[Bug target/109987] New: ICE in in rs6000_emit_le_vsx_store on ppc64le with -Ofast -mno-power8-vector

[Bug target/109987] New: ICE in in rs6000_emit_le_vsx_store on ppc64le with -Ofast -mno-power8-vector

[jamborm at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109987

            Bug ID: 109987
           Summary: ICE in in rs6000_emit_le_vsx_store on ppc64le with
                    -Ofast -mno-power8-vector
           Product: gcc
           Version: 14.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: jamborm at gcc dot gnu.org
  Target Milestone: ---
              Host: x86_64-linux
            Target: ppc64le-linux-gnu

With a cross compiler (revision r14-1248-gd156c605420023)
configured with:

/home/worker/buildworker/tiber-gcc-trunk-ppc64le/build/configure
--enable-languages=c,c++,fortran,rust,m2 --disable-bootstrap
--disable-libsanitizer --disable-multilib --enable-checking=release
--prefix=/home/worker/cross --target=ppc64le-linux-gnu
--with-as=/usr/bin/powerpc64le-suse-linux-as

and our existing testcase gcc/testsuite/gcc.target/powerpc/pr102347.c, running
it as

~/cross/bin/ppc64le-linux-gnu-gcc
/home/worker/buildworker/tiber-option-juggler/build/gcc/testsuite/gcc.target/powerpc/pr102347.c
-Ofast -mno-power8-vector

Results in ICE:

during RTL pass: split2
/home/worker/buildworker/tiber-option-juggler/build/gcc/testsuite/gcc.target/powerpc/pr102347.c:
In function ‘main’:
/home/worker/buildworker/tiber-option-juggler/build/gcc/testsuite/gcc.target/powerpc/pr102347.c:15:1:
internal compiler error: in rs6000_emit_le_vsx_store, at
config/rs6000/rs6000.cc:10547
   15 | }
      | ^
0x63d7f7 rs6000_emit_le_vsx_store(rtx_def*, rtx_def*, machine_mode)
       
/home/worker/buildworker/tiber-gcc-trunk-ppc64le/build/gcc/config/rs6000/rs6000.cc:10547
0x1293d37 gen_movv16qi(rtx_def*, rtx_def*)
       
/home/worker/buildworker/tiber-gcc-trunk-ppc64le/build/gcc/config/rs6000/vector.md:161
0x8938b7 rtx_insn* insn_gen_fn::operator()<rtx_def*, rtx_def*>(rtx_def*,
rtx_def*) const
        /home/worker/buildworker/tiber-gcc-trunk-ppc64le/build/gcc/recog.h:407
0x8938b7 emit_move_ccmode
        /home/worker/buildworker/tiber-gcc-trunk-ppc64le/build/gcc/expr.cc:4043
0x8938b7 emit_move_insn_1(rtx_def*, rtx_def*)
        /home/worker/buildworker/tiber-gcc-trunk-ppc64le/build/gcc/expr.cc:4188
0x893c9d emit_move_insn(rtx_def*, rtx_def*)
        /home/worker/buildworker/tiber-gcc-trunk-ppc64le/build/gcc/expr.cc:4339
0x12c925b gen_split_707(rtx_insn*, rtx_def**)
       
/home/worker/buildworker/tiber-gcc-trunk-ppc64le/build/gcc/config/rs6000/mma.md:498
0x136dcda split_14
       
/home/worker/buildworker/tiber-gcc-trunk-ppc64le/build/gcc/config/rs6000/mma.md:494
0x136dcda split_18
       
/home/worker/buildworker/tiber-gcc-trunk-ppc64le/build/gcc/config/rs6000/rs6000.md:471
0x861170 try_split(rtx_def*, rtx_insn*, int)
       
/home/worker/buildworker/tiber-gcc-trunk-ppc64le/build/gcc/emit-rtl.cc:3804
0xb5af71 split_insn
       
/home/worker/buildworker/tiber-gcc-trunk-ppc64le/build/gcc/recog.cc:3385
0xb60221 split_all_insns()
       
/home/worker/buildworker/tiber-gcc-trunk-ppc64le/build/gcc/recog.cc:3489
0xb60308 execute
       
/home/worker/buildworker/tiber-gcc-trunk-ppc64le/build/gcc/recog.cc:4413

[Bug target/109987] ICE in in rs6000_emit_le_vsx_store on ppc64le with -Ofast -mno-power8-vector

[linkw at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109987

Kewen Lin <linkw at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever confirmed|0                           |1
           Keywords|                            |ice-on-valid-code
                 CC|                            |bergner at gcc dot gnu.org,
                   |                            |linkw at gcc dot gnu.org,
                   |                            |segher at gcc dot gnu.org
   Last reconfirmed|                            |2023-05-29

--- Comment #1 from Kewen Lin <linkw at gcc dot gnu.org> ---
Confirmed, it's similar to what the issue was found in PR103627 #c4, previously
I made a patch to make MMA feature require power9-vector, see
https://gcc.gnu.org/pipermail/gcc-patches/2021-December/587310.html. But Segher
thought power9-vector is a workaround option, we should make it go away, so
just guard it under vsx, see his comment
https://gcc.gnu.org/pipermail/gcc-patches/2022-January/589303.html.

Unfortunately this issue is specified another workaround option
-mno-power8-vector, I think we probably need to put -mpower{8,9}-vector removal
in a high priority.

[Bug target/109987] ICE in in rs6000_emit_le_vsx_store on ppc64le with -Ofast -mno-power8-vector

[linkw at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109987

Kewen Lin <linkw at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |fkastl at suse dot cz

--- Comment #2 from Kewen Lin <linkw at gcc dot gnu.org> ---
*** Bug 113115 has been marked as a duplicate of this bug. ***

[Bug target/109987] ICE in in rs6000_emit_le_vsx_store on ppc64le with -Ofast -mno-power8-vector

[linkw at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109987

Kewen Lin <linkw at gcc dot gnu.org> changed:

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

--- Comment #3 from Kewen Lin <linkw at gcc dot gnu.org> ---
As discussed in PR113115, I'm going to give option power{8,9}-vector removal a
shot.

[Bug target/109987] ICE in in rs6000_emit_le_vsx_store on ppc64le with -Ofast -mno-power8-vector

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109987

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
   Target Milestone|---                         |14.0
             Status|ASSIGNED                    |RESOLVED

--- Comment #4 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Fixed via r14-9129-g438ef143679e84 (I don't know why the commit message didn't
show up here).