milkymist-pfpu: fix potential integer overflow

Since the lm32 is a 32 bit architecture, just return a 32 bit value which
is then converted to a 64 bit value.

Spotted by coverity, CID 1005506.
Signed-off-by: NMichael Walle <michael@walle.cc>
Signed-off-by: NMichael Tokarev <mjt@tls.msk.ru>

hw/misc/milkymist-pfpu.c

@@ -137,7 +137,7 @@ struct MilkymistPFPUState {
 };
 typedef struct MilkymistPFPUState MilkymistPFPUState;
 
-static inline hwaddr
+static inline uint32_t
 get_dma_address(uint32_t base, uint32_t x, uint32_t y)
 {
     return base + 8 * (128 * y + x);