diff --git a/mono/arch/arm/arm-codegen.h b/mono/arch/arm/arm-codegen.h
index e7dc99f..2460fc4 100644
--- a/mono/arch/arm/arm-codegen.h
+++ b/mono/arch/arm/arm-codegen.h
@@ -1084,6 +1084,16 @@ typedef union {
 #define ARM_MOVT_REG_IMM_COND(p, rd, imm16, cond) ARM_EMIT(p, (((cond) << 28) | (3 << 24) | (4 << 20) | ((((guint32)(imm16)) >> 12) << 16) | ((rd) << 12) | (((guint32)(imm16)) & 0xfff)))
 #define ARM_MOVT_REG_IMM(p, rd, imm16) ARM_MOVT_REG_IMM_COND ((p), (rd), (imm16), ARMCOND_AL)

+/* MCR */
+#define ARM_DEF_MCR_COND(coproc, opc1, rt, crn, crm, opc2, cond)	\
+	ARM_DEF_COND ((cond)) | ((0xe << 24) | (((opc1) & 0x7) << 21) | (0 << 20) | (((crn) & 0xf) << 16) | (((rt) & 0xf) << 12) | (((coproc) & 0xf) << 8) | (((opc2) & 0x7) << 5) | (1 << 4) | (((crm) & 0xf) << 0))
+
+#define ARM_MCR_COND(p, coproc, opc1, rt, crn, crm, opc2, cond)	\
+	ARM_EMIT(p, ARM_DEF_MCR_COND ((coproc), (opc1), (rt), (crn), (crm), (opc2), (cond)))
+
+#define ARM_MCR(p, coproc, opc1, rt, crn, crm, opc2) \
+	ARM_MCR_COND ((p), (coproc), (opc1), (rt), (crn), (crm), (opc2), ARMCOND_AL)
+
 #ifdef __cplusplus
 }
 #endif
diff --git a/mono/mini/cpu-arm.md b/mono/mini/cpu-arm.md
index 1b253dd..3894753 100644
--- a/mono/mini/cpu-arm.md
+++ b/mono/mini/cpu-arm.md
@@ -46,7 +46,7 @@
 #
 # See the code in mini-x86.c for more details on how the specifiers are used.
 #
-memory_barrier: len:4
+memory_barrier: len:8 clob:a
 nop: len:4
 relaxed_nop: len:4
 break: len:4
diff --git a/mono/mini/mini-arm.c b/mono/mini/mini-arm.c
index c42bb50..72d84aa 100644
--- a/mono/mini/mini-arm.c
+++ b/mono/mini/mini-arm.c
@@ -3257,6 +3257,8 @@ mono_arch_output_basic_block (MonoCompile *cfg, MonoBasicBlock *bb)

 		switch (ins->opcode) {
 		case OP_MEMORY_BARRIER:
+			ARM_MOV_REG_IMM8 (code, ARMREG_R0, 0);
+			ARM_MCR (code, 15, 0, ARMREG_R0, 7, 10, 5);
 			break;
 		case OP_TLS_GET:
 #ifdef HAVE_AEABI_READ_TP
diff --git a/mono/utils/mono-membar.h b/mono/utils/mono-membar.h
index 13ed6e7..1c343b0 100644
--- a/mono/utils/mono-membar.h
+++ b/mono/utils/mono-membar.h
@@ -129,7 +129,7 @@ static inline void mono_memory_write_barrier (void)
 #elif defined(__arm__)
 static inline void mono_memory_barrier (void)
 {
-	__asm__ __volatile__ ("" : : : "memory");
+	__asm__ __volatile__ ("mcr p15, 0, %0, c7, c10, 5" : : "r" (0) : "memory");
 }

 static inline void mono_memory_read_barrier (void)