2006/2007 Mac Pro (1,1/2,1) and macOS Sierra

[hwojtek]

Yupp, I believe a single 2,1 would do, as a 2,1 reflash should be considered a minimum requirement for tinkering with the Mac Pro.
As said, I can help with the arrangements if need be.

 

[dougp59]

Yupp, I believe a single 2,1 would do, as a 2,1 reflash should be considered a minimum requirement for tinkering with the Mac Pro.
As said, I can help with the arrangements if need be.

Can someone get a hold of Piker Alpha on this question. Reason is, there are people in the forum running 1,1 who can't afford to upgrade or can't do the upgrade. If Piker Alpha can do his work with just a 2,1 fine, we can roll with that and make sure it's a sweet rig.

ps. Thanks to nathan for his donation! First one (besides me).
[doublepost=1474576586][/doublepost]An updated Go Fund Me URL for the Piker Alpha campaign.
https://www.gofundme.com/pamacpro
https://www.gofundme.com/pamacpro

 

[F1Mac]

Can someone get a hold of Piker Alpha on this question. Reason is, there are people in the forum running 1,1 who can't afford to upgrade or can't do the upgrade. If Piker Alpha can do his work with just a 2,1 fine, we can roll with that and make sure it's a sweet rig.

Just send him a pm and tell him about your idea, you'll get a straight answer from the boss himself

Also let's not get our hopes up just yet. We have to be realistic there is unfortunately a good chance that Sierra will never run on these machines. But the world will not end so there's that

 

[dougp59]

Just send him a pm and tell him about your idea, you'll get a straight answer from the boss himself

Also let's not get our hopes up just yet. We have to be realistic there is unfortunately a good chance that Sierra will never run on these machines. But the world will not end so there's that

Where do I IM him?

 

[TT01]

Pike is extremely knowledgeable and if he does not see a possibility I do not think he will work on Sierra. If he is working on Sierra adoption on Mac Pro 1,1 and 2,1, I think, it means there is a possibility.

Concerning buying Pike a Mac Pro 2,1 sounds like a good idea and I thought about to suggest it last year when he mentioned he cannot help some of us as he does not have a Mac Pro. If he wants one, I will be very happy to support you.

However, I gave up on my idea last year, when I saw how his work on El Capitan was progressing. Therefore, I honestly do not think Pike will want it. It is his decision of course. I think this is because, he seems to be very busy and has many interests. He, most likely, has no time to play with a Mac Pro. I believe, what he needs is support from our community to compile, test, give him the results and help him out. This is more like how the projects are developed among groups in my time. One person does not do everything. The work is shared among the group. We also have wonderful people who have excellent knowledge and really help him and make great contributions. He has acknowledged their contributions. In my opinion, I may be wrong, of course, that Pike will want to spend his available time on working on the boot.efi and may be other components that may be needed to make Sierra work. I have a good feeling that he will be successful.

I would like to add here that, how much I appreciate his and his teams great work whenever I boot up and use my Mac Pro 2,1 running El Capitan 10.11.6 powered by his boot.efi 3.1, and, also, how the working recovery disk running his boot.efi has saved my kernel panic stricken Mac Pro when I installed a new not fully prime time software.

After the kernel panic problem solved, I actually wanted to send him a donation again but he says "I do not accept donation at this time" at his github development site. I hope he opens it up soon.

 

[flyinmac]

Pike is extremely knowledgeable and if he does not see a possibility I do not think he will work on Sierra. If he is working on Sierra adoption on Mac Pro 1,1 and 2,1, I think, it means there is a possibility.

Concerning buying Pike a Mac Pro 2,1 sounds like a good idea and I thought about to suggest it last year when he mentioned he cannot help some of us as he does not have a Mac Pro. If he wants one, I will be very happy to support you.

However, I gave up on my idea last year, when I saw how his work on El Capitan was progressing. Therefore, I honestly do not think Pike will want it. It is his decision of course. I think this is because, he seems to be very busy and has many interests. He, most likely, has no time to play with a Mac Pro. I believe, what he needs is support from our community to compile, test, give him the results and help him out. This is more like how the projects are developed among groups in my time. One person does not do everything. The work is shared among the group. We also have wonderful people who have excellent knowledge and really help him and make great contributions. He has acknowledged their contributions. In my opinion, I may be wrong, of course, that Pike will want to spend his available time on working on the boot.efi and may be other components that may be needed to make Sierra work. I have a good feeling that he will be successful.

I would like to add here that, how much I appreciate his and his teams great work whenever I boot up and use my Mac Pro 2,1 running El Capitan 10.11.6 powered by his boot.efi 3.1, and, also, how the working recovery disk running his boot.efi has saved my kernel panic stricken Mac Pro when I installed a new not fully prime time software.

After the kernel panic problem solved, I actually wanted to send him a donation again but he says "I do not accept donation at this time" at his github development site. I hope he opens it up soon.

I would agree with your perspective regarding community involvement.

I've worked with developers before where they wrote software without access to even the same computer platform. They'd write it up, send me the code. I'd compile it and test it on the hardware and give them feedback on what issues were found.

A good programmer often knows the machine well enough to code without having to have the real hardware. And sometimes they prefer to stay on the code side, and not test it themselves.

Another benefit to having someone else do the testing is that bugs are found faster. For example, back when I used to do some programming, I remember one of the funniest tests.

I had written and tested this program that I had spent a week on. Something simple. I'd spent hours going over the code, worked out all the bugs, and thought it was time for someone else to test it.

So I sat my brother at the keyboard and let him try it. About 3 key presses in, he had crashed it.

Now, the program did do everything the way it was supposed to. Because I knew how to use the program. I wrote it.

Why did it break? He pressed a wrong key, and the program didn't like that key.

Thing is, when we test our own work, we use the program as intended. Someone else might do something different than we anticipated, and bring it crashing down.

That is the benefit of having someone else do all the testing. We tell them what it does, and let them stumble through it, to see what breaks. They tell us, and we say wow, hadn't even thought someone might do it that way. Let's make sure it can't do that again.

 

[randyoo]

For what it's worth, I'm currently in Europe, and I've got a 2006 Mac Pro (1,1 flashed to 2,1) that I'd be willing to contribute to Piker, assuming he's actually interested in having it. It's been upgraded to 8-core, 8800GT graphics (boot screens work), Airport (wifi), 16GB RAM, a small SSD (60GB?), and a 3TB HDD, currently running 10.11.6 perfectly.

Thing is, I don't really need it any more, and I'd prefer it went to a good cause, rather than try to sell it for max profit--I'd be willing to let it go for just $250... So the GoFundMe goal could probably be substantially reduced, unless you really wanted to have a big bundle of cash to include, as well.

 

[flyinmac]

For what it's worth, I'm currently in Europe, and I've got a 2006 Mac Pro (1,1 flashed to 2,1) that I'd be willing to contribute to Piker, assuming he's actually interested in having it. It's been upgraded to 8-core, 8800GT graphics (boot screens work), Airport (wifi), 16GB RAM, a small SSD (60GB?), and a 3TB HDD, currently running 10.11.6 perfectly.

Thing is, I don't really need it any more, and I'd prefer it went to a good cause, rather than try to sell it for max profit--I'd be willing to let it go for just $250... So the GoFundMe goal could probably be substantially reduced, unless you really wanted to have a big bundle of cash to include, as well.

Sounds like a fair and generous offer. And, I'm sure it would suit the purpose quite well. So, all that remains is for someone to contact Piker to find out if he's even interested in having such a machine, or if he prefers to continue using the method he's been using.

 

[jt_69.V]

What about this solution + boot.efi patch?
http://dosdude1.com/sierrapatch.html

I'll try this method on my mp1,1 on this weekend just for fun

I tried it this morning on a 1.1 (patched to 2.1) Mac Pro, but it didn't boot.

I installed the patched macOS Sierra system on an external USB drive (SSD) and checked that it was booting on a 2008 iMac (which means that the patch worked).

Then I replaced both boot.efi files with Piker's ones. The Mac seems to start booting ok, displaying the Apple logo, and the led of the USB drive blinks for a while (maybe 1 min). But then the led turns off, the Mac Pro keeps displaying the Apple logo, and nothing more happens

 

[dougp59]

Sounds like a fair and generous offer. And, I'm sure it would suit the purpose quite well. So, all that remains is for someone to contact Piker to find out if he's even interested in having such a machine, or if he prefers to continue using the method he's been using.

I have IM'd Piker Alpha and am awaiting a reply.
[doublepost=1474637474][/doublepost]

For what it's worth, I'm currently in Europe, and I've got a 2006 Mac Pro (1,1 flashed to 2,1) that I'd be willing to contribute to Piker, assuming he's actually interested in having it. It's been upgraded to 8-core, 8800GT graphics (boot screens work), Airport (wifi), 16GB RAM, a small SSD (60GB?), and a 3TB HDD, currently running 10.11.6 perfectly.

Thing is, I don't really need it any more, and I'd prefer it went to a good cause, rather than try to sell it for max profit--I'd be willing to let it go for just $250... So the GoFundMe goal could probably be substantially reduced, unless you really wanted to have a big bundle of cash to include, as well.

That's awesome Randy. Shipping to Spain from your country would be how much? I believe the beast weighs 50lbs.

 

[flyinmac]

I tried it this morning on a 1.1 (patched to 2.1) Mac Pro, but it didn't boot.

I installed the patched macOS Sierra system on an external USB drive (SSD) and checked that it was booting on a 2008 iMac (which means that the patch worked).

Then I replaced both boot.efi files with Piker's ones. The Mac seems to start booting ok, displaying the Apple logo, and the led of the USB drive blinks for a while (maybe 1 min). But then the led turns off, the Mac Pro keeps displaying the Apple logo, and nothing more happens

Yep, gotta find a way around the SSE 4 code. My thoughts are either replacement library files (if it's practical), or a translation layer.

Somewhere in the history of this thread, someone mentioned that they found the SSE 4 calls in library files. If that's where it's at, it might be possible to produce modified library files for each one that has SSE calls in them.

Most software these days calls on system software, or libraries to perform tasks. The days of commercial software calling directly on the hardware has mostly passed.

So replacing the library files and system files that have SSE 4 calls might be an option.

The question is how many files is that? Each would have to be modified and equivalent code injected to replace the code that is removed.

The downside to this approach, is the fact that it would have to be repeated every time Apple issued an update.

It is doable though. Think about how in the hobbyist markets they port an operating system to a whole different CPU and platform. It can be done. And given that all we'd be replacing is SSE 4 calls, it would be simpler than a complete platform change.

But it really comes down to time, effort, and the level of interest someone has in doing all the work.

The fact that Sierra will likely see 5 to 6 major updates, plus security patches, and minor updates, means that this approach would not be a one time effort.
[doublepost=1474662431][/doublepost]

I have IM'd Piker Alpha and am awaiting a reply.
[doublepost=1474637474][/doublepost]

That's awesome Randy. Shipping to Spain from your country would be how much? I believe the beast weighs 50lbs.

Hopefully we'll hear some good news from Piker.

 

[F1Mac]

Yep, gotta find a way around the SSE 4 code. My thoughts are either replacement library files (if it's practical), or a translation layer.

Somewhere in the history of this thread, someone mentioned that they found the SSE 4 calls in library files. If that's where it's at, it might be possible to produce modified library files for each one that has SSE calls in them.

Most software these days calls on system software, or libraries to perform tasks. The days of commercial software calling directly on the hardware has mostly passed.

So replacing the library files and system files that have SSE 4 calls might be an option.

The question is how many files is that? Each would have to be modified and equivalent code injected to replace the code that is removed.

The downside to this approach, is the fact that it would have to be repeated every time Apple issued an update.

It is doable though. Think about how in the hobbyist markets they port an operating system to a whole different CPU and platform. It can be done. And given that all we'd be replacing is SSE 4 calls, it would be simpler than a complete platform change.

But it really comes down to time, effort, and the level of interest someone has in doing all the work.

The fact that Sierra will likely see 5 to 6 major updates, plus security patches, and minor updates, means that this approach would not be a one time effort.

Geez that sounds like an awful amount of time to waste on this. Remember, in 12 months the next macOS is being released...and a whole new set of troubles for older hardware.

Based on my recent experience with Sierra, unless someone desperately needs to talk to Siri or pasting stuff from an iPhone because they're too lazy to do the whole thing directly on the Mac, I'd say trying to modify the OS like this is big waste of time and effort - but that's just me of course. I mean it's way more than patching an installer and one file (boot.efi), which took a good month if not more last year before Pike got it working. One month. One file. How many library files are we talking about in Sierra? Anyway, I guess I'm just thinking out loud here so I'll go back to my cave now

 

[randyoo]

I have IM'd Piker Alpha and am awaiting a reply.
[doublepost=1474637474][/doublepost]

That's awesome Randy. Shipping to Spain from your country would be how much? I believe the beast weighs 50lbs.

Yeah, it's a beast, indeed! A quick search indicates it'll come in around $50 for shipping. If Piker responds with interest, then I'll see if I can find a better offer.

 

[blindpcguy]

if this goes through and the donations havnt ended by next friday il definatly pitch in if he acepts his is a good cause to keep our old beasts up and running on latest os

 

[Ultracyclist]

Geez that sounds like an awful amount of time to waste on this. Remember, in 12 months the next macOS is being released...and a whole new set of troubles for older hardware.

Based on my recent experience with Sierra, unless someone desperately needs to talk to Siri or pasting stuff from an iPhone because they're too lazy to do the whole thing directly on the Mac, I'd say trying to modify the OS like this is big waste of time and effort - but that's just me of course. I mean it's way more than patching an installer and one file (boot.efi), which took a good month if not more last year before Pike got it working. One month. One file. How many library files are we talking about in Sierra? Anyway, I guess I'm just thinking out loud here so I'll go back to my cave now

We don't need 10.12 if we can just make it so that we can run new updates the require 10.12 such as the new iWorks update that only runs on 10.12. Or if Final cut updates to a version that requires 10.12.

Maybe this wil be much easier to fix.

To be honest, I would really love to se Sierra on my cMP but if you ask me if I NEED it the clear answer is NO, I just want to be able to update iWorks and al my other apps if it comes to that.

 

[flyinmac]

We don't need 10.12 if we can just make it so that we can run new updates the require 10.12 such as the new iWorks update that only runs on 10.12. Or if Final cut updates to a version that requires 10.12.

Maybe this wil be much easier to fix.

To be honest, I would really love to se Sierra on my cMP but if you ask me if I NEED it the clear answer is NO, I just want to be able to update iWorks and al my other apps if it comes to that.

Those could be as simple as editing the "allowed" list in each app. For example, deleting certain lines in the boot camp installer will permit the newer boot camp drivers to install in Windows 10 on Machines that Apple says are not permitted.

 

[Ultracyclist]

Those could be as simple as editing the "allowed" list in each app. For example, deleting certain lines in the boot camp installer will permit the newer boot camp drivers to install in Windows 10 on Machines that Apple says are not permitted.

Indeed that used to work. But not with all applications anymore.

 

[dougp59]

Hi all. Piker Alpha responded to my IM. He was greatly flattered at this effort but has politely and graciously declined. He said, if anything, accepting the gift would only add pressure on him to deliver a solution and he's not sure if and when he would have time to do that.

I will close down the Go Fund Me page today and refund the donations. Thanks for the support and we'll see what happens in this thread 30 days after Sierra general release.

 

[hrutkaymods]

Hi all. Piker Alpha responded to my IM. He was greatly flattered at this effort but has politely and graciously declined. He said, if anything, accepting the gift would only add pressure on him to deliver a solution and he's not sure if and when he would have time to do that.

I will close down the Go Fund Me page today and refund the donations. Thanks for the support and we'll see what happens in this thread 30 days after Sierra general release.

Yeah right now I'm playing me just wait game. Hopefully we'll see something by November

 

[owbp]

So far I've tried Clover, Enoch and VirtualBox.
All above give the same results as current boot.efi - nothing...

 

[hwojtek]

As of now, there is no piece of software that will make Sierra run on a 1,1 and 2,1 Mac Pro, be it boot.efi, Clover, Enoch, any virtualization software (it talks directly to the processor after all) - none.
There are checks for SSE 4,1 instructions and our processors do not support it. There is no software to circumvent this. Apart for the 32-bit EFI we need to bypass, this is the new minimum supported instruction set and we have to face it. As you have already seen, the boot.efi kindly provided by Pike R Alpha, works - the system starts up and begins to load the libraries. It's at the later stage of bootup when MacOS checks for supported processor instructions and doesn't find any.

Possible solutions as I see them (do not read "viable"):

1. a systemwide SSE4,1 emulation layer (Pros: transparent. Cons: will slow the computer to a halt if MacOS does SSE 4,1 excessively, and I can imagine it does with Siri)
2. patching each and every library using SSE4,1 instructions with pre-SSE4,1 code (Pros: a bit faster in executing the code than option 1. Cons: you want to rewrite the whole OS using older and less than optimal code)
3. holding our breath until it comes up that the Sierra crash people are experiencing is the single use of SSE4,1 in the whole system and can be patched with a couple of NOOPs instead of the original code (Pros: would make the 1,1/2,1 work as intended. Cons: extremely unlikely that Apple only made use of these instructions in order to keep us off Sierra).
4. creating a firmware flash for 1,1/2,1 with microcode that would allow the old computer to run next generation of Xeons (LGA 771-socketed Harpertown Xeons based on Penryn microarchitecture), which support SSE4,1. (Pros: you get a 3,1 Mac for 1,1 price. Cons: even more extremely unlikely, as 1: the Penryn microcode from an Intel BIOS would need to be decrypted first, which - given the current state of cryptography - would happen somewhen in 26th century, 2: somebody with a thorough knowledge of Intel trade secrets would need to put together a firmware for a Mac Pro).

Given all the above: options 1 and 2 will be effectively slowing down the computer VERY MUCH. Option 3 would be a blessing, and we need to wait until the whole bootup process and open source files provided by Apple are examined. I am afraid, though, that SSE4,1 code has been found in Sierra in numerous places. Option 4 would be perfect - however it just won't happen. It won't.

So unless it's option 3, we're in about same situation when we were left with our dual G5's, only a couple of years later.

 

[Tower-Union]

Thank you for the excellent insight Wojtek! Looks like it is likely time to face the music and start saving for a hackintosh.

 

[uezi]

It's not only us, who suffer from lack of SSE4.1 support. SSE4.1 are intel specific instructions while in the "wild" a lot of people have AMDs as well. So from time to time, a companies release games, where they forget about this (e.g. "No Man's Sky" or "Metal Gear Solid V: Phantom Pain").
In an interesting link I found ( http://gearnuke.com/no-mans-sky-pc-fix-crashes-sse4-1-support/ ) there's a workaround how to start such a game with SSE4.1 emulation (links to intel developer site to download a cpu emulator for Windows).

Now as a PoC, it should be possible to start a MacPro 1,1/2,1 in Windows, execute VirtualBox with intel's CPU emulator an try to boot Sierra... Sure, not performing well but at least to check if it is even feasible and what the impact to performance could be.

 

[Draeconis]

It's not only us, who suffer from lack of SSE4.1 support. SSE4.1 are intel specific instructions while in the "wild" a lot of people have AMDs as well. So from time to time, a companies release games, where they forget about this (e.g. "No Man's Sky" or "Metal Gear Solid V: Phantom Pain").
In an interesting link I found ( http://gearnuke.com/no-mans-sky-pc-fix-crashes-sse4-1-support/ ) there's a workaround how to start such a game with SSE4.1 emulation (links to intel developer site to download a cpu emulator for Windows).

Now as a PoC, it should be possible to start a MacPro 1,1/2,1 in Windows, execute VirtualBox with intel's CPU emulator an try to boot Sierra... Sure, not performing well but at least to check if it is even feasible and what the impact to performance could be.

Don't get me wrong, I enjoy the prospect of someone finding out a way to get this to work, but for many reasons, including those echoed here, I'm not sure this will ever work well.

You've also got to stack the efforts required against the reward. If you read the review Ars Technica did on Sierra, you can see that apart from APFS, there are a few minor security enhancements, and that's about it. I doubt anyone running a 1,1 or a 2,1 is all that concerned with things like Continuity or Siri (please correct me if I'm wrong).

So, APFS is looks great, or at the very least it's good to see the back of HFS+, and I guess some would consider that alone enough of a reason to move, but if you're getting down to the level of ripping OS X apart completely, it's going a bit far, isn't it?

El Cap is the current OS you can run on these machines, which history would suggest would mean you'll still get another 2 years of security upgrades. After that, on average those who bought at launch would have had their machines 11-12 years! It's safe to say barring APFS there's not much more Apple could do to improve performance on your machine. Adding emulators is only going to add overhead and slow things down, which isn't what people want to deal with.

As I say, don't want to dampen anyone's spirits or dissuade anyone from investigating this; when I heard about the Tiamo thunk between EFI64 and EFI32, that was pretty magical. Pushing hardware to its limits and getting things to do things they were never designed for is fascinating to watch. But in this case, I fear the solution would only make these machines slower.

Happy to be proved wrong though

 

[VAGDesign]

To tell you the truth, I prefer Yosemite to El Capitan on my Mac Pro. The extra security they added and locked out modifications by the user, makes me think, Sierra will have more added.

So I don't really need Sierra. I can have it on my MBP and I didn't installed it yet, maybe never will. I see it like Mountain Lion version. Nothing really new added to force me use it. Siri to me is deactivated even on iOS devices, don't need it.

Mac Pro 1,1 is probably the only Mac that can work with OSX 10.4 to 10.11.
EIGHT generations of OSX! Think about it...

I will welcome any try but don't need this time.
Cheers and keep the hope alive even if the need is not there