From mboxrd@z Thu Jan 1 00:00:00 1970 From: Luc Van Oostenryck Subject: [RFC PATCH 0/2] use a single multiplication instruction Date: Sun, 18 Feb 2018 10:15:17 +0100 Message-ID: <20180218091519.6708-1-luc.vanoostenryck@gmail.com> Return-path: Received: from mail-wm0-f66.google.com ([74.125.82.66]:50845 "EHLO mail-wm0-f66.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751238AbeBRJP0 (ORCPT ); Sun, 18 Feb 2018 04:15:26 -0500 Received: by mail-wm0-f66.google.com with SMTP id k87so10067809wmi.0 for ; Sun, 18 Feb 2018 01:15:26 -0800 (PST) Sender: linux-sparse-owner@vger.kernel.org List-Id: linux-sparse@vger.kernel.org To: linux-sparse@vger.kernel.org Cc: Luc Van Oostenryck These two patches replace the instrcutions for signed & unsigned multiplication (OP_MULS & OP_MULU) by a single one: OP_MUL. There is no reasons to have both since the IR doesn't do full/widening multiplication. Luc Van Oostenryck (2): unsigned multiplication is also associative no need for signed & unsigned multiplication Documentation/IR.md | 7 ++----- cse.c | 4 ++-- example.c | 5 ++--- linearize.c | 9 ++++----- linearize.h | 2 +- opcode.c | 3 +-- simplify.c | 15 +++++---------- sparse-llvm.c | 6 +----- validation/int128.c | 2 +- 9 files changed, 19 insertions(+), 34 deletions(-) -- 2.16.0