Clover General discussion

[Sherlocks]

Your code above on iMac17,1 is 'Product Name', not 'Family'. I think the 'Product Name' should be 'iMac17,1' (This is correct.) and 'Family' should be 'iMac'. (Not 'iMac17,1')

Also. I saw in Clover Configurator, the 'Family' is 'iMac'. May be not correct.

Snip20161222_25.png

 

No.

iMac17,1 use same name both Product and Family name.

 

System Information

Manufacturer: Apple Inc.

Product Name: iMac17,1

Version: 1.0

Serial Number: XXXXXXXXXXXX

UUID: XXXXXXX-XXXX-XXXX-XXXXX-XXXXXXXXXXXX

Wake-up Type: Power Switch

SKU Number:  

Family: iMac17,1

 

check this. please check

SMBIOS.txt

 

also has already correct data from iMac17,1 smbios data

https://sourceforge.net/p/cloverefiboot/code/3967/tree//rEFIt_UEFI/Platform/platformdata.c#l494

 

what do you think about correct data between darwin dump smbios data and clover configurator?

[PMheart]

No.

iMac17,1 use same name both Product and Family name.

 

check this. please check

SMBIOS.txt

 

also has already correct data from iMac17,1 smbios data

https://sourceforge.net/p/cloverefiboot/code/3967/tree//rEFIt_UEFI/Platform/platformdata.c#l494

 

what do you think about correct data between darwin dump smbios data and clover configurator?

I see. In this case Apple has changed this entry. Thanks for the info.

Yes. This means Clover Configurator is incorrect.

[vusun123]

I see. In this case Apple has changed this entry. Thanks for the info.

Yes. This means Clover Configurator is incorrect.

Use Clover Cloud Editor, it's more up-to-date if you are lazy to edit the config.plist manually

[maleorderbride]

Slice (or any dev), are you aware that enabling Above 4G Decoding in the BIOS will not allow MacOS to start up? It is an old issue, but I cannot remember if I have seen it discussed in this thread. I have tried with both Clover & Chameleon, so this is more of a feature request than a bug. 

 

You can see ASUS's description of 4G Decoding here: https://www.asus.com/support/FAQ/1004170/

 

However, is there any way to address this within the bootloader, or is this purely a kernel problem? Theoretically, any 64-bit OS should be capable of this, correct?

 

 

 

The reason I ask is that if you have 2x GTX 9xx series cards, as well as working Thunderbolt, then the motherboard will fail to POST until you enable 4G Decoding. At that point Windows 7 or higher will load, but MacOS will now restart while loading.

 

 

I am in 10.12.2 now and have noticed that I can now boot with Above 4G Decoding enabled in the BIOS with a single GPU. However, if I add a second GPU then I can only boot with the Web Drivers disabled (GTX 980 Ti's).

Booting with both installed and the Web Driver enabled results in a black screen, sometimes a mouse cursor on the black screen, and sometimes a network presence. If I leave it there for long enough I eventually get a Web Driver error:

 

NVDA(Graphics): Channel Timeout!

NVDA: Channel Info [0, 0x2, 0x3, 0x2]

 

I cannot actually connect to the device via the network, nor is this the typical AppleGraphicsControl-caused black screen as I have patched that as well as tried a DSDT that has GFX1 for both GPUs.

 

Is any of this related to Clover, or does this seem to be purely an Nvidia web driver issue?

 

Thanks!

[Slice]

Hi Sherlocks!

  "MacBook-Aluminum",   // MBP132
  "MacBook-Aluminum",   // MBP133

Not aluminum, but aluminium!

[Sherlocks]

Hi Sherlocks!

  "MacBook-Aluminum",   // MBP132
  "MacBook-Aluminum",   // MBP133

Not aluminum, but aluminium!

 

sorry. i don't really thinks this mistake.

old smbios has this name...

https://sourceforge.net/p/cloverefiboot/code/3917/tree//rEFIt_UEFI/Platform/platformdata.c

 

i just copy and paste. 

 

i will add imac 13,3 education model with fix.

education model has difference in other web site.

most of mac info sites has model name is imac13,1

http://www.everymac.com/systems/apple/imac/specs/imac-core-i3-3.3-21-inch-aluminum-early-2013-education-specs.html

 

but geekbench shown model name imac13,3

https://browser.primatelabs.com/geekbench3/2339913

you can see imac13,3 in motherboard entry

how do you think?

 

thank you:)

 

add. strange

everymac has "aluminum"

http://www.everymac.com/systems/apple/macbook_pro/specs/macbook-pro-core-i7-2.9-15-late-2016-retina-display-touch-bar-specs.html

 

 

This notebook is offered in either silver colored aluminum or a darker "Space Gray" colored aluminum housing (but not other colors) and weighs just a bit more than 4 pounds (1.83 kg).

 

what is correct?

 

i saw both aluminum and aluminium are using in dictionary

[RehabMan]

sorry. i don't really thinks this mistake.

old smbios has this name...

https://sourceforge.net/p/cloverefiboot/code/3917/tree//rEFIt_UEFI/Platform/platformdata.c

 

i just copy and paste. 

 

i will add imac 13,3 education model with fix.

 

thank you:)

 

add. strange

everymac has "aluminum"

http://www.everymac.com/systems/apple/macbook_pro/specs/macbook-pro-core-i7-2.9-15-late-2016-retina-display-touch-bar-specs.html

 

what is correct?

 

i saw both aluminum and aluminium are using in dictionary

 

Aluminum is the correct spelling as per North American English.

Given the location of Apple headquarters in Cupertino, California, seems expected.

[PMheart]

Use Clover Cloud Editor, it's more up-to-date if you are lazy to edit the config.plist manually

Thanks. But it has been a long time that I didn't use Clover...

[Sherlocks]

added imac13,3 model.

 

many apple product info site inform imac13,3 is not exist, only shown imac13,1. so, i skipped this model.

http://www.everymac.com/systems/apple/imac/specs/imac-core-i3-3.3-21-inch-aluminum-early-2013-education-specs.html

 

like PMHeart mentioned, found this model's exist

first geekbench site

http://browser.primatelabs.com/geekbench3/search?utf8=%E2%9C%93&q=Mac-7DF2A3B5E5D671ED

 

more correct information in apple site

https://support.apple.com/ko-kr/HT201518

iMac (21.5-inch, Early 2013) iMac13,3 IM131.010A.B09 (2015-002)

 

as result, everymac and other sites has wrong information.

i added this model.

 

PMheart also Mentioned iMac15,2. but there is no iMac15,2 product name. only exist 15,1 name.

this is wrong in RevoBoot source

https://github.com/Piker-Alpha/RevoBoot/blob/clang/i386/libsaio/smbios/model_data.h#L133

 

board-id is Mac-FA842E06C61E91C5.

 

geekbench

http://browser.primatelabs.com/v4/cpu/1404453

it shown 15,1 in board entry.

 

kernel panic log in apple site. we sure it from this

https://discussions.apple.com/thread/6840233?start=0&tstart=0

 

other models are no problem.

 

also I changed the information by board ID. you can see what i said.

 

additionally, some imac models use mobile cpu.

here is list

 

  "Mac-81E3E92DD6088272",  // IM144   Intel Core i5-4260U @ 1.40 GHz 

  "Mac-A369DDC4E67F1C45",  // IM161   Intel Core i5-5250U @ 1.60 GHz 

 

because of this fact, i changed

 

      gSettings.ChassisType = MiscChassisTypeLapTop; //09;

      if((Model == iMac162) || (Model == iMac171)) {

          gSettings.Mobile      = FALSE;

          break;

      }

      gSettings.Mobile      = TRUE;

      break;

 

added

 

      gSettings.ChassisType = MiscChassisTypeAllInOne; //13;

      if(Model == iMac144) {

          gSettings.Mobile      = TRUE;

          break;

      }

 

tested build in latest r3968

 

 

you did this in r3963

 

chflags hidden "${mntpt}/${NVRAMFilename}"

 

changed

 

 

#chflags hidden "${mntpt}/${NVRAMFilename}"

 

we always see nvram file in EFI root and Mac HDD root.

 

i wonder that rc script is testing in latest commit?

 

i just wonder nvram file's hidden or not

 

thank you.

platformdata-update.zip

[PMheart]

added imac13,3 model.

 

many apple product info site inform imac13,3 is not exist, only shown imac13,1. so, i skipped this model.

http://www.everymac.com/systems/apple/imac/specs/imac-core-i3-3.3-21-inch-aluminum-early-2013-education-specs.html

 

like PMHeart mentioned, found this model's exist

first geekbench site

http://browser.primatelabs.com/geekbench3/search?utf8=%E2%9C%93&q=Mac-7DF2A3B5E5D671ED

 

more correct information in apple site

https://support.apple.com/ko-kr/HT201518

iMac (21.5-inch, Early 2013) iMac13,3 IM131.010A.B09 (2015-002)

 

as result, everymac and other sites has wrong information.

i added this model.

 

PMheart also Mentioned iMac15,2. but there is no iMac15,2 product name. only exist 15,1 name.

this is wrong in RevoBoot source

https://github.com/Piker-Alpha/RevoBoot/blob/clang/i386/libsaio/smbios/model_data.h#L133

 

board-id is Mac-FA842E06C61E91C5.

 

geekbench

http://browser.primatelabs.com/v4/cpu/1404453

it shown 15,1 in board entry.

 

kernel panic log in apple site. we sure it from this

https://discussions.apple.com/thread/6840233?start=0&tstart=0

 

other models are no problem.

 

also I changed the information by board ID. you can see what i said.

 

additionally, some imac models use mobile cpu.

here is list

 

because of this fact, i changed

 

added

 

tested build in latest r3968

 

 

you did this in r3963

 

changed

 

we always see nvram file in EFI root and Mac HDD root.

 

i wonder that rc script is testing in latest commit?

 

i just wonder nvram file's hidden or not

 

thank you.

Great job!

But I don't think we can rely on GeekBench etc.

Ah. Yes. Some iMac and Macmini models use mobile CPUs. But what else?

Cuz `#chflags hidden "${mntpt}/${NVRAMFilename}"` has been commented, and thus the /nvram.plist should be visible?

[Sherlocks]

Great job!

But I don't think we can rely on GeekBench etc.

Ah. Yes. Some iMac and Macmini models use mobile CPUs. But what else?

Cuz `#chflags hidden "${mntpt}/${NVRAMFilename}"` has been commented, and thus the /nvram.plist should be visible?

Ofc. I dont want to rely on GeekBench. But we can get more correct data if we avoid hack build information.

 

I also wonder commented visible file like you said.

 

Now, we are still testing or not about nvram work? Rc script has problem in r3961. rc script is no problem for me in lastest commit. I wonder whether Rc script resolve or not.

 

나의 LG-F410S 의 Tapatalk에서 보냄

[PMheart]

Ofc. I dont want to rely on GeekBench. But we can get more correct data if we avoid hack build information.

 

I also wonder commented visible file like you said.

 

Now, we are still testing or not about nvram work? Rc script has problem in r3961. rc script is no problem for me in lastest commit. I wonder whether Rc script resolve or not.

 

나의 LG-F410S 의 Tapatalk에서 보냄

Some GeekBench results might be uploaded by hackintosh builds...

Ah sorry. I have no need for rc scpt and thus I can't test...

BTW. It seems that InjectKext and ForceKextsToLoad are attached... I want to use ForceKextsToLoad but don't want to inject Kexts in some cases. I just want to use ForceKextsToLoad. What should I do in this case?

[TheRacerMaster]

iMac13,3 does exist, it's the 21.5-inch, Early 2013 education-only model (with an i3-3225, using HD 4000). iMac15,1 is the only model that does exist; however, iMac15,1 uses three different board IDs (depending on the GPU); same thing with iMac17,1.

[PMheart]

iMac13,3 does exist, it's the 21.5-inch, Early 2013 education-only model (with an i3-3225, using HD 4000). iMac15,1 is the only model that does exist; however, iMac15,1 uses three different board IDs (depending on the GPU); same thing with iMac17,1.

Yes. And it seems that iMac13,3 is the only model which Apple uses Intel Core i3 Processor.

[nmano]

It is so wonderful what you have been doing and merry Christmas to All.

[smolderas]

I'm sorry, but the wiki for compiling clover seems outdated.

I'm getting following error with `ebuild.sh`, even though the new "mtoc.NEW" is installed:

TOOLCHAIN_DIR: <redacted>/clover/src/edk2/Clover/../../opt/local
Installing mtoc
Password:
Installing mtoc.NEW
NASM_PREFIX: <redacted>/clover/src/edk2/Clover/../../opt/local/bin/
NASM_VER: 2.12.02
./ebuild.sh: line 336: [: 2: integer expression expected
./ebuild.sh: line 337: [: 2: integer expression expected
./ebuild.sh: line 338: [: 2: integer expression expected
your nasm is not good to build Clover!

Any idea?

 

Edit: I solved my issue by changing all comparison operators respectively:

-ge - >=

-gt - >

-eq - ==

 

 

Edit2: Well, it still won't build:

[CC] ScanMem32Wrapper
Usage: /usr/local/bin/libtool [OPTION]... [MODE-ARG]...
Try 'libtool --help' for more information.
Usage: /usr/local/bin/libtool [OPTION]... [MODE-ARG]...
Try 'libtool --help' for more information.
[CC] PcdLib
libtool:   error: unrecognised option: '-static'

I'm using edk2@23520

 

Edit3:

I have "libtool 2.4.6". Is it too new? Is there a specific version to use with clover?

Edited December 25, 2016 by smolderas

[Slice]

Did you apply Patches_for_EDK2?

I am using system libtool as it set by default Xcode installation from 10.7 up to recent 10.12.3

[smolderas]

Did you apply Patches_for_EDK2?

I am using system libtool as it set by default Xcode installation from 10.7 up to recent 10.12.3

Yes, of course.

[Sherlocks]

@Slice

 

i found brightness load is not work after reboot. always hold on brightness after boot.

 

I mistook the brightness value for the change.

 

So in a recent rc script I tracked down the problem.

 

latest rc script code(hold on brightness value)

 

Supreme-MBP:~ supreme$ sudo /etc/rc.shutdown.d/80.save_nvram_plist.local

NVRAM saved on root of disk0s1

Supreme-MBP:~ supreme$ 

 

# echo "$rootDevice"

    # If NVRAMDevice is not specify use rootDevice

    #NVRAMDevice=${NVRAMDevice:-$rootDevice}

    NVRAMDevice=$(findFirstESP)

 

# echo "$NVRAMDevice"

    # Normalize NVRAMDevice (keep only the device name (remove /dev/)

    NVRAMDevice=${NVRAMDevice##*/}

 

# echo "normal $NVRAMDevice"

    if [[ ! "$NVRAMDevice" =~ ^disk[0-9]*s[0-9]*$ ]]; then

        # NVRAMDevice is a whole disk certainly a RAID device

        # try to save nvram.plist on the first Apple Boot Partition

 

        local AppleBootDevice=${rootDevice}

# echo "$AppleBootDevice"

        if [[ -n "$AppleBootDevice" ]]; then

            # Try to save nvram.plist on the Apple Boot device

            mountAndSaveNVRAM "$AppleBootDevice"

            [[ $? -eq 0 ]] && return

        fi

    fi

    

 

brightness work code

 

Supreme-MBP:~ supreme$ sudo /etc/rc.shutdown.d/80.save_nvram_plist.local

Password:

NVRAM saved to '/nvram.plist' [disk0s4]

Supreme-MBP:~ supreme$ 

 

# echo "$rootDevice"

    # If NVRAMDevice is not specify use rootDevice

    #NVRAMDevice=${NVRAMDevice:-$rootDevice}

    NVRAMDevice=${rootDevice}

 

# echo "$NVRAMDevice"

    # Normalize NVRAMDevice (keep only the device name (remove /dev/)

    NVRAMDevice=${NVRAMDevice##*/}

 

# echo "normal $NVRAMDevice"

    if [[ ! "$NVRAMDevice" =~ ^disk[0-9]*s[0-9]*$ ]]; then

        # NVRAMDevice is a whole disk certainly a RAID device

        # try to save nvram.plist on the first Apple Boot Partition

 

        local AppleBootDevice=$(findFirstESP)

# echo "$findFirstESP"

        if [[ -n "$AppleBootDevice" ]]; then

            # Try to save nvram.plist on the Apple Boot device

            mountAndSaveNVRAM "$AppleBootDevice"

            [[ $? -eq 0 ]] && return

        fi

    fi

 

i tried to change other method to hold on make nvram.plist in EFI root with bightness load work.

 

but i can't. according to this reslove, you need to find new method.

 

merry christmas.

 

thank you.

[Israel Fernandes]

Good afternoon, I own an EVGA X58 motherboard, and an X5660 xeon processor, and I wanted to use smbios for macpro 5.1, but the TyMCEDriver patch does not work. Does anyone have the correct patches?

[TheRacerMaster]

Try adding -nehalem_error_disable as a boot arg.

[Israel Fernandes]

Try adding -nehalem_error_disable as a boot arg.

Worked Tanks

[Israel Fernandes]

Worked on el capitan, sierra still on black screen with parked mouse.

[smolderas]

Did you apply Patches_for_EDK2?

I am using system libtool as it set by default Xcode installation from 10.7 up to recent 10.12.3

With the default Apple built libtool I finally can compile clover again. The newest one of libtool lacks the option `-static`.

[PMheart]

Worked on el capitan, sierra still on black screen with parked mouse.

This could be not related to AppleTyMCEDriver. It seemed that AppleTyMCEDriver just caused KP on non-ECC memory. For further discussion, we may discuss at installation section? 

 

PMheart