[Bug tree-optimization/113736] New: ICE: verify_gimple failed: incompatible types in 'PHI' argument 0 with _BitInt() struct copy to __seg_fs/gs

[Bug tree-optimization/113736] New: ICE: verify_gimple failed: incompatible types in 'PHI' argument 0 with _BitInt() struct copy to __seg_fs/gs

[zsojka at seznam dot cz]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113736

            Bug ID: 113736
           Summary: ICE: verify_gimple failed: incompatible types in 'PHI'
                    argument 0 with _BitInt() struct copy to __seg_fs/gs
           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
  Target Milestone: ---
              Host: x86_64-pc-linux-gnu
            Target: x86_64-pc-linux-gnu

Created attachment 57303
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=57303&action=edit
reduced testcase (from gcc.target/i386/pr90856.c)

Compiler output:
$ x86_64-pc-linux-gnu-gcc -O2 testcase.c 
testcase.c: In function 'clearS2_n':
testcase.c:12:1: error: incompatible types in 'PHI' argument 0
   12 | clearS2_n (__seg_gs S2 *p)
      | ^~~~~~~~~
vector(2) unsigned long * {ref-all}

<address-space-2> struct S2 *

vectp_p.4_14 = PHI <p_2(D)(2), vectp_p.4_15(5)>
during GIMPLE pass: vect
testcase.c:12:1: internal compiler error: verify_gimple failed
0x1567165 verify_gimple_in_cfg(function*, bool, bool)
        /repo/gcc-trunk/gcc/tree-cfg.cc:5666
0x13cf7e4 execute_function_todo
        /repo/gcc-trunk/gcc/passes.cc:2088
0x13cfd3e 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-r14-8771-20240203001826-g4b7d4d8a4af-checking-yes-rtl-df-extra-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
--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-r14-8771-20240203001826-g4b7d4d8a4af-checking-yes-rtl-df-extra-amd64
Thread model: posix
Supported LTO compression algorithms: zlib zstd
gcc version 14.0.1 20240203 (experimental) (GCC)

[Bug tree-optimization/113736] ICE: verify_gimple failed: incompatible types in 'PHI' argument 0 with _BitInt() struct copy to __seg_fs/gs

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113736

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2024-02-03
                 CC|                            |rguenth at gcc dot gnu.org
     Ever confirmed|0                           |1
             Status|UNCONFIRMED                 |NEW

--- Comment #1 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
struct S { _BitInt(710) a; };
struct T { struct S b[2]; };

void
foo (__seg_gs struct T *p)
{
  struct S s;
  p->b[0] = s;
}

Bitint lowering changes here
  MEM <<address-space-2> _BitInt(768)> [(<address-space-2> struct T *)p_2(D)] =
s_4(D);
to
  VIEW_CONVERT_EXPR<unsigned long[12]>(MEM <<address-space-2> _BitInt(768)>
[(<address-space-2> struct T *)p_2(D)])[_5] = s_7(D);
accesses in a loop.  Is that invalid and should have <address-space-2> also in
the VCE type?  Or is this just a vectorizer bug?

[Bug tree-optimization/113736] ICE: verify_gimple failed: incompatible types in 'PHI' argument 0 with _BitInt() struct copy to __seg_fs/gs

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113736

--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
struct S { int a[4]; };
struct T { struct S b[16]; };

void
foo (__seg_gs struct T *p, int q)
{
  for (int i = 0; i < 16; ++i)
    {
      p->b[i].a[0] = q;
      p->b[i].a[1] = q;
      p->b[i].a[2] = q;
      p->b[i].a[3] = q;
    }
}

is vectorized right though (uses <address-space-2> vector(4) int * vectp_p.4).

[Bug tree-optimization/113736] ICE: verify_gimple failed: incompatible types in 'PHI' argument 0 with _BitInt() struct copy to __seg_fs/gs

[rguenther at suse dot de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113736

--- Comment #3 from rguenther at suse dot de <rguenther at suse dot de> ---
On Sat, 3 Feb 2024, jakub at gcc dot gnu.org wrote:

> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113736
> 
> Jakub Jelinek <jakub at gcc dot gnu.org> changed:
> 
>            What    |Removed                     |Added
> ----------------------------------------------------------------------------
>    Last reconfirmed|                            |2024-02-03
>                  CC|                            |rguenth at gcc dot gnu.org
>      Ever confirmed|0                           |1
>              Status|UNCONFIRMED                 |NEW
> 
> --- Comment #1 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
> struct S { _BitInt(710) a; };
> struct T { struct S b[2]; };
> 
> void
> foo (__seg_gs struct T *p)
> {
>   struct S s;
>   p->b[0] = s;
> }
> 
> Bitint lowering changes here
>   MEM <<address-space-2> _BitInt(768)> [(<address-space-2> struct T *)p_2(D)] =
> s_4(D);
> to
>   VIEW_CONVERT_EXPR<unsigned long[12]>(MEM <<address-space-2> _BitInt(768)>
> [(<address-space-2> struct T *)p_2(D)])[_5] = s_7(D);
> accesses in a loop.  Is that invalid and should have <address-space-2> also in
> the VCE type?  Or is this just a vectorizer bug?

I think that's OK, I will have a look.

[Bug tree-optimization/113736] ICE: verify_gimple failed: incompatible types in 'PHI' argument 0 with _BitInt() struct copy to __seg_fs/gs

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113736

--- Comment #4 from Richard Biener <rguenth at gcc dot gnu.org> ---
(In reply to rguenther@suse.de from comment #3)
> On Sat, 3 Feb 2024, jakub at gcc dot gnu.org wrote:
> > Bitint lowering changes here
> >   MEM <<address-space-2> _BitInt(768)> [(<address-space-2> struct T *)p_2(D)] =
> > s_4(D);
> > to
> >   VIEW_CONVERT_EXPR<unsigned long[12]>(MEM <<address-space-2> _BitInt(768)>
> > [(<address-space-2> struct T *)p_2(D)])[_5] = s_7(D);
> > accesses in a loop.  Is that invalid and should have <address-space-2> also in
> > the VCE type?  Or is this just a vectorizer bug?
> 
> I think that's OK, I will have a look.

I stand corrected - it isn't correct.  The address-space needs to be on
all types involved in a memory reference (RTL expansion is later quite
forgiving though).

This needs better documentation and maybe even IL checking I guess.

[Bug tree-optimization/113736] ICE: verify_gimple failed: incompatible types in 'PHI' argument 0 with _BitInt() struct copy to __seg_fs/gs

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113736

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

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

--- Comment #5 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Created attachment 57322
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=57322&action=edit
gcc14-pr113736.patch

Untested fix.

[Bug tree-optimization/113736] ICE: verify_gimple failed: incompatible types in 'PHI' argument 0 with _BitInt() struct copy to __seg_fs/gs

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113736

--- 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:483c061d699309d58a1b28ce5c00ee9b55a7365c

commit r14-8824-g483c061d699309d58a1b28ce5c00ee9b55a7365c
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Tue Feb 6 12:58:55 2024 +0100

    lower-bitint: Encode address space qualifiers in VIEW_CONVERT_EXPRs
[PR113736]

    As discussed in the PR, e.g. build_fold_addr_expr needs TYPE_ADDR_SPACE
    on the outermost reference rather than just on the base, so the
    following patch makes sure to propagate the address space from
    the accessed var to the MEM_REFs and/or VIEW_CONVERT_EXPRs used to
    access those.

    2024-02-06  Jakub Jelinek  <jakub@redhat.com>

            PR tree-optimization/113736
            * gimple-lower-bitint.cc (bitint_large_huge::limb_access): Use
            var's address space for MEM_REF or VIEW_CONVERT_EXPRs.

            * gcc.dg/bitint-86.c: New test.

[Bug tree-optimization/113736] ICE: verify_gimple failed: incompatible types in 'PHI' argument 0 with _BitInt() struct copy to __seg_fs/gs

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113736

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

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

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