From e62bdd4df1d573d51bd2483a5d5cd2aa0b26e7b2 Mon Sep 17 00:00:00 2001
From: "Field G. Van Zee" <field@cs.utexas.edu>
Date: Mon, 18 Feb 2019 18:35:30 -0600
Subject: [PATCH] Disable TBM, XOP, LWP instructions in AMD configs.

Details:
- Added -mno-tbm -mno-xop -mno-lwp to CKVECFLAGS in bulldozer,
  piledriver, steamroller, and excavator configurations to explicitly
  disable AMD's bulldozer-era TBM, XOP, and LWP instruction sets in an
  attempt to fix the invalid instruction error that has plagued Travis
  CI builds since 6a014a3. Thanks to Devin Matthews for pointing out
  that the offending instruction was part of TBM (issue #300).
- Restored -O3 to piledriver configuration's COPTFLAGS.
---
 config/bulldozer/make_defs.mk   | 4 ++--
 config/excavator/make_defs.mk   | 4 ++--
 config/piledriver/make_defs.mk  | 6 +++---
 config/steamroller/make_defs.mk | 4 ++--
 4 files changed, 9 insertions(+), 9 deletions(-)

diff --git a/config/bulldozer/make_defs.mk b/config/bulldozer/make_defs.mk
index 6a75d199e..06f7d8251 100644
--- a/config/bulldozer/make_defs.mk
+++ b/config/bulldozer/make_defs.mk
@@ -63,10 +63,10 @@ endif
 # Flags specific to optimized kernels.
 CKOPTFLAGS     := $(COPTFLAGS)
 ifeq ($(CC_VENDOR),gcc)
-CKVECFLAGS     := -mfpmath=sse -mavx -mfma4 -march=bdver1
+CKVECFLAGS     := -mfpmath=sse -mavx -mfma4 -march=bdver1 -mno-tbm -mno-xop -mno-lwp
 else
 ifeq ($(CC_VENDOR),clang)
-CKVECFLAGS     := -mfpmath=sse -mavx -mfma4 -march=bdver1
+CKVECFLAGS     := -mfpmath=sse -mavx -mfma4 -march=bdver1 -mno-tbm -mno-xop -mno-lwp
 else
 $(error gcc or clang are required for this configuration.)
 endif
diff --git a/config/excavator/make_defs.mk b/config/excavator/make_defs.mk
index 40a552198..48de8ed50 100644
--- a/config/excavator/make_defs.mk
+++ b/config/excavator/make_defs.mk
@@ -63,10 +63,10 @@ endif
 # Flags specific to optimized kernels.
 CKOPTFLAGS     := $(COPTFLAGS)
 ifeq ($(CC_VENDOR),gcc)
-CKVECFLAGS     := -mfpmath=sse -mavx -mfma -mno-fma4 -march=bdver4
+CKVECFLAGS     := -mfpmath=sse -mavx -mfma -march=bdver4 -mno-fma4 -mno-tbm -mno-xop -mno-lwp
 else
 ifeq ($(CC_VENDOR),clang)
-CKVECFLAGS     := -mfpmath=sse -mavx -mfma -mno-fma4 -march=bdver4
+CKVECFLAGS     := -mfpmath=sse -mavx -mfma -march=bdver4 -mno-fma4 -mno-tbm -mno-xop -mno-lwp
 else
 $(error gcc or clang are required for this configuration.)
 endif
diff --git a/config/piledriver/make_defs.mk b/config/piledriver/make_defs.mk
index 87b1b8644..425040692 100644
--- a/config/piledriver/make_defs.mk
+++ b/config/piledriver/make_defs.mk
@@ -57,16 +57,16 @@ endif
 ifeq ($(DEBUG_TYPE),noopt)
 COPTFLAGS      := -O0
 else
-COPTFLAGS      := -O2
+COPTFLAGS      := -O3
 endif
 
 # Flags specific to optimized kernels.
 CKOPTFLAGS     := $(COPTFLAGS)
 ifeq ($(CC_VENDOR),gcc)
-CKVECFLAGS     := -mfpmath=sse -mavx -mfma -mno-fma4 -march=bdver2
+CKVECFLAGS     := -mfpmath=sse -mavx -mfma -march=bdver2 -mno-fma4 -mno-tbm -mno-xop -mno-lwp
 else
 ifeq ($(CC_VENDOR),clang)
-CKVECFLAGS     := -mfpmath=sse -mavx -mfma -mno-fma4 -march=bdver2
+CKVECFLAGS     := -mfpmath=sse -mavx -mfma -march=bdver2 -mno-fma4 -mno-tbm -mno-xop -mno-lwp
 else
 $(error gcc or clang are required for this configuration.)
 endif
diff --git a/config/steamroller/make_defs.mk b/config/steamroller/make_defs.mk
index 656076fd9..58315d7f5 100644
--- a/config/steamroller/make_defs.mk
+++ b/config/steamroller/make_defs.mk
@@ -63,10 +63,10 @@ endif
 # Flags specific to optimized kernels.
 CKOPTFLAGS     := $(COPTFLAGS)
 ifeq ($(CC_VENDOR),gcc)
-CKVECFLAGS     := -mfpmath=sse -mavx -mfma -mno-fma4 -march=bdver3
+CKVECFLAGS     := -mfpmath=sse -mavx -mfma -march=bdver3 -mno-fma4 -mno-tbm -mno-xop -mno-lwp
 else
 ifeq ($(CC_VENDOR),clang)
-CKVECFLAGS     := -mfpmath=sse -mavx -mfma -mno-fma4 -march=bdver3
+CKVECFLAGS     := -mfpmath=sse -mavx -mfma -march=bdver3 -mno-fma4 -mno-tbm -mno-xop -mno-lwp
 else
 $(error gcc or clang are required for this configuration.)
 endif