[Bug tree-optimization/113406] New: ICE: tree check: expected class 'type', have 'exceptional' (error_mark) in useless_type_conversion_p, at gimple-expr.cc:85 with -fstrub=internal and _BitInt()

[Bug tree-optimization/113406] New: ICE: tree check: expected class 'type', have 'exceptional' (error_mark) in useless_type_conversion_p, at gimple-expr.cc:85 with -fstrub=internal and _BitInt()

[zsojka at seznam dot cz]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113406

            Bug ID: 113406
           Summary: ICE: tree check: expected class 'type', have
                    'exceptional' (error_mark) in
                    useless_type_conversion_p, at gimple-expr.cc:85 with
                    -fstrub=internal and _BitInt()
           Product: gcc
           Version: 14.0
            Status: UNCONFIRMED
          Keywords: ice-on-valid-code
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: zsojka at seznam dot cz
                CC: jakub at gcc dot gnu.org
  Target Milestone: ---
              Host: x86_64-pc-linux-gnu
            Target: x86_64-pc-linux-gnu

Created attachment 57092
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=57092&action=edit
reduced testcase

Compiler output:
$ x86_64-pc-linux-gnu-gcc -fstrub=internal testcase.c 
during GIMPLE pass: bitintlower0
testcase.c: In function 'foo':
testcase.c:2:1: internal compiler error: tree check: expected class 'type',
have 'exceptional' (error_mark) in useless_type_conversion_p, at
gimple-expr.cc:85
    2 | foo(void)
      | ^~~
0x89127d tree_class_check_failed(tree_node const*, tree_code_class, char
const*, int, char const*)
        /repo/gcc-trunk/gcc/tree.cc:9002
0x77c702 tree_class_check(tree_node*, tree_code_class, char const*, int, char
const*)
        /repo/gcc-trunk/gcc/tree.h:3767
0x77c702 useless_type_conversion_p(tree_node*, tree_node*)
        /repo/gcc-trunk/gcc/gimple-expr.cc:85
0x15540a4 verify_gimple_return
        /repo/gcc-trunk/gcc/tree-cfg.cc:4888
0x15540a4 verify_gimple_stmt
        /repo/gcc-trunk/gcc/tree-cfg.cc:5125
0x1558311 verify_gimple_in_cfg(function*, bool, bool)
        /repo/gcc-trunk/gcc/tree-cfg.cc:5602
0x13c83d4 execute_function_todo
        /repo/gcc-trunk/gcc/passes.cc:2088
0x13c892e execute_todo
        /repo/gcc-trunk/gcc/passes.cc:2142
Please submit a full bug report, with preprocessed source (by using
-freport-bug).
Please include the complete backtrace with any bug report.
See <https://gcc.gnu.org/bugs/> for instructions.

$ x86_64-pc-linux-gnu-gcc -v
Using built-in specs.
COLLECT_GCC=/repo/gcc-trunk/binary-latest-amd64/bin/x86_64-pc-linux-gnu-gcc
COLLECT_LTO_WRAPPER=/repo/gcc-trunk/binary-trunk-r14-7255-20240115165927-g6c703b4eb68-checking-yes-rtl-df-extra-nobootstrap-amd64/bin/../libexec/gcc/x86_64-pc-linux-gnu/14.0.1/lto-wrapper
Target: x86_64-pc-linux-gnu
Configured with: /repo/gcc-trunk//configure --enable-languages=c,c++
--enable-valgrind-annotations --disable-nls --enable-checking=yes,rtl,df,extra
--disable-bootstrap --with-cloog --with-ppl --with-isl
--build=x86_64-pc-linux-gnu --host=x86_64-pc-linux-gnu
--target=x86_64-pc-linux-gnu --with-ld=/usr/bin/x86_64-pc-linux-gnu-ld
--with-as=/usr/bin/x86_64-pc-linux-gnu-as --disable-libstdcxx-pch
--prefix=/repo/gcc-trunk//binary-trunk-r14-7255-20240115165927-g6c703b4eb68-checking-yes-rtl-df-extra-nobootstrap-amd64
Thread model: posix
Supported LTO compression algorithms: zlib zstd
gcc version 14.0.1 20240115 (experimental) (GCC)

[Bug middle-end/113406] ICE: tree check: expected class 'type', have 'exceptional' (error_mark) in useless_type_conversion_p, at gimple-expr.cc:85 with -fstrub=internal and _BitInt()

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113406

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2024-01-15
             Status|UNCONFIRMED                 |NEW
     Ever confirmed|0                           |1
          Component|tree-optimization           |middle-end

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Confirmed, scrub is changing the IR to:
```
  void * .strub.watermark.2;
  _BitInt(146) retval.1;

  <bb 2> [local count: 10000]:
  __builtin___strub_enter (&.strub.watermark.2);
  retval.1_3 = foo.strub.0 (&.strub.watermark.2);
  __builtin___strub_leave (&.strub.watermark.2);
  .strub.watermark.2 ={v} {CLOBBER};
  return retval.1_3;

```
Which looks fine but bitintlower is not expecting a gimple ssa of _BitInt(N) in
the GIMPLE_RETURN .

Here is a reduced _GIMPLE testcase without the need for -fstrub=internal :
```
_BitInt(146)
foo1(void);

_BitInt(146)
__GIMPLE()
foo(void) 
{
  _BitInt(146) t;
  t = foo1();
  return t;
}
```

[Bug middle-end/113406] ICE: tree check: expected class 'type', have 'exceptional' (error_mark) in useless_type_conversion_p, at gimple-expr.cc:85 with -fstrub=internal and _BitInt() (and __GIMPLE with _BitInt return)

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113406

--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
strub should be honoring aggregate_value_p, see e.g. PR112941

[Bug middle-end/113406] ICE: tree check: expected class 'type', have 'exceptional' (error_mark) in useless_type_conversion_p, at gimple-expr.cc:85 with -fstrub=internal and _BitInt() (and __GIMPLE with _BitInt return)

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113406

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|unassigned at gcc dot gnu.org      |jakub at gcc dot gnu.org
             Status|NEW                         |ASSIGNED

--- Comment #3 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Created attachment 57099
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=57099&action=edit
gcc14-pr113406.patch

Seems ipa-strub.cc contains a copy of the expand_thunk I've fixed some weeks
ago.

[Bug middle-end/113406] ICE: tree check: expected class 'type', have 'exceptional' (error_mark) in useless_type_conversion_p, at gimple-expr.cc:85 with -fstrub=internal and _BitInt() (and __GIMPLE with _BitInt return)

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113406

--- Comment #4 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Jakub Jelinek from comment #3)
> Created attachment 57099 [details]
> gcc14-pr113406.patch
> 
> Seems ipa-strub.cc contains a copy of the expand_thunk I've fixed some weeks
> ago.

But the gimple testcase still fails ...

[Bug middle-end/113406] ICE: tree check: expected class 'type', have 'exceptional' (error_mark) in useless_type_conversion_p, at gimple-expr.cc:85 with -fstrub=internal and _BitInt() (and __GIMPLE with _BitInt return)

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113406

--- Comment #5 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
(In reply to Andrew Pinski from comment #4)
> (In reply to Jakub Jelinek from comment #3)
> > Created attachment 57099 [details]
> > gcc14-pr113406.patch
> > 
> > Seems ipa-strub.cc contains a copy of the expand_thunk I've fixed some weeks
> > ago.
> 
> But the gimple testcase still fails ...

Because it is incorrect.  I think it needs to use <result> or something like
that.

[Bug middle-end/113406] ICE: tree check: expected class 'type', have 'exceptional' (error_mark) in useless_type_conversion_p, at gimple-expr.cc:85 with -fstrub=internal and _BitInt() (and __GIMPLE with _BitInt return)

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113406

--- Comment #6 from GCC Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Jakub Jelinek <jakub@gcc.gnu.org>:

https://gcc.gnu.org/g:2c74d5c0a0f21ae1d0c195ca0d81a3d1032f27e3

commit r14-8183-g2c74d5c0a0f21ae1d0c195ca0d81a3d1032f27e3
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Wed Jan 17 13:54:44 2024 +0100

    ipa-strub: Fix handling of _BitInt returns [PR113406]

    Seems pass_ipa_strub::execute contains a copy of the expand_thunk
    code I've changed for _BitInt in r14-6805 PR112941 - larger _BitInts
    are aggregate_value_p even when they are is_gimple_reg_type.

    2024-01-17  Jakub Jelinek  <jakub@redhat.com>

            PR middle-end/113406
            * ipa-strub.cc (pass_ipa_strub::execute): Check aggregate_value_p
            regardless of whether is_gimple_reg_type (restype) or not.

            * gcc.dg/bitint-70.c: New test.

[Bug middle-end/113406] ICE: tree check: expected class 'type', have 'exceptional' (error_mark) in useless_type_conversion_p, at gimple-expr.cc:85 with -fstrub=internal and _BitInt() (and __GIMPLE with _BitInt return)

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113406

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

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

--- Comment #7 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Fixed.

[Bug middle-end/113406] ICE: tree check: expected class 'type', have 'exceptional' (error_mark) in useless_type_conversion_p, at gimple-expr.cc:85 with -fstrub=internal and _BitInt() (and __GIMPLE with _BitInt return)

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113406

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |14.0