target-mips: fix for incorrect multiplication with MULQ_S.PH

The change corrects sign-related issue with MULQ_S.PH. It also includes
extension to the already existing test which will trigger the issue.
Signed-off-by: NPetar Jovanovic <petarj@mips.com>
Signed-off-by: NAurelien Jarno <aurelien@aurel32.net>
(cherry picked from commit 9c19eb1e)
Signed-off-by: NMichael Roth <mdroth@linux.vnet.ibm.com>

target-mips/dsp_helper.c

@@ -652,7 +652,7 @@ static inline int32_t mipsdsp_sat16_mul_q15_q15(uint16_t a, uint16_t b,
         temp = 0x7FFF0000;
         set_DSPControl_overflow_flag(1, 21, env);
     } else {
-        temp = ((uint32_t)a * (uint32_t)b);
+        temp = (int16_t)a * (int16_t)b;
         temp = temp << 1;
     }
 

tests/tcg/mips/mips32-dspr2/mulq_s_ph.c

@@ -6,6 +6,21 @@ int main()
     int rd, rs, rt, dsp;
     int result, resultdsp;
 
+    rs = 0x80000000;
+    rt = 0x0ffc0000;
+    result = 0xF0040000;
+    resultdsp = 0;
+
+    __asm
+        ("mulq_s.ph %0, %2, %3\n\t"
+         "rddsp %1\n\t"
+         : "=r"(rd), "=r"(dsp)
+         : "r"(rs), "r"(rt)
+        );
+    dsp = (dsp >> 21) & 0x01;
+    assert(rd  == result);
+    assert(dsp == resultdsp);
+
     rs = 0x80001234;
     rt = 0x80004321;
     result = 0x7FFF098B;