[Bug c++/99062] New: [10/11 Regression] ICE in tree_to_uhwi, at tree.h:4579

[Bug c++/99062] New: [10/11 Regression] ICE in tree_to_uhwi, at tree.h:4579

[gscfq@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99062

            Bug ID: 99062
           Summary: [10/11 Regression] ICE in tree_to_uhwi, at tree.h:4579
           Product: gcc
           Version: 11.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: gscfq@t-online.de
  Target Milestone: ---

Changed between 20190512 and 20190519 :


$ cat z1.cc
enum { a = 1 << 31 };
void *f () __attribute__ ((assume_aligned (a, 4)));


$ g++-11-20210207 -c z1.cc
z1.cc:2:50: internal compiler error: in tree_to_uhwi, at tree.h:4579
    2 | void *f () __attribute__ ((assume_aligned (a, 4)));
      |                                                  ^
0x8835a6 tree_to_uhwi(tree_node const*)
        ../../gcc/tree.h:4579
0x8835a6 handle_assume_aligned_attribute
        ../../gcc/c-family/c-attribs.c:3565
0x823311 decl_attributes(tree_node**, tree_node*, int, tree_node*)
        ../../gcc/attribs.c:724
0x6e6fc7 cplus_decl_attributes(tree_node**, tree_node*, int)
        ../../gcc/cp/decl2.c:1596
0x6272a2 grokfndecl
        ../../gcc/cp/decl.c:9983
0x6d5261 grokdeclarator(cp_declarator const*, cp_decl_specifier_seq*,
decl_context, int, tree_node**)
        ../../gcc/cp/decl.c:13818
0x6da5f8 start_decl(cp_declarator const*, cp_decl_specifier_seq*, int,
tree_node*, tree_node*, tree_node**)
        ../../gcc/cp/decl.c:5333
0x79bffb cp_parser_init_declarator
        ../../gcc/cp/parser.c:21660
0x77cf3a cp_parser_simple_declaration
        ../../gcc/cp/parser.c:14381
0x7a0ea5 cp_parser_declaration
        ../../gcc/cp/parser.c:14078
0x7a03db cp_parser_translation_unit
        ../../gcc/cp/parser.c:4936
0x7a03db c_parse_file()
        ../../gcc/cp/parser.c:45179
0x86d982 c_common_parse_file()
        ../../gcc/c-family/c-opts.c:1218

[Bug c++/99062] [10/11 Regression] ICE in tree_to_uhwi, at tree.h:4579

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99062

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mpolacek at gcc dot gnu.org,
                   |                            |msebor at gcc dot gnu.org
   Last reconfirmed|                            |2021-02-10
           Keywords|                            |ice-on-valid-code
             Status|UNCONFIRMED                 |NEW
     Ever confirmed|0                           |1
   Target Milestone|---                         |10.3

--- Comment #1 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Started with r271338.

[Bug c++/99062] [10/11 Regression] ICE in tree_to_uhwi, at tree.h:4579

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99062

--- Comment #2 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Must be this part:

@@ -2935,8 +2936,8 @@ handle_assume_aligned_attribute (tree *node, tree name,
tree args, int,
          /* The misalignment specified by the second argument
             must be non-negative and less than the alignment.  */
          warning (OPT_Wattributes,
-                  "%qE attribute argument %E is not in the range [0, %E)",
-                  name, val, align);
+                  "%qE attribute argument %E is not in the range [0, %wu]",
+                  name, val, tree_to_uhwi (align) - 1);
          *no_add_attrs = true;
          return NULL_TREE;
        }

[Bug c++/99062] [10/11 Regression] ICE in tree_to_uhwi, at tree.h:4579

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99062

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|unassigned at gcc dot gnu.org      |mpolacek at gcc dot gnu.org
             Status|NEW                         |ASSIGNED
           Priority|P3                          |P2

--- Comment #3 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
So we need to check that ALIGN is sane.

[Bug c++/99062] [10/11 Regression] ICE in tree_to_uhwi, at tree.h:4579

[msebor at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99062

--- Comment #4 from Martin Sebor <msebor at gcc dot gnu.org> ---
There already is a test in the attribute handler to validate the alignment: it
calls integer_pow2p (val).  Val here is:
  <integer_cst 0x7fffea7f7ed0 type <integer_type 0x7fffea8105e8 int> constant
-2147483648>
but integer_pow2p (val) returns true.  I'd expect the function to return false
since INT_MIN is not a power of two.  Anyway, the test integer_pow2p (val)
should probably be changed to tree_int_cst_sgn (val) < 0 || integer_pow2p
(val).

Separately, the warning should be enhanced to understand that [x, y) is a
half-open range and not complain; then the code can go back to the original [0,
%E) without any additional checking.

[Bug c++/99062] [10/11 Regression] ICE in tree_to_uhwi, at tree.h:4579

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99062

--- Comment #5 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
(In reply to Martin Sebor from comment #4)
> There already is a test in the attribute handler to validate the alignment:
> it calls integer_pow2p (val).  Val here is:
>   <integer_cst 0x7fffea7f7ed0 type <integer_type 0x7fffea8105e8 int>
> constant -2147483648>
> but integer_pow2p (val) returns true.  I'd expect the function to return
> false since INT_MIN is not a power of two.  Anyway, the test integer_pow2p
> (val) should probably be changed to tree_int_cst_sgn (val) < 0 ||
> integer_pow2p (val).

About to post a patch that hoists the tree_int_cst_sgn check so that we don't
have to repeat it.

[Bug c++/99062] [10/11 Regression] ICE in tree_to_uhwi, at tree.h:4579

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99062

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

https://gcc.gnu.org/g:30a4d95bf76b0a0fdb66ac0211589a4434c83af3

commit r11-7259-g30a4d95bf76b0a0fdb66ac0211589a4434c83af3
Author: Marek Polacek <polacek@redhat.com>
Date:   Wed Feb 10 15:24:25 2021 -0500

    c-family: ICE with assume_aligned attribute [PR99062]

    We ICE in handle_assume_aligned_attribute since r271338 which added

    @@ -2935,8 +2936,8 @@ handle_assume_aligned_attribute (tree *node, tree
name, tree args, int,
              /* The misalignment specified by the second argument
                 must be non-negative and less than the alignment.  */
              warning (OPT_Wattributes,
    -                  "%qE attribute argument %E is not in the range [0, %E)",
    -                  name, val, align);
    +                  "%qE attribute argument %E is not in the range [0,
%wu]",
    +                  name, val, tree_to_uhwi (align) - 1);
              *no_add_attrs = true;
              return NULL_TREE;
            }
    because align is INT_MIN and tree_to_uhwi asserts tree_fits_uhwi_p -- which
    ALIGN does not and the prior tree_fits_shwi_p check is fine with it, as
    well as the integer_pow2p check.

    Since neither of the arguments to assume_aligned can be negative, I've
    hoisted the tree_int_cst_sgn check.  And add the missing "argument"
    word to an existing warning.

    gcc/c-family/ChangeLog:

            PR c++/99062
            * c-attribs.c (handle_assume_aligned_attribute): Check that the
            alignment argument is non-negative.  Tweak a warning message.

    gcc/testsuite/ChangeLog:

            PR c++/99062
            * gcc.dg/attr-assume_aligned-4.c: Adjust dg-warning.
            * g++.dg/ext/attr-assume-aligned.C: New test.

[Bug c++/99062] [10 Regression] ICE in tree_to_uhwi, at tree.h:4579

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99062

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[10/11 Regression] ICE in   |[10 Regression] ICE in
                   |tree_to_uhwi, at            |tree_to_uhwi, at
                   |tree.h:4579                 |tree.h:4579

--- Comment #7 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Fixed on trunk so far.

[Bug c++/99062] [10 Regression] ICE in tree_to_uhwi, at tree.h:4579

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99062

--- Comment #8 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-10 branch has been updated by Richard Biener
<rguenth@gcc.gnu.org>:

https://gcc.gnu.org/g:9d020d111fbeae228b20d75a8c832b214e561ab3

commit r10-9553-g9d020d111fbeae228b20d75a8c832b214e561ab3
Author: Marek Polacek <polacek@redhat.com>
Date:   Wed Feb 10 15:24:25 2021 -0500

    c-family: ICE with assume_aligned attribute [PR99062]

    We ICE in handle_assume_aligned_attribute since r271338 which added

    @@ -2935,8 +2936,8 @@ handle_assume_aligned_attribute (tree *node, tree
name, tree args, int,
              /* The misalignment specified by the second argument
                 must be non-negative and less than the alignment.  */
              warning (OPT_Wattributes,
    -                  "%qE attribute argument %E is not in the range [0, %E)",
    -                  name, val, align);
    +                  "%qE attribute argument %E is not in the range [0,
%wu]",
    +                  name, val, tree_to_uhwi (align) - 1);
              *no_add_attrs = true;
              return NULL_TREE;
            }
    because align is INT_MIN and tree_to_uhwi asserts tree_fits_uhwi_p -- which
    ALIGN does not and the prior tree_fits_shwi_p check is fine with it, as
    well as the integer_pow2p check.

    Since neither of the arguments to assume_aligned can be negative, I've
    hoisted the tree_int_cst_sgn check.  And add the missing "argument"
    word to an existing warning.

    gcc/c-family/ChangeLog:

            PR c++/99062
            * c-attribs.c (handle_assume_aligned_attribute): Check that the
            alignment argument is non-negative.  Tweak a warning message.

    gcc/testsuite/ChangeLog:

            PR c++/99062
            * gcc.dg/attr-assume_aligned-4.c: Adjust dg-warning.
            * g++.dg/ext/attr-assume-aligned.C: New test.

    (cherry picked from commit 30a4d95bf76b0a0fdb66ac0211589a4434c83af3)

[Bug c++/99062] [10 Regression] ICE in tree_to_uhwi, at tree.h:4579

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99062

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
             Status|ASSIGNED                    |RESOLVED
      Known to fail|                            |10.2.0

--- Comment #9 from Richard Biener <rguenth at gcc dot gnu.org> ---
Fixed.