Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

macOS boot.efi developer's thread

randyoo said:
I'm not sure if I did something wrong, but the boot.efi output when I upgraded so it would compile with 2015 didn't work for me, and I didn't have time to try again yet with 2013.
Click to expand...
Look what Peter said. Debug is broken.

That, and me not having unsupported hardware, makes it almost impossible for me to fix this problem.

@splifingate,

I took care of one of the compiler issues that you reported.

I also moved the legacy installer detection, so this has to be re-verified.
 
Last edited:
Pike R. Alpha said:
Look what Peter said. Debug is broken.

That, and me not having unsupported hardware, makes it almost impossible for me to fix this problem.

@splifingate,

I took care of one of the compiler issues that you reported.

I also moved the legacy installer detection, so this has to be re-verified.
Click to expand...

thx

mheh . . . that sure is a large load of code adjustment ;)

I just updated an EL Cap install from createinstallmedia from 10.11.0 > 10.11.1, and it failed because of the .efi replacement (still have yet to fixor that).

Attached is commit 2961e70e for others' use

I am in the process of placing said into current, and 10.11.1 installations (if I can actually get all the server shares to synch).

First is re-placement of legacy .efi's...

[edit]

Boot into legacy environ successful--everything performs, as it did earlier (and, as expected).

btw, Pike: code analysis on commit 2961e70e produces no errors, warnings or messages

b-btw, successful boot into 10.11.1 El Cap produced from 10.11.0 createinstallmedia, upgraded to 10.11.1 internally, using commit 2961e70e (I have debugging, verbose, and black active).
[/edit]
 

Attachments

  • boot_commit-2961e70e.zip
    206.7 KB · Views: 189
  • image.jpeg
    image.jpeg
    3 MB · Views: 211
  • image.jpeg
    image.jpeg
    2.9 MB · Views: 202
Last edited:
Pike R. Alpha said:
Look what Peter said. Debug is broken.

That, and me not having unsupported hardware, makes it almost impossible for me to fix this problem.

@splifingate,

I took care of one of the compiler issues that you reported.

I also moved the legacy installer detection, so this has to be re-verified.
Click to expand...

Thanks, Pike & Peter. Got it working.

Commit 2961e7, boots to installer GUI using stick created with "createinstallmedia". No time to create and test on a legacy installer tonight, but maybe someone else will? Attached is commit 2961e70e75b9e43719cd4f2248fdda4edb0202bb boot.efi build.
 

Attachments

  • boot.commit 2961e70e75b9e43719cd4f2248fdda4edb0202bb.zip
    204.4 KB · Views: 155
Taking care of a compiler issue. Black and grey boot.efi.

Oops! I've just seen that splifingate and randyoo compiled the same thing while I was doing it. Oh, well. Here you have a choice of colours.

PeterHolbrook said:
To everyone interested in this project:

[P]eople must be aware that [this and other] interim future versions are NOT intended as a replacement for the official repository versions. Until further notice, those of you who want to use Pike's boot.efi ought to go to http://piker-alpha.github.io/macosxbootloader/ and download either the "black" version or the "grey" one, according to your particular preference (the change is purely cosmetic; otherwise, they are exactly the same; the choice is irrelevant as far as the operating system is concerned). Pike alone will decide when such repository versions will be updated with a newer version.

Please, notice that the [enclosed and other] upcoming experimental versions might contain bugs that could cripple your ability to boot your old Mac. So, unless you are absolutely certain of what you are doing and know how to reverse such undesirable situations, KEEP AWAY FROM THEM. In general terms, [these] versions ARE NOT FOR YOU!
Click to expand...
 

Attachments

  • boot 2961e70e75b9e43719cd4f2248fdda4edb0202bb.zip
    412.4 KB · Views: 183
PeterHolbrook said:
Taking care of a compiler issue. Black and grey boot.efi.

Oops! I've just seen that splifingate and randyoo compiled the same thing while I was doing it. Oh, well. Here you have a choice of colours.
Click to expand...

Haha! Yeah, I think splifingate must have beat me to it by less than a minute. :)

Pike, I meant to share this with you yesterday when I saw it, it made me think of you: http://imgur.com/NaDrX1e

If there's anything I can do or try to help fix the debug compile, just let me know. It's the default when you load the "solution", so that was indeed my problem. Release seems to be working perfectly.
 
randyoo said:
Haha! Yeah, I think splifingate must have beat me to it by less than a minute. :)

Pike, I meant to share this with you yesterday when I saw it, it made me think of you: http://imgur.com/NaDrX1e

If there's anything I can do or try to help fix the debug compile, just let me know. It's the default when you load the "solution", so that was indeed my problem. Release seems to be working perfectly.
Click to expand...

Hey, r . . . I've not built the Debug version of anything since I first tried to make sense of all this back when Tiamo first suggested I give VSE a spin for myself to see if things worked . . . I just compiled 'Debug: Win32', and the build was successful . . . is this what needs fixing? . . . I haven't installed it, yet . . . hmmm, let's see how that sits with Yose...

[e]r, I need to walk the dogs, first....
 
randyoo said:
Haha! Yeah, I think splifingate must have beat me to it by less than a minute. :)

Pike, I meant to share this with you yesterday when I saw it, it made me think of you: http://imgur.com/NaDrX1e

If there's anything I can do or try to help fix the debug compile, just let me know. It's the default when you load the "solution", so that was indeed my problem. Release seems to be working perfectly.
Click to expand...
Man. That is one heck of a number plate. You know. If you make long hours like me – I usually only sleep 4 hours at night – anything like that looks the same.

@splifingate,

Compiling it is not the issue, and I think that the idea was to connect to it from another computer/virtual machine.
 
Last edited:
splifingate said:
Hey, r . . . I've not built the Debug version of anything since I first tried to make sense of all this back when Tiamo first suggested I give VSE a spin for myself to see if things worked . . . I just compiled 'Debug: Win32', and the build was successful . . . is this what needs fixing? . . . I haven't installed it, yet . . . hmmm, let's see how that sits with Yose...

[e]r, I need to walk the dogs, first....
Click to expand...

I compiled 'Debug: Win32', and the build seemed to be successful, but the resulting file didn't boot the machine. When I changed to Release, the resulting boot.efi file worked. I saw Pike saying to comment certain lines, but then it seems changes were made since that suggestion, so I'm just waiting to help try to solve this issue.
 
randyoo said:
I compiled 'Debug: Win32', and the build seemed to be successful, but the resulting file didn't boot the machine. When I changed to Release, the resulting boot.efi file worked. I saw Pike saying to comment certain lines, but then it seems changes were made since that suggestion, so I'm just waiting to help try to solve this issue.
Click to expand...
If you are talking about post #654 then please forget that. Was a test only.
 
Pike R. Alpha said:
If you are talking about post #654 then please forget that. Was a test only.
Click to expand...
Yes, I was talking about that post.

Do you know which commit caused the regression? Or am I confusing the "debug" option during the build process for the verbose mode and debugging output generated during most of recent test builds?

Anyway, if it's important to you to create builds using the debug option, then just tell me if I can be of assistance. I could comment lines or blocks of code, rebuild, and test all in one go now.
 
randyoo said:
Yes, I was talking about that post.

Do you know which commit caused the regression? Or am I confusing the "debug" option during the build process for the verbose mode and debugging output generated during most of recent test builds?

Anyway, if it's important to you to create builds using the debug option, then just tell me if I can be of assistance. I could comment lines or blocks of code, rebuild, and test all in one go now.
Click to expand...
The verbose debug data you may have seen was all done from a regular release version of boot.efi Not one was made with the debug option.

For the Visual Studio compiler debug means "optimisation off" (read slower) whereas release versions are made with speed in mind. To speed up the boot process so to speak.
 
  • Like
Reactions: randyoo
Pike R. Alpha said:
Man. That is one heck of a number plate. You know. If you make long hours like me – I usually only sleep 4 hours at night – anything like that looks the same.

@splifingate,

Compiling it is not the issue, and you think that the idea was to connect to it from another computer/virtual machine.
Click to expand...

Aye, it all fits, now. As tiamo said in

https://forums.macrumors.com/threads/mac-pro-2-1-and-os-x-mavericks.1598176/page-3#post-18433273

"...you CAN DEBUG this bootloader over COM/1394 with WinDbg! yes SOURCE level debugging(this is another story:D)"

I'll let that field lay fallow, if you know what I mean ;)
 
o.k., I created new partitions for the 10.11.1 createinstallmedia, and 10.11.1 legacy (as seen in the first attachment).

In each of every partition, I only replaced the boot.efi's with commit 2961e70e

Boot into: 10.11.0 createinstallmedia, 10.11.0 legacy, and 10.11.1 legacy, drives me straight into the installer [Version 1.0 (733)] as seen in the second attachment. "System Information..." is present in menu > Utilities

Boot into 10.11.1 createinstallmedia (named '10.11.1-cim', though it is labeled as "Install OS X El Capitan" on Option-boot (see fifth attachment)) takes me to 'OS X Utilities' (Restore, Get Help, Install, Disk Utility) as seen in the third attachment. 'System Information...' is missing...

Selecting "Install OS X" from that, takes me to the installer, as seen in the fourth attachment.

Odd thing is that this installer is newly versioned [Version 1.7.33 (1133)]!

I have successfully re-booted into both the 10.11.0 and 10.11.1 createinstallmedia environments (using commit 2961e70e) multiple times with no corruption to either 'Install OS X El Capitan.app'.

That being said, I must admit to not having performed an installation from either.

Successful boot into working (daily-driver) Yosemite installation using commit 2961e70e is resounding, as is the same for El Capitan (which I updated internally to 10.11.1).

Just a few things for all'y'all to chew-on in the interim...this 'ol body needs some rest....
 

Attachments

  • image.jpeg
    image.jpeg
    3 MB · Views: 178
  • image.jpeg
    image.jpeg
    3 MB · Views: 192
  • image.jpeg
    image.jpeg
    1.3 MB · Views: 163
  • image.jpeg
    image.jpeg
    2.9 MB · Views: 148
  • image.jpeg
    image.jpeg
    2.4 MB · Views: 159
I was wondering, Pike, if build 2961e70e75b9e43719cd4f2248fdda4edb0202bb is pre-version 3.1, version 3.1 itself or post-version 3.1. Can you clarify the current situation?
 
  • Like
Reactions: PeterHolbrook
Add Black and Grey Release configurations to Solution.

Pike, unless you indicate otherwise, in future I'll only compile the "black release", except when we are about to actually publish a new "official" version (such as v3.2, v3.5 or whatever).

Edit: I hadn't seen your previous post. Sorry. How is that thing compiled now? I don't suppose it's automatically compiled whenever we go to that web page, is it? I suppose you do the actual compilation.

PeterHolbrook said:
To everyone interested in this project:

[P]eople must be aware that [this and other] interim future versions are NOT intended as a replacement for the official repository versions. Until further notice, those of you who want to use Pike's boot.efi ought to go to http://piker-alpha.github.io/macosxbootloader/ and download either the "black" version or the "grey" one, according to your particular preference (the change is purely cosmetic; otherwise, they are exactly the same; the choice is irrelevant as far as the operating system is concerned). Pike alone will decide when such repository versions will be updated with a newer version.

Please, notice that the [enclosed and other] upcoming experimental versions might contain bugs that could cripple your ability to boot your old Mac. So, unless you are absolutely certain of what you are doing and know how to reverse such undesirable situations, KEEP AWAY FROM THEM. In general terms, [these] versions ARE NOT FOR YOU!
Click to expand...
 

Attachments

  • boot d9d770c57e79dfc733ef083481a94a9db6f3827b.zip
    412.5 KB · Views: 179
Last edited:
PeterHolbrook said:
Edit: Sorry. How is that thing compiled now? I don't suppose it's automatically compiled whenever we go to that web page, is it? I suppose you do the actual compilation.
Click to expand...

Actually, yeah, it's automatically compiled! If you look at the "console" log, you can see it pulling the git repo, then compiling on the command line. Way better than what you (and I) had been doing manually!
 
  • Like
Reactions: PeterHolbrook
randyoo said:
Actually, yeah, it's automatically compiled! If you look at the "console" log, you can see it pulling the git repo, then compiling on the command line. Way better than what you (and I) had been doing manually!
Click to expand...
OK. If so, I suppose I won't be compiling anymore. I'll just come here to see if I can contribute some comment and see your reports on how the new builds behave on old hardware. Let me know if I can be of further assistance.
 
@randyoo

Can you confirm that the generated boot.efi's work on actual hardware? I wasn't able to test them myself when I set things up on AppVeyor.

@PeterHolbrook

I had no intention of stealing your thunder regarding the builds, but when I realized you guys were having to do this via a VM within OS X, I figured there had to be a better way.
 
PeterHolbrook said:
OK. If so, I suppose I won't be compiling anymore. I'll just come here to see if I can contribute some comment and see your reports on how the new builds behave on old hardware. Let me know if I can be of further assistance.
Click to expand...
clawfinger said:
@randyoo

Can you confirm that the generated boot.efi's work on actual hardware? I wasn't able to test them myself when I set things up on AppVeyor.
Click to expand...

Yep, it seems to work fine. I was able to boot into an installer, as well as an installed copy of 10.11.0, and didn't notice any issues.
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back