diff --git a/core/combo/arch/arm/armv7-a-neon.mk b/core/combo/arch/arm/armv7-a-neon.mk
index 5dc1fecdd..5d5b05009 100644
--- a/core/combo/arch/arm/armv7-a-neon.mk
+++ b/core/combo/arch/arm/armv7-a-neon.mk
@@ -6,6 +6,8 @@ ARCH_ARM_HAVE_VFP               := true
 ARCH_ARM_HAVE_VFP_D32           := true
 ARCH_ARM_HAVE_NEON              := true
 
+local_arch_has_lpae := false
+
 ifneq (,$(filter cortex-a15 krait denver,$(TARGET_$(combo_2nd_arch_prefix)CPU_VARIANT)))
 	# TODO: krait is not a cortex-a15, we set the variant to cortex-a15 so that
 	#       hardware divide operations are generated. This should be removed and a
@@ -13,9 +15,7 @@ ifneq (,$(filter cortex-a15 krait denver,$(TARGET_$(combo_2nd_arch_prefix)CPU_VA
 	#       core/clang/arm.mk.
 	arch_variant_cflags := -mcpu=cortex-a15
 
-	# Fake an ARM compiler flag as these processors support LPAE which GCC/clang
-	# don't advertise.
-	arch_variant_cflags += -D__ARM_FEATURE_LPAE=1
+	local_arch_has_lpae := true
 	arch_variant_ldflags := \
 		-Wl,--no-fix-cortex-a8
 else
@@ -26,6 +26,8 @@ ifeq ($(strip $(TARGET_$(combo_2nd_arch_prefix)CPU_VARIANT)),cortex-a8)
 else
 ifneq (,$(filter cortex-a7 cortex-a53 cortex-a53.a57,$(TARGET_$(combo_2nd_arch_prefix)CPU_VARIANT)))
 	arch_variant_cflags := -mcpu=cortex-a7
+
+	local_arch_has_lpae := true
 	arch_variant_ldflags := \
 		-Wl,--no-fix-cortex-a8
 else
@@ -37,6 +39,16 @@ endif
 endif
 endif
 
+ifeq (true,$(local_arch_has_lpae))
+	# Fake an ARM compiler flag as these processors support LPAE which GCC/clang
+	# don't advertise.
+	# TODO This is a hack and we need to add it for each processor that supports LPAE until some
+	# better solution comes around. See Bug 27340895
+	arch_variant_cflags += -D__ARM_FEATURE_LPAE=1
+endif
+
+local_arch_has_lpae :=
+
 arch_variant_cflags += \
     -mfloat-abi=softfp \
     -mfpu=neon