Clover General discussion

[naiclub]

9 hours ago, Slice said:

 

 

I do multi-boot between Mac and Windows. When I go to Windows and I get a black screen, what is the reason?
There are 2 SSDs, 1 of which is SATA640GB to the Mac. With the other child is NVMe m.2 WDBlack1TB installed in Windows 10 questions are that Why does it enter Windows and the screen is black? Any friends who have a way to do it, please tell us.

Follow this thread

Thank you very much

[Matgen84]

Hi @Slice @Jief_Machak


Is it possible to enhance the menu in the new BuildMe. To avoid space with the header. Please.

 

Spoiler

 

[AntarcticP]

Warning for users of Clover Configurator.

The latest version (5.18.2.0) messes up my config.plist file --> changes KernelToPatchwhich is needed for my AMD.

And removes InjectKexts in SystemParameters.

Result: system will not boot. So best not to use at this moment.

[Jief_Machak]

12 hours ago, Slice said:

You should use common header.

Of course, there is only one header.

 

12 hours ago, Slice said:

.h can be included many times.

 

By many times, I assume "by many different compilation units".

Because if you include twice a header (not guarded) in the same compilation unit (a .cpp or it's included files), you'll get "error: redefinition of 'xxx'" for classes.
That why we guard all headers with :

#ifndef PLATFORM_SELF_H_
#define PLATFORM_SELF_H_

blabla

#endif

It's true that for a header containing only extern definition (variables and functions), you could include it more than once in the same compilation unit as it is legit to re-declare extern vars and functions. At the time of pure C, it was mostly true.

 

We know that declaring a var (not extern, the var itself) or a function implementation in a file that you include in more than one compilation unit will give "duplicate symbol" errors. But it's not true for member init and method implementation. It's legit to put class member initialisation and method implementation in headers. For example, XPointer::isEmpty() implementation is in the header file. Or XPointer constructor.

That said, I don't how the compiler manages that. Maybe just deleting all duplicate symbol to keep only one.

It seems to currently compile fine with braced initialisation (or realisation), so that just proves it.

 

I have million of class member initialisation in headers. Like in the big SETTINGS_DATA object and all it's inner classes. At that time, I was using more something like 

      XStringWArray                     DisabledAML = XStringWArray();

But then I was reading about braced-list-initializer here : https://en.cppreference.com/w/cpp/language/list_initialization and 

      XStringWArray                     DisabledAML {};

does the same, but quicker to type, and less code in non-optimized compilation.

 

16 hours ago, Slice said:

/Users/sergey/src/CloverBootloader/rEFIt_UEFI/refit/../Settings/Self.h:55:14: warning: type of 'reInitialize' does not match original declaration [-Wlto-type-mismatch]

 

14 hours ago, Jief_Machak said:

/Users/sergey/src/CloverBootloader/rEFIt_UEFI/Settings/../Settings/../Settings/Self.h:19:7: note: a different type is defined in another translation unit

 

16 hours ago, Slice said:

/Users/sergey/src/CloverBootloader/rEFIt_UEFI/Settings/../Settings/../Settings/Self.h:25:21: note: a field with different name is defined in another translation unit

 

16 hours ago, Slice said:

/Users/sergey/src/CloverBootloader/rEFIt_UEFI/refit/../gui/../libeg/../Settings/Self.h:33:36: note: the first difference of corresponding definitions is field 'm_SelfSimpleVolume'

These errors looks more like an other compilation unit use previous definitions from a previous compilation. The edk makefile system does not always recompile everything needed. I'd say that when you fixed the unused vars, it probably triggered a rebuild of this compilation unit that was using an old version, making it using the right version.

[matxpa]

On 5/22/2021 at 9:50 AM, matxpa said:

@Slice & @Jief_Machak

 

Hello.

On r5135 release why « All checks have failed » on « Build Clover Pack » from commitID 8da3efae to 2b84ee6 and from 4bd3434 to 4c30ccd6 ?

Regards

 

 

Nice... we can see it's fixed

 

Spoiler

 

[LAbyOne]

8 hours ago, Matgen84 said:

Hi @Slice @Jief_Machak


Is it possible to enhance the menu in the new BuildMe. To avoid space with the header. Please.

 

  Hide contents

 

Well with your great surprise, it is done, since that is caused by : 

1) the length of a sentence, 

2) number of items into the menu list,

So what you are seeing is the "enhanced" version,

Now since the question arose,  I'm announcing here, and you all will find the post in Main Clover Thread, build me is taking a very different way (style) from what you were used to ... I invite you all to take a look there, and give your opinion...

 

Link to Download is there

Edited May 24, 2021 by LAbyOne
Added link to new Clover Suite Builder

[Jief_Machak]

Where does it download things ? I saw something about a "src" folder but I don't know where.

[LAbyOne]

3 hours ago, Jief_Machak said:

Where does it download things ? I saw something about a "src" folder but I don't know where.

it depends on what things are you referring to..

normally if tools the directory auto opens right after it is done

kexts and other files are all stored into Extra/Kexts, Extra/Tools, etc

If you are referring to the actual Download / build  directory its src/tmp hidden

I should probably specify this too... 

open $HOME/src/tmp

Edited May 24, 2021 by LAbyOne

[MacKonsti]

Hi everyone, @Slice and @Jief_Machak I've installed r5135 on my NUC; downloaded the Catalina update 2021-003 and upon first reboot, the "Data" partition was automatically selected OK and timed-out (I have 10 sec in config) and booted.

However, upon the needed second reboot, Clover did not timeout and was waiting for me to select (again) the "Data" AFPS partition to continue and complete the update.

Just FYI. This was also happening with previous Clover versions, cannot remember since when (I am sorry).
Not sure what is flagged or marked in that "Data" partition and Clover detects it at first but not at second reboot.

Moreover, my disk is an NVMe with 1 AFPS partition and 2 containers, the System and Data. Nothing complex...

Thank you!

[Jief_Machak]

True, it does that since a while...

[LAbyOne]

Latest Clover Suite Builder 

Download is ready

Release 1.3

*this one can also be available as menuItem

you can leave comments on thread in signature if interested

Thank you

Edited May 28, 2021 by LAbyOne

[Slice]

On 5/25/2021 at 2:58 PM, MacKonsti said:

Hi everyone, @Slice and @Jief_Machak I've installed r5135 on my NUC; downloaded the Catalina update 2021-003 and upon first reboot, the "Data" partition was automatically selected OK and timed-out (I have 10 sec in config) and booted.

However, upon the needed second reboot, Clover did not timeout and was waiting for me to select (again) the "Data" AFPS partition to continue and complete the update.

Just FYI. This was also happening with previous Clover versions, cannot remember since when (I am sorry).
Not sure what is flagged or marked in that "Data" partition and Clover detects it at first but not at second reboot.

Moreover, my disk is an NVMe with 1 AFPS partition and 2 containers, the System and Data. Nothing complex...

Thank you!

Debug will be useful to continue discussion.

It may depends on NVMe file path which I have no to check. With my configuration Big Sur updates run automatically from begin to end.

[MacKonsti]

1 hour ago, Slice said:

Debug will be useful to continue discussion.

It may depends on NVMe file path which I have no to check. With my configuration Big Sur updates run automatically from begin to end.

Hi @Slice thanks, will have to wait to new OS update then, as I cannot reproduce this now obviously.

However please confirm if the BOOT section "debug" key is the one you request to be enabled, whenever a new update arrives, so I can trace the actions for you?

Also please remind me where to find the debug logs once created? I haven't used debug for a long time... Thank you.

[Slice]

23 minutes ago, MacKonsti said:

Hi @Slice thanks, will have to wait to new OS update then, as I cannot reproduce this now obviously.

However please confirm if the BOOT section "debug" key is the one you request to be enabled, whenever a new update arrives, so I can trace the actions for you?

Also please remind me where to find the debug logs once created? I haven't used debug for a long time... Thank you.

Yes, debug key in Boot section.

The log with long name will be created in /EFI/CLOVER/misc/ folder. If you will do several restarts then there will be several logs with different names corresponding the time of the restart.

[JorgeMax]

@SliceI upgraded from 11.4 to beta 11.5. In the process, the first restart was not automatic, there was no countdown and I had to press the "Enter" key. The .zip file has the logs of the 3 restarts

 

misc.zip

[Slice]

5 hours ago, JorgeMax said:

@SliceI upgraded from 11.4 to beta 11.5. In the process, the first restart was not automatic, there was no countdown and I had to press the "Enter" key. The .zip file has the logs of the 3 restarts

 

misc.zip 38.21 kB · 2 downloads

Thank you for the report. I will study it in the next week as I am on trip now.

[yapan4]

r5136

[SavageAUS]

Clover 5136 reported as 5135

[Slice]

5 minutes ago, SavageAUS said:

Clover 5136 reported as 5135

 

This is a bug caused as I think by new build system.

I will look later.

[yapan4]

On 6/1/2021 at 4:19 PM, SavageAUS said:

Clover 5136 reported as 5135

In my case, the correct version is displayed. Installed through the pkg installer as an update from version 5135

 

 

Spoiler

 

Edited June 3, 2021 by yapan4

[Slice]

Second compilation resolved the problem

 

CLOVERX64.efi.zip

[MacKonsti]

14 hours ago, Slice said:

Second compilation resolved the problem

Hi @Slice thank you for the update.

Will you replace the Git Hub files in your release then with the new compiled ones?

 

Also, do you consider r5136 more steady or mature compared to r5135 ? Because some posts ago, you had mentioned that the (previous) release was not fully ready yet and we should wait for a new one/update.

Many thanks.

[SavageAUS]

Clover 5136 AudioDxe.efi still has audio problem with AMD rig.


Sent from my iPhone using Tapatalk

[Slice]

Yes, AudioDxe is still wrong.

8 hours ago, MacKonsti said:

Hi @Slice thank you for the update.

Will you replace the Git Hub files in your release then with the new compiled ones?

 

Also, do you consider r5136 more steady or mature compared to r5135 ? Because some posts ago, you had mentioned that the (previous) release was not fully ready yet and we should wait for a new one/update.

Many thanks.

Always new release is better than old and all releases are not perfect.

[naiclub]

 

Spoiler

 

 

Spoiler

 

Edited June 2, 2021 by naiclub