Version 2.2.5b3 available (better use of NVRAM) |
December, 26, 2002 4:32 PM |
OSXGuru |
Version 2.2.5b3 of XPostFacto is available for testing. Here is the URL: http://eshop.macsales.com/OSXCenter/XPostFacto/Download/XPostFacto2.2.5b3.sit The new feature in version 2.2.5 is that XPostFacto is now aware of the "device number" and "function number" for PCI devices. This helps with certain SCSI cards which differentiate between SCSI busses by using different function numbers. What version 2.2.5b3 adds is that it checks to see whether we are going to use too much space in NVRAM. If so, it uses only the device number and function number, rather than the name of the card. So, for instance, instead of something like: pci1/ACARD6280@E/@0:9 you would have: pci1/@E/@0:9 So this may be helpful for those of you who have had trouble with running out of space in NVRAM. |
. |
RE: Version 2.2.5b5 out now |
December, 28, 2002 8:59 PM |
OSXGuru |
. |
I think that version 2.2.5b5 should help with the UltraTek-based cards. I'm making the assumption that all the affected cards have Open Firmware names that begin with "UltraTek". I'm not absolutely sure this is true, but it does simplify things considerably :-) Here's the URL: http://eshop.macsales.com/OSXCenter/XPostFacto/Download/XPostFacto2.2.5b5.sit |
. |
E: Version 2.2.5b3 available (better use of NVRAM) |
December, 28, 2002 8:07 PM |
OSXGuru |
. |
It looks like there are two problems with 2.2.5b3. One is that it is not getting the Open Firmware name for the Radeon 7000 correct. I've been working on that, and have it fixed now. In doing so, I've now got a better algorithm for coming up with the Open Firmware names in general. So for that problem, version 2.2.5b4 should help--and you can get it from here: http://eshop.macsales.com/OSXCenter/XPostFacto/Download/XPostFacto2.2.5b4.sit The other problem is specific to the UltraTek-based ATA cards. What's happening there is a little complex. The UltraTek cards pretend to be SCSI cards in Mac OS 9, but are ATA cards in Mac OS X. So if you ask "which Open Firmware node corresponds to the UltraTek card?", you get a different answer in Mac OS 9 and Mac OS X. In Mac OS 9, you get an Open Firmware node that is a SCSI device, and in Mac OS X you get an Open Firmware node that is an ATA device. This explains the transformation that you noticed, joevt. pci1/@F/UltraTek133/@0:9 points to the SCSI node. The firmware on the UltraTek card sees this, and changes it to the corresponding ATA node (which is pci1/@F/UltraTek133/FrmTk-0/@0:9 in your case). The problem is that the UltraTek firmware does not make this transformation if the boot-device is pci1/@F/@2/@0:9 instead. Technically, this boot-device settings is "correct", in the sense that it corresponds to the SCSI node. But the UltraTek firmware doesn't recognize that it needs to do the transformation to the ATA node. There are a couple of options for fixing this. One would be to never abbreviate UltraTek nodes, so that the UltraTek firmware knows to make the transformation. The other would be for XPostFacto to just make the transformation itself. The problem is that then XPostFacto has to know what transformation to make. That isn't a difficulty in the cases I know about, but there may be other cases (or changes in the future). So, I'll fix it one way or the other. BTW, the reason that saving 3 bytes (" -v") allows for another 24 is that XPostFacto isn't particularly sophisticated about how much abbreviating to do. If not abbreviating would require more NVRAM space than is available, then XPostFacto abbreviates everything. So if the " -v" pushes you over the edge, then everything gets abbreviated, which saves a lot more than 3 bytes. |
. |
RE: Single line spacing |
December, 28, 2002 3:18 PM |
nick.ashton |
. |
To single space lines you need to put < br > at the end of each line. Don't include the spaces around the br string - I only put them in so it would print instead of being interpreted as html. |
. |
RE: Version 2.2.5b3 available (better use of NVRAM |
December, 27, 2002 8:50 PM |
joevt |
. |
That was better. One more time (viewing the source for this page would have shown the indenting in the original message). One last time: FF831B50: /bandit@F2000000 FF841660: /ATY,RV100Parent@D FF8637E8: /ATY,RV100ad_A@0,0 FF864A28: /ATY,RV100ad_B@0,0 FF865E98: /pci-bridge@F FF866B00: /pci16b8,12@0 FF866D70: /pci16b8,12@0,1 FF866FE0: /pci16b8,21@0,2 FF867250: /pci16b8,1@1 FF8674E8: /UltraTek133@2 FF891970: /FrmTk-0@10,0 FF891DD0: /sd@0 FF892440: /FrmTk-1@11,0 FF8928A0: /sd@0 Now, how do you make the lines single spaced? |
. |
E: Version 2.2.5b3 available (better use of NVRAM) |
December, 27, 2002 8:03 PM |
joevt |
. |
Ryan, you don't have a Radeon 7000 or Tempo Trio to test with? I get repeated "can't OPEN:" errors being sent to the modem port because the output-device set by XPostFacto 2.2.5b3 is wrong. If I use auto-boot? false and type boot in Open Firmware, only one "can't OPEN:" error is displayed without leaving Open Firmware. Here is the device tree for my Power Mac 8600: FF831B50: /bandit@F2000000 FF841660: /ATY,RV100Parent@D FF8637E8: /ATY,RV100ad_A@0,0 FF864A28: /ATY,RV100ad_B@0,0 FF865E98: /pci-bridge@F FF866B00: /pci16b8,12@0 FF866D70: /pci16b8,12@0,1 FF866FE0: /pci16b8,21@0,2 FF867250: /pci16b8,1@1 FF8674E8: /UltraTek133@2 FF891970: /FrmTk-0@10,0 FF891DD0: /sd@0 FF892440: /FrmTk-1@11,0 FF8928A0: /sd@0 XPostFacto 2.2.5b3 sets the output-device and boot-device to pci1/ATY,RV100Parent/@D pci1/pci-bridge/@2/@0:9 If I turn off verbose mode "-v " then it sets them to pci1/ATY,RV100Parent/ATY,RV100ad_A@D pci1/pci-bridge/UltraTek133@2/@0:9 How does removing three bytes for "-v " allow for an additional 24 bytes for "ATY,RV100ad_A" and "UltraTek133"? Is the @2 necessary if there is no other pci1/pci-bridge/UltraTek133? XPostFacto 2.2.4 sets them to pci1/ATY,RV100Parent/ATY,RV100ad_A pci1/pci-bridge/UltraTek133/@0:9 (btw Mac OS 9.2.2 Startup Disk control panel set the boot-device to / AAPL,ROM which makes it useless) XPostFacto 2.2.5 b3 should have set the output-device to pci1/@D/ATY,RV100ad_A I corrected that in Open Firmware and tried to boot. I still got the "Can't OPEN:" error but the output-device now worked. It seems that the boot-device pci1/pci-bridge/@2/@0:9 doesn't work even though it looks like it should. pci1/@F/@2/@0:9 did not work either. The following does work: pci1/@F/UltraTek133/@0:9 When I booted into OS X, and restarted into Open Firmware, the boot- device was somehow changed to the more specific form pci1/@F/UltraTek133/FrmTk-0/@0:9 I tried pci1/@F/@2/@10/@0:9 and it worked and was not changed after restart. I tried the long form used by XPostFacto 2.2.5b3 (when verbose mode is off) pci1/pci-bridge/UltraTek133@2/@0:9 by using setenv in Open Firmware (I have auto-boot? set to false). I typed printenv to verify, then forced a restart using command control power key, got back into Open Firmware, typed printenv to verify again. Open Firmware had changed it to pci1/pci-bridge/UltraTek133@2/FrmTk-0/@0:9 which works. |
. |
RE: Version 2.2.5b3 available (better use of NVRAM |
December, 27, 2002 4:40 PM |
hermane |
. |
OK Ryan, right again. The drive was formated initially in OSX. When the "no bootable HFS partition" showed up, I reformated the drive in 9.2.2, never considering that 9.1 might be different. So based on your hint, reformated the drive in 9.1 and cloned my 10.2.3/9.2.2 SCSI volume to a single 77GB partition attached to the Trio, resulting in a bootable configuration. I can use either XPF 2.2.4 or 2.2.5b3 to boot the partition - no "out of NVRAM" with 2.2.4 now. The only difference between the two is that 2.2.5b3 doesn't activate the display until the blue screen phase. Boot settings are Trio drive as "boot device", "keyboard" as input, "display port A" as output, and throttle is "none". Thanks again for the fine effort and support. Looking forward to v3 and L3 cache support. "OWC Cache Config" sounds fine to me. |
. |
E: Version 2.2.5b3 available (better use of NVRAM) |
December, 27, 2002 8:38 AM |
OSXGuru |
. |
The "no bootable HFS partition" message may mean your volume has an HFS Plus filesystem with no HFS wrapper. The Mac OS ROMs on older machines only understand HFS, so Apple used to always create HFS Plus volumes inside an HFS wrapper. However, it seems that the Disk Utility in Mac OS X creates HFS Plus volumes without a wrapper. Perhaps the Drive Setup in 9.2.2 does this as well--I haven't tested that. All this to say that if you formatted the drive in Mac OS X, then reformatting in Mac OS 9.x may help. And if you formatted in 9.2.2, then reformatting from 9.1 may help. |
. |
RE: Version 2.2.5b3 available(better use of NVRAM) |
December, 27, 2002 12:05 AM |
hermane |
. |
Good News feedback (my baseline configuration): PM9600 w/Sonnet 450MHz G4, 1GB RAM, original internal 4GB SCSI HD has OS9 and OSX; internal 9GB and 18GB SCSI HDs for additional storage and older OSs, RadeonPCI 7000 display card (v208 firmware update), Keyspan 4-port serial card, older OrangeLink 2xFW/2xUSB card. Running OS 9.2.2 (using OS9Helper) and OSX v10.2.3 with XpostFacto v2.2.4; entire configuration works admirably and runs flawlessly - Thanks Ryan! With XPF 2.2.4, display is activated during the entire boot sequence but 2.2.5b3 does not activiate the display until the blue screen phase of the boot. Not so Good News (my experimental configuration): Recently added a Sonnet TempoTrio with an 80GB IDE HD to get additional fast/cheap storage, but there are some issues with this last tweak: Trio FW and USB2 ports work fine, except FW won't boot an external device; initially tried a single ~80GB partition using Drive Setup to partition. Loaded OS9.2.2, which boots and runs OK; cloned OSX v10.2.2 to the drive, but on XPF boot, OSX displays a grey screen and the following text message repeated multiple times: "no bootable HFS partition Can't open deblocker package" and "can't OPEN: pci1/pci-bridge/UItraTek100+/FrmTk-0/@0:9 Can't open deblocker package" Tried an XPF (both v2.2.4 and v 2.2.5b3) install of OSX v10.2 but machine hung during reboot; with 2.2.5b3 however, no longer get "out of NVRAM error" with keyboard and display port A selected. Tried two partitions, one 7.9 GB and the remaining ~70 GB, guessing that the problem might be the <8GB first partition issue with Old World ROMs (not certain this applies for this configuration). OS 9.2.2 boots and runs fine from both partitions but I'm still not able to install OSX on either partition with XPF (v2.2.4 or v2.2.5b3). Still getting a hang on reboot to the OSX installer, and cloning another OSX system to either partition results in the same boot error as above. DFA checkdisk indicates both partitions are normal. I'm ready to go for an 800 MHz processor upgrade as soon as the Trio is booting 10.2 and the L3 cache activation is possible. |
. |
RE: Version 2.2.5b3 available (better use of NVRAM |
December, 26, 2002 9:37 PM |
egonzales21 |
. |
Thanks for reply. You are right. All selections reverted back to none. However, I reselected everything in open firmware menu. I even made sure the throttle was the same. In spite of this the screen does not reappear until the blue screen appears. This happens both to my 10.2.3 partition as well as my 10.2.2 partition. |
. |
E: Version 2.2.5b3 available (better use of NVRAM) |
December, 26, 2002 9:00 PM |
OSXGuru |
. |
Ah, one thing I didn't anticipate about version 2.2.5b3 is that you'll need to reselect the video port in the open firmware menu--your preference will probably have reverted to "none". Does it work OK if you reselect it? Or had you done that already? |
. |
RE: Version 2.2.5b3 available (better use of NVRAM |
December, 26, 2002 8:41 PM |
egonzales21 |
. |
Tried using it to see if it would help me with a problem I am having trying to get the following 3 cards to coexist: 1 Tempo Trio 2 Radeon 7000 3 Atto UL2D. The Atto card is connected to 2 hard drives which serve as my 10.2.3 and 9.1 boot disk respectively. Sorry to say that this did not help my situation. The three cards do fine in 9.1 but the computer will not boot into 10.2.3. I just get an initial blank screen, the screen light turns green but the screen remains the same. To reboot I have to remove the Tempo Trio before the system will reboot. By the why the only difference I see with the 2.2.5b3 version is that my Radeon 7000 does not turn on the screen until the blue screen reappears even when selecting the correct video port in the open firmware menu. Screen was coming on early with all other version of 2.2.4 and 2.2.5 |
|
|