amd/blis
1
0
Fork 0
mirror of https://github.com/amd/blis.git synced 2026-05-13 10:35:38 +00:00
Code Issues Packages Projects Releases Wiki Activity

Disable TBM, XOP, LWP instructions in AMD configs.

Browse Source 
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.
This commit is contained in:
Field G. Van Zee
2019-02-18 18:35:30 -06:00
committed by Devrajegowda, Kiran
parent e7b73bf1ed
commit e62bdd4df1
4 changed files with 9 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
config/bulldozer/make_defs.mk
View File

@@ -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

4
config/excavator/make_defs.mk
View File

@@ -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

6
config/piledriver/make_defs.mk
View File

@@ -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

4
config/steamroller/make_defs.mk
View File

@@ -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