Who's Online |
61 user(s) are online ( 42 user(s) are browsing Forums)
Members: 1
Guests: 60
sailor,
more...
|
|
|
|
Can AGP and PCI-66 be used simultaneously?
|
Posted on: 2007/9/13 20:51
#2401
|
Home away from home
|
This is related to this thread on ATI's recent decision to open up their hardware documentation. There was a discussion a while a go about the PCI-66 slot and whether you could use it and an AGP card at the same time. I don't think that there ever was a definite conclusion. My interest in it is because, while none of the recent AGP cards will work in our AGP 2x slot (wrong voltage), there are a few PCI cards available. If it were functional, then we could possibly use these cards. however, for development it would help if an existing AGP card were also present as a backup. So, does anyone have any experience with using this slot? Can AGP and PCI-66 be used simultaneouslty. Does the slot even work? Hans
|
|
|
|
Re: New ATI drivers?
|
Posted on: 2007/9/13 20:01
#2402
|
Home away from home
|
@billt
Ok, so that would be usable for now. I assume that it plugs into the PCI 66MHz slot. Would it be possible to have both an AGP and a PCI-66 card plugged in and just switch between the two? That would enable experimentation without having to swap cards all the time.
I remember there being a discussion about the 66MHz PCI slot, but I don't think that there was ever a conclusion about whether it could have something plugged in at the same time as AGP.
Hans
|
|
|
|
Re: Memory protection and tasks/processes
|
Posted on: 2007/9/13 18:45
#2403
|
Home away from home
|
@centaurz Quote: centaurz wrote: @Hans
AFAIK, at the moment you'd better use MEMF_SHARED memory between your different processes because parent process memory flagged as MEMF_PRIVATE is not shared with children (NP_Child) (someone correct me if I'm wrong).
When you say process, do you mean a DOS process, or a task? We're going to have to be very specific with our terminology, or things are going to get very confusing. I'm asking about multiple tasks created by one DOS process. Hans
|
|
|
|
Memory protection and tasks/processes
|
Posted on: 2007/9/13 18:25
#2404
|
Home away from home
|
With the new MEMF_PRIVATE flag for memory allocations, I was wondering how this related to tasks/processes. On other OSes, processes have separate memory spaces, but threads within a process can share data.
So, if I declare memory MEMF_PRIVATE, can I access them from all tasks created by the same DOS process, or is it private to that task? It would be nice to have threads with the same memory protection model as other systems (i.e., protect the process, but allow threads to share data freely).
I'm about to start work on multi-threading, and it would like to use what memory protection facilities are available.
Hans
|
|
|
|
Re: New ATI drivers?
|
Posted on: 2007/9/13 18:19
#2405
|
Home away from home
|
@orgin
Unfortunately they're AGP 8x only, which means that they use lower voltage signaling than the Amigaone AGP bus does (1.5V instead of 3.3V). Hence, putting one of these in your machine, would fry stuff.
Hans
|
|
|
|
Re: New ATI drivers?
|
Posted on: 2007/9/13 17:02
#2406
|
Home away from home
|
@orgin
Now would definitely be a good time for PCI-Express Amiga hardware to arrive. AFAIK, the most recent card that can be put into an A1 (without an expensive PCI-to-PCIe riser) is a Radeon 9800.
Hans
|
|
|
|
Re: The Memory Protection Debacle
|
Posted on: 2007/9/10 18:11
#2407
|
Home away from home
|
@LiveForIt
I think that you and Rogue are talking about slightly different things. You want to use AmigaOS4 as a host and run a different OS within it. Rogue mentioned that "ExecSG's HAL can be made to work with a virtualizer service, essentially making AmigaOS 4.0 a hosted system", i.e., AmigaOS4 could be modified to be hosted; Rogue never said anything about it being used as a host.
Running another OS hosted would definitely be great. Mac OSX (PowerPC) would be a nice candidate, although I'd imagine that writing the necessary drivers would require considerable effort.
Hans
|
|
|
|
Re: The Memory Protection Debacle
|
Posted on: 2007/9/5 5:29
#2408
|
Home away from home
|
@Atheist
Rogue has already stated that old software will run in a restricted subenvironment. That way they'll still run, but won't take down the whole system if they screw up.
Hans
|
|
|
|
Re: ACK proven untrustworthy
|
Posted on: 2007/8/26 21:56
#2409
|
Home away from home
|
@Rogue
Any idea what kind of message-passing system the new Amiga OS will have? I hope that it won't require copying data into and out of a buffer, etc., but allow the memory involved to temporarily be shared with the target process.
The other thing I'm wondering about is how this would affect BOOPSI. BOOPSI passes data by reference all over the place. Could this be solved simply by having an intuition thread within a process, or will BOOPSI be deprecated too?
Hans
|
|
|
|
Re: AmiTV 0.4 (CX2388X and Bt8x8)
|
Posted on: 2007/8/25 0:55
#2410
|
Home away from home
|
@LiveForIt Quote: LiveForIt wrote: @Hans
mobilephone and use a SD-CARD reader
Sorry, I don't have a cell phone with a built in camera. I do have a webcam but, to be honest, I don't have any proper PAL video feeds. Without a proper PAL signal, AmiTV 0.5 won't decode colour. Hans
|
|
|
|
Re: AmiTV 0.4 (CX2388X and Bt8x8)
|
Posted on: 2007/8/24 20:50
#2411
|
Home away from home
|
@LiveForIt Quote: LiveForIt wrote: @all
Can some post some screenshots? I like to have a look
Unfortunately I don't know of any program that grabs screenshots + the overlay. video is displayed in an overlay window. Hans
|
|
|
|
Re: AmiTV 0.4 (CX2388X and Bt8x8)
|
Posted on: 2007/8/24 2:23
#2412
|
Home away from home
|
@Cap s-video worked for me. Well, according to xawtv, my card seems to mirror the svideo as a composite signal on composite-2/3 somewhere. A summary of my findings is over on AW.net ( here). It's nice to see it working though. Hans
|
|
|
|
Re: AmiTV 0.4 (CX2388X and Bt8x8)
|
Posted on: 2007/8/23 21:04
#2413
|
Home away from home
|
@Spirantho
I'll try it out as soon as I get home.
Hans
|
|
|
|
Re: Sam440ep "The Red One"
|
Posted on: 2007/8/22 21:45
#2414
|
Home away from home
|
@Ferry
This sounds promising. It would be really nice if they could release Amiga OS4 for this board at the same time. However, we'll just have to wait and see.
Hans
|
|
|
|
Re: Conexant TV card test program!
|
Posted on: 2007/8/18 15:43
#2415
|
Home away from home
|
@Spirantho
Hmm, should I get a connexant based card. The problem is, I haven't found one that has a dual NTSC/PAL tuner. I'm currently in Canada (an NTSC zone) but I will be returning to New Zealand later (a PAL zone). I really don't feel like buying hardware that I'll have to dump less than a year from now.
Ah well, at least you're making progress, and, AmiTV should be rather advanced by the time the SDK arrives. BTW, is the RISC program really going to be larger than 2k (didn't Rogue say that the allocation block size was supposed to be 4k?).
Hans
|
|
|
|
Re: Conexant TV card test program!
|
Posted on: 2007/8/17 21:01
#2416
|
Home away from home
|
@Spirantho
I'll have to wait until you release a BT878 version to test. Have you solved the interrupt problem yet? If you're getting a usable image without interrupts, perhaps you could do the same with the BT878 that the Connexant chip is based on.
Hans
|
|
|
|
Re: Future OS4 direction
|
Posted on: 2007/8/17 17:21
#2417
|
Home away from home
|
@tonyw
I would imagine that exec functions such as SuperState() and Disable() will dissapear too.
Hans
|
|
|
|
Re: When does DMA not DMA?
|
Posted on: 2007/8/17 15:15
#2418
|
Home away from home
|
@Spirantho
That's good news. The interrupt thing is strange though.
Hans
|
|
|
|
Re: When does DMA not DMA?
|
Posted on: 2007/8/16 19:59
#2419
|
Home away from home
|
@Spirantho Quote: Spirantho wrote: This all succeeds, no problemo. DMA 'starts' up ok on each line otherwise it yells loudly.
Then I program the Connexant chip:
DWORD *tPointer = programBase;
*tPointer++ = SwitchEndian( OPCODE_WRITE | 0x0300 ); // This writes 0x300 bytes....
*tPointer++ = SwitchEndian( mFrameBufferLine[ currentLine ] ); // To this address.
That's more or less it, anyway.
This used to work. It doesn't now!
Don't you have to have a SYNC instruction first? Does the Cx2388x have some additional registers over the BT878/879 that need correct initialization first? Quote: However, my data is never being changed. It's DMAing to it, but the target address never changes from 0xDEDEDEDE (my default value).
Is this target address in a register on the card? Or elsewhere. It's a little confusing that you say that it think's that it's writing to the right place when the target address never gets updated. I hope that you can get this working. I also hope that your code will still be able to drive the BT878/879, because that's what my TV card is. Hans
|
|
|
|
Re: Future OS4 direction
|
Posted on: 2007/8/16 0:51
#2420
|
Home away from home
|
@Rogue
I think that a list of programming recommendations that will minimise the porting effort from the current OS4 to the version with full memory protection would be helpful. For example, are you planning on replacing the message passing system completely? Or will we need to declare memory for messages as MEMF_SHARED? Or is there something else that needs to be done?
I don't know if you can answer these questions right now, as I have no idea how far planning/design is for these changes. However, having a list of recommended programming practises would be appreciated.
Hans
|
|
|
|