INIC-1620TA/1622TA2/1623TA2 SATA controllers from initio

Hello, fellow ATA developers.

I've just bought this brand-new/quite affordable INIC-1622TA2 controller. AFAIK, there's no driver yet. Initio has kindly posted document of this controller on their website[1]. I skimmed through the doc and it seems to be a pretty decent piece of hardware with NCQ support.

I'm gonna work on this driver probably from next week or the week after.
So, please let me know if someone is already working or planning to
work on it. I'll be happy to be a tester in that case. The other way around would be helpful, too. If you have this type of controller and feeling like to participate in testing, please let me know.

Thanks.

[1] http://initio.com/products/sata.htm

tejun

Hi Tejun,

It may be a good idea to ask Jeff to update his excellent status web-page:

http://linux.yyz.us/sata/sata-status.html

I guess that patch updating html file is fine but verify that with Jeff.

Bartlomiej

Yes, patches to those web pages are welcome...

Jeff

Jeff Garzik <[email protected]> :
[...]

@@ -14,7 +14,7 @@ urchinTracker();
 
 <center>
 <h1>Serial ATA (SATA) Linux status report</h1>
-<h2>January 25, 2006</h2>
+<h2>February 1, 2006</h2>
 </center>
 
 <P>

@@ -42,6 +42,7 @@ google_ad_channel ="9598368507";
 <li><a href="#recent_updates">Recent updates</a>
 <li><a href="#hw_support">Hardware support</a>
 	<ul>
+	<li><a href="#inic">Initio INI-1623</a></li>
 	<li><a href="#ich5">Intel ICH "IDE mode"</a>
 	<li><a href="#ahci">AHCI (newer Intel ICH, ULi, others)</a>
 	<li><a href="#tx2">Promise TX2/TX4</a>

@@ -77,7 +78,18 @@ Recent updates have concentrated on bug 
 <hr>
 <h2><a name="hw_support">2. Hardware support</a></h2>
 
-
+<h3><a name="inic">Initio INI-1623</a></h3>
+<p>
+Tejun Heo coordinates the development. See the
+<a href="http://marc.theaimsgroup.com/?l=linux-ide">linux-ide</a> mailing-list
+archive to get in touch with him. The hardware documentation is available:
+<ul>
+<li><a href="http://initio.com/images/products/sata/Initio INIC-1623 Databook v1.3.pdf">INIC-1623 1st generation</a></li>
+<li><a
+href="http://initio.com/pdfs/Initio INIC-1620TA2_1622TA2_1623TA2 Databook v1.4.pdf">INIC-1623TA2 2nd generation</a></li>
+</ul>
+</p>
+<hr />
 <h3><a name="ich5">Intel ICH "IDE" mode</a></h3>
 
 Summary: No TCQ/NCQ. Looks like a PATA controller, but with a few added,

Any progress? I acquired two of those beasts today and *really* would like to get them up&running ASAP. :-/

There seems to be a driver for Linux 2.4 at
http://www.sunix.com.tw/it/en/down/driver/SATA/SATA2100/Linux%202.4.zip
which might be helpful in writing a 2.6 driver.

The documentation on that page doesn't mention the 1622 any more, only the 1623. Does anybody still have the "old" data sheet?

Which of the sata_*.c drivers would be a good sample code / starting point for writing a new driver?

Thanks,

Matthias Urlichs   |   {M:U} IT Design @ m-u-it.de   |  [email protected]
Disclaimer: The quote was selected randomly. Really. | http://smurf.noris.de
 - -
The worst men often give the best advice.
					-- Bailey

Yeah, I saw that. The problem with the driver is that it doesn't contain any license notice. I'm not sure if it would be okay to read the driver and use the acquired info to implement GPL driver. I asked sunix about it but got no useful answer.

For controllers which support NCQ, ahci and sata_sil24 are pretty good examples.

tejun

Its clearly based on some of the GPL IDE controller code and blocks of it are cut/pasted.

If you want to write the driver then I'm happy to write the description of the device. It's basically an SFF device with MMIO on BAR5

Alan

Ok I noiw have a copy of the driver which is clearly marked as GPL and since it credits me anyway makes it clear enough for anyone I'd think

libata driver will follow, just needs a test victim...

Alan, I have the hardware and I have half-finished driver lying around somewhere. If you send me the GPL driver, I can do it too.

Thanks.

tejun

Hi,

Alan:

*raises hand* I've got a freshly-installed Debian box here, which should be an ideal test candidate over the weekend.

Matthias Urlichs   |   {M:U} IT Design @ m-u-it.de   |  [email protected]
Disclaimer: The quote was selected randomly. Really. | http://smurf.noris.de
 - -
"When are you BUTTHEADS gonna learn that you can't oppose Gestapo
tactics *with* Gestapo tactics?"
					-- Reuben Flagg

Hi

How did this go? I doubt I'm the only one who fell afoul of the "has linux drivers" implication on initio's website...

I know its early days, but any idea when this might be ready for production?

Regards & thanks,
--

View this message in context: http://www.nabble.com/-COORDINATE--INIC-1620TA-1622TA2-1623TA2-SATA-controllers-from-initio-tf987765.html#a7827233
Sent from the linux-ide mailing list archive at Nabble.com.

I sent Tejun the data book but we've yet to succeed in cloning him or getting him a time machine.

I had a look at it. The chip is a little weird so it'll take someone a while to write a driver for it and they'll need the hardware handy.

Hi,

The chip really is weird. For some reason, I can't do ATA PIO from mmio TF but can do so using BAR{0,2}. It does have sff-compliant BAR0-3 but no BMDMA at BAR4. And it can't do more than 256 sects at a time. Ah... also, SRST nor phy reset via SControl doesn't give valid device signature. Only hardreset via IDMA_CTRL works. Bah...

Anyways, I'm getting there. My 1622 is happily transferring 76M file without problem in and out of a seagate harddrive using pio mode. So, please standby.

tejun