[Bug target/124335] invalid "vmovsh" operand with __builtin_ia32_loadsh_mask() and -masm=intel since r9-6397

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=124335

--- Comment #6 from GCC Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-15 branch has been updated by Jakub Jelinek
<jakub@gcc.gnu.org>:

https://gcc.gnu.org/g:e8edba9551d2996ec0b85cee9a1c468df6a74eaf

commit r15-10908-ge8edba9551d2996ec0b85cee9a1c468df6a74eaf
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Tue Mar 3 09:50:44 2026 +0100

    i386: Fix up *avx512f_load<mode>_mask for -masm=intel [PR124335]

    The Intel syntax part is missing % before 3, so it always prints {3}
    rather than {k1} or similar.

    Fixed thusly.

    2026-03-03  Jakub Jelinek  <jakub@redhat.com>

            PR target/124335
            * config/i386/sse.md (*avx512f_load<mode>_mask): Use %{%3%} instead
of
            %{3%} for -masm=intel syntax.

            * gcc.target/i386/avx512fp16-pr124335.c: New test.

    (cherry picked from commit b3502a668627caa7f49609a36dbeaf088f17037c)