]> git.proxmox.com Git - mirror_qemu.git/commit - tcg/tcg-op-gvec.c
tcg: Restrict check_size_impl to multiples of the line size
authorRichard Henderson <richard.henderson@linaro.org>
Mon, 9 Jul 2018 13:51:34 +0000 (14:51 +0100)
committerPeter Maydell <peter.maydell@linaro.org>
Mon, 9 Jul 2018 13:51:34 +0000 (14:51 +0100)
commit499748d7683198a765d17b4fdf6901ab9dca920c
tree44f8d687d5cc7dd7ed8ac565cf71f8ce8f9e6fb0
parent2f95a3b09aebdcb5c9152a7ac434a5d57441fe82
tcg: Restrict check_size_impl to multiples of the line size

Normally this is automatic in the size restrictions that are placed
on vector sizes coming from the implementation.  However, for the
legitimate size tuple [oprsz=8, maxsz=32], we need to clear the final
24 bytes of the vector register.  Without this check, do_dup selects
TCG_TYPE_V128 and clears only 16 bytes.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Tested-by: Alex Bennée <alex.bennee@linaro.org>
Message-id: 20180705191929.30773-2-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
tcg/tcg-op-gvec.c