Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Resolved 2006/2007 Mac Pro (1,1/2,1) and OS X Yosemite
- Thread starter Mr. Zarniwoop
- Start date
- Sort by reaction score
Rephrase all you want, you are still mistaken. For those who use Little Snitch, it is a HUGE issue.
Your argumentative responses always remind me of that famous Mark Twain quote. So I'll bow to your superior experience and leave it there.
wait a sec, I though that he always tried to run Mavericks with LS 3.4.1 and Pike's boot.efi??? only in the end he ran the whole thing on Yosemite and it worked flawless..!? did I get it wrong?
You may have got it wrong according to this report:
I have no personal experience one way or another. I haven't restarted my Mac Pro 1,1->2,1 in weeks. Since nobody seems to have tested my latest "debug" build, I plan to install LS 3.4.1 tonight (UK time) and see what happens when I install that "debug" version of Pike's boot.efi in Mavericks. If it constantly reboots, as reported, I can easily boot into Snow Leopard and repair the harm done by simply copying Tiamo's boot.efi on top. In any case, if any of you wants to test drive my debug build, you are welcome to report failure or success in either Mavericks or Yosemite, or both.
Last edited:
okay, I missed that. but: he mentions "others"... would be interesting to hear from them. because the way I understand it, gpatpandp did not have problems anymore as soon as he tried it with Yosemite. maybe gpatpandp could clarify?
Certainly...sorry for the confusion. My findings have been that I experienced no issues nor any problems in OS X 10.10 Yosemite Pubic Beta 5, with the Pike Alpha Build 6c697848ef installed, and then installing Little Snitch v.3.4.1 from scratch (not updating from a previous Little Snitch version).
The problem seems to manifest itself when updating from a previous Little Snitch version.
Can you mention any others that have had similar difficulties (continuous boot?/inability to boot?) with LS 3.4.1 and Pike's boot.efi in Yosemite? Can you test drive my "debug" build in Yosemite with LS 3.4.1? Does it exhibit the same problem as the "release" builds?
None whatsoever. It simply occurred to me that, since Tiamo's boot.efi is a debug build, perhaps debug builds aren't as picky as release ones. That's all. It's a long shot, but worth trying. If that succeeds, great. If it doesn't, we should compile Tiamo's boot.efi with VS 2013 and the latest version of NASM in Release/Win32 mode and see if it exhibits the same problem with LS 3.4.1 and Mavericks.
2006/2007 Mac Pro (1,1/2,1) and OS X Yosemite
I agree, was just curious.
According to the person who originally posted about this issue, he was unable to update LS while using Mavericks and Tiamo's boot.efi. It was also causing a continous reboot.
I agree, was just curious.
According to the person who originally posted about this issue, he was unable to update LS while using Mavericks and Tiamo's boot.efi. It was also causing a continous reboot.
Hi Hennessie,
Here is "the other" and "the person".
You are completely right Hennessie and gpatpandp also when he wrote:
"The problem seems to manifest itself when updating from a previous Little Snitch version."
I'm planning to update to LS 3.4.1 myself (in Mavericks) later today. My computer hasn't been rebooted in a long time. After the update and the reboot, I'll substitute Tiamo's trusted boot.efi for "my" debug version of Pike's latest commit and see if I get a continuous reboot or all is well. If disaster strikes, it should be relatively easy to repair the harm done. Be that as it may, in case you are able to assess this before I do (about five more hours before I shut down my Mac Pro), it would be great if someone could verify if the debug build is any more tolerant to LS 3.4.1 in Yosemite as I won't be able to test that myself today.
2006/2007 Mac Pro (1,1/2,1) and OS X Yosemite
So the bug is independent of the boot.efi file used (besides the official Apple boot.efi where the bug doesn't exist).
Something must be corrupting the boot.efi. Most likely it is something in the thunk between 32bit and 64bit. Once the thunk is broken the computer can no longer boot.
Ps: sorry, I couldn't remember how to spell your user name.
So the bug is independent of the boot.efi file used (besides the official Apple boot.efi where the bug doesn't exist).
Something must be corrupting the boot.efi. Most likely it is something in the thunk between 32bit and 64bit. Once the thunk is broken the computer can no longer boot.
Ps: sorry, I couldn't remember how to spell your user name.
Last edited:
What am I missing?
After reading through this entire thread (40 pages!) and grabbing the latest boot.efi, I used a 2011 Macbook Air to install Yosemite to an SSD in a USB case. I successfully rebooted the Air on this external disk. Then I moved the SSD to a dock on a Hackintosh, where I moved the SSD's two copies of boot.efi to boot.efi.Apple, and replaced both with the latest boot.efi.
Trying to boot this on a MacPro1,1 failed miserably, but back in the Hack, I used "bless -info" to discern the problem: blessing is by inode, and it was still trying to use the original boot.efi (now renamed boot.efi.Apple). I fixed this with bless -folder blahblah -file boot.efi, confirmed the fix with bless -info, and put the SSD back into the 1,1.
This time, it started booting, throwing up a grey screen with an apple logo and a progress indicator... which made it not quite halfway and hung. Rebooting produced identical results.
So the I booted in verbose, and after most of a screenfull of reasonable looking stuff, it spit this out, and then hung:
Google has very little about this, but one post on insanelymac suggests this happens when it can't find the kernel. And yet, I see /System/Library/Caches/com.apple.kext.caches/Startup/kernelcache. Well, maybe it's bad? So I pulled the one out of basesystem.dmg on the Yosemite installer (which is 7MB larger), stuck that in instead, and now... slight progress. After the two bash errors (which I assume are innocuous, in which case they should have used 2&>/dev/null) the kernel clearly boots, probing the various network interfaces.
But then it hangs at "Waiting for DSMOS...". (WTF? This isn't a hackintosh!) After 15 minutes or more, it spits out "IOGraphics flags 0x43" but still no "DSMOS has arrived" and no further boot.
So, does anyone have any idea what is going on here? This is supposed to be easy. :-(
After reading through this entire thread (40 pages!) and grabbing the latest boot.efi, I used a 2011 Macbook Air to install Yosemite to an SSD in a USB case. I successfully rebooted the Air on this external disk. Then I moved the SSD to a dock on a Hackintosh, where I moved the SSD's two copies of boot.efi to boot.efi.Apple, and replaced both with the latest boot.efi.
Trying to boot this on a MacPro1,1 failed miserably, but back in the Hack, I used "bless -info" to discern the problem: blessing is by inode, and it was still trying to use the original boot.efi (now renamed boot.efi.Apple). I fixed this with bless -folder blahblah -file boot.efi, confirmed the fix with bless -info, and put the SSD back into the 1,1.
This time, it started booting, throwing up a grey screen with an apple logo and a progress indicator... which made it not quite halfway and hung. Rebooting produced identical results.
So the I booted in verbose, and after most of a screenfull of reasonable looking stuff, it spit this out, and then hung:
Code:
bash: /etc/rc.server: No such file or directory
bash: /etc/rc.installer_cleanup: No such file or directoryGoogle has very little about this, but one post on insanelymac suggests this happens when it can't find the kernel. And yet, I see /System/Library/Caches/com.apple.kext.caches/Startup/kernelcache. Well, maybe it's bad? So I pulled the one out of basesystem.dmg on the Yosemite installer (which is 7MB larger), stuck that in instead, and now... slight progress. After the two bash errors (which I assume are innocuous, in which case they should have used 2&>/dev/null) the kernel clearly boots, probing the various network interfaces.
But then it hangs at "Waiting for DSMOS...". (WTF? This isn't a hackintosh!) After 15 minutes or more, it spits out "IOGraphics flags 0x43" but still no "DSMOS has arrived" and no further boot.
So, does anyone have any idea what is going on here? This is supposed to be easy. :-(
2006/2007 Mac Pro (1,1/2,1) and OS X Yosemite
If your plan is to use this in the Mac Pro then use the Macbook Air to build an installer on an 8GB or larger thumb drive by using my guide then put the SSD in the Mac Pro and install it directly. I think you have over complicated the whole thing.
If your plan is to use this in the Mac Pro then use the Macbook Air to build an installer on an 8GB or larger thumb drive by using my guide then put the SSD in the Mac Pro and install it directly. I think you have over complicated the whole thing.
I am aware that that can be done (and in fact I built and successfully used a Mavericks installer that way) but what I actually did is much simpler, and several people have reported success in this very thread using the same technique. In short, it is:
1) Install to new disk on working Yosemite Mac
2) Replace boot.efi in two locations
3) Install disk into Macpro1,1, boot.
That is supposed to be all it takes. The other stuff I did was an attempt to fix it. I'm not hugely surprised the fixes didn't work, but the original attempt should have.
I understand what process you were using, I have done that before as well. But, it obviously didn't work for you so why not try a different method...
No, as I have said several times, my Mac Pro is a production machine, and I can't afford to reboot it very often. It's been working for weeks or months now without ever stopping. I'll shut it down tonight just to test this on Mavericks, once I update to LS 3.4.1. I'll report whatever I find.
Pike posted a comment about the KEXT of Little Snitch looking for OS X's kernel file. now if we boot Mavericks with Pike's new boot.efi it might be possible that LS is looking for the kernel in the wrong place (kernel has moved from /mach_kernel to /System/Library/Kernels/kernel).
In my case i boot Mavericks with Tiamo and Yosemite with Pike, and the LS updating issue is met in these 2 situations, i never boot Mavericks with Pike.
That's too bad. I'll try myself later today for completeness' sake.
Yes, I saw that earlier today, but I doubt that is the explanation. After all, LS 3.4.1 is supposed to be an update designed especially for Yosemite, so I'm sure its developers are well aware of the location of the new kernel.
Edit: And Pike's boot.efi, without LS 3.4.1, is supposed to successfully locate the Mavericks kernel, isn't it?
----------
Do I interpret correctly that when you install LS 3.4.1 or update to it in both operating systems, you get a continuos boot? Meaning, in the case of Mavericks, that Tiamo's boot.efi also gets confused or corrupted? Can you compare the presumably corrupted boot.efi with a non-corrupted copy and see if there are any differences between a before-LS 3.4.1- and an after-LS 3.4.1-state?
Last edited:
Really i do not know how to explain anymore ...
This issue occurs with any LS version using Mavericks and Tiamo or/and using Yosemite and Pike, and only when you update LS.