@Hans

You're a detective Sounds good to me.

_________________
ikir

That VIA compatibilty bit is for really old VIA chipsets. I tried it anyway, though. :)

I've played around with AmiTV a bit more this morning...

Something's changed in the interrupts somewhere - the AmiTV version you're using never calls the interrupt handler because there's an interrupt bit set on the card which doesn't get set under PR3 or before. I have no idea why.

I've found - after a lot of trial and error - that I can clear the interrupts in the IntuiTick and get a very choppy picture - but it does work. However, clearing interrupts in the interrupt handler itself doesn't do this.

There's something weeird going on... could take a while. I've had one crash while playing with it (although that did have test code in it too), but managed to successfully go from S4C to BBC 1 (that's channel 21 to channel 32 I think round here) so it didn't fall over immediately it lost a good signal.

@Spirantho

Quote:

That VIA compatibilty bit is for really old VIA chipsets. I tried it anyway, though. :)

So VIA have a history of making buggy chipsets? It's pretty sad when other hardware companies start installing hardware workarounds in their devices for bugs in your chipsets.

Anyway, it sounds like your making progress which is good.

Hans

Last edited by Hans on 24-Mar-2006 at 02:00 PM.

_________________
http://hdrlab.org.nz/ - Amiga OS 4 projects, programming articles and more. Home of the RadeonHD driver for Amiga OS 4.x project.
https://keasigmadelta.com/ - More of my work.

All chipsets have bugs. If there were no bugs, you'd never have to update your BIOS. :)

The only difference is that because VIA used to be so much smaller than Intel, they had a lot less support from other people like the motherboard manufacturers. Now they're large themselves, the bugs are worked around in the same way as Intel ones are.

@Spirantho

Quote:

All chipsets have bugs. If there were no bugs, you'd never have to update your BIOS. :)

True. But one does hope that there are no real show-stoppers that make it into released products. If this lockup problem can't be solved with a software workaround then it is a show-stopper (i.e. the only real solution is to buy a new computer).

Hans

_________________
http://hdrlab.org.nz/ - Amiga OS 4 projects, programming articles and more. Home of the RadeonHD driver for Amiga OS 4.x project.
https://keasigmadelta.com/ - More of my work.

@Hans

The VIA 686B Southbridge used in the A1 is used in a lot of other machines of the P-III/Athlon Thunderbird class, I think it was. It's a very common chip.

That doesn't mean to say that it doesn't have bugs, in it, though - it just means that VIA provided the necessary workarounds to the motherboard manufacturers.... and I'd put money on that same info not being provided to the guys writing the BIOS on the A1.

@Spirantho

I dont know is this just coincidental but tv-cards happen to have troubles on Pegasos I too and it uses VIA southbridge. I think it sounds reasonable because it is the only common nominator there...

_________________
Amiga Developer
Amiga 500, Efika, Mac Mini and PowerBook

@Spirantho

please don't give up!

@Spirantho

Please continue developing AmiTV !

_________________
Avatar babe - Monica Bellucci.

@Hans

Quote:

I'm assuming that this interrupt is being triggered like crazy when there's no signal and causing the system to fall over. So, forcing a delay between one interrupt and the next may help, if it is feasible.

Don't know if this info is still current, but EntilZha wrote in August, 2004:
Quote:

What you describe is exactly the reason why they are implemented like this in OS4. Soft interrupts are generated by a PPC 'sc' instruction, i.e. a system call. Since the A1 does not have a possibility to generate a delayed interrupt, they are executed synchronously.

#6

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

@Spirantho

Quote:

The VIA 686B Southbridge used in the A1 is used in a lot of other machines of the P-III/Athlon Thunderbird class, I think it was. It's a very common chip. That doesn't mean to say that it doesn't have bugs, in it, though - it just means that VIA provided the necessary workarounds to the motherboard manufacturers.... and I'd put money on that same info not being provided to the guys writing the BIOS on the A1.

I'd put money on -that- too.

Also, since it has been said many times that the problems often relate to the interaction
of more than "one" part, this post from @Samwel probably has some relevance:
Quote:

The use of the northbridge with its bad performance and the extremely buggy and badly documented southbridge can't be called good news now can it? We don't really need anymore of those problem boards for this small market. The Artcia S and VIA686B combination has been, and still gives, nothing but trouble for the developers.

Perhaps the post is a bit harsh, but it illustrates the "combination" issue.

#6

Last edited by number6 on 25-Mar-2006 at 04:49 PM.

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

@number6

Quote:

I'd put money on -that- too.

That's a pretty safe bet. In fact, IIRC, there were comments about the general lack of support from VIA with regard to both hardware and software issues a year or two ago. If they had given some support then maybe my computer would have the on-board audio chips soldered in.

Hans

_________________
http://hdrlab.org.nz/ - Amiga OS 4 projects, programming articles and more. Home of the RadeonHD driver for Amiga OS 4.x project.
https://keasigmadelta.com/ - More of my work.

@Hans

Quote:

In fact, IIRC, there were comments about the general lack of support from VIA with regard to both hardware and software issues a year or two ago.

Even Alan Redhouse commented on this about 2 1/2 years ago:
Quote:

VIA were particularly unhelpful in coming forward with documentation about what was obviously a well known and solved problem in the x86 world.

@thread

Edit(Added):If you type VIA686B into Google, I guarantee you will spend more time
reading links to bugs than you possibly have time in this life to read about.

Some interesting reading here too:
http://www.viaarena.com/default.aspx?PageID=3&FCat=18

#6

Last edited by number6 on 25-Mar-2006 at 06:00 PM.

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

@number6

This one is even better:
Quote:

All my devices are assigned to one IRQ. Why? If you are using an older chipset (such as VT82C686b) then this is because of ACPI Power Management. This can be disabled in the BIOS but it is not recommended.

And the "funny" thing is thatour southbridge is the VT82C686b

_________________
Best regards,
hnl_dk - Henning Nielsen Lund [Denmark]

Please send no PM to me, email me if you want to contact me. See you somewhere else.

@hnl_dk

I noticed the "not recommended", but iirc that particular comment referred to
this need re:laptops.

#6

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

@hnl_dk

Quote:

All my devices are assigned to one IRQ. Why? If you are using an older chipset (such as VT82C686b) then this is because of ACPI Power Management. This can be disabled in the BIOS but it is not recommended.

IMHO, this doesn't apply for the AmigaOne. The 82C686B southbridge isn't configured to enable the ACPI Power Management (at least under A1 Linux) and in the case of the A1, ACPI doesn't influence the interrupt assignment. Only the advanced interrupt controller (APIC) could influence the interrupt assignement, but it isn't used at all on the AmigaOne (APIC is x86 specific hardware, IIRC for Pentium2 and above), because every interrupt is routed to the cascaded i8259 interrupt controllers.

AFAIK PCI devices can choose between four interrupt lines (INTA-INTD). The routing of these interrupt lines can be specified in the southbridge (normally they are assigned to IRQ7, IRQ9, IRQ10, IRQ11 - u-boot settings), but not all interrupt lines are available for every PCI device/slot. Some devices can only use INTA and INTB for example (e.g. graphics card), which would result in some interrupt lines being used twice (graphics card on AGP bus and sound card on PCI bus). That should not be a problem, because PCI interrupt are defined as level triggered.

_________________
A1SE: G3@600MHz, 2GB, 1GBit network card
A1XE: G4@933MHz, 2GB, refitted AC'97 codec
microA1: G3@800MHz, 1GB

- A1 Linux support -

@itix

Quote:

I dont know is this just coincidental but tv-cards happen to have troubles on Pegasos I too and it uses VIA southbridge. I think it sounds reasonable because it is the only common nominator there...

Note, that PegasosI uses another southbridge than A1, but they use the same northbridge (ArticiaS). IIRC TV card software doesn't really work with PegI. The problems with sound were mostly fixed though

@Spirantho

Please keep working on AmiTV, I think it's a must have app for OS4. Got my TV-card now so I really want a stable version. Thanks!

/shoe

@brotheris

I know I know But sshhhhh...!

_________________
Amiga Developer
Amiga 500, Efika, Mac Mini and PowerBook

@shoe
Quote:

Please keep working on AmiTV, I think it's a must have app for OS4. Got my TV-card now so I really want a stable version. Thanks!

We all want a stable version.. that's what's holding it up. :)

I need to continue experimenting to see if I can find a work-around, but it could take some time...

Hopefully when a new OS4 compatible machine appears it'll work better!