Add the __qdbl intrinsic to the arm_acle.h header

[mroth]

Summary

Hi all,

ACLE section 9.4.2 defines the __qdbl intrinsic:

int32_t qdbl(int32_t);
Doubles a signed 32-bit number, with saturation. qdbl(x) is equal to __qadd(x,x) except that the argument x is evaluated only once.
Sets the Q flag if the addition saturates.

Since we already have an arm_acle.h header file with qadd implemented, it's trivial to implement this. I've also added a test that should make sure the argument is only evaluated once - this would catch e.g. if someone decided to do #define qdbl(X) __qadd(X,X) instead, which evaluates X twice. Does this look reasonable?

Cheers
Moritz

[t.p.northover]

Hi Moritz,

LGTM! Thanks for extending the ACLE support.

Cheers.

Tim.

[rengolin]

LGTM too.

--renato

[compnerd]

+1

[rengolin]

r212264

lib/Headers/arm_acle.h

   __qsub(int32_t t, int32_t v) {
   return __builtin_arm_qsub(t, v);
 }
+
+static __inline__ int32_t __attribute__((always_inline, nodebug))
+__qdbl(int32_t t) {
+  return __builtin_arm_qadd(t, t);
+}
 #endif
 
 /* CRC32 intrinsics */

test/CodeGen/arm_acle.c

 int32_t test_qsub(int32_t a, int32_t b) {
   return __qsub(a, b);
 }
+
+extern int32_t f();
+// AArch32-LABEL: test_qdbl
+// AArch32: [[VAR:%[a-z0-9]+]] = {{.*}} call {{.*}} @f
+// AArch32-NOT: call {{.*}} @f
+// AArch32: call i32 @llvm.arm.qadd(i32 [[VAR]], i32 [[VAR]])
+int32_t test_qdbl() {
+  return __qdbl(f());
+}
 #endif
 
 /* CRC32 intrinsics */