Clover General discussion

[Slice]

47 minutes ago, Jief_Machak said:

A refactor of the logic, to make it simpler, would be to use “version start” - “version stop” logic ?

The “version stop” would NOT match.

11 : matches only 11

11.2-11.3 : matches 11.2 11.2.1 etc.

11.2.1-11.3.4 matches from 11.2.1 until 11.3.3

11.4-12.5 matches 11.4.x to 12.4.x

12- matches 12 and greater

 

we can then apply the same logic for next folders.

Great idea! 

To not make new troubles I will propose to keep backward compatibility. Yes, I agree that 12 should correspond only 12 and not 12.1. But we already have kexts folder "12" not "12.x". Keep for folders old logic?

They are de-facto different for folders and for patching.

On the other hand nobody keeps OS version 12. All of us already updated to 12.0.1, 12.1 or 12.2.

[PG7]

On 12/24/2021 at 2:18 PM, Andres ZeroCross said:

I got kernel panic with CLOVER 5142 in Legacy Mode . but if i use CLOVER 5119 Legacy Mode then there is no kernel panic.

I use latest Lilu, AppleALC.kext, VirtualSMC.kext, RestrictEvent.kext, WhateverGreen.kext.

The KP logs show about Lilu and RestrictEvent.kext, i tried to delete RestrictEvent.kext but then KP Logs will show KP about Lilu and VirtualSMC,kext.

Then i revert to Clover 5119 with same version of Lilu, AppleALC.kext, VirtualSMC.kext, RestrictEvent.kext, WhateverGreen.kext and booting is fine.

Any special trick to inject kext in CLOVER 5142???

 

takes the OpenRuntime.efi driver and UEFI file and puts it into Bios

test it, I have this same error here and using this formula it works without problems (I don't have it at all, only on the HP DV6)

On 12/6/2021 at 9:47 PM, PG7 said:

@Jief_Machak @Slice

 

Good afternoon night

 

I have a question:

- it would be possible to put this code of OpenCore Legacy Patcher in Clover config

I use some Hackintosh that I need to patch to activate the graphics NOT metal

to be able to use this patch we must use an ARG "amfi_get_out_of_my_way=1" which blocks Micro and Webcam authorizations on Chrome, Edge and more places

with this last code we can use without problems

I use Google Translate but any doubts I can try to use other translate to facilitate

 

since already thanks

            <key>4D1EDE05-38C7-4A6A-9CC6-4BCCA8B38C14</key>
			<dict>
				<key>DefaultBackgroundColor</key>
				<data>AAAAAA==</data>
				<key>OC_BID</key>
				<string>Mac-F60DEB81FF30ACF6</string>
			</dict>


           Code Work

			<key>4D1FDA02-38C7-4A6A-9CC6-4BCCA8B30102</key>
			<dict>
				<key>rtc-blacklist</key>
				<data></data>
				<key>OCLP-Settings</key>
				<string>-allow_amfi -allow_fv</string>
				<key>OCLP-Version</key>
				<string>0.3.2</string>
			</dict>

 

I tested it and it's fixed

thanks

 @Slice @Jief_Machak

relaunch

thanks 

[Andres ZeroCross]

5 hours ago, PG7 said:

takes the OpenRuntime.efi driver and UEFI file and puts it into Bios

test it, I have this same error here and using this formula it works without problems (I don't have it at all, only on the HP DV6)

Thanks

[Jief_Machak]

8 hours ago, Slice said:

Great idea! 

To not make new troubles I will propose to keep backward compatibility. Yes, I agree that 12 should correspond only 12 and not 12.1. But we already have kexts folder "12" not "12.x". Keep for folders old logic?

They are de-facto different for folders and for patching.

On the other hand nobody keeps OS version 12. All of us already updated to 12.0.1, 12.1 or 12.2.

If we want to keep backward compatibility, we can't unified the logic (12 means 12.x.x in kext folder and means only 12 in patches).

If we can't unified the logic, maybe we should keep what we already have ?

 

I would be against a special intermediary version that warns users that their kext folder "12" should be renamed "12.x" or "12-13" (for only 12), or "12-" for 12 and more.

After few weeks, we switch to a proper unified logic...

[SavageAUS]

Value 51397  makes  every everything even  faster
sysctl -a | grep freq                       
hw.busfrequency: 48000000
hw.busfrequency_max: 48000000
hw.busfrequency_min: 48000000
hw.cpufrequency: 1850000000
hw.cpufrequency_max: 1850000000
hw.cpufrequency_min: 1850000000
hw.tbfrequency: 1000000000
machdep.tsc.frequency: 873749000
 
boot-log.txt

It seems like the lower the value the faster the animations and the higher value slows everything down.

[LockDown]

6 hours ago, PG7 said:

and UEFI file and puts it into Bios

You mean  all the .efi in UEFI folder?

[Guest 5T33Z0]

On 12/26/2021 at 8:30 AM, Slice said:

@5T33Z0

Look please your instructions about MatchOS.There should be masked strings 12.x, 11.5.x etc.

Ok, I've changed the description of "MatchOS":

 

https://github.com/5T33Z0/Clover-Crate/blob/main/Kernel_And_Kext_Patches/README.md#about-patching-masks

 

I will add a chapter about the way Clover handles kexts in the kext folder once this is decided for macOS 12.

[Slice]

On 12/27/2021 at 7:03 AM, SavageAUS said:

Value 51397  makes  every everything even  faster
sysctl -a | grep freq                       
hw.busfrequency: 48000000
hw.busfrequency_max: 48000000
hw.busfrequency_min: 48000000
hw.cpufrequency: 1850000000
hw.cpufrequency_max: 1850000000
hw.cpufrequency_min: 1850000000
hw.tbfrequency: 1000000000
machdep.tsc.frequency: 873749000
 
boot-log.txt

It seems like the lower the value the faster the animations and the higher value slows everything down.

No, the value must exactly correspond hardware.

I think you need additional kernel patches for AMD Ryzen.

 

[SavageAUS]

No, the value must exactly correspond hardware.
I think you need additional kernel patches for AMD Ryzen.
 

Clover was working perfectly prior to this issue which may have coincided with the upgrade of parts so I think it’s an underlying issue somewhere rather than the patches.


Sent from my iPhone using Tapatalk

[MacKonsti]

Hello @Slice and @Jief_Machak and everyone, season greetings. I updated to r5143 on my NUC8 and it works well. I have been browsing the boot log (in Hackintool) and also compared my config to the sample in the package Slice, and wanted to report my findings for possible fixing:

 

1. Logging during boot

0:105  0:000  === [ GetDevices ] ==============================
[...]
0:105  0:000  Legacy MAC address of LAN #0= FF:FF:FF:FF:FF:FF:
0:105  0:000  _checkOEMPath Look for oem dir at path '\EFI\CLOVER\OEM\NUC8i7BEH--FF-FF-FF-FF-FF-FF'. Dir doesn't exist.
0:105  0:000  _checkOEMPath Look for oem dir at path '\EFI\CLOVER\OEM\NUC8BEB\UEFI'. Dir doesn't exist.
0:105  0:000  _checkOEMPath Look for oem dir at path '\EFI\CLOVER\OEM\NUC8i7BEH'. Dir doesn't exist.
0:106  0:000  _checkOEMPath Look for oem dir at path '\EFI\CLOVER\OEM\NUC8i7BEH-2300'. Dir doesn't exist.
0:106  0:000  _checkOEMPath Look for oem dir at path '\EFI\CLOVER\OEM\NUC8BEB'. Dir doesn't exist.
0:106  0:000  _checkOEMPath Look for oem dir at path '\EFI\CLOVER\OEM\NUC8BEB-2300'. Dir doesn't exist.
0:106  0:000  === [ GetUserSettings ] =========================

I think you may want to remove the word _checkOEMPath being displayed, and just leave "Look for oem dir at path" perhaps consider improving it to "Looking for OEM folder at path"

0:106  0:000  === [ GetUserSettings ] =========================
0:106  0:000  GetDefaultModel()=iMac18,3
0:107  0:000  Using config.plist at path: \EFI\CLOVER
0:108  0:000  Cannot find smbios.plist at path '\EFI\CLOVER' : Not Found
0:108  0:000  '\EFI\CLOVER\smbios.plist' not loaded. Efi error Not Found

You may want to consider to make the error reported only once (not found SMBIOS.plist file) ?

38:424  0:000  === [ ACPIPatchedAML ] ==========================
38:424  0:000  Unsorted
38:424  0:000  Inserting SSDT-APPLE.aml from \EFI\CLOVER\ACPI\patched: size=236 OCABC: AllocPages 1 0x6F5BC000 (1) - Success
... Success
38:424  0:000  Inserting SSDT-AWAC.aml from \EFI\CLOVER\ACPI\patched: size=81 OCABC: AllocPages 1 0x6F5BB000 (1) - Success
... Success
38:424  0:000  Inserting SSDT-EC-USBX.aml from \EFI\CLOVER\ACPI\patched: size=377 OCABC: AllocPages 1 0x6F5BA000 (1) - Success
... Success
38:425  0:000  Inserting SSDT-HPTE.aml from \EFI\CLOVER\ACPI\patched: size=74 OCABC: AllocPages 1 0x6F5B9000 (1) - Success
... Success
38:425  0:000  Inserting SSDT-JHL6340.aml from \EFI\CLOVER\ACPI\patched: size=1430 OCABC: AllocPages 1 0x6F5B8000 (1) - Success
... Success
38:425  0:000  Inserting SSDT-NAMES.aml from \EFI\CLOVER\ACPI\patched: size=199 OCABC: AllocPages 1 0x6F5B7000 (1) - Success
... Success
38:425  0:000  Disabled: SSDT-PLUG.aml, skip
38:425  0:000  Inserting SSDT-PMCR.aml from \EFI\CLOVER\ACPI\patched: size=134 OCABC: AllocPages 1 0x6F5B6000 (1) - Success
... Success
38:426  0:000  Inserting SSDT-SBUS.aml from \EFI\CLOVER\ACPI\patched: size=189 OCABC: AllocPages 1 0x6F5B5000 (1) - Success
... Success
38:426  0:000  Inserting SSDT-XOSI.aml from \EFI\CLOVER\ACPI\patched: size=190 OCABC: AllocPages 1 0x6F5B4000 (1) - Success
... Success

Same for this one, there is double confirmation not really necessary (I refer to the second ... Success string) or is it different?

 

2. Missing or not-yet-updated Sample config parameters

 

I know you did not create/add this but it would be useful to include DisabledAML in your sample, Slice, after DisableASPM ?

		<key>#DisabledAML</key>
		<array>
			<string>SSDT-PLUG.aml</string>
			<string>SSDT-MINE.aml</string>
			<string>SSDT-OTHER.aml</string>
		</array>

Also the RenameDevices you made us be careful in structure, is still the old version:

		<key>#RenameDevices</key>
		<array>
			<dict>
				<key>_SB.PCI0.RP09.PXSX</key>
				<string>SSD0</string>
			</dict>
		</array>

3. Quick Questions

 

a) In the definition of SMBIOS, the key ChassisType accepts integers and hex as string?

		<key>ChassisType</key>
		<integer>16</integer>
or
		<key>ChassisType</key>
		<string>0x10</string>

These are both valid and correct?

 

b) What does your value "MLB" stand for (means) in the key LocationInChassis? Should we replace it with something custom of ours?
 

c) The value of <key>SmbiosVersion</key> as <string>0x0300</string> is only for iMac? I cannot find reference of value for Macmini8,1... and not sure where it is used/displayed ? Or is it for all Apple firmwares this version?

 

d) There is a bootlog entry 37:907  0:616  Boot option Boot0000 not found in my list. Any idea please where this is coming from?
 

Thank you for reading and considering.

 

Edited December 29, 2021 by MacKonsti

[Slice]

@MacKonsti

I will account this as TODO but this moment I can't even think about your questions.

2. Yes, it should be updated.

3a) as always in Clover. <integer> decimal and <string> hexadecimal acceptable.

3b) MLB is MotherLogicBoard serial number.

3c) It appear to be no matter.

 

[Guest 5T33Z0]

@MacKonsti I used the Rename Devices feature in Clover Configurator and it created the Entries in the correct format:

<key>RenameDevices</key>
 <array>
  <dict>
	<key>_SB.PCI0.SAT1</key>
	<string>SATA</string>
  </dict>

 

[MacKonsti]

Yes thanks @5T33Z0 it's just that the sample config in the packaged Clover still has the old format without <array> it was a note for Slice or Jief to update on GitHub.

However, it seems all this time, I have had LocationInChassis with a wrong value, it was supposed to be the same MLB as in RtVariables and not the word "MLB"? It is still unclear...

[PG7]

Hi

 

Clover r5143 is different at boot in Legacy
until r5142 I could install clover in boot6 and it worked in Legacy mode
on r5143 if you do it in boot6 it just stays on screen 6_ and doesn't come out of it
already with Boot7 it works but the mouse is unusable

 

thank you to all of the team

[Slice]

19 hours ago, MacKonsti said:

Yes thanks @5T33Z0 it's just that the sample config in the packaged Clover still has the old format without <array> it was a note for Slice or Jief to update on GitHub.

However, it seems all this time, I have had LocationInChassis with a wrong value, it was supposed to be the same MLB as in RtVariables and not the word "MLB"? It is still unclear...

Yes, it is the bug and I don't remember its origin. Correcting now.

[Slice]

4 hours ago, PG7 said:

Hi

 

Clover r5143 is different at boot in Legacy
until r5142 I could install clover in boot6 and it worked in Legacy mode
on r5143 if you do it in boot6 it just stays on screen 6_ and doesn't come out of it
already with Boot7 it works but the mouse is unusable

 

thank you to all of the team

it looks like the new bug. I always recommend to use an old version of boot6 because it never changed in functions.

[Slice]

On 12/30/2021 at 3:43 PM, PG7 said:

Hi

 

Clover r5143 is different at boot in Legacy
until r5142 I could install clover in boot6 and it worked in Legacy mode
on r5143 if you do it in boot6 it just stays on screen 6_ and doesn't come out of it
already with Boot7 it works but the mouse is unusable

 

thank you to all of the team

Test, please, this version. I probably know what is wrong

 

boot6

[denjohan]

I was trying to make a Clover USB with BootDiskUtility_v2.1.2020.028b for legacy booting on an HP xw6600 (BIOS 1.46A) but it gets stuck at the "6_" prompt. Unfortunately, there are few older Clover versions with images readable by BDU so I had to make the USB on a Mac with Clover v2.3k_r3911 (newer versions didn't boot either). I replaced the r5143 boot6 created by BDU with the version provided by Slice but it still gets stuck at the "6_" prompt, even if I press "6". The r3911 USB made on the Mac boots into Clover but I can't boot from the non-TPM Windows11 installation USB made with Rufus 3.17.1846. Once I get the legacy booting and Windows11 to work, I'm planning to install MacOS too, hopefully Catalina.

[LockDown]

Is ExtendedFirmwareFeatures & ExtendedFirmwareFeaturesMask are just random characters as long its 35/64 bits?

Edited January 5, 2022 by LockDown

[joevt]

1 hour ago, LockDown said:

Is ExtendedFirmwareFeatures & ExtendedFirmwareFeaturesMask are just random characters as long its 35/64 bits?

0-35 = 36 bits.

 https://github.com/acidanthera/OpenCorePkg/blob/1cb8490fd60f6b5f983d42e5bbf08e3d2d43d8f6/Include/Apple/IndustryStandard/AppleFeatures.h

 

See the showfirmwarefeatures command in the gfxutil.sh script at https://gist.dreamtobe.cn/joevt/477fe842d16095c2bfd839e2ab4794ff

 

[PG7]

On 1/2/2022 at 1:37 PM, Slice said:

Test, please, this version. I probably know what is wrong

 

boot6 439.5 kB · 10 downloads

@Slice 

 

hello, good year to the whole team

 

Slice with this boot did NOT work

 

but then i followed your idea and tested it with the r5142 boot6 and it worked

it would be possible to fix the next Clover with this boot6 that is working on pkg r5142

 

thanks for the solution

 

* google translate

logs boot 

2022-01-05_20-03_CLOVERX64.EFI.log preboot.log

Edited January 5, 2022 by PG7

[mnfesq]

I use Clover on my Windows desktop because my motherboard does not support NVMe drives.  The NVMe driver in Clover loads and then allows me to use an NVMe drive as my Windows system drive.  I can't tell you how amazing that is to me and I am very grateful to the Clover team who made that possible.  I have a question:  If Clover can enable an unsupported motherboard to load an NVMe drive, might Clover also be capable of being modified to include something that would allow Windows 11 to be run on a system with a CPU that does not have all of the minimum features required to install Windows 11?  My motherboard supports TPM 2.0 but I am using a Haswell (4th Gen.) CPU and Windows 11 requires an 8th Gen. CPU.  Thanks.

[Avery B]

1 hour ago, Hervé said:

I invite you to refer to the dedicated Win11 thread of the Windows section where you'll find existing answers to your particular Win11 query. And no, there's no need of Clover for this.

 

As for your motherboard and NVME, what you state seems impossible to me. By its own very nature, an NVME SSD is basically a PCIe SSD connected over an M.2 physical interface. If your motherboard does not support NVME SSDs, I would expect it supports M.2 SATA SSDs only. Of course, I assume here that your motherboard has M.2 slots for SSD, unless you tell us you actually use some M.2-to-SATA adapter here.

 

There are 2 x types of M.2 SSD:

1. NVME -> PCIe SSDs that use n x  lanes (usually 2 or 4) of given PCIe versions (usually 3.0 or 4.0). One usually needs to compare the motherboard's specs to the NVME SSD's specs in order to verify that PCIe versions and number of lanes are aligned. For instance, it's a bit of a waste to fit a PCIe 4.0 x4 NVME SSD to a motherboard with M.2 slots that only supports PCIe 3.0 x2.
2. SATA -> regular SATA-III SSDs of M.2 form but otherwise identical to a regular 2.5" SATA SSD. They can usually be fitted to any motherboard that supports M.2 disks, SATA being usually universally supported on those (it's normally SATA or SATA+NVME, rarely NVME only).

Happy to get bashed on the matter of course but I cannot see how Clover would be able to provide support for hardware unsupported by your motherboard. If it does, Slice & other Clover devs should be on a rapid way to amazing fortune. You're probably just missing the necessary driver in your Win (Win10?) bootstrap EFI partition.

It could be that the motherboard does not have a UEFI driver for NVMe SSDs, but the hardware supports NVMe SSDs. So adding the NVMe driver through Clover/OC would work.

[mnfesq]

Thanks for your input.  Actually, I have two PCI-e adapter cards - one is for an mSATA SSD drive and one is a M.2 NVMe drive.  I also have SATA connectors to the motherboard.  Clover is installed on the mSATA card and Windows 10 is installed on the NVMe drive.  BTW, my desktop has the MSI H87-G43 motherboard, the i7-4790 CPU and 32 GB DDR3 RAM.  

 

@Hervé I will check out the Windows 11 thread.  I have been in the fast lane of the Windows Insider Program for years and would love to put Windows 11 on my desktop and the 3 virtual machines I run on it.  I am aware of patches that have attempted to bypass with system check that Windows 11 installer uses but, like with macOS, it would be nice to make all the changes in the boot loader so that I could install vanilla versions of Windows 11 and not have to patch each preview version I test out.

[Slice]

On 1/13/2022 at 10:24 PM, mnfesq said:

 I have a question:  If Clover can enable an unsupported motherboard to load an NVMe drive, might Clover also be capable of being modified to include something that would allow Windows 11 to be run on a system with a CPU that does not have all of the minimum features required to install Windows 11?  My motherboard supports TPM 2.0 but I am using a Haswell (4th Gen.) CPU and Windows 11 requires an 8th Gen. CPU.  Thanks.

This time no. 

To use unsupported CPU in macOS Clover is able to modify macOS kernel. But about Windows 11 there is quite another story. I am not able to modify somehow Windows 11 to change its behaviour.