug c/49801] New: df_live_verify_transfer_functions fails with to use of CC_REGNUM and checking enabled in rx backend

[Paulo.Matos at csr dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49801

           Summary: df_live_verify_transfer_functions fails with to use of
                    CC_REGNUM and checking enabled in rx backend
           Product: gcc
           Version: 4.6.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: Paulo.Matos@csr.com


Created attachment 24802
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=24802
files resulting from -da -fdump-tree-all-all

Compiled the rx backend enough to get a cc1:

$ ../gcc-4.6.1/configure  --enable-languages=c --target=rx-linux-elf
--disable-libssp --disable-shared --disable-libquadmath --enable-checking=all
--prefix=$HOME/local/rx-gcc-461 && make -j5

...

<fails at libgcc>

$ cat shift.c 
unsigned long foo(unsigned int a, unsigned int b)
{
  if(a < b)
    return a << b;
  else
    return a >> b;
}

$ ~/tmp/rx-gcc461-build/gcc/cc1 -Os -da -fdump-tree-all-all shift.c 
 {GC 953k -> 896k} foo {GC 900k -> 899k}
Analyzing compilation unit
 {GC 1092k -> 959k} {GC 959k -> 959k}Performing interprocedural optimizations
 <*free_lang_data> {GC 959k -> 959k} {GC 959k -> 959k} {GC 959k -> 959k} {GC
959k -> 959k} {GC 959k -> 959k} {GC 959k -> 959k} <visibility> {GC 959k ->
959k} {GC 959k -> 959k} {GC 959k -> 959k} {GC 959k -> 959k} {GC 959k -> 959k}
{GC 959k -> 959k} <early_local_cleanups> {GC 959k -> 959k} {GC 959k -> 959k}
{GC 959k -> 959k} {GC 959k -> 959k} {GC 959k -> 959k} {GC 963k -> 963k} {GC
963k -> 963k} {GC 963k -> 963k} {GC 964k -> 963k} {GC 963k -> 963k} {GC 965k ->
963k} {GC 964k -> 963k} {GC 963k -> 963k} <emutls> {GC 963k -> 963k} {GC 963k
-> 963k} {GC 964k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} <whole-program>
{GC 963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC
963k -> 963k} {GC 963k -> 963k} <ipa-profile> {GC 963k -> 963k} {GC 963k ->
963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} <cp> {GC 963k ->
963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k}
{GC 963k -> 963k} <inline> {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k ->
963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} <pure-const> {GC
963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k ->
963k} <static-var> {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC
963k -> 963k} {GC 963k -> 963k}Assembling functions:
 foo {GC 965k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k}
{GC 965k -> 965k} {GC 966k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC
968k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC 965k ->
965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k}
{GC 965k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC
967k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC 966k ->
966k} {GC 966k -> 966k} {GC 966k -> 966k} {GC 966k -> 966k} {GC 966k -> 965k}
{GC 972k -> 969k} {GC 969k -> 969k} {GC 970k -> 969k} {GC 969k -> 969k} {GC
969k -> 969k} {GC 969k -> 969k} {GC 974k -> 973k} {GC 973k -> 973k} {GC 973k ->
973k} {GC 973k -> 971k} {GC 971k -> 971k} {GC 972k -> 972k} {GC 972k -> 972k}
{GC 973k -> 972k} {GC 972k -> 972k} {GC 972k -> 972k} {GC 972k -> 972k} {GC
972k -> 972k} {GC 972k -> 972k} {GC 972k -> 972k} {GC 973k -> 972k} {GC 972k ->
972k} {GC 972k -> 972k} {GC 972k -> 972k} {GC 974k -> 972k} {GC 972k -> 972k}
shift.c: In function ‘foo’:
shift.c:7:1: internal compiler error: in df_live_verify_transfer_functions, at
df-problems.c:1816
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.

[Bug c/49801] df_live_verify_transfer_functions fails with to use of CC_REGNUM and checking enabled in rx backend

[Paulo.Matos at csr dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49801

--- Comment #1 from Paulo J. Matos <Paulo.Matos at csr dot com> 2011-07-21 09:03:06 UTC ---
I pinpointed this issue to the bitmaps computed in
df_live_verify_transfer_functions (df-problems.c).

bitmap saved_gen has no bit set while bb_info->gen has bit 16 set
(corresponding to the CC reg). Something seems to be confusing function
computing these values but I haven't been able yet to pinpoint the problem.

[Bug middle-end/49801] df_live_verify_transfer_functions fails with to use of CC_REGNUM and checking enabled in rx backend

[Paulo.Matos at csr dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49801

Paulo J. Matos <Paulo.Matos at csr dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |major

[Bug middle-end/49801] df_live_verify_transfer_functions fails with to use of CC_REGNUM and checking enabled in rx backend

[nickc at redhat dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49801

Nick Clifton <nickc at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |nickc at redhat dot com

--- Comment #2 from Nick Clifton <nickc at redhat dot com> 2011-10-04 09:08:57 UTC ---
Hi Paulo,

  Is this bug still reproducible for you ?

  I tried building an rx-elf toolchain from the current 4.6 branch sources and
I could not reproduce the problem...

Cheers
  Nick

[Bug middle-end/49801] df_live_verify_transfer_functions fails with to use of CC_REGNUM and checking enabled in rx backend

[Paulo.Matos at csr dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49801

--- Comment #3 from Paulo J. Matos <Paulo.Matos at csr dot com> 2011-10-05 11:08:36 UTC ---
(In reply to comment #2)
>   Is this bug still reproducible for you ?
> 
>   I tried building an rx-elf toolchain from the current 4.6 branch sources and
> I could not reproduce the problem...

Hi Nick,

Yes, this is still reproducible.

Here's what I did.
$ wget
ftp://ftp.mirrorservice.org/sites/sourceware.org/pub/gcc/snapshots/4.6-20110930/gcc-4.6-20110930.tar.bz2
$ mkdir gcc-build && cd gcc-build
$ ../gcc-4.6-20110930/configure  --enable-languages=c --target=rx-linux-elf
--disable-libssp --disable-shared --disable-libquadmath --enable-checking=all
--prefix=$HOME/local/rx-gcc-461-head && make -j5
<fails when building libgcc>
$ <create shift.c as in initial report>
$ gcc/cc1 -Os -da -fdump-tree-all-all shift.c 
 {GC 953k -> 896k} foo {GC 900k -> 899k}
Analyzing compilation unit
 {GC 1092k -> 959k} {GC 959k -> 959k}Performing interprocedural optimizations
 <*free_lang_data> {GC 959k -> 959k} {GC 959k -> 959k} {GC 959k -> 959k} {GC
959k -> 959k} {GC 959k -> 959k} {GC 959k -> 959k} <visibility> {GC 959k ->
959k} {GC 959k -> 959k} {GC 959k -> 959k} {GC 959k -> 959k} {GC 959k -> 959k}
{GC 959k -> 959k} <early_local_cleanups> {GC 959k -> 959k} {GC 959k -> 959k}
{GC 959k -> 959k} {GC 959k -> 959k} {GC 959k -> 959k} {GC 963k -> 963k} {GC
963k -> 963k} {GC 963k -> 963k} {GC 964k -> 963k} {GC 963k -> 963k} {GC 965k ->
963k} {GC 964k -> 963k} {GC 963k -> 963k} <emutls> {GC 963k -> 963k} {GC 963k
-> 963k} {GC 964k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} <whole-program>
{GC 963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC
963k -> 963k} {GC 963k -> 963k} <ipa-profile> {GC 963k -> 963k} {GC 963k ->
963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} <cp> {GC 963k ->
963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k}
{GC 963k -> 963k} <inline> {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k ->
963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} <pure-const> {GC
963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k ->
963k} <static-var> {GC 963k -> 963k} {GC 963k -> 963k} {GC 963k -> 963k} {GC
963k -> 963k} {GC 963k -> 963k}Assembling functions:
 foo {GC 965k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k}
{GC 965k -> 965k} {GC 966k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC
968k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC 965k ->
965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k}
{GC 965k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC
967k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC 965k -> 965k} {GC 966k ->
966k} {GC 966k -> 966k} {GC 966k -> 966k} {GC 966k -> 966k} {GC 966k -> 965k}
{GC 972k -> 969k} {GC 969k -> 969k} {GC 970k -> 969k} {GC 969k -> 969k} {GC
969k -> 969k} {GC 969k -> 969k} {GC 974k -> 973k} {GC 973k -> 973k} {GC 973k ->
973k} {GC 973k -> 971k} {GC 971k -> 971k} {GC 972k -> 971k} {GC 972k -> 972k}
{GC 973k -> 972k} {GC 972k -> 972k} {GC 972k -> 972k} {GC 972k -> 972k} {GC
972k -> 972k} {GC 972k -> 972k} {GC 972k -> 972k} {GC 972k -> 972k} {GC 972k ->
972k} {GC 972k -> 972k} {GC 972k -> 972k} {GC 974k -> 972k} {GC 972k -> 972k}
shift.c: In function ‘foo’:
shift.c:7:1: internal compiler error: in df_live_verify_transfer_functions, at
df-problems.c:1816
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.

[Bug middle-end/49801] df_live_verify_transfer_functions fails with to use of CC_REGNUM and checking enabled in rx backend

[nickc at redhat dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49801

--- Comment #4 from Nick Clifton <nickc at redhat dot com> 2011-10-06 14:21:57 UTC ---
Created attachment 25430
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=25430
workaround for bb live register checking problem

[Bug middle-end/49801] df_live_verify_transfer_functions fails with to use of CC_REGNUM and checking enabled in rx backend

[nickc at redhat dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49801

--- Comment #5 from Nick Clifton <nickc at redhat dot com> 2011-10-06 14:25:16 UTC ---
Hi Paulo,

  Thanks for the step by step guide.  I can now reproduce the problem.

  It looks to me like a generic problem in the live register analysis code. 
Which is a but beyond my debugging capabilities, so I have uploaded a simple
workaround to avoid the issue.  Please note, this is definitely a workaround,
not a patch, so use it at your discretion.  In the meantime I will continue to
investigate and see if I can get my head around the df-core.c code.

Cheers
  Nick

[Bug middle-end/49801] df_live_verify_transfer_functions fails with to use of CC_REGNUM and checking enabled in rx backend

[rth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49801

Richard Henderson <rth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2011-10-06
                 CC|                            |rth at gcc dot gnu.org
     Ever Confirmed|0                           |1

--- Comment #6 from Richard Henderson <rth at gcc dot gnu.org> 2011-10-06 21:37:29 UTC ---
Confirmed for 4.7 as well.

[Bug middle-end/49801] df_live_verify_transfer_functions fails with to use of CC_REGNUM and checking enabled in rx backend

[rth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49801

Richard Henderson <rth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
  Attachment #25430|0                           |1
        is obsolete|                            |
             Status|NEW                         |ASSIGNED
         AssignedTo|unassigned at gcc dot       |rth at gcc dot gnu.org
                   |gnu.org                     |

--- Comment #7 from Richard Henderson <rth at gcc dot gnu.org> 2011-10-06 21:59:22 UTC ---
Created attachment 25435
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=25435
workaround for live-register update

I think this patch proves I have no idea how the df interfaces work.
In particular, why the call to df_analyze is required *after* doing
the work.  Seems like for a simple problem like liveness, this ought
to be kept up-to-date without extra help...

[Bug middle-end/49801] df_live_verify_transfer_functions fails with to use of CC_REGNUM and checking enabled in rx backend

[bonzini at gnu dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49801

Paolo Bonzini <bonzini at gnu dot org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |bonzini at gnu dot org

--- Comment #8 from Paolo Bonzini <bonzini at gnu dot org> 2011-10-07 07:34:27 UTC ---
Setting TODO_df_finish should be enough.  You should need none of:

- deferred rescanning (ultimately emit_insn_after_1 calls df_insn_rescan),
since recursive splits are relatively rare.

- df_analyze, since df_insn_rescan tells liveness verification to skip dirty
blocks, no matter if it is called by on-the-fly or deferred rescanning

- df_live_add_problem, since it should be okay to use df_lr at -O1 (unlike
df_live, and like traditional flow.c, df_lr treats uninitialized registers as
live).

In fact, probably you don't need df_live_add_problem in compare-elim.c if you
use df_get_live_out and df_get_live_in instead of &DF_LIVE_BB_INFO (bb)->out
and &DF_LIVE_BB_INFO (bb)->in.  (My fault during review).

[Bug middle-end/49801] df_live_verify_transfer_functions fails with to use of CC_REGNUM and checking enabled in rx backend

[nickc at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49801

--- Comment #9 from Nick Clifton <nickc at gcc dot gnu.org> 2011-10-10 13:31:10 UTC ---
Author: nickc
Date: Mon Oct 10 13:31:03 2011
New Revision: 179749

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=179749
Log:
    PR middle-end/49801
    * compare-elim.c (find_comparisons_in_bb): Use df_get_live_in and
    df_get_live_out instead of accessing the bitmaps directly.
    (execute_compare_elim_after_reload): Remove calls to df_set_flags,
    df_live_add_problem and df_analyze.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/compare-elim.c

[Bug middle-end/49801] df_live_verify_transfer_functions fails with to use of CC_REGNUM and checking enabled in rx backend

[nickc at redhat dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49801

--- Comment #10 from Nick Clifton <nickc at redhat dot com> 2011-10-10 13:33:45 UTC ---
Hi Paulo,

  This should be fixed now.

Cheers
  Nick

[Bug middle-end/49801] df_live_verify_transfer_functions fails with to use of CC_REGNUM and checking enabled in rx backend

[nickc at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49801

--- Comment #11 from Nick Clifton <nickc at gcc dot gnu.org> 2011-10-11 07:49:11 UTC ---
Author: nickc
Date: Tue Oct 11 07:49:04 2011
New Revision: 179787

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=179787
Log:
    PR middle-end/49801
    * compare-elim.c (find_comparisons_in_bb): Use df_get_live_in and
    df_get_live_out instead of accessing the bitmaps directly.
    (execute_compare_elim_after_reload): Remove calls to df_set_flags,
    df_live_add_problem and df_analyze.

Modified:
    branches/gcc-4_6-branch/gcc/ChangeLog
    branches/gcc-4_6-branch/gcc/compare-elim.c

[Bug middle-end/49801] df_live_verify_transfer_functions fails with to use of CC_REGNUM and checking enabled in rx backend

[Paulo.Matos at csr dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49801

--- Comment #12 from Paulo J. Matos <Paulo.Matos at csr dot com> 2011-10-17 15:26:51 UTC ---
Sorry for the time taken to reply.
I have tested snapshots of 4.6 and 4.7 and both seem happy now.

Looks like it's sorted. Thanks.

[Bug middle-end/49801] df_live_verify_transfer_functions fails with to use of CC_REGNUM and checking enabled in rx backend

[bonzini at gnu dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49801

Paolo Bonzini <bonzini at gnu dot org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED

--- Comment #13 from Paolo Bonzini <bonzini at gnu dot org> 2012-01-20 07:53:16 UTC ---
per comment 12.