size optimization of the IEC support

c64/kernal/iec/iec_tx_command.s

@@ -19,17 +19,10 @@ iec_tx_command:
 	;; and it is going to be a command (pulled ATN)
 	jsr iec_pull_atn_clk_release_data
 
-	;; Give devices time to respond (response is mandatory!)
-	jsr iec_wait1ms
+	;; Give devices time to respond (response is mandatory!) by pulling DATA
+	jsr iec_wait1ms_for_data
+	bmi iec_tx_command_dev_not_found
 
-	;; Did at least one device respond by pulling DATA?
-	lda CIA2_PRA
-	and #BIT_CIA2_PRA_DAT_IN ; XXX try to optimize this, move to separate routine
-	beq +
-
-	;; No devices present on the bus, so we can immediately return with device not found	
-	jmp iec_return_DEVICE_NOT_FOUND
-*
 	;; At least one device responded, but they are still allowed to stall
 	;; (can be busy processing something), we have to wait till they are all
 	;; ready (or bored with DOS attack...)
@@ -46,17 +39,14 @@ iec_tx_command:
 	;; According to https://www.pagetable.com/?p=1135 there is some complicated and dangerous
 	;; flow here, but http://www.zimmers.net/anonftp/pub/cbm/programming/serial-bus.pdf (page 6)
 	;; advices to just wait 1ms and check the DATA status
-	jsr iec_wait1ms
-	lda CIA2_PRA
-	and #BIT_CIA2_PRA_DAT_IN ; XXX try to optimize this, move to separate routine
-	beq +
 
-	;; XXX possible optimization of the flow above: for many commands it is enough
-	;; to wait for the DATA being pulled, as confirmation is expected from
-	;; a single device only - but not sure if it's worth the trouble
+	jsr iec_wait1ms_for_data
+	bmi iec_tx_command_dev_not_found
 
-	jmp iec_return_DEVICE_NOT_FOUND
-*
 	;; Done
 
 	jmp iec_return_success
+
+iec_tx_command_dev_not_found:
+
+	jmp iec_return_DEVICE_NOT_FOUND

c64/kernal/iec/iec_wait1ms_for_data.s

@@ -1,6 +1,6 @@
 
 ;;
-;; Wait 1 ms (1000 usec)
+;; Wait 1 ms (1000 usec) - afterwards can use 'bpl' to check that DATA is pulled
 ;;
 ;; On NTSC this means 1023 cycles (slightly less on PAL, but we have to be sure).
 ;; We can't count on badlines, as the screen might be disabled.
@@ -8,7 +8,7 @@
 ;; 63 cycles, we need at least 17 full rasters to be 100% sure we wait enough
 
 
-iec_wait1ms:
+iec_wait1ms_for_data:
 	ldx #18 ; we probably won't start with the beginning of raster
 	lda VIC_RASTER
 *
@@ -17,4 +17,6 @@ iec_wait1ms:
 	lda VIC_RASTER
 	dex
 	bne -
+
+	lda CIA2_PRA ; pulled DATA = highest bit 0 = 'plus'
 	rts