- Added Thumb assembly optimizations
diff --git a/include/polarssl/bn_mul.h b/include/polarssl/bn_mul.h
index a6a2c65..f57d3a7 100644
--- a/include/polarssl/bn_mul.h
+++ b/include/polarssl/bn_mul.h
@@ -498,7 +498,57 @@
#if defined(__arm__)
-#if !defined(__thumb__)
+#if defined(__thumb__)
+
+#define MULADDC_INIT \
+ asm( "ldr r0, %0 " :: "m" (s)); \
+ asm( "ldr r1, %0 " :: "m" (d)); \
+ asm( "ldr r2, %0 " :: "m" (c)); \
+ asm( "ldr r3, %0 " :: "m" (b)); \
+ asm( "lsr r7, r3, #16 " ); \
+ asm( "mov r9, r7 " ); \
+ asm( "lsl r7, r3, #16 " ); \
+ asm( "lsr r7, r7, #16 " ); \
+ asm( "mov r8, r7 " );
+
+#define MULADDC_CORE \
+ asm( "ldmia r0!, {r6} " ); \
+ asm( "lsr r7, r6, #16 " ); \
+ asm( "lsl r6, r6, #16 " ); \
+ asm( "lsr r6, r6, #16 " ); \
+ asm( "mov r4, r8 " ); \
+ asm( "mul r4, r6 " ); \
+ asm( "mov r3, r9 " ); \
+ asm( "mul r6, r3 " ); \
+ asm( "mov r5, r9 " ); \
+ asm( "mul r5, r7 " ); \
+ asm( "mov r3, r8 " ); \
+ asm( "mul r7, r3 " ); \
+ asm( "lsr r3, r6, #16 " ); \
+ asm( "add r5, r5, r3 " ); \
+ asm( "lsr r3, r7, #16 " ); \
+ asm( "add r5, r5, r3 " ); \
+ asm( "add r4, r4, r2 " ); \
+ asm( "mov r2, #0 " ); \
+ asm( "adc r5, r2 " ); \
+ asm( "lsl r3, r6, #16 " ); \
+ asm( "add r4, r4, r3 " ); \
+ asm( "adc r5, r2 " ); \
+ asm( "lsl r3, r7, #16 " ); \
+ asm( "add r4, r4, r3 " ); \
+ asm( "adc r5, r2 " ); \
+ asm( "ldr r3, [r1] " ); \
+ asm( "add r4, r4, r3 " ); \
+ asm( "adc r2, r5 " ); \
+ asm( "stmia r1!, {r4} " );
+
+#define MULADDC_STOP \
+ asm( "str r2, %0 " : "=m" (c)); \
+ asm( "str r1, %0 " : "=m" (d)); \
+ asm( "str r0, %0 " : "=m" (s) :: \
+ "r0", "r1", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9" );
+
+#else
#define MULADDC_INIT \
asm( "ldr r0, %0 " :: "m" (s)); \