Atto UL2D firmware summary - Question for OSXGuru |
October, 26, 2002 7:03 PM |
jeglin |
[This was all done using a PTP with a G3/500 PowerLogix upgrade card, and XPF 2.2.4b1] Using the link to the site found from the recent Atto thread... http://www.focal.co.jp/1_PRODUCT/ATTO/ ... to get the older firmware revisions, and the latest version from Atto themselves, I have tested 1.3.5(f2), 1.3.6(f0), 1.4.4(f1), 1.5(f0), 1.6.2(f2), and 1.6.4(f0). All except 1.6.x will boot on OS9. None will boot OSX :-( The problem on OS9 with the 1.6.x versions is still a little unclear to me. The the casual viewer will see that, after boot and just before the disks are mounted on the desktop, there is a System Error requiring reboot. MacsBug show that there is a null pointer that hoses the system; that would make it officially a bug, but whose I am not sure. Digging around in MacsBug eventually finds me inside a bunch of calls of the name MPSomething_or_other. I guessed that maybe this was the Multiprocessing stuff but removing just the MP plug-in from the extensions folder did not fix it. If one boots without extensions, there is *no* crash, but then XPF cannot run because it cannot call GetComponentResource (part of InterfaceLib). OTOH, running only OS9 Base agains causes the crash. By turning off additional extensions from OS9 Base, one can actually boot into OS9 with firmware 1.6.x and still run XPF. I did not further sleuth which extension(s) are key because most unfortunately, I found that with 1.6.x the "can't open" error for the Atto is still present. So, finally the question for OSXGuru: any clues on whether this Atto boot issue is more likely an Atto problem or an OSX problem? (I'm assuming it's not an XPF problem :-) |
. |
Atto UL2D firmware summary - Question for OSXGuru |
December, 10, 2002 8:45 PM |
OSXGuru |
. |
I'll get the "official" version 2.2.5 out the door soon, and that will have support for shortening these boot-device strings some more. Hopefully that will deal with the remaining NVRAM size issues--we'll see. |
. |
XPostFacto 2.2.5b2 and NVRAM |
December, 07, 2002 6:09 AM |
joevt |
. |
I am unable to boot in OS X with XPostFacto 2.2.5b2 even though it adds only a couple characters to the boot-device string. When I push the restart button in XPostFacto, it boots into OS 9 instead of OS X and my IDE drive partitions do not appear until I restart into OS 9 again. Sonnet Tempo Trio: before (2.2.4): pci1/pci-bridge/UltraTek133/@0:9 after (2.2.5b2): pci1/pci-bridge/UltraTek133@2/@0:9 I think the limit on the boot-device string is only 32 characters on Old World Macs. If XPostFacto can't figure out how to make shorter names then maybe it should have an option to allow manual entry of the name (input, output and boot; though only boot and output tend to be long). |
. |
NVRAM error solved killing ATI 7000 |
December, 06, 2002 9:13 AM |
muri |
. |
I tried to boot from the built-in video card and I got no NVRAM error ... maybe selecting the ATI radeon 7000 increases the size of what is written to NVRAM ... but how can I enable the ATI card now that I boot in X? Thank you |
. |
OSXGURU: NVRAM error in log |
December, 03, 2002 3:47 PM |
muri |
. |
This is the last part of the log from my install: Restarting ... Boot-device: pci1/ATTO,ExpressPCIProUL2D@E,1/@3:6 Boot-file: -i Boot-command: 0 bootr -v rd=*scsi-int/@3:9 input-device: kbd output-device: pci1/ATY,RV100Parent/ATY,RV100ad_A NVRAM size limits exceeded any idea? |
. |
Atto's response to my UL2D firmware inquiry |
December, 03, 2002 3:19 PM |
Tony.Scaminaci |
. |
Here's the tail end of a string of emails between Atto tech support and me over the past week. I don't seem to be making a lot of progress in convincing them to support the UL2D cards, but I took one last stab at it anyway: Mark, Once again, thanks for your detailed reply. I must say that, as a loyal customer who owns 4 of your cards, I feel slighted. I don't know why your management made the decision to drop all support for the UL2 series cards when there are so many of these cards out there. The only reason I started out with this card is that Apple put it into my G4 when I requested SCSI support. It's been a flawless performer for three years now and I've purchased three more of your SCSI cards because of the UL2D's performance and reliability. Atto has been regarded as a top-notch supplier for SCSI cards - something that your competitors can't really claim. Being an engineer myself, I can understand why support must move on to newer designs. However, I haven't heard of any trade-up program being offered to move existing customers from the UL2 series cards to the UL3 series. As an end-user, it makes no sense to trash both of my UL2D cards just because Atto has end-of-lifed them. They work just fine, with the exception of the sleep issue under OS X. Most vendors either extend support for older, but still usable equipment, or offer customers a trade-up program in order to pull the older equipment off of the market. Most vendors don't just drop support for perfectly good equipment without offering their customers a viable, common sense alternative. Paying full retail price for UL3D cards so that my system could sleep in OS X would be ludicrous. But I might be willing to upgrade the cards as part of a trade-up program. Please forward my comments to your management. I would like to hear directly from them why they are alienating good customers. I can't believe that sleep support in the UL2 series could be very different (if at all) from that of the UL3 series. If the intent is to force existing customers to buy UL3 cards and throw out the UL2 cards, I can assure you that won't happen. Regards, Tony techsupp@attotech.com wrote: > > Hi Tony, > We are sorry, but as I mentioned in my previous email that there will > be no further firmware being developed for the UL2 series cards. This > includes support for deep sleep. Yes, we just released a new plus driver > for the PSC, and UL3 series cards that will support sleep mode in OSX, but > the UL2 cards have been End of Life for about a year now. I can certainly > bring your request to the attention of the people who make those decisions > but I know for sure there are no plans on changing the status of the UL2 > series cards. > The following is a standard email regarding sleep support for the UL2 > cards, now with the addition of the Plus driver the PSC and the UL3 cards > have support. > > With regards to Power Management Support, please refer to the table > below. The UL2D, UL3S, and UL3D are the only cards that support Power > Management - and must be in OS 9 and in a G4 Model machine ONLY! No Fibre > Cards or other SCSI Adapters are supported at all. Additionally, nothing > is supported in OS X as of today. Though there is not an official > statement yet, I understand that we will be including Power Management > Support in future releases. > > Power Management Support > ______________________________________________________________________ ______ > PSC DC UL2S UL2S UL2D UL3S UL3D FibreChannel > OS 9.x NO NO NO NO YES* YES* YES* NO > OS X NO NO NO NO NO NO NO NO > > * = G4 Models ONLY > > If you have any further questions, feel free to ask. > > Regards, > > Mark Polito, Technical Support Technician > ATTO Technology, Inc. · 155 CrossPoint Parkway · Amherst, New York 14068 > PHONE: 716-691-1999 ext. 242 - FAX: 716-691-9353 > EMAIL: TechSupp@attotech.com > www.attotech.com |
. |
still can't boot (problem writing to NVRAM) |
December, 03, 2002 5:42 AM |
muri |
. |
Both 225b1 and 225b2 keeps on telling me that there is a problem writing to NVRAM ... what does this mean? |
. |
RE: Atto UL2D firmware summary - Question for OSXG |
November, 27, 2002 9:51 PM |
brian.stroud |
. |
Well I can confirm that 2.2.5b2 does indeed seem to solve the boot problem with this card. Under 10.2 it will also work with the latest 1.6.4(f0) firmware installed, although as discussed previously this causes a problem when booted back into 9. I am still unable to install 10.2 on my 37GB second drive, and I think that this may be a drive problem, as it fails on both the ATTO UL2D card and the ATTO PSC card. After using the Install option in XPostFacto the machine restarts, data is read from the 37GB drive as though it is about to boot, but then the reads become less and less until there is no drive activity - I guess about 5 seconds in all, thoughout this time there is no video signal. Any ideas on this one, does the boot drive have to be on SCSI ID 0 ? |
. |
RE: Atto UL2D firmware summary - Question for OSXG |
November, 27, 2002 2:24 PM |
jeglin |
. |
OSXGuru! I've been away, I didn't see all your posts until just now. I'm gonna track this rabidly, hopefully starting late tonight. I also have an Apple developer utility to sleuth the device tree in OS9, so I would be happy to troubleshoot this extensively with you. Thanks!!! |
. |
... and version 2.2.5b2 |
November, 27, 2002 7:13 AM |
OSXGuru |
. |
It looks like version 2.2.5b1 may be mixing up the two channels. So I've posted a version 2.2.5b2 that swaps the names--this may work for the moment, though it's a little inelegant. In any event, if 2.2.5b1 doesn't work, try this instead: http://eshop.macsales.com/OSXCenter/XPostFacto/Download/XPostFacto2.2.5b2.sit And let me know if either one works :-) |
. |
XPostFacto version 2.2.5b1 |
November, 26, 2002 7:16 PM |
OSXGuru |
. |
I've posted XPostFacto version 2.2.5b1. It is meant to get the boot-device correct for the Atto card. Ideally, you should get a boot-device with "@E,1" at the end (or something like that--the "E" part can vary). Here's the URL for download: http://eshop.macsales.com/OSXCenter/XPostFacto/Download/XPostFacto2.2.5b1.sit Let me know if it helps--if it does, I'll release it formally. If not, I'll try again. |
. |
OK, I've figured out one issue |
November, 26, 2002 5:01 PM |
OSXGuru |
. |
I should have been able to figure this out faster than this, but better late than never ... It looks like the Atto cards create two entries in the device tree, which differ only in their function number (the ",1" at the end of the Open Firmware name). Basically, XPostFacto would have been unable to set the boot-device properly for anything on the ",1" channel, since it didn't know how to figure out what the function number was. The good news is that in figuring this out, I realized that it is absurdly simple to get the device number and function number--they are encoded into the assigned-address property in the device tree. (I should have figured this out long ago, too). So it will not be difficult to get XPostFacto to generate the correct boot-device. (And a side benefit is that if NVRAM space is low, it is possible to omit the name altogether and just use the device number and function number). One problem is that it won't be easy for XPostFacto to match volumes with the correct device-tree entry. Hopefully it will just get it right--if not, I'll figure something out. So, my apologies for not figuring this out sooner, and I should have something for you to test soon (this weekend, if not sooner). I don't know whether it will fix all of the problems, but it will certainly get rid of one roadblock. |
. |
Atto UL2D firmware summary - Question for OSXGuru |
November, 26, 2002 3:26 PM |
OSXGuru |
. |
Fascinating--the @E,1 at the end is very interesting indeed. Ordinarily, you don't absolutely need the @E--it tells Open Firmware which slot the card is in, but it can figure that out from the name. But the ,1 is a parameter of some kind. I wonder whether it is a reference to the SCSI channel? Are there two channels on this card? It might be helpful if I could see the device-tree entry for the card. The easiest way to get it would be to use this command in the terminal: ioreg -p IODeviceTree -l Perhaps you could e-mail the output to me? |
. |
Atto UL2D firmware summary - Question for OSXGuru |
November, 26, 2002 2:15 PM |
kbata |
. |
boot-device pci1/ATTO,ExpressPCIProUL2D@E,1/@0:6 |
. |
RE: boot-device settings |
November, 26, 2002 1:48 PM |
OSXGuru |
. |
I've noted that some of you are able to boot from the ATTO card if you set up the boot process with the Startup Disk preference pane in Mac OS X. However, if you try to use XPostFacto, it doesn't work. What this suggests to me is that XPostFacto is getting the boot-device wrong. If someone could do a little experiment for me, it might help. When you have successfully booted from the ATTO card, use the following command in a Terminal window: nvram boot-device On my machine, that results in something like: boot-device pci1/ACARD,6260/@0:6 You'll get something different--write it down for me. Then, reboot into Mac OS 9. Launch XPostFacto and select the ATTO-connected volume as the disk to boot from. The XPostFacto window should say what it thinks the proper boot-device setting is. Note it down. In theory, it should be the same as (or similar to) the setting that Mac OS X uses. If it is different, perhaps I can fix XPostFacto to get it right. Of course, only try this if you know how to successfully get back to Mac OS X on the ATTO card after booting into Mac OS 9. |
. |
RE: Atto UL2D firmware summary - Question for OSXG |
November, 22, 2002 12:33 AM |
burnside |
. |
I had the same problem of a start up crash with the UL2D/1.6.4 card on a 9500/Powerlogix G4/450MHz/Quantum IV 18GB 80pin (Apple formatted) with OS 8.6, 9.2.1 and 9.2.2 (patched) . This traced to Foreign File Access extension as it would boot fine with this disabled.It also booted fine before I upgraded the firmware. Unable to boot 10.2 on the IBM 36GB 68pin (HDST) or 10.1.3 on the Quantum after the upgrade, didn't try before. Both drives and card were from OWC. |
. |
RE: Atto UL2D firmware summary - Question for OSXG |
November, 21, 2002 10:00 AM |
jeglin |
. |
kbata: see my original post at the top :-) Firmware 1.4.4 and 1.5 are good for booting OS9.x with the UL2D card. |
. |
RE: Atto UL2D firmware summary - Question for OSXG |
November, 21, 2002 9:10 AM |
gregoryy |
. |
I don't think you want to use Apple Startup Disk control panel if you ever want to choose OS X, then use System Disk 3.3.1 or 2.6.2. Apple's will change the nvram settings that XPF made I think, the others are safe to use and have worked. |
. |
RE: Atto UL2D firmware summary - Question for OSXG |
November, 20, 2002 6:07 PM |
muri |
. |
Atto told me that last version supporting OldWorld Macs is 1.4.4 ... I guess that with NewWorld any version goes. |
. |
RE: Atto UL2D firmware summary - Question for OSXG |
November, 20, 2002 6:07 PM |
muri |
. |
Atto told me that latest version supporting OldWorld Macs is 1.4.4 ... I guess that with NewWorld any version goes. |
. |
Atto UL2D firmware summary - Question for OSXGuru |
November, 20, 2002 10:53 AM |
kbata |
. |
Does anyone know if this card will boot into 9.1 with any of the firmware versions? |
. |
Atto UL2D firmware summary - Question for OSXGuru |
November, 20, 2002 10:40 AM |
kbata |
. |
Maybe it would be a good idea if everyone on this forum contacted Atto to encourage them to change their minds and develop one more update to the firmware for this card. In the email they sent me they said the reason for stopping development was that the card was discontinued before OS X was released. That may be true but the problem is more with OS 9 than X. |
. |
Atto UL2D firmware summary - Question for OSXGuru |
November, 16, 2002 7:36 PM |
earlyd |
. |
Opps. In my posting of November, 16, 2002Â 5:27 PM, the last sentence in the 1st paragraph should have read: "I've never tried to install it on the ATTO drive directly." Sorry 'bout that. --Dwight |
. |
RE: Atto UL2D firmware summary - Question for OSXG |
November, 16, 2002 6:10 PM |
paul_findley |
. |
Just confirmed that selecting the LVD drive with OSX on the ATTO, from a motherboard's 50-pin drive's booted OSX, system preference/startup pane, allows the LVD drive to boot OS X on my 8500. So when ATTO says U2LD is not bootable for OS X, it seems they are being a bit modest. As noted, G3's and above have no problem, and it looks like we have a minor problem in how the drive is tagged for booting in the legacy machines. I'll try copying over the 9.2.2 startup disk panel onto my 9.1 installation, and see if that solves the problem. It's a pain having two OSX installs, one just to choose the other one. |
. |
Atto UL2D firmware summary - Question for OSXGuru |
November, 16, 2002 5:27 PM |
earlyd |
. |
Oh, Oh. I just thought of something. I always test a new version of the 10.x by installing it on my internal (built-in) SCSI drive. I then use the copy utility CCL to copy the partition to my unpartitioned Atlas drive on the ATTO UL2D SCSI chain. I've tried to install it on the ATTO drive directly. I would suggest that if you can boot into 9.x with the UL2d, try my method to see if you can boot from the UL2D. --Dwight |