[Bug rtl-optimization/58960] New: [4.9 regression] ICE in bmp_iter_set_init

[Bug rtl-optimization/58960] New: [4.9 regression] ICE in bmp_iter_set_init

[schwab@linux-m68k.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58960

            Bug ID: 58960
           Summary: [4.9 regression] ICE in bmp_iter_set_init
           Product: gcc
           Version: 4.9.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: rtl-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: schwab@linux-m68k.org

gcc/xg++ -Bgcc/ ../gcc/testsuite/g++.dg/opt/pr48272.C -nostdinc++ -I
ia64-suse-linux/libstdc++-v3/include/ia64-suse-linux
-Iia64-suse-linux/libstdc++-v3/include -I../libstdc++-v3/libsupc++
-I../libstdc++-v3/include/backward -I../libstdc++-v3/testsuite/util
-std=gnu++11 -O3 -ftracer -fsched-pressure -S
../gcc/testsuite/g++.dg/opt/pr48272.C: In constructor
‘std::iostreams::zlib_error::zlib_error(int)’:
../gcc/testsuite/g++.dg/opt/pr48272.C:128:5: internal compiler error:
Segmentation fault
     }
     ^
0x4000000000be814f crash_signal
        ../../gcc/toplev.c:334
0x40000000012a428f bmp_iter_set_init
        ../../gcc/bitmap.h:360
0x40000000012a428f initiate_reg_pressure_info
        ../../gcc/haifa-sched.c:982
0x40000000012a600f initiate_bb_reg_pressure_info
        ../../gcc/haifa-sched.c:1030
0x40000000012a600f sched_setup_bb_reg_pressure_info(basic_block_def*, rtx_def*)
        ../../gcc/haifa-sched.c:3034
0x4000000000b46c3f schedule_region
        ../../gcc/sched-rgn.c:3015
0x4000000000b46fff schedule_insns()
        ../../gcc/sched-rgn.c:3390
0x4000000000b4718f rest_of_handle_sched
        ../../gcc/sched-rgn.c:3583
0x4000000000b4718f execute
        ../../gcc/sched-rgn.c:3647


Program received signal SIGSEGV, Segmentation fault.
bmp_iter_set_init (bit_no=<optimized out>, start_bit=0, map=<optimized out>, 
    bi=<optimized out>) at ../../gcc/bitmap.h:341
341       bi->elt1 = map->first;
(gdb) bt
#0  bmp_iter_set_init (bit_no=<optimized out>, start_bit=0, 
    map=<optimized out>, bi=<optimized out>) at ../../gcc/bitmap.h:341
#1  initiate_reg_pressure_info (live=0x40) at ../../gcc/haifa-sched.c:982
#2  0x40000000012a6010 in initiate_bb_reg_pressure_info (bb=<optimized out>)
    at ../../gcc/haifa-sched.c:1030
#3  sched_setup_bb_reg_pressure_info (bb=0x2000000000a42de8, 
    after=0x20000000009ee6c0) at ../../gcc/haifa-sched.c:3034
#4  0x4000000000b46c40 in schedule_region (rgn=3) at ../../gcc/sched-rgn.c:3015
#5  0x4000000000b47000 in schedule_insns () at ../../gcc/sched-rgn.c:3390
#6  0x4000000000b47190 in rest_of_handle_sched () at ../../gcc/sched-rgn.c:3583
#7  (anonymous namespace)::pass_sched::execute (
    this=0x4000000000a60600 <execute_one_pass(opt_pass*)+2384>)
    at ../../gcc/sched-rgn.c:3647
#8  0x4000000000a60600 in execute_one_pass (pass=0x600000000029bbc0)
    at ../../gcc/passes.c:2215
#9  0x4000000000a60dc0 in execute_pass_list (pass=0x600000000029bbc0)
    at ../../gcc/passes.c:2267
#10 0x4000000000a60e10 in execute_pass_list (pass=0x600000000029ac00)
    at ../../gcc/passes.c:2268
#11 0x40000000005a4f20 in expand_function (node=0x2000000000a80980)
    at ../../gcc/cgraphunit.c:1753
#12 0x40000000005a94a0 in expand_all_functions ()
    at ../../gcc/cgraphunit.c:1858
#13 0x40000000005aa180 in compile () at ../../gcc/cgraphunit.c:2195
#14 0x40000000005aa500 in finalize_compilation_unit ()
    at ../../gcc/cgraphunit.c:2272
#15 0x400000000022cf30 in cp_write_global_declarations ()
    at ../../gcc/cp/decl2.c:4401
#16 0x4000000000be9500 in compile_file () at ../../gcc/toplev.c:559
#17 0x4000000000bee3c0 in do_compile () at ../../gcc/toplev.c:1894
#18 toplev_main (argc=16279, argv=0x600ffffffffeefe8)
    at ../../gcc/toplev.c:1970
#19 0x40000000013ee4b0 in main (argc=34, argv=0x600ffffffffeefe8)
    at ../../gcc/main.c:36

Reverting r204194 avoids the crash.  Looks similar to bug48272.
>From gcc-bugs-return-433257-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org Fri Nov 01 12:51:10 2013
Return-Path: <gcc-bugs-return-433257-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Delivered-To: listarch-gcc-bugs@gcc.gnu.org
Received: (qmail 8093 invoked by alias); 1 Nov 2013 12:51:10 -0000
Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-bugs.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-help@gcc.gnu.org>
Sender: gcc-bugs-owner@gcc.gnu.org
Delivered-To: mailing list gcc-bugs@gcc.gnu.org
Received: (qmail 8069 invoked by uid 48); 1 Nov 2013 12:51:05 -0000
From: "afenkart at gmail dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug c/58961] New: missing "<variable> may be used uninitialized in this function"
Date: Fri, 01 Nov 2013 12:51:00 -0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: new
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: c
X-Bugzilla-Version: 4.7.2
X-Bugzilla-Keywords:
X-Bugzilla-Severity: normal
X-Bugzilla-Who: afenkart at gmail dot com
X-Bugzilla-Status: UNCONFIRMED
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags:
X-Bugzilla-Changed-Fields: bug_id short_desc product version bug_status bug_severity priority component assigned_to reporter
Message-ID: <bug-58961-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-SW-Source: 2013-11/txt/msg00034.txt.bz2
Content-length: 432

http://gcc.gnu.org/bugzilla/show_bug.cgi?idX961

            Bug ID: 58961
           Summary: missing "<variable> may be used uninitialized in this
                    function"
           Product: gcc
           Version: 4.7.2
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
          Assignee: unassigned at gcc dot gnu.org
          Reporter: afenkart at gmail dot com

[Bug rtl-optimization/58960] [4.9 regression] ICE in bmp_iter_set_init

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58960

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.9.0

[Bug rtl-optimization/58960] [4.9 regression] ICE in bmp_iter_set_init

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58960

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P1
                 CC|                            |pinskia at gcc dot gnu.org

[Bug rtl-optimization/58960] [4.9 regression] ICE in bmp_iter_set_init

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58960

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jakub at gcc dot gnu.org

--- Comment #1 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
ia64-*-* is neither primary nor secondary target, why is this P1?

[Bug rtl-optimization/58960] [4.9 regression] ICE in bmp_iter_set_init

[mpolacek at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58960

Marek Polacek <mpolacek at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P1                          |P2
                 CC|                            |mpolacek at gcc dot gnu.org

--- Comment #2 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
I also wondered why is this P1.  Decreasing to P2...

[Bug rtl-optimization/58960] [4.9 regression] ICE in bmp_iter_set_init

[abel at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58960

Andrey Belevantsev <abel at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |abel at gcc dot gnu.org

--- Comment #3 from Andrey Belevantsev <abel at gcc dot gnu.org> ---
I've tried to reproduce this with a simple cross (no includes, just ./gcc/xg++
-Bgcc/ -nostdinc++ -nostdinc++ -std=gnu++11 -O3 -ftracer -fsched-pressure -S
~/develop/trunk/gcc/testsuite/g++.dg/opt/pr48272.C -S); this doesn't fail but
instead valgrind spills out a number of ira-related warnings of uninitialized
values -- does this still fails for you?  Do I need to try the native build?

[Bug rtl-optimization/58960] [4.9 regression] ICE in bmp_iter_set_init

[schwab@linux-m68k.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58960

--- Comment #4 from Andreas Schwab <schwab@linux-m68k.org> ---
Still fails as of r206866.

http://gcc.gnu.org/ml/gcc-testresults/2014-01/msg01656.html

[Bug rtl-optimization/58960] [4.9 regression] ICE in bmp_iter_set_init

[abel at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58960

Andrey Belevantsev <abel at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |ASSIGNED
   Last reconfirmed|                            |2014-01-22
           Assignee|unassigned at gcc dot gnu.org      |abel at gcc dot gnu.org
     Ever confirmed|0                           |1

--- Comment #5 from Andrey Belevantsev <abel at gcc dot gnu.org> ---
Created attachment 31915
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=31915&action=edit
proposed patch

[Bug rtl-optimization/58960] [4.9 regression] ICE in bmp_iter_set_init

[abel at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58960

Andrey Belevantsev <abel at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |vmakarov at gcc dot gnu.org

--- Comment #6 from Andrey Belevantsev <abel at gcc dot gnu.org> ---
Well, I guess one is responsible for what one has tamed, so reproduced now.

The issue is actually simple -- we add new blocks when scheduling (they are
recovery blocks for speculation insns) and for them we do not calculate
DF_LIVE_IN bitmaps so we segfault when we try to read those bitmaps in the
sched pressure info initialization code.  I see the following possible fixes:

1) calculate liveness for the new blocks via the DF interface.  Now it doesn't
look like anyone is using the partial analysis interface via df_set_blocks
nowadays, and the new blocks do not form a loop so df_analyze_loop Richard has
added would not work.  What can be done is gather the new blocks in a separate
bitmap and add their preds/succs which should hopefully have the right liveness
info, so partial df_analyze on those can do the trick.

2) manually calculate liveness for the new blocks.  Again DF doesn't offer much
for this, you'd need to df_grow_block_info for df_live and then try deriving
the liveness information from the original block's successor.  This will need
some manual propagation through the new block as we do in the selective
scheduler.

3) decide that the recovery code is supposed to be cold enough so we don't care
about applying register pressure sensitive code to it, thus reset
sched_pressure to SCHED_PRESSURE_NONE for all newly created regions.  The patch
from the previous comment does just that; we only need to free the
sched-pressure data immediately at this point as we later fail to do that and
ICE in IRA later.  I had to factorize the finalization code in a separate
function for that.  
The patch supports only blocks ending up in the new region, not the same
region, but it looks like the current code always puts recovery blocks in the
new region.

Vlad, what's your opinion on this?

[Bug rtl-optimization/58960] [4.9 regression] ICE in bmp_iter_set_init

[schwab@linux-m68k.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58960

--- Comment #7 from Andreas Schwab <schwab@linux-m68k.org> ---
Testresults: http://gcc.gnu.org/ml/gcc-testresults/2014-01/msg01781.html

No regressions, and g++.dg/opt/pr48272.C now passes.

[Bug rtl-optimization/58960] [4.9 regression] ICE in bmp_iter_set_init

[abel at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58960

--- Comment #8 from Andrey Belevantsev <abel at gcc dot gnu.org> ---
(In reply to Andreas Schwab from comment #7)
> Testresults: http://gcc.gnu.org/ml/gcc-testresults/2014-01/msg01781.html
> 
> No regressions, and g++.dg/opt/pr48272.C now passes.

Thank you, my own run is still working.  If that would be fine and Vlad would
agree with this option, I will install the patch.

[Bug rtl-optimization/58960] [4.9 regression] ICE in bmp_iter_set_init

[abel at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58960

Andrey Belevantsev <abel at gcc dot gnu.org> changed:

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

--- Comment #10 from Andrey Belevantsev <abel at gcc dot gnu.org> ---
Fixed on trunk.

[Bug rtl-optimization/58960] [4.9 regression] ICE in bmp_iter_set_init

[abel at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58960

--- Comment #9 from Andrey Belevantsev <abel at gcc dot gnu.org> ---
Author: abel
Date: Tue Feb 18 05:41:29 2014
New Revision: 207832

URL: http://gcc.gnu.org/viewcvs?rev=207832&root=gcc&view=rev
Log:
        PR rtl-optimization/58960
        * haifa-sched.c (alloc_global_sched_pressure_data): New,
        factored out from ...
        (sched_init): ... here.
        (free_global_sched_pressure_data): New, factored out from ...
        (sched_finish): ... here.
        * sched-int.h (free_global_sched_pressure_data): Declare.
        * sched-rgn.c (nr_regions_initial): New static global.
        (haifa_find_rgns): Initialize it.
        (schedule_region): Disable sched-pressure for the newly
        generated regions.


Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/haifa-sched.c
    trunk/gcc/sched-int.h
    trunk/gcc/sched-rgn.c