20-Apr-2024 03:09 GMT.		 UNDER CONSTRUCTION
Anonymous, there are 15 items in your selection
[News] Bounty 11 - New ATA DeviceANN.lu
Posted on 06-Jul-2004 19:28 GMT by One Man Army15 comments
View flat
View list
Aros-Exec Reports that AROS now has a new ATA Device.
Bounty 11 - New ATA Device : Comment 1 of 15ANN.lu
Posted by Amon_Re on 06-Jul-2004 18:03 GMT
Only PIO Modes? No dma? (Hah!) :P
Bounty 11 - New ATA Device : Comment 2 of 15ANN.lu
Posted by Michal Schulz on 06-Jul-2004 18:30 GMT
In reply to Comment 1 (Amon_Re):
As you can see DMA support is one of the nearest-future improvements. However, I wanted to release the code alredy. And no, it will not take year or two to implement DMA.
Bounty 11 - New ATA Device : Comment 3 of 15ANN.lu
Posted by Amon_Re on 06-Jul-2004 20:27 GMT
In reply to Comment 2 (Michal Schulz):
Relax, i wasn't having a dig at you, it's a joke ;) Keep up the good work
Bounty 11 - New ATA Device : Comment 4 of 15ANN.lu
Posted by JoannaK on 06-Jul-2004 22:07 GMT
I just tried Aros Live-CD for a first time.. (20 March version) Looks real nice, seems highly stable, fast etc...
Bounty 11 - New ATA Device : Comment 5 of 15ANN.lu
Posted by Anonymous on 07-Jul-2004 02:23 GMT
I love the irony... several well-known AROS developers have been spending a fair bit of time criticizing Hyperion for taking so long to write a DMA-capable IDE driver. Eventually AROS gets its own IDE driver, and what's left out? DMA. It will be *really* funny if the first DMA-capable release of the AROS driver corrupts data.

Now, this does not apply to *all* AROS devs, most are pretty decent guys, but I can't help thinking that if a certain individual with initials FA had kept his trap shut, there's be less sniggers and more congratulations. Bottom line, people living in glass houses should not throw stones.
Bounty 11 - New ATA Device : Comment 6 of 15ANN.lu
Posted by brotheris on 07-Jul-2004 04:25 GMT
In reply to Comment 5 (Anonymous):
Yes, x86 motherboard makers use the same chipset everywhere =]
Bounty 11 - New ATA Device : Comment 7 of 15ANN.lu
Posted by Elwood on 07-Jul-2004 04:53 GMT
Hi,

Am I right to say that if we want to know about DMA, an AROS version for the AmigaOne would be the only way ? i.e. if you do a DMA-capable driver on a x86 machine, it won't prove anything related to the potential DMA problem of the A1.
Bounty 11 - New ATA Device : Comment 8 of 15ANN.lu
Posted by Anonymous on 07-Jul-2004 06:14 GMT
In reply to Comment 6 (brotheris):
Is that why they decided to support *no* chipset whatsoever? Too many to choose from so they pick none? I suppose that it's a lot easier to slag the work of people than it is to do it yourself...
Bounty 11 - New ATA Device : Comment 9 of 15ANN.lu
Posted by Anonymous on 07-Jul-2004 07:42 GMT
In reply to Comment 7 (Elwood):
>it won't prove anything related to the potential DMA problem of the A1.

Tss, Elwood :-)

There is no "potential DMA problem of the A1."
There is a DMA problem of the A1.

Q: Will you make believe that data corruption in not a problem ?
A: Linux sucks

Q: But it works with Peg I + April
A: April is a doggle

Q: but it works without slow down!
A: dunno what you could invite now :-)

PS: ok there is a bit of trolling, but, Elwood seems to think that its computer works...

Bye
Bounty 11 - New ATA Device : Comment 10 of 15ANN.lu
Posted by Fabio Alemagna on 07-Jul-2004 08:48 GMT
In reply to Comment 5 (Anonymous):
> I love the irony... several well-known AROS developers have been spending a
> fair bit of time criticizing Hyperion for taking so long to write a
> DMA-capable IDE driver. Eventually AROS gets its own IDE driver, and what's
> left out? DMA. It will be *really* funny if the first DMA-capable release of
> the AROS driver corrupts data.

It takes a fool to recognize irony in this, because no one in the AROS team ever criticized them for "taking so long". The discussions have solely been about the AmigaONE, not their ability to get something done.

As Michal said, he just wanted to get the device out as soon as possible, DMA is going to be implemented in the next few days. I have been one of the guys who begged him to commit the driver as it is now, he didn't want to, at first.

> Now, this does not apply to *all* AROS devs, most are pretty decent guys, but
> I can't help thinking that if a certain individual with initials FA had kept
> his trap shut, there's be less sniggers and more congratulations. Bottom line,
> people living in glass houses should not throw stones.

At least I am not an anonymous coward. Look, you say whataver you want about me, but at lest put a name over there, so I know who's the fool is talking to me .-)
Bounty 11 - New ATA Device : Comment 11 of 15ANN.lu
Posted by Anonymous on 07-Jul-2004 08:49 GMT
In reply to Comment 4 (JoannaK):
JoannaK: try with a nightly build!
You are using a pretty obsolet version...
For instance, without this new ATA driver...
Bounty 11 - New ATA Device : Comment 12 of 15ANN.lu
Posted by Fabio Alemagna on 07-Jul-2004 08:52 GMT
In reply to Comment 11 (Anonymous):
The current night ly build doesn't have this device activated, there's still the old (and pretty damn slow) ide.device. This is so that other AROS devs can test the device and report problems, if any, before the casual user risks to destroy his HDD :-)
Bounty 11 - New ATA Device : Comment 13 of 15ANN.lu
Posted by Graham_NLI on 07-Jul-2004 14:48 GMT
Well done, and when DMA is done it will be a major step forward in usability.

Still a long way to go, but I might try a nightly build once the new driver is in. What other software comes on the nightly build livecd?
Bounty 11 - New ATA Device : Comment 14 of 15ANN.lu
Posted by hammer on 07-Jul-2004 20:15 GMT
In reply to Comment 8 (Anonymous):
Note that, the Microsoft’s generic DMA(upto UDMA Mode5) IDE driver works on nForceI/II/III's IDE controllers.

Generic MS IDE drivers are;
_atapi.sys; (Microsoft's generic IDE port driver).
_pciide.sys; (Microsoft's generic controller minidriver; This handles the hardware-dependent aspects of the controller driver-minidriver pair and which can be used with most IDE controller hardware.)
_pciidex.sys; (Microsoft's generic PnP and power management)
Bounty 11 - New ATA Device : Comment 15 of 15ANN.lu
Posted by Michal Schulz on 11-Jul-2004 17:05 GMT
In reply to Comment 5 (Anonymous):
I love the irony... several well-known AROS developers have been spending a fair bit of time criticizing Hyperion for taking so long to write a DMA-capable IDE driver. Eventually AROS gets its own IDE driver, and what's left out? DMA. It will be *really* funny if the first DMA-capable release of the AROS driver corrupts data.

As I promised, I have released today an UDMA-capable version of ata.device. Until now I haven't noticed any data corruptions allthough some intensive tests. Perhaps I will try to move bigger data and do MD5 on it.
Anonymous, there are 15 items in your selection
Back to Top