WhyCan Forum(哇酷开发者社区)

驱动里有用到float类型的变量,在编译的时候出现了
ERROR: "__aeabi_fcmpgt"   undefined!
ERROR: "__aeabi_fcmplt" undefined!
ERROR: "__aeabi_ddiv"  undefined!
ERROR: "__aeabi_fcmpge"  undefined!
ERROR: "__aeabi_d2f"  undefined!
ERROR: "__aeabi_dadd"   undefined!
ERROR: "__aeabi_dmul"  undefined!
ERROR: "__aeabi_f2d" undefined!
ERROR: "__aeabi_fsub"  undefined!
ERROR: "__aeabi_fcmple"  undefined!
这些错误,提示没有浮点库. makefile文件如下:

KERNELDIR := /home/jonsen/f1c100s/linux-nano-5.2-flash
CURRENT_PATH := $(shell pwd)
CROSS_COMPILE:= arm-linux-
ARCH = arm
obj-m := test.o

CFLAGS += -mfloat-abi=softfp

build: kernel_modules

kernel_modules:
    $(MAKE) ARCH=$(ARCH) CROSS_COMPILE=$(CROSS_COMPILE)   -C $(KERNELDIR) M=$(CURRENT_PATH) modules
clean:
    $(MAKE) -C $(KERNELDIR) M=$(CURRENT_PATH) clean

请问怎么处理....让编译支持软件浮点.