[Bug middle-end/60469] New: simple cilk plus program ICEs

[Bug middle-end/60469] New: simple cilk plus program ICEs

[andi-gcc at firstfloor dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60469

            Bug ID: 60469
           Summary: simple cilk plus program ICEs
           Product: gcc
           Version: unknown
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: middle-end
          Assignee: unassigned at gcc dot gnu.org
          Reporter: andi-gcc at firstfloor dot org

gcc version 4.9.0 20140209 (experimental) (GCC) 

gcc -fcilkplus test.c

cilk.c: In function 'main':
tcilk.c:14:5: internal compiler error: Segmentation fault
 int main()
     ^
0x8a5df7 crash_signal
        ../../gcc/gcc/toplev.c:337
0x9048e3 get_frame_type
        ../../gcc/gcc/tree-nested.c:234
0x9048e3 get_chain_decl
        ../../gcc/gcc/tree-nested.c:342
0x9056bc get_chain_decl
        ../../gcc/gcc/tree-nested.c:851
0x9056bc get_nonlocal_debug_decl
        ../../gcc/gcc/tree-nested.c:856
0x905932 convert_nonlocal_reference_op
        ../../gcc/gcc/tree-nested.c:934
0xa61f7c walk_tree_1(tree_node**, tree_node* (*)(tree_node**, int*, void*),
void*, pointer_set_t*, tree_node* (*)(tree_node**, int*, tree_node*
(*)(tree_node**, int*, void*), void*, pointer_set_t*))
        ../../gcc/gcc/tree.c:10920
0x725263 walk_gimple_op(gimple_statement_base*, tree_node* (*)(tree_node**,
int*, void*), walk_stmt_info*)
        ../../gcc/gcc/gimple-walk.c:221
0x725476 walk_gimple_stmt(gimple_stmt_iterator*, tree_node*
(*)(gimple_stmt_iterator*, bool*, walk_stmt_info*), tree_node* (*)(tree_node**,
int*, void*), walk_stmt_info*)
        ../../gcc/gcc/gimple-walk.c:535
0x725629 walk_gimple_seq_mod(gimple_statement_base**, tree_node*
(*)(gimple_stmt_iterator*, bool*, walk_stmt_info*), tree_node* (*)(tree_node**,
int*, void*), walk_stmt_info*)
        ../../gcc/gcc/gimple-walk.c:577
0x725629 walk_gimple_seq_mod(gimple_statement_base**, tree_node*
(*)(gimple_stmt_iterator*, bool*, walk_stmt_info*), tree_node* (*)(tree_node**,
int*, void*), walk_stmt_info*)
        ../../gcc/gcc/gimple-walk.c:58
0x725551 walk_gimple_stmt(gimple_stmt_iterator*, tree_node*
(*)(gimple_stmt_iterator*, bool*, walk_stmt_info*), tree_node* (*)(tree_node**,
int*, void*), walk_stmt_info*)
        ../../gcc/gcc/gimple-walk.c:577
0x725629 walk_gimple_seq_mod(gimple_statement_base**, tree_node*
(*)(gimple_stmt_iterator*, bool*, walk_stmt_info*), tree_node* (*)(tree_node**,
int*, void*), walk_stmt_info*)
        ../../gcc/gcc/gimple-walk.c:58
0x725521 walk_gimple_stmt(gimple_stmt_iterator*, tree_node*
(*)(gimple_stmt_iterator*, bool*, walk_stmt_info*), tree_node* (*)(tree_node**,
int*, void*), walk_stmt_info*)
        ../../gcc/gcc/gimple-walk.c:545
0x725629 walk_gimple_seq_mod(gimple_statement_base**, tree_node*
(*)(gimple_stmt_iterator*, bool*, walk_stmt_info*), tree_node* (*)(tree_node**,
int*, void*), walk_stmt_info*)
        ../../gcc/gcc/gimple-walk.c:58
0x904127 walk_body
        ../../gcc/gcc/tree-nested.c:603
0x9060ee walk_function
        ../../gcc/gcc/tree-nested.c:614
0x9060ee walk_all_functions
        ../../gcc/gcc/tree-nested.c:679
0x90739f lower_nested_functions(tree_node*)
        ../../gcc/gcc/tree-nested.c:2724
Please submit a full bug report,
with preprocessed source if appropriate.
Please include the complete backtrace with any bug report.
See <http://gcc.gnu.org/bugs.html> for instructions.


test program:


void foo() { asm("" ::: "memory"); }

#define ALEN 1024


void f(int *a, int *b, int *c)
{
        int i;
#pragma simd
        for (i = 0; i < ALEN; i++)
                a[i] = b[i] * c[i];
}

int main()
{
        int a[ALEN], b[ALEN], c[ALEN];

        b[:] = 100;
        c[0:ALEN:2] = 99;
        c[1:ALEN:2] = 101;

        f(a, b, c);

        asm("" ::: "memory"); /* Make sure a,b,c, are not optimized away */

        _Cilk_spawn foo();
        _Cilk_spawn foo();
        _Cilk_sync;
        return 0;
}

[Bug middle-end/60469] simple cilk plus program ICEs

[izamyatin at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60469

Igor Zamyatin <izamyatin at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |izamyatin at gmail dot com

--- Comment #2 from Igor Zamyatin <izamyatin at gmail dot com> ---
ICE could be seen even for

void foo() { asm("" ::: "memory"); }

#define ALEN 1024

int main(int argc, char* argv[])
{
    int b[ALEN];

    b[:] = 100;
    _Cilk_spawn foo();
    return 0;
}

The "bad" VAR_DECL here is the initial variable for the loop that is created
during array annotation expression expanding. The problem seems to be in the
way how this var_decl is created.
For C++ case create_temp_var is used where DECL_CONTEXT is set up. For C case
build_decl is used where no DECL_CONTEXT is filled

[Bug middle-end/60469] simple cilk plus program ICEs

[andi-gcc at firstfloor dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60469

--- Comment #3 from Andi Kleen <andi-gcc at firstfloor dot org> ---
I've tried a couple of things to fix this:

- Fill in DECL_CONTEXT to current_fn_decl in cilk
- Fill in DECL_CONTEXT for VAR_DECLs when creating the nested wrapper

No banana so far. The first causes other errors and the second still has the
same segfault

[Bug middle-end/60469] simple cilk plus program ICEs

[andi-gcc at firstfloor dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60469

--- Comment #5 from Andi Kleen <andi-gcc at firstfloor dot org> ---
You're right. It works in C++.

That's similar to my earlier patch, but I didn't comment out the other check
like you did. If commenting out the check work it would seem right to me.

Can you post it as a RFC?

[Bug middle-end/60469] simple cilk plus program ICEs

[izamyatin at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60469

--- Comment #6 from Igor Zamyatin <izamyatin at gmail dot com> ---
Yes, I was going to post it after complete testing

[Bug middle-end/60469] simple cilk plus program ICEs

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60469

--- Comment #7 from H.J. Lu <hjl.tools at gmail dot com> ---
(In reply to Igor Zamyatin from comment #6)
> Yes, I was going to post it after complete testing

You should set DECL_SEEN_IN_BIND_EXPR_P when setting
DECL_CONTEXT, similar to gimple_add_tmp_var.

[Bug middle-end/60469] simple cilk plus program ICEs

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60469

--- Comment #8 from H.J. Lu <hjl.tools at gmail dot com> ---
(In reply to H.J. Lu from comment #7)
> (In reply to Igor Zamyatin from comment #6)
> > Yes, I was going to post it after complete testing
> 
> You should set DECL_SEEN_IN_BIND_EXPR_P when setting
> DECL_CONTEXT, similar to gimple_add_tmp_var.

Or we can use create_tmp_var.

[Bug middle-end/60469] simple cilk plus program ICEs

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60469

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

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

--- Comment #9 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
(In reply to H.J. Lu from comment #8)
> (In reply to H.J. Lu from comment #7)
> > (In reply to Igor Zamyatin from comment #6)
> > > Yes, I was going to post it after complete testing
> > 
> > You should set DECL_SEEN_IN_BIND_EXPR_P when setting
> > DECL_CONTEXT, similar to gimple_add_tmp_var.
> 
> Or we can use create_tmp_var.

That is much better idea, it will handle tons of other things, like setting
DECL_ARTIFICIAL/DECL_IGNORED_P flags etc.  In C++ FE, cp-array-notation.c
apparently uses get_temp_regvar, which is also fine (but only defined in C++
FE).

[Bug middle-end/60469] simple cilk plus program ICEs

[izamyatin at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60469

--- Comment #10 from Igor Zamyatin <izamyatin at gmail dot com> ---
(In reply to Jakub Jelinek from comment #9)
> (In reply to H.J. Lu from comment #8)
> > (In reply to H.J. Lu from comment #7)
> > > (In reply to Igor Zamyatin from comment #6)
> > > > Yes, I was going to post it after complete testing
> > > 
> > > You should set DECL_SEEN_IN_BIND_EXPR_P when setting
> > > DECL_CONTEXT, similar to gimple_add_tmp_var.
> > 
> > Or we can use create_tmp_var.
> 
> That is much better idea, it will handle tons of other things, like setting
> DECL_ARTIFICIAL/DECL_IGNORED_P flags etc.  In C++ FE, cp-array-notation.c
> apparently uses get_temp_regvar, which is also fine (but only defined in C++
> FE).

 Yes, I tried create_tmp_var but it was undefined so I thought it's not a good
idea...
Will try further with it then

[Bug middle-end/60469] simple cilk plus program ICEs

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60469

--- Comment #11 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
You need to include gimple-expr.h header for that.  But, if you look e.g. at
c/c-typeck.c or c-family/cilk.c, it is already used in the FEs.

[Bug middle-end/60469] simple cilk plus program ICEs

[izamyatin at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60469

--- Comment #12 from Igor Zamyatin <izamyatin at gmail dot com> ---
Thanks, will post a patch after the testing

[Bug middle-end/60469] simple cilk plus program ICEs

[kyukhin at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60469

--- Comment #13 from Kirill Yukhin <kyukhin at gcc dot gnu.org> ---
Author: kyukhin
Date: Sat Apr 12 17:57:15 2014
New Revision: 209336

URL: http://gcc.gnu.org/viewcvs?rev=209336&root=gcc&view=rev
Log:
gcc/c/
    PR middle-end/60469
    * c-array-notation.c (fix_builtin_array_notation_fn): Use
    create_tmp_var instead build_decl for creating temps.
    (build_array_notation_expr): Likewise.
    (fix_conditional_array_notations_1): Likewise.
    (fix_array_notation_expr): Likewise.
    (fix_array_notation_call_expr): Likewise.

gcc/testsuite/
    PR middle-end/60469
    * c-c++-common/cilk-plus/CK/pr60469.c: New test.


Added:
    trunk/gcc/testsuite/c-c++-common/cilk-plus/CK/pr60469.c
Modified:
    trunk/gcc/c/ChangeLog
    trunk/gcc/c/c-array-notation.c
    trunk/gcc/testsuite/ChangeLog

[Bug middle-end/60469] simple cilk plus program ICEs

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60469

--- Comment #14 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Please don't forget to backport the Cilk+ bugfixes to 4.9 branch after 4.9.0 is
released, as long as they are small and non-risky.

[Bug middle-end/60469] simple cilk plus program ICEs

[kyukhin at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60469

--- Comment #15 from Kirill Yukhin <kyukhin at gcc dot gnu.org> ---
Author: kyukhin
Date: Tue Apr 15 06:27:07 2014
New Revision: 209400

URL: http://gcc.gnu.org/viewcvs?rev=209400&root=gcc&view=rev
Log:
gcc/c/
    PR middle-end/60469
    * c-array-notation.c (fix_builtin_array_notation_fn): Use
    create_tmp_var instead build_decl for creating temps.
    (build_array_notation_expr): Likewise.
    (fix_conditional_array_notations_1): Likewise.
    (fix_array_notation_expr): Likewise.
    (fix_array_notation_call_expr): Likewise.

gcc/testsuite/
    PR middle-end/60469
    * c-c++-common/cilk-plus/CK/pr60469.c: New test.


Added:
    branches/gcc-4_9-branch/gcc/testsuite/c-c++-common/cilk-plus/CK/pr60469.c
Modified:
    branches/gcc-4_9-branch/gcc/c/ChangeLog
    branches/gcc-4_9-branch/gcc/c/c-array-notation.c
    branches/gcc-4_9-branch/gcc/testsuite/ChangeLog

[Bug middle-end/60469] simple cilk plus program ICEs

[andi-gcc at firstfloor dot org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=60469

Andi Kleen <andi-gcc at firstfloor dot org> changed:

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

--- Comment #16 from Andi Kleen <andi-gcc at firstfloor dot org> ---
Fixed since some time