diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index 2bf1244b0ac8..2ab99334e660 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -25,6 +25,7 @@ config ARM
 	select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
 	select ARCH_HAVE_CUSTOM_GPIO_H
 	select ARCH_HAVE_NMI_SAFE_CMPXCHG if CPU_V7 || CPU_V7M || CPU_V6K
+	select ARCH_HAVE_TRACE_MMIO_ACCESS if AMLOGIC_DEBUG_IOTRACE
 	select ARCH_HAS_GCOV_PROFILE_ALL
 	select ARCH_KEEP_MEMBLOCK
 	select ARCH_MIGHT_HAVE_PC_PARPORT
diff --git a/arch/arm/include/asm/io.h b/arch/arm/include/asm/io.h
index 79d246ac93ab..9df35303c0f7 100644
--- a/arch/arm/include/asm/io.h
+++ b/arch/arm/include/asm/io.h
@@ -115,7 +115,6 @@ static inline u32 __raw_readl(const volatile void __iomem *addr)
 		     : "Qo" (*(volatile u32 __force *)addr));
 	return val;
 }
-
 /*
  * Architecture ioremap implementation.
  */
@@ -288,6 +287,8 @@ extern void _memset_io(volatile void __iomem *, int, size_t);
  * IO port primitives for more information.
  */
 #ifndef readl
+/* use the macro definitions in include/asm-generic/io.h */
+#if !IS_ENABLED(CONFIG_AMLOGIC_DEBUG_IOTRACE)
 #define readb_relaxed(c) ({ u8  __r = __raw_readb(c); __r; })
 #define readw_relaxed(c) ({ u16 __r = le16_to_cpu((__force __le16) \
 					__raw_readw(c)); __r; })
@@ -313,6 +314,7 @@ extern void _memset_io(volatile void __iomem *, int, size_t);
 #define writesb(p,d,l)		__raw_writesb(p,d,l)
 #define writesw(p,d,l)		__raw_writesw(p,d,l)
 #define writesl(p,d,l)		__raw_writesl(p,d,l)
+#endif
 
 #ifndef __ARMBE__
 static inline void memset_io(volatile void __iomem *dst, unsigned c,
@@ -338,7 +340,6 @@ static inline void memcpy_toio(volatile void __iomem *to, const void *from,
 	mmiocpy((void __force *)to, from, count);
 }
 #define memcpy_toio(to,from,count) memcpy_toio(to,from,count)
-
 #else
 #define memset_io(c,v,l)	_memset_io(c,(v),(l))
 #define memcpy_fromio(a,c,l)	_memcpy_fromio((a),c,(l))
diff --git a/arch/arm64/include/asm/io.h b/arch/arm64/include/asm/io.h
index 3995652daf81..8f6e6980b5c7 100644
--- a/arch/arm64/include/asm/io.h
+++ b/arch/arm64/include/asm/io.h
@@ -112,6 +112,9 @@ static inline u64 __raw_readq(const volatile void __iomem *addr)
 #define __iowmb()		dma_wmb()
 #define __iomb()		dma_mb()
 
+#if IS_BUILTIN(CONFIG_AMLOGIC_DEBUG_IOTRACE) || \
+    ((IS_MODULE(CONFIG_AMLOGIC_DEBUG_IOTRACE)) && defined(MODULE))
+#else
 /*
  * Relaxed I/O memory access primitives. These follow the Device memory
  * ordering rules but do not guarantee any ordering relative to Normal memory
@@ -141,6 +144,7 @@ static inline u64 __raw_readq(const volatile void __iomem *addr)
 #define writew(v,c)		({ __iowmb(); writew_relaxed((v),(c)); })
 #define writel(v,c)		({ __iowmb(); writel_relaxed((v),(c)); })
 #define writeq(v,c)		({ __iowmb(); writeq_relaxed((v),(c)); })
+#endif
 
 /*
  *  I/O port access primitives.
diff --git a/include/asm-generic/io.h b/include/asm-generic/io.h
index c154199269b1..6e64abad09a6 100644
--- a/include/asm-generic/io.h
+++ b/include/asm-generic/io.h
@@ -18,6 +18,13 @@
 #include <asm/mmiowb.h>
 #include <asm-generic/pci_iomap.h>
 
+/* get the macro ' _THIS_IP_' definition */
+#if IS_ENABLED(CONFIG_AMLOGIC_DEBUG_IOTRACE)
+#ifdef CONFIG_ARM
+#include <linux/kernel.h>
+#endif
+#endif
+
 #ifndef __io_br
 #define __io_br()      barrier()
 #endif