[Bug target/113611] New: [14 Regression] gcc.dg/pr110279-1.c fails on cross build since gcc-14-5779-g746344dd538

[Bug target/113611] New: [14 Regression] gcc.dg/pr110279-1.c fails on cross build since gcc-14-5779-g746344dd538

[thiago.bauermann at linaro dot org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113611

            Bug ID: 113611
           Summary: [14 Regression] gcc.dg/pr110279-1.c fails on cross
                    build since gcc-14-5779-g746344dd538
           Product: gcc
           Version: 14.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: thiago.bauermann at linaro dot org
                CC: dizhao at os dot amperecomputing.com
  Target Milestone: ---
            Target: arm-linux-gnueabihf

After commit g:746344dd5380 ("swap ops in reassoc to reduce cross backedge
FMA") the following failure started appearing on cross builds of 32 bits Arm:

Running gcc:gcc.dg/dg.exp ...
FAIL: gcc.dg/pr110279-1.c scan-tree-dump-times widening_mul "Generated FMA" 3

We're seeing it with toolchains built with --host=x86_64-linux-gnu and targets
arm-linux-gnueabihf and arm-none-eabi. Both targets with and without
--with-mode=thumb.

Interestingly, with a native compiler (with --host=arm-linux-gnueabihf and
--target=arm-linux-gnueabihf) I can't reproduce the problem.

I tested on today's trunk (commit ffeab69e1ffc) and the failures are still
present.

Here's how to reproduce on an x86_64-linux machine with
--target=arm-linux-gnueabihf:

1. Build and install GCC:

$ ~/src/gcc/configure" \
    SHELL=/bin/bash \
    --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 \
    --disable-multilib \
    --with-float=hard \
    --with-fpu=vfpv3-d16 \
    --with-tune=cortex-a9 \
    --with-arch=armv7-a \
    --enable-threads=posix \
    --enable-multiarch \
    --enable-libstdcxx-time=yes \
    --enable-gnu-indirect-function \
    --with-sysroot=/var/tmp/sysroot-arm-linux-gnueabihf \
    --enable-checking=yes \
    --disable-bootstrap \
    --enable-languages=default \
    --prefix=/tmp/arm-linux-gnueabihf \
    --build=x86_64-pc-linux-gnu \
    --host=x86_64-pc-linux-gnu \
    --target=arm-linux-gnueabihf \
&& make \
        SHELL=/bin/bash \
        -w \
        -j $(nproc) \
        CFLAGS_FOR_BUILD="-pipe -g -O2" \
        CXXFLAGS_FOR_BUILD="-pipe -g -O2" \
        LDFLAGS_FOR_BUILD="-static-libgcc" \
        MAKEINFOFLAGS=--force \
        BUILD_INFO="" \
        MAKEINFO=echo \
&& make install

2. Finally, use it to compile the problematic .c file:

$ /tmp/arm-linux-gnueabihf/bin/arm-linux-gnueabihf-gcc
/home/bauermann/src/gcc/gcc/testsuite/gcc.dg/pr110279-1.c
-fdiagnostics-plain-output -Ofast --param avoid-fma-max-bits=512 --param
tree-reassoc-width=4 -fdump-tree-widening_mul-details -S -o pr110279-1.s
$ grep "Generated FMA" pr110279-1.c.215t.widening_mul || echo FAIL
FAIL

[Bug testsuite/113611] [14 Regression] gcc.dg/pr110279-1.c fails on cross build since gcc-14-5779-g746344dd538

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113611

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |testsuite-fail
          Component|target                      |testsuite

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
I suspect this is just a testcase issue.

[Bug testsuite/113611] [14 Regression] gcc.dg/pr110279-1.c fails on cross build since gcc-14-5779-g746344dd538

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113611

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

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

[Bug testsuite/113611] [14 Regression] gcc.dg/pr110279-1.c fails on cross build since gcc-14-5779-g746344dd538

[rearnsha at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113611

Richard Earnshaw <rearnsha at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |INVALID
             Status|UNCONFIRMED                 |RESOLVED

--- Comment #2 from Richard Earnshaw <rearnsha at gcc dot gnu.org> ---
I don't see how this can be a regression.

> --with-fpu=vfpv3-d16 

FMA was added in vfpv4.  If I change the fpu to add this then the test
generates the relevant comments in the dump file.

Arguably this test should check that the target has FMA instructions before
running, but that's a different issue.