[Bug c++/115439] New: [Regression] ICEs after 7ca35f2e43 on master-thumb_m55_hard_eabi

[Bug c++/115439] New: [Regression] ICEs after 7ca35f2e43 on master-thumb_m55_hard_eabi

[carlos.seo at linaro dot org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115439

            Bug ID: 115439
           Summary: [Regression] ICEs after 7ca35f2e43 on
                    master-thumb_m55_hard_eabi
           Product: gcc
           Version: 15.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: carlos.seo at linaro dot org
  Target Milestone: ---

Commit 7ca35f2e43 (internal-fn: Do not force vcond_mask operands to reg) causes
the following crashes in the testsuite on master-thumb_m55_hard_eabi

FAIL: 18 regressions

regressions.sum:
                === g++ tests ===

Running g++:g++.dg/dg.exp ...
FAIL: c-c++-common/vector-compare-3.c -std=gnu++14 (internal compiler error: in
extract_insn, at recog.cc:2812)
FAIL: c-c++-common/vector-compare-3.c -std=gnu++14 (test for excess errors)
FAIL: c-c++-common/vector-compare-3.c -std=gnu++17 (internal compiler error: in
extract_insn, at recog.cc:2812)
FAIL: c-c++-common/vector-compare-3.c -std=gnu++17 (test for excess errors)
FAIL: c-c++-common/vector-compare-3.c -std=gnu++20 (internal compiler error: in
extract_insn, at recog.cc:2812)
FAIL: c-c++-common/vector-compare-3.c -std=gnu++20 (test for excess errors)
FAIL: c-c++-common/vector-compare-3.c -std=gnu++98 (internal compiler error: in
extract_insn, at recog.cc:2812)
FAIL: c-c++-common/vector-compare-3.c -std=gnu++98 (test for excess errors)
FAIL: g++.dg/opt/pr79734.C -std=gnu++14 (internal compiler error: in
extract_insn, at recog.cc:2812)
FAIL: g++.dg/opt/pr79734.C -std=gnu++14 (test for excess errors)
FAIL: g++.dg/opt/pr79734.C -std=gnu++17 (internal compiler error: in
extract_insn, at recog.cc:2812)
FAIL: g++.dg/opt/pr79734.C -std=gnu++17 (test for excess errors)
FAIL: g++.dg/opt/pr79734.C -std=gnu++20 (internal compiler error: in
extract_insn, at recog.cc:2812)
FAIL: g++.dg/opt/pr79734.C -std=gnu++20 (test for excess errors)
FAIL: g++.dg/opt/pr79734.C -std=gnu++98 (internal compiler error: in
extract_insn, at recog.cc:2812)
FAIL: g++.dg/opt/pr79734.C -std=gnu++98 (test for excess errors)
                === gcc tests ===

Running gcc:gcc.dg/dg.exp ...
FAIL: c-c++-common/vector-compare-3.c -Wc++-compat  (internal compiler error:
in extract_insn, at recog.cc:2812)
FAIL: c-c++-common/vector-compare-3.c -Wc++-compat  (test for excess errors)

Details:
https://ci.linaro.org/job/tcwg_gnu_embed_check_gcc--master-thumb_m55_hard_eabi-build/436/artifact/artifacts/notify/mail-body.txt/*view*/

[Bug target/115439] [15 Regression] ICEs after r15-638 on master-thumb_m55_hard_eabi

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115439

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Target|                            |arm-eabi
          Component|c++                         |target
            Summary|[Regression] ICEs after     |[15 Regression] ICEs after
                   |7ca35f2e43 on               |r15-638 on
                   |master-thumb_m55_hard_eabi  |master-thumb_m55_hard_eabi
           Keywords|                            |ice-on-valid-code
   Target Milestone|---                         |15.0

[Bug target/115439] [15 Regression] ICEs after r15-638 on master-thumb_m55_hard_eabi

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115439

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
https://gcc.gnu.org/pipermail/gcc-patches/2024-June/654086.html

[Bug target/115439] [15 Regression] ICEs after r15-638 on master-thumb_m55_hard_eabi

[rdapp at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115439

Robin Dapp <rdapp at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |rdapp at gcc dot gnu.org

--- Comment #2 from Robin Dapp <rdapp at gcc dot gnu.org> ---
We already discussed that briefly on the mailing list.  Richard Sandiford
suspected a target bug.  What options do I need to configure GCC in order to
reproduce that locally?

[Bug target/115439] [15 Regression] ICEs after r15-638 on master-thumb_m55_hard_eabi

[carlos.seo at linaro dot org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115439

--- Comment #3 from Carlos Eduardo Seo <carlos.seo at linaro dot org> ---
(In reply to Robin Dapp from comment #2)
> We already discussed that briefly on the mailing list.  Richard Sandiford
> suspected a target bug.  What options do I need to configure GCC in order to
> reproduce that locally?

You can reproduce exactly what is being done by the CI by following:
https://git-us.linaro.org/toolchain/ci/interesting-commits.git/plain/gcc/sha1/7ca35f2e430081d6ec91e910002f92d9713350fa/tcwg_gnu_embed_check_gcc/master-thumb_m55_hard_eabi/reproduction_instructions.txt

If you prefer to do it manually:

configure  SHELL=/bin/bash \
       
--with-mpc=/home/tcwg-buildslave/workspace/tcwg_gnu_1/abe/builds/destdir/x86_64-pc-linux-gnu
\
       
--with-mpfr=/home/tcwg-buildslave/workspace/tcwg_gnu_1/abe/builds/destdir/x86_64-pc-linux-gnu
\
       
--with-gmp=/home/tcwg-buildslave/workspace/tcwg_gnu_1/abe/builds/destdir/x86_64-pc-linux-gnu
\
        --with-gnu-as \
        --with-gnu-ld \
        --disable-libmudflap \
        --enable-lto \
        --enable-shared \
        --without-included-gettext \
        --enable-nls \
        --with-system-zlib \
        --disable-sjlj-exceptions \
        --enable-gnu-unique-object \
        --enable-linker-build-id \
        --disable-libstdcxx-pch \
        --enable-c99 \
        --enable-clocale=gnu \
        --enable-libstdcxx-debug \
        --enable-long-long \
        --with-cloog=no \
        --with-ppl=no \
        --with-isl=no \
        --enable-threads=no \
        --disable-multiarch \
        --disable-multilib \
        --with-mode=thumb \
        --with-arch=armv8.1-m.main+mve.fp+fp.dp \
        --with-float=hard \
       
--with-sysroot=/home/tcwg-buildslave/workspace/tcwg_gnu_1/abe/builds/destdir/x86_64-pc-linux-gnu/arm-eabi
\
        --with-headers=yes \
        --with-native-system-header-dir=/include \
        --with-newlib \
        --enable-checking=yes \
        --disable-bootstrap \
        --enable-languages=c,c++,lto \
       
--prefix=/home/tcwg-buildslave/workspace/tcwg_gnu_1/abe/builds/destdir/x86_64-pc-linux-gnu
\
        --build=x86_64-pc-linux-gnu \
        --host=x86_64-pc-linux-gnu \
        --target=arm-eabi

[Bug target/115439] [15 Regression] ICEs after r15-638 on master-thumb_m55_hard_eabi

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115439

--- Comment #4 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
I think something like this:
```
[apinski@xeond2 arm]$ git diff mve.md
diff --git a/gcc/config/arm/mve.md b/gcc/config/arm/mve.md
index 9fe51298cdc..8af4797e935 100644
--- a/gcc/config/arm/mve.md
+++ b/gcc/config/arm/mve.md
@@ -6897,6 +6897,12 @@ (define_expand "vcond_mask_<mode><MVE_vpred>"
          (match_operand:MVE_VLD_ST 2 "s_register_operand")))]
   "TARGET_HAVE_MVE"
 {
+  if (!register_operand (operands[3], <MVE_VPRED>mode))
+    operands[3] = force_reg (operands[3]);
+  if (!register_operand (operands[1], <MODE>mode))
+    operands[1] = force_reg (operands[1]);
+  if (!register_operand (operands[2], <MODE>mode))
+    operands[1] = force_reg (operands[2]);
   switch (GET_MODE_CLASS (<MODE>mode))
     {
       case MODE_VECTOR_INT:

```
Signed-off-by: Andrew Pinski  <quic_apinski@quicinc.com>

[Bug target/115439] [15 Regression] ICEs after r15-638 on master-thumb_m55_hard_eabi

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115439

--- Comment #5 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Andrew Pinski from comment #4)
> I think something like this:
> ```
> [apinski@xeond2 arm]$ git diff mve.md
> diff --git a/gcc/config/arm/mve.md b/gcc/config/arm/mve.md
> index 9fe51298cdc..8af4797e935 100644
> --- a/gcc/config/arm/mve.md
> +++ b/gcc/config/arm/mve.md
> @@ -6897,6 +6897,12 @@ (define_expand "vcond_mask_<mode><MVE_vpred>"
>           (match_operand:MVE_VLD_ST 2 "s_register_operand")))]
>    "TARGET_HAVE_MVE"
>  {
> +  if (!register_operand (operands[3], <MVE_VPRED>mode))
> +    operands[3] = force_reg (operands[3]);
> +  if (!register_operand (operands[1], <MODE>mode))
> +    operands[1] = force_reg (operands[1]);
> +  if (!register_operand (operands[2], <MODE>mode))
> +    operands[1] = force_reg (operands[2]);

copy and pasto:`operands[1] = force_reg (operands[1]);` is the correct line for
above.

>    switch (GET_MODE_CLASS (<MODE>mode))
>      {
>        case MODE_VECTOR_INT:
> 
> ```
> Signed-off-by: Andrew Pinski  <quic_apinski@quicinc.com>

[Bug target/115439] [15 Regression] ICEs after r15-638 on master-thumb_m55_hard_eabi

[rdapp at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115439

--- Comment #6 from Robin Dapp <rdapp at gcc dot gnu.org> ---
Looks reasonable.  That's what we were doing before in internal-fn.cc before
expanding (except operands[2]).

Are you going to post a patch?

[Bug target/115439] [15 Regression] ICEs after r15-638 on master-thumb_m55_hard_eabi

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115439

--- Comment #7 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Robin Dapp from comment #6)
> Are you going to post a patch?

I was going to let the arm folks take care of it since I have no way to test
it.