sparclite: correct wssr[0] reset state

src/devices/bus/saitek_osa/sparc.cpp

@@ -66,6 +66,8 @@ void saitekosa_sparc_device::device_reset()
 		set_ram_mask(ioport("RAM")->read() ? 22 : 20);
 		m_installed = true;
 	}
+
+	host_io_w(0, 0x1c00);
 }
 
 

src/devices/cpu/sparc/sparc.cpp

@@ -17,8 +17,6 @@
 //      - Extended-precision FPU support
 //      - Coprocessor support
 //      - Finish SPARClite peripherals
-//      - SPARClite m_wssr[0] should be 0x7ff40 at reset (32 wait cycles for CS0),
-//        but that is problematic for Saitek Renaissance + Sparc module
 //
 //================================================================
 
@@ -675,9 +673,11 @@ void mb86930_device::device_reset()
 	m_ssctrl = 0x08;
 	m_spmr = 0;
 	m_spmr_mask = ~0ULL;
-	std::fill_n(&m_wssr[0], 3, 0);
 	m_last_masked_addr = 0ULL;
 
+	std::fill_n(&m_wssr[0], 3, 0);
+	m_wssr[0] = 0x1ffd << 6;
+
 	m_arsr[0] = (9 << 23);
 	m_amr[0] = (0x1f << 1);
 

src/mame/drivers/saitek_renaissance.cpp

@@ -23,7 +23,8 @@ so a chessboard display + 7seg info. It's on a small drawer that can be
 pushed in to hide the chessboard display.
 
 TODO:
-- fart noise at boot if maestroa or sparc module is inserted
+- fart noise at boot if maestroa module is inserted
+- weird beep at boot if sparc module is inserted (related to above?)
 - make it a subdriver of saitek_leonardo.cpp? or too many differences
 - same TODO list as saitek_leonardo.cpp