Clover General discussion

[jsl2000]

On 7/26/2020 at 6:16 PM, Matgen84 said:

 

@jsl2000 You're using mod Clover r5119  to boot BS Beta 2 or Beta 3. With official Clover, I don't believe that the size of OcQuirks matters: in both cases, impossible to boot from USB Installer or external Hard disk, or internal hard disk on my Z390 system. At this time, of course.

Please check my EFI for Clover for your reference which worked for Big Sur beta3 and Catalina in AMD FX-6300 hackintosh, but not worked for USB installer yet.

EFI.zip

[Matgen84]

22 hours ago, Slice said:

Delete your debug.log from misc folder. Else it grown unrestricted.

And I see no debug messages in your log. 

See kernel_patcher.cpp line 26.

There should be 

#define KERNEL_DEBUG 1

if you want to help me to see your problems.

 

Thanks @Slice kernel_patcher.cpp line 26 is already #define KERNEL_DEBUG 1 in my local repo. 

 

I upload a "clean" debug.log file r5120 commit 0724be9. If you can help me to boot Big Sur  (Z390 system)

 

EDIT: I test r5119 modified by kuckkuck and his KernelToPatch in config.plist, I can boot Big Sur Beta 3 on my Z390 Aorus Master system from USB disk

 

debug.log

Edited July 28, 2020 by Matgen84

[Matgen84]

2 minutes ago, jsl2000 said:

Please check my EFI for Clover for your reference which worked for Big Sur beta3 and Catalina in AMD FX-6300 hackintosh, but not worked for USB installer yet.

 

Thanks. 

 

I'm using Coffee Lake system not AMD. I carry out my tests from a USB key, BS is installed on an external drive.

[Matgen84]

16 minutes ago, mifjpn said:

Hello @Slice

I can boot Big Sur Develpers Beta 3 by  0724bed9-bulid.

BigSurDevBeta3 boot confirmation info on Clover (latest build)

In the early days of that Build, I think I saw Kext's allocation error.
Maybe there is a problem with gcc's library, as Badruzeus thinks.
Good luck

 

Hello @mifjpn

 

Can you boot Big Sur Develpers Beta 3 by  0724bed9-bulid from USB disk or external disk !  I can't on my Z390 system. 

[Matgen84]

19 hours ago, mifjpn said:

Thank you for rapid reply.

I boot Big Sur on internal SSD(with same EFI).

I noticed...
I wonder if everyone is up to date with Kext's Bullid.
Please use "Kexts Update.command.zip" made by MaLd0n on the "Olarila Big Sur Vanilla Image / USB Installation" web page.
That command will now bulid almost all Kexts in the Lilu4All/Release folder on your desktop.
I think the latest build of Kext is essential for Big Sur.

 

 

Thanks @mifjpn I already use "Kexts Update.command.zip" made by MaLd0n. It's a great tool.

 

Please, complete your signature with your config. It's helpful for users.

[pkdesign]

17 hours ago, Matgen84 said:

 

You're right. The new way for Clover is to include Quirks parameters in config.plist. Do you update Clover 5120 to latest commit ? 

 

Try to make two changes (Dortania Guide for Coffee Lake):

 

SetupVirtualMap to YES

EnableWriteUnprotector to FALSE

 

Honestly, I'm kind of over trying. It works as is. I have updated to 5120 but I'll just leave Quirks where they are. There is no need to change at this point.

[jmacie]

Do you upgrade your clover to 5120, then sub OcQuirks.efi and OpenRuntime.efi for aptio in Drivers folder. Then you add the quirks for your cpu from Dortania into your config .plist. Is that it, so far? Do you need vsmcgen=1 in boot arg? I know they said with the newest Lilu commit that you don't need that boot arg. I need clover, as I have spent too much time failing to make an opencore loader for my x99 , and am reticent to get blasted asking opencore questions on their page. Thanks all, johnm

Edited July 29, 2020 by jmacie

[fusion71au]

On 7/25/2020 at 4:02 PM, crazybirdy said:

 

Clover r5119 mod load old /System/Library/PrelinkedKernels/prelinkedkernel

But Clover 5120 load new /System/Library/KernelCollections/BootKernelExtensions.kc

 

Maybe BootKernelExtensions.kc is harder to hack.

 

On 7/28/2020 at 12:57 AM, Matgen84 said:

I can't boot r5120 commit 0724be9 on Big Sur (works fine for Catalina)
Can't boot on Big Sur Installer and Installed Big Sur: stuck this time at apfs_module:2411 

• Before, I see "couldn't load" kexts and "couldn't alloc" messages

  Reveal hidden contents

 

debug.log

 

On my legacy desktop (GA-P5aUD3, system 2 in signature), I had similar issues for Clover_r5120_0724be9 in default "Kernel Collections mode" ie when trying to boot Big_Sur Beta3 Installer through BootKernelExtensions.kc pathway ---> failed kext injection and kext patching....

Spoiler

 

The good news is that Clover_r5120_0724be9 works perfectly for booting my older macOS High Sierra 10.13.6 and Catalina 10.15.6.

 

Also can boot successfully into already installed Big_Sur Beta3 on external USB HD provided we boot through the old "prelinkedkernel pathway" (by setting the following variables into NVRAM in High Sierra terminal)...

 

sudo nvram booter-fileset-kernel="%00"
sudo nvram booter-fileset-basesystem="%00"

Spoiler

bootlog_Clover_r5120_BigSur_Beta3_FullInstall.txt

 

 

However with the above nvram booter-fileset changes, this prevents booting into the Big_Sur Installer via KernelCollections ---> kernel panic...

Spoiler

 

[Matgen84]

48 minutes ago, fusion71au said:

 

 

On my legacy desktop (GA-P5aUD3, system 2 in signature), I had similar issues for Clover_r5120_0724be9 in default "Kernel Collections mode" ie when trying to boot Big_Sur Beta3 Installer through BootKernelExtensions.kc pathway ---> failed kext injection and kext patching....

  Reveal hidden contents

 

The good news is that Clover_r5120_0724be9 works perfectly for booting my older macOS High Sierra 10.13.6 and Catalina 10.15.6.

 

Also can boot successfully into already installed Big_Sur Beta3 on external USB HD provided we boot through the old "prelinkedkernel pathway" (by setting the following variables into NVRAM in High Sierra terminal)...

 

sudo nvram booter-fileset-kernel="%00"
sudo nvram booter-fileset-basesystem="%00"

  Reveal hidden contents

bootlog_Clover_r5120_BigSur_Beta3_FullInstall.txt

 

 

However with the above nvram booter-fileset changes, this prevents booting into the Big_Sur Installer via KernelCollections ---> kernel panic...

  Reveal hidden contents

 

 

Thanks for these explanations  It can help me to boot Installed BS Beta 3. A temporary solution while waiting for @Slice to continue its work for BS.

Yesterday in one of my posts, I wrote: I test r5119 modified by kuckkuck and his KernelToPatch in config.plist, I can boot Big Sur Beta 3 on my Z390 Aorus Master system from USB disk and external drive. Maybe your ideas and those of kuckkuck will allow @Slice to continue the improvements of Clover for BS. I hope.

Clover r5120 commit 0724be9 works too for booting Catalina on my Z390 system.

 

Edited July 29, 2020 by Matgen84

[Common_Sense]

It would be very good with a NVRAM section in CLOVER where variables could be defined. In case NVRAM gets cleared or if it is a new installation where variables are not yet defined. That would help with defining:

sudo nvram booter-fileset-kernel="%00"
sudo nvram booter-fileset-basesystem="%00"

SystemParameters section in config.plist would be very good for defining NVRAM variables.

 

Just a thought.

[MICKHAEL]

18 minutes ago, Common_Sense said:

It would be very good with a NVRAM section in CLOVER where variables could be defined. In case NVRAM gets cleared or if it is a new installation where variables are not yet defined. That would help with defining:

sudo nvram booter-fileset-kernel="%00"
sudo nvram booter-fileset-basesystem="%00"

SystemParameters section in config.plist would be very good for defining NVRAM variables.

 

Just a thought.

Yes. But this will not be already OC?

[Common_Sense]

4 minutes ago, MICKHAEL said:

Yes. But this will not be already OC?

Correct, OC already has the possibility to add NVRAM variables in config.plist. I have OC working as well on a USB stick, but I prefer Clover since I cannot get Windows to boot with OC. Also I find Clover to be "just what I need, no more no less", faster quicker and easier to use and setup.

[Common_Sense]

18 hours ago, mifjpn said:

Hello everyone.
I tried booting BigSur Developers Beta 4 with Clober.
Even if I define the following NV-RAM variables
sudo nvram booter-fileset-kernel="%00"
sudo nvram booter-fileset-basesystem="%00"

I see "couldn't load" kexts and "couldn't alloc" messages.
And can't boot on Big Sur stuck this time at apfs_module:2411

 

OC works but you have to remove booter-fileset-kernel, booter-fileset-basesystem and make a NVRAM reset. Also make sure you have the latest version of OC.

 

I had the exact same problem as you before I updated to the latest version of OC and removed those 2 variables. OC was updated yesterday, don't know if the latest version is required or not, have not tested with older versions.

 

Clover does not seem to work on beta 4, I think it might need similar changes as was made to OC in yesterdays update.

[Matgen84]

Hi @Slice

Test r5120 commit 76c0e3fe: I can't on my Z390 Aorus Master

 

• ACPI error RP01.PXSX
• Couldn't or failed to load Kexts
• Couldn't alloc class 

Stuck at apfs_module:2411

[Jief_Machak]

Hi everyone,

 

Long time no see ! Was busy for quite some time. That included a long stay in a hospital to fight COVID-19. Please be careful everyone, even if you think "I'm young enough, no lung or heart condition, etc., I should have mild or no symptoms" : it can still go suddenly all very bad. So please, be careful.

I'm ok now so enough of that.

 

In the last few days, I did some rework about APFS bootable volume detection. I'd like to share so if there is a regression, you know where it may come from.

 

Before there was an "APFSUUIDBank" where all folder from Preboot and Recovery were stored. Problem with that was wrong version information, wrong icon, if, for example, your Preboot contains 3 different macOs version to boot. Title weren't great as it was something like "Boot Mac OS X from Preboot".

 

Now, APFS detection is made at ScanVolume. APFS Preboot and recovery can contains folders whose name are an UUID. I called this UUID a "APFS target UUID". Booting through one of these folder end up having the volume with this "target UUID" as root volume. These target UUID are stored in ApfsTargetUUIDArray in REFIT_VOLUME structure.
In ScanLoader(), a loader entry is created for each target UUID from ApfsTargetUUIDArray.

 

Let me know if something is wrong.
 

[Cyberdevs]

@Jief_Machak

 

Welcome back and I'm glad that you are ok now.

[Jief_Machak]

Thanks.

[Matgen84]

@Jief_Machak

Glad to see you again. And the fight against covid-19 is over 

[Jief_Machak]

Covid-19 are dead, but I'm still connected to an oxygen machine. I have to rest after having gone to the toilet, because my lungs are not able to replace the oxygen that I used to stand up for 3 minutes ! Seems that it's because my lungs are surrounded by dead cells (covid and anti-bodies), which means it's supposed to clear with time.

So yes, ware is over, but recovering will take time.

Good news is : I can sit in front of my computer 

 

On my last commit, I removed the hard coded Preboot and Recovery partition names. So now, you can rename these partition. New name will appear at Clover menu. For example : Boot Mac OS X from [target partition] via [preboot partition name].

[Slice]

@Jief_Machak

Glad to see you alive!

Hope you will help us with Clover problems.

[Jief_Machak]

1 hour ago, Slice said:

Glad to see you alive!

Thanks.

1 hour ago, Slice said:

Hope you will help us with Clover problems.

Sure. I'm a bit disconnected of what the problems are now, but please tell me.

 

Oh, I have a question : line 812 of lib.cpp, I had to change "Volume->DiskKind = DISK_KIND_NODISK;" to "Volume->DiskKind = DISK_KIND_INTERNAL;" to be able to get the red cross on recovery partition. The condition is "if DevicePathType(DevicePath) == MEDIA_DEVICE_PATH && DevicePathSubType(DevicePath) == MEDIA_VENDOR_DP"

Does anyone know what it means ?

Should we make an exception if it's an APFS volume ?

[Slice]

2 hours ago, Jief_Machak said:

Thanks.

Sure. I'm a bit disconnected of what the problems are now, but please tell me.

 

Oh, I have a question : line 812 of lib.cpp, I had to change "Volume->DiskKind = DISK_KIND_NODISK;" to "Volume->DiskKind = DISK_KIND_INTERNAL;" to be able to get the red cross on recovery partition. The condition is "if DevicePathType(DevicePath) == MEDIA_DEVICE_PATH && DevicePathSubType(DevicePath) == MEDIA_VENDOR_DP"

Does anyone know what it means ?

Should we make an exception if it's an APFS volume ?

MEDIA_VENDOR_DP was used in recent commit 18b35c87f42313f0c9bc680269065e16841ed929 for Catalina. But I have no answer to your question. I have to remember all algo for this.

[Slice]

2 hours ago, Jief_Machak said:

 

Sure. I'm a bit disconnected of what the problems are now, but please tell me.

 

The problem was discussed here 

 

[Leonardo Assunção]

Hi, everyone!

Could anyone please teach me how to build Clover with the latest commit? I have Xcode installed.

 

Thanks! =)

[Matgen84]

7 hours ago, Leonardo Assunção said:

Hi, everyone!

Could anyone please teach me how to build Clover with the latest commit? I have Xcode installed.

 

Thanks! =)

 

Download Clover from repo and launch Buildme.

Xcode command line tools must be installed 

Edited August 9, 2020 by Matgen84