LSPlayer (CIA version) by Leonard

Rotschi · 22 June 2024, 16:49

As far as I understand the two subroutines LSP_MainIrq and LSP_DMAconIrq are called in turns, i.e., the 1st one sets int6 to the 2nd one and vice versa.

I am curious about the first line in both subroutines:

.LSP_MainIrq:
btst.b #0,$bfdd00 int6 triggered by timerA

LSP_DMAconIrq:
btst.b #1,$bfdd00 int6 triggered by timerB

First I thought - since we have to read out $bfdd00 anyway for clearing the pending of the CIA int - that this may be a kind of extra check for ensuring that both subroutines are called in the correct order (killing two birds with one stone), but on second thought I am not sure about this.
Okay, to make a long story short: What is the btst check good for and why not a simple tst.b $bfdd00 w/o branch?

hooverphonique · 22 June 2024, 17:45

As you may have noticed, they check different bits - one for each timer.
This verifies that it's the correct interrupt source that triggered the ISR.

Rotschi · 25 June 2024, 15:02

sure, but isn't it redundant here

a/b · 25 June 2024, 16:18

It's redundant if you only use timers A and B, and nothing else that can trigger CIAB interrupts. Those 2 timer callbacks are serialized so no problem there, but if something else is running in parallel (enabled after your call to player init) and triggers an interrupt then you have a problem.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
JOIN THE ZOOM CHAT: Leonard Tramiel - April 20, 2023 - Commodore, Atari, and Science!	Marcio D.	Amiga scene	1	30 April 2023 05:22
Interesting Leonard Tramiel interviews fresh on youtube	ImmortalA1000	Nostalgia & memories	8	12 December 2022 09:46
Leonard Tramiel TED talk about Commodore etc..	eXeler0	Nostalgia & memories	1	17 March 2017 13:20
example of a CIA timer interrupt in assembler using cia.resource	Apollo	Coders. Asm / Hardware	3	05 July 2013 08:40
CIA test	Toni Wilen	Coders. General	13	03 March 2007 22:14

22 June 2024, 16:49	#1
Rotschi Registered User Join Date: Apr 2023 Location: "Hamcastle" Posts: 25	LSPlayer (CIA version) by Leonard As far as I understand the two subroutines LSP_MainIrq and LSP_DMAconIrq are called in turns, i.e., the 1st one sets int6 to the 2nd one and vice versa. I am curious about the first line in both subroutines: .LSP_MainIrq: btst.b #0,$bfdd00 int6 triggered by timerA LSP_DMAconIrq: btst.b #1,$bfdd00 int6 triggered by timerB First I thought - since we have to read out $bfdd00 anyway for clearing the pending of the CIA int - that this may be a kind of extra check for ensuring that both subroutines are called in the correct order (killing two birds with one stone), but on second thought I am not sure about this. Okay, to make a long story short: What is the btst check good for and why not a simple tst.b $bfdd00 w/o branch?

22 June 2024, 17:45	#2
hooverphonique ex. demoscener "Bigmama" Join Date: Jun 2012 Location: Fyn / Denmark Posts: 1,643	As you may have noticed, they check different bits - one for each timer. This verifies that it's the correct interrupt source that triggered the ISR.

25 June 2024, 15:02	#3
Rotschi Registered User Join Date: Apr 2023 Location: "Hamcastle" Posts: 25	sure, but isn't it redundant here

25 June 2024, 16:18	#4
a/b Registered User Join Date: Jun 2016 Location: europe Posts: 1,066	It's redundant if you only use timers A and B, and nothing else that can trigger CIAB interrupts. Those 2 timer callbacks are serialized so no problem there, but if something else is running in parallel (enabled after your call to player init) and triggers an interrupt then you have a problem.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)