MP 1,1-5,1 Latebloom - An experimental workaround for the 11.3+ race condition

[sfalatko]

Some help please

I guess the latebloom delay is input in the boot-args section in config.plist. As this is missing from my config could somebody guide me or post their config.plist?

Thanks

eVasilis - everything is explained in the first post. My recommendation is that if you can't figure it out from that post wait on testing until it matures.

Regards.

 

[startergo]

you can do the log show in apple script with the line

note the \

display dialog (do shell script "log show --predicate  'sender == \"latebloom\"' --start $(date \"+%Y-%m-%d\") --debug")

edit: attached App

Nice.
Here is an enhancement :

display dialog (do shell script "log show --predicate  'sender == \"latebloom\" AND composedMessage CONTAINS \"Hook placed successfully\"' --start $(date \"+%Y-%m-%d\") --debug")

 

[eVasilis]

eVasilis - everything is explained in the first post. My recommendation is that if you can't figure it out from that post wait on testing until it matures.

Regards.

Thanks. My problem is that there is nowhere in my config.plist that I can input/change any boot-args. The first post is explanatory but my config does not allow for such changes.

 

[xumucane]

Thanks a lot guys for you amazing work... Live on! the Mac Pro 3,1-5,1

 

[Ausdauersportler]

Update:

Model	iMac11,3
CPU	X4370
GPU	RX480 mobile
Ethernet	Broadcom 57765-B0
Wireless	Broadcom BCM943602CDP
Bluetooth	Broadcom BCM943602CDP
Storage	1x SSD (SATA II)
Optical	1x ODD PATA
Memory	2x 8 GB DDR3 SODIMM 1333 MHz
Boot ROM	99.0.0.0.0 (latest)
latebloom	100, 200
lb_range	20, 40
lb_debug	1, 0

After upgrading to Monetary Beta 2 I experienced failed cold and warm starts - after a while I could not reboot into Monterey, again. Installing latebloom with the first parameter set fixed this immediately and completely.

The first setting was (100 +- 20, lb_debug=1). After noticing the Monterey Beta 3 bug I started to change values. Just disabling lb_debug from 1 to 0 lead immediately to failed booting again, got not a single successful one. There were two ways to recover from that:

1. enabling lb_debug=1

or

2. choosing higher values 200 +- 40 with lb_debug=0

or

3. choosing values like 120 +- 30 with lb_debug=1

I stopped the experiments directly after finding the second stable plateau within this mine field. Focussed on reboots and new cold starts: Have not had a single failure with both sets.

EDIT (07/31/2021)
3. Tried to clean internal SSD and external USB devices from pre Big Sur APFS partitions to check a connection between this and the problem. Spent a complete Sunday to finish three complete Monterey installations and 50+ boots/reboots without latebloom unless I faced a situation where I consistently ended up with a failed boot at the same very point in the boot sequence (USB bus enumeration). Had simply to enable latebloom to successfully boot, again.
So there is no way around this tool right now.

Spoiler: kernel extensions

AGPM-Overide.kext
AirportBrcmFixup.kext
AppleALC.kext
CatalinaBCM5701Ethernet.kext
CPUFriend.kext
CPUFriendDataProvider.kext
latebloom.kext
Lilu.kext
NightShiftEnabler.kext
SMC-Spoof.kext
USB-Map.kext
WhateverGreen.kext

EDIT (09/14/2021)
Updating to Big Sur 11.6 was a hard one on an iMac11,1 using 120+-30 and lb_debug=1 and verbose mode enabled. The prints of the verbose mode to the screen add other additional delays to the boot process and will interfere with the latebloom functionality.
Possibly I have to enter another search round to get it stable.

 

[eVasilis]

Different config.plist (Martin-Lo's). Is this correct?

<key>7C436110-AB2A-4BBB-A880-FE41995C9F82</key>
            <dict>
                <key>boot-args</key>
                <string>-lilubetaall -wegbeta agdpmod=pikera shikigva=80 unfairgva=1 mbasd=1 -wegtree -no_compat_check no32exec=0 latebloom=100 lb_range=20</string>
                <key>csr-active-config</key>

 

[JoeOIVOV]

Some help please

I guess the latebloom delay is input in the boot-args section in config.plist. As this is missing from my config could somebody guide me or post their config.plist?

Thanks

I'd grab a fresh config.plist from 0.7.1, It wouldn't let me attach mine, but its not much different then stock with the latebloom line.

Activate AMD hardware acceleration

I created a new "easy to install package" as per many requested (based on the official OpenCore). [The attached package updated to 1.0.0] For those who only need HWAccel, post #594 has everything you need. This new package mainly offer HWAccel + boot screen. Also, this package should work in...

forums.macrumors.com

 

[eVasilis]

Activate AMD hardware acceleration

I created a new "easy to install package" as per many requested (based on the official OpenCore). [The attached package updated to 1.0.0] For those who only need HWAccel, post #594 has everything you need. This new package mainly offer HWAccel + boot screen. Also, this package should work in...

forums.macrumors.com

Use this post, this will work without any additional configuration. Just replace your EFI folder with the latest version 0.7.1.

Thanks. I am already on 0.7.1. I just changed the config.plist to Martin's one and added the arguments as per my post above.

 

[Macschrauber]

Activate AMD hardware acceleration

I created a new "easy to install package" as per many requested (based on the official OpenCore). [The attached package updated to 1.0.0] For those who only need HWAccel, post #594 has everything you need. This new package mainly offer HWAccel + boot screen. Also, this package should work in...

forums.macrumors.com

Use this post, this will work without any additional configuration. Just replace your EFI folder with the latest version 0.7.1.

Is latebloom now part of that package?

 

[HuRR]

Is latebloom now part of that package?

I don't think so. I guess the question @eVasilis was asking is whether the latebloom boot-args were in the correct position after he switched to the OC Package? Can someone confirm?

 

[JoeOIVOV]

Is latebloom now part of that package?

Not yet, its too new, I'm sure eventually it will as it seems to work great!

Just talking about missing the NVRAM section to add the boot args. Grabbing a new config and latest version of OC Configurator should surely fix that. I wouldn't want to accidentally make a typo in that file.

 

[startergo]

This script will add Latebloom in your config plist as 3rd element under Kernel.Add (EFI must be mounted):

plutil -insert Kernel.Add.3 -xml '<dict>
    <key>Arch</key>
    <string>x86_64</string>
    <key>BundlePath</key>
    <string>latebloom.kext</string>
    <key>Comment</key>
    <string>PCI delay for BS/Monterey</string>
    <key>Enabled</key>
    <true/>
    <key>ExecutablePath</key>
    <string>Contents/MacOS/latebloom</string>
    <key>MaxKernel</key>
    <string></string>
    <key>MinKernel</key>
    <string>20.4.0</string>
    <key>PlistPath</key>
    <string>Contents/Info.plist</string>
</dict>' /Volumes/EFI/EFI/OC/config.plist

Also you can create an application in Automator. Just paste it in the shell script window and choose File/Convert to/ Application.

 

[eVasilis]

This script will add Latebloom in your config list as 3rd element under Kernel.Add (EFI must be mounted):
ate an application in Automator. Just paste it in the shell script window and choose File/Convert to/ Application.

Is the order important?

 

[startergo]

Is the order important?

For some kexts it is. For instance Lilu has to be first, VirtualSMC second etc. In general kext dependencies have to be after the kext they depend on.

 

[khronokernel]

Is the order important?

latebloom has no other dependancies in your EFI so could be the first, last or anywhere in between. An easy way to tell is by looking at the OSBundleLibraries to better see if any kexts in your EFI are needed. As you see below, all the kexts latebloom needs is within macOS. The info.plist is located under latebloom.kext/Contents/Info.plist

 

[macnrolla]

Thanks again, i just updated from 11.2.3 to 11.4 without issues, what never happened before since 11.3 . Without the kext, the installer always hangs after the first reboot, now everything is fine. I just injected the kext via OpenCore (0.7.1) without boot-args. I use the great config from the forum. Amazing, now there is even Monterey possible. ???

 

[tsialex]

I've being testing latebloom kext with my Mac Pros at home/office for some time now (most of my tests were with 0.15 and 0.16) and found that it's efficacy is directly dependent with the Mac Pro hardware configuration and the latebloom value needs to be tailored to the hardware config. More PCIe devices installed / dual processor / slower processor, more delay is needed to overcome the KPs.

While I did most of my tests with my home test Mac Pro (early-2009 with single X5680 and no PCIe switches), these are the values that I'm using for each config trying to achieve the best reliability possible. My intention is to never get a corrupt disk again and the additional boot time is irrelevant, so I'm trying to be conservative.

Year model + CPU	GPU:	PCIe AHCI/NVMe:	USB:	AirPort Extreme:	latebloom value:
early-2009 with dual X5680	eVGA GTX 680 Mac Edition	1 SM951-512	Orico FL1100	BCM94322MC	200
early-2009 with single X5680	eVGA GTX 680 Mac Edition	1 SM951-256	Orico FL1100	BCM94322MC	80
mid-2010 with single W3680	Sapphire Pulse RX 560	1 PM961-256	Caldigit with PCIe v2.0 switch	BCM94360CD	175
mid-2012 with single X5680	Sapphire Pulse RX 580	SSD7101A-1 v101 with 4 PM961-256 blades + 980 PRO-512	Caldigit with PCIe v2.0 switch	BCM94360CD	250

One thing to note, I still have very occasional KPs when cold booting with my mid-2012 with single X5680 and with my early-2009 with dual X5680 in normal usage while for the other two with less PCIe devices installed, it's almost perfect for cold boots not so much for warm boots.

If you have slower Xeons, you'll probably have to increase the latebloom value.

 

[eVasilis]

For some kexts it is. For instance Lilu has to be first, VirtualSMC second etc. In general kext dependencies have to be after the kext they depend on.

Mine is lilu, whatever green, radeon boost and all the rest after

 

[HuRR]

Thanks again, i just updated from 11.2.3 to 11.4 without issues, what never happened before since 11.3 . Without the kext, the installer always hangs after the first reboot, now everything is fine. I just injected the kext via OpenCore (0.7.1) without boot-args. I use the great config from the forum. Amazing, now there is even Monterey possible. ???

That's good too hear. I'm on Martin's package which has been great for me. I'm going to try in a few minutes. So you kept the default ms delay?

 

[startergo]

Mine is lilu, whatever green, radeon boost and all the rest after

@khronokernel answered:

latebloom has no other dependancies in your EFI so could be the first, last or anywhere in between.

 

[eVasilis]

if someone can verify that I put the boot-args in the correct place, I am go.
(as per my post above)

 

[macnrolla]

That's good too hear. I'm on Martin's package which has been great for me. I'm going to try in a few minutes. So you kept the default ms delay?

Yes. But i don't have many devices nor dual CPU. I think, testing and playing with the values is important, but the kext definitiv helps.

 

[khronokernel]

@Syncretic Have a small suggestion for latebloom, specifically the idea of defining arguments as stand alone variables in NVRAM instead of just boot-args. Idea is that the user won't clutter their boot-args as much, similar to how RtcMemoryFixup allows you to set the RTC Blacklist either with the boot-arg rtcfx_exclude=xxx or with the stand alone variable rtc-blacklist

For reference, the Lilu/OpenCore GUID is 4D1FDA02-38C7-4A6A-9CC6-4BCCA8B30102 however we use it for OCLP to set the patcher version and could be used by latebloom as well. Would recommend not using 7C436110-AB2A-4BBB-A880-FE41995C9F82 as that's Apple's Boot Variable GUID which Apple may be used to store undocumented variables potentially matching yours.

 

[sfalatko]

Question - so we have to run with SIP disabled but don't we have to boot into Big Sur recovery to disable SIP? So if we have boot problems can we do that (already testing 11.4)? Thoughts?

 

[Ausdauersportler]

Question - so we have to run with SIP disabled but don't we have to boot into Big Sur recovery to disable SIP? So if we have boot problems can we do that (already testing 11.4)? Thoughts?

Answer: There is no need to disable SIP when using OpenCore as described on post #1.

Anyway: SIP with OpenCore will be disabled within the config.plist. And installing the kext within Big Sur would surely need to boot into the Big Sur recovery. This is not a real option....