diff --git a/arch/arm/include/asm/arch-sunxi/dram_sunxi_dw.h b/arch/arm/include/asm/arch-sunxi/dram_sunxi_dw.h
index 03fd46b..66e206d 100644
--- a/arch/arm/include/asm/arch-sunxi/dram_sunxi_dw.h
+++ b/arch/arm/include/asm/arch-sunxi/dram_sunxi_dw.h
@@ -13,6 +13,8 @@
 #ifndef _SUNXI_DRAM_SUN8I_H3_H
 #define _SUNXI_DRAM_SUN8I_H3_H
 
+#include <linux/bitops.h>
+
 struct sunxi_mctl_com_reg {
 	u32 cr;			/* 0x00 control register */
 	u32 cr_r1;		/* 0x04 rank 1 control register (R40 only) */
@@ -211,7 +213,6 @@ struct sunxi_mctl_ctl_reg {
  * the 32-bit wide access consists of. Also three control signals can be
  * adjusted individually.
  */
-#define BITS_PER_BYTE		8
 #define NR_OF_BYTE_LANES	(32 / BITS_PER_BYTE)
 /* The eight data lines (DQn) plus DM, DQS and DQSN */
 #define LINES_PER_BYTE_LANE	(BITS_PER_BYTE + 3)
diff --git a/arch/arm/mach-kirkwood/include/mach/gpio.h b/arch/arm/mach-kirkwood/include/mach/gpio.h
index aa8c5da..ac2397d 100644
--- a/arch/arm/mach-kirkwood/include/mach/gpio.h
+++ b/arch/arm/mach-kirkwood/include/mach/gpio.h
@@ -15,10 +15,6 @@
 #ifndef __KIRKWOOD_GPIO_H
 #define __KIRKWOOD_GPIO_H
 
-/* got from kernel include/linux/bitops.h */
-#define BITS_PER_BYTE 8
-#define BITS_TO_LONGS(nr)	DIV_ROUND_UP(nr, BITS_PER_BYTE * sizeof(long))
-
 #define GPIO_MAX		50
 #define GPIO_OFF(pin)		(((pin) >> 5) ? 0x0040 : 0x0000)
 #define GPIO_OUT(pin)		(MVEBU_GPIO0_BASE + GPIO_OFF(pin) + 0x00)
diff --git a/drivers/gpio/kw_gpio.c b/drivers/gpio/kw_gpio.c
index 43b27e3..cc26cc1 100644
--- a/drivers/gpio/kw_gpio.c
+++ b/drivers/gpio/kw_gpio.c
@@ -14,7 +14,7 @@
  */
 
 #include <common.h>
-#include <asm/bitops.h>
+#include <linux/bitops.h>
 #include <asm/io.h>
 #include <asm/arch/soc.h>
 #include <asm/arch/gpio.h>
diff --git a/drivers/mtd/nand/nand_base.c b/drivers/mtd/nand/nand_base.c
index 5bb4ea8..efe3e4f 100644
--- a/drivers/mtd/nand/nand_base.c
+++ b/drivers/mtd/nand/nand_base.c
@@ -901,8 +901,6 @@ static int nand_wait(struct mtd_info *mtd, struct nand_chip *chip)
 	return status;
 }
 
-#define BITS_PER_BYTE 8
-
 /**
  * nand_check_erased_buf - check if a buffer contains (almost) only 0xff data
  * @buf: buffer to test
diff --git a/include/linux/bitops.h b/include/linux/bitops.h
index 576b15d..a47f6d1 100644
--- a/include/linux/bitops.h
+++ b/include/linux/bitops.h
@@ -5,9 +5,16 @@
 #include <asm-generic/bitsperlong.h>
 #include <linux/compiler.h>
 
+#ifdef	__KERNEL__
 #define BIT(nr)			(1UL << (nr))
+#define BIT_ULL(nr)		(1ULL << (nr))
 #define BIT_MASK(nr)		(1UL << ((nr) % BITS_PER_LONG))
 #define BIT_WORD(nr)		((nr) / BITS_PER_LONG)
+#define BIT_ULL_MASK(nr)	(1ULL << ((nr) % BITS_PER_LONG_LONG))
+#define BIT_ULL_WORD(nr)	((nr) / BITS_PER_LONG_LONG)
+#define BITS_PER_BYTE		8
+#define BITS_TO_LONGS(nr)	DIV_ROUND_UP(nr, BITS_PER_BYTE * sizeof(long))
+#endif
 
 /*
  * Create a contiguous bitmask starting at bit position @l and ending at
diff --git a/include/usb/lin_gadget_compat.h b/include/usb/lin_gadget_compat.h
index 4a01585..d0d71f7 100644
--- a/include/usb/lin_gadget_compat.h
+++ b/include/usb/lin_gadget_compat.h
@@ -10,12 +10,10 @@
 #ifndef __LIN_COMPAT_H__
 #define __LIN_COMPAT_H__
 
+#include <linux/bitops.h>
 #include <linux/compat.h>
 
 /* common */
-#define BITS_PER_BYTE				8
-#define BITS_TO_LONGS(nr) \
-	DIV_ROUND_UP(nr, BITS_PER_BYTE * sizeof(long))
 #define DECLARE_BITMAP(name, bits) \
 	unsigned long name[BITS_TO_LONGS(bits)]