[Bug c/110506] New: ice: tree check: expected none of vector_type, have vector_type in get_value_for_expr, at tree-ssa-ccp.cc:686

[Bug c/110506] New: ice: tree check: expected none of vector_type, have vector_type in get_value_for_expr, at tree-ssa-ccp.cc:686

[dcb314 at hotmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110506

            Bug ID: 110506
           Summary: ice: tree check: expected none of vector_type,  have
                    vector_type in get_value_for_expr, at
                    tree-ssa-ccp.cc:686
           Product: gcc
           Version: unknown
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
          Assignee: unassigned at gcc dot gnu.org
          Reporter: dcb314 at hotmail dot com
  Target Milestone: ---

For this C code:

struct {
  long *sp;
  long *csp
} neko_interp_loop_vm;
int neko_interp_loop_vm_2;
void neko_interp_loop() {
  long pc[] = {&&LabelAccGlobal, &&LabelPhysCompare, &&LabelTailCall,
               &&LabelLoop, &&LabelMakeArray2};
  long *sp, *csp = neko_interp_loop_vm.csp;
LabelAccGlobal:
  neko_interp_loop_vm.sp = sp;
  neko_interp_loop_vm.csp = csp;
  goto * 0;
LabelTailCall:
  csp = sp -= neko_interp_loop_vm_2;
LabelMakeArray2:
LabelPhysCompare:
LabelLoop:
  goto * 0;
}

compiled by g:9757e4440bd8755d, does this:

$ ~/gcc/results.20230628.asan.ubsan/bin/gcc -c -w -O2 bug936.c
during GIMPLE pass: ccp
/home/dcb38/rpmbuild/BUILD/neko-2-3-0/vm/interp.c: In function
‘neko_interp_loop’:
/home/dcb38/rpmbuild/BUILD/neko-2-3-0/vm/interp.c:613:9: internal compiler
error: tree check: expected none of vector_type, have vector_type in
get_value_for_expr, at tree-ssa-ccp.cc:686
  613 | int_val neko_interp_loop( neko_vm *VM_ARG, neko_module *m, int_val
_acc, int_val *_pc ) {
      |         ^~~~~~~~~~~~~~~~
0x11656f9 tree_not_check_failed(tree_node const*, char const*, int, char
const*, ...)
        ../../trunk.year/gcc/tree.cc:8936
0xf75774 tree_not_check(tree_node*, char const*, int, char const*, tree_code)
        ../../trunk.year/gcc/tree.h:3581
0xf75774 get_value_for_expr(tree_node*, bool)
        ../../trunk.year/gcc/tree-ssa-ccp.cc:686

It was fine with a version from a couple of days earlier:

foundBugs $ ~/gcc/results.20230626.asan.ubsan/bin/gcc -c -w -O2 bug936.c
foundBugs $ ~/gcc/results.20230626.asan.ubsan/bin/gcc -v
Using built-in specs.
COLLECT_GCC=/home/dcb38/gcc/results.20230626.asan.ubsan/bin/gcc
COLLECT_LTO_WRAPPER=/home/dcb38/gcc/results.20230626.asan.ubsan/libexec/gcc/x86_64-pc-linux-gnu/14.0.0/lto-wrapper
Target: x86_64-pc-linux-gnu
Configured with: ../trunk.year/configure
--prefix=/home/dcb38/gcc/results.20230626.asan.ubsan --disable-multilib
--disable-bootstrap --with-build-config=bootstrap-asan
--with-build-config=bootstrap-ubsan --with-pkgversion=3a39a31b8ae9c646
--enable-checking=df,extra,fold,rtl,yes --enable-languages=c,c++,fortran
Thread model: posix
Supported LTO compression algorithms: zlib zstd
gcc version 14.0.0 20230626 (experimental) (3a39a31b8ae9c646)

[Bug tree-optimization/110506] [14 Regression] Ice: tree check: expected none of vector_type, have vector_type in get_value_for_expr, at tree-ssa-ccp.cc:686

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110506

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|c                           |tree-optimization
           Keywords|                            |ice-on-valid-code
            Summary|ice: tree check: expected   |[14 Regression] Ice: tree
                   |none of vector_type,  have  |check: expected none of
                   |vector_type in              |vector_type,  have
                   |get_value_for_expr, at      |vector_type in
                   |tree-ssa-ccp.cc:686         |get_value_for_expr, at
                   |                            |tree-ssa-ccp.cc:686
   Target Milestone|---                         |14.0

[Bug tree-optimization/110506] [14 Regression] Ice: tree check: expected none of vector_type, have vector_type in get_value_for_expr, at tree-ssa-ccp.cc:686

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110506

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|needs-bisection             |

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
No reason to do a bisect as it was introduced by the extra checking commit.

[Bug tree-optimization/110506] [14 Regression] Ice: tree check: expected none of vector_type, have vector_type in get_value_for_expr, at tree-ssa-ccp.cc:686

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110506

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2023-06-30
             Status|UNCONFIRMED                 |NEW
                 CC|                            |rguenth at gcc dot gnu.org
     Ever confirmed|0                           |1

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Confirmed.

[Bug tree-optimization/110506] [14 Regression] Ice: tree check: expected none of vector_type, have vector_type in get_value_for_expr, at tree-ssa-ccp.cc:686

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110506

--- Comment #3 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Andrew Pinski from comment #1)
> No reason to do a bisect as it was introduced by the extra checking commit.

That is r14-2150 .

[Bug tree-optimization/110506] [14 Regression] Ice: tree check: expected none of vector_type, have vector_type in get_value_for_expr, at tree-ssa-ccp.cc:686

[dcb314 at hotmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110506

--- Comment #4 from David Binderman <dcb314 at hotmail dot com> ---
This C code might be related:

typedef unsigned uint32_t;
typedef uint32_t uint32x4 __attribute__((vector_size(16)));
typedef struct {
  uint32x4 b, d
} prng_t;
prng_t prng_rand_128_r_x;
main_flags;
main() {
  uint32_t ref_crc[] = {7, 3};
  uint32x4 e = (prng_rand_128_r_x.b << 27) + (prng_rand_128_r_x.b >> 32 - 27);
  prng_rand_128_r_x.d = e;
  if (ref_crc[main_flags])
    __assert_fail();
}

$ ~/gcc/results/bin/gcc -c -w -O2 bug936B.c
during GIMPLE pass: slp
bug936B.c: In function ‘main’:
bug936B.c:8:1: internal compiler error: tree check: expected none of
vector_type
, have vector_type in vect_recog_rotate_pattern, at tree-vect-patterns.cc:3729
    8 | main() {
      | ^~~~
0x1166129 tree_not_check_failed(tree_node const*, char const*, int, char
const*,
 ...)
        ../../trunk.year/gcc/tree.cc:8936
0x1cc8da3 vect_recog_rotate_pattern(vec_info*, _stmt_vec_info*, tree_node**)
        ../../trunk.year/gcc/tree-vect-patterns.cc:0

[Bug tree-optimization/110506] [14 Regression] Ice: tree check: expected none of vector_type, have vector_type in get_value_for_expr, at tree-ssa-ccp.cc:686

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110506

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
           Assignee|unassigned at gcc dot gnu.org      |rguenth at gcc dot gnu.org
            Version|unknown                     |14.0

--- Comment #5 from Richard Biener <rguenth at gcc dot gnu.org> ---
I will have a look.

[Bug tree-optimization/110506] [14 Regression] Ice: tree check: expected none of vector_type, have vector_type in get_value_for_expr, at tree-ssa-ccp.cc:686

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110506

--- Comment #6 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Richard Biener <rguenth@gcc.gnu.org>:

https://gcc.gnu.org/g:bd7e9856fe5bbeb487797476c4fffb660f63cf4f

commit r14-2255-gbd7e9856fe5bbeb487797476c4fffb660f63cf4f
Author: Richard Biener <rguenther@suse.de>
Date:   Mon Jul 3 08:28:17 2023 +0200

    tree-optimization/110506 - bogus non-zero mask in CCP for vector types

    get_value_for_expr was blindlessly using TYPE_PRECISION to produce
    a mask for vector typed entities which the new tree checking now
    catches.

            PR tree-optimization/110506
            * tree-ssa-ccp.cc (get_value_for_expr): Check for integral
            type before relying on TYPE_PRECISION to produce a nonzero mask.

            * gcc.dg/pr110506.c: New testcase.

[Bug tree-optimization/110506] [14 Regression] Ice: tree check: expected none of vector_type, have vector_type in get_value_for_expr, at tree-ssa-ccp.cc:686

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110506

--- Comment #7 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Richard Biener <rguenth@gcc.gnu.org>:

https://gcc.gnu.org/g:000590c074615cbfffb6ad854a6474e623801460

commit r14-2256-g000590c074615cbfffb6ad854a6474e623801460
Author: Richard Biener <rguenther@suse.de>
Date:   Mon Jul 3 09:17:06 2023 +0200

    tree-optimization/110506 - ICE in pattern recog with TYPE_PRECISION

    The following re-orders checks to make sure we check TYPE_PRECISION
    on an integral type.

            PR tree-optimization/110506
            * tree-vect-patterns.cc (vect_recog_rotate_pattern): Re-order
            TYPE_PRECISION access with INTEGRAL_TYPE_P check.

            * gcc.dg/pr110506-2.c: New testcase.

[Bug tree-optimization/110506] [14 Regression] Ice: tree check: expected none of vector_type, have vector_type in get_value_for_expr, at tree-ssa-ccp.cc:686

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110506

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

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

--- Comment #8 from Richard Biener <rguenth at gcc dot gnu.org> ---
Fixed.