OpenCore General Discussion

[zisqo]

4 hours ago, REKTIMU2 said:

 

I just wanted to make a package installer, so I asked to use the Logo. After some reading I believed we could not only use the Logo or "OpenCore" but I believed the package names as well so I asked Fritz what I needed to do to make the installer and if I had to I would consider forking.

 

My takeaway was that I or anyone else had no right to distribute OpenCore binaries without permission unless it was forked and renamed.

 

But if anyone has the legal right to distribute source/binaries without the logo or "OpenCore", I think everything is fine then?

 

and what the N-D-K fork??? 

[mhaeuser]

On 6/23/2020 at 6:58 AM, REKTIMU2 said:

But if anyone has the legal right to distribute source/binaries without the logo or "OpenCore", I think everything is fine then?

You have the legal right when complying to BSD-3 and copyright law, yes. Yet we do not like to see this for obvious and often discussed reasons.

[MacPato]

@Download-Fritz how's the figuring Opencore out? seems its broken for now for macOS 11

[mhaeuser]

@MacPato It's a bloody mess, expect nothing this week.

[eSaF]

I now can clearly see why this Thread has gained the notoriety as the 'Angry Thread' title in some quarters. Why are ideas or suggestions shot down with such dismissive and derogatory remarks by some, some of the best innovations and inventions today are based on initial seemly bad but good intentioned ideas and forked into something better or worthwhile. When I first joined IM, I spent most of the time lurking and browsing, reading as many posts as possible, learning and gathering knowledge about the Hackintosh Scene from folks I thought were just brilliant giving insightful tips and fixes. Now unfortunately it seems to me it's a case of 'No your idea or suggestion is not part of our planning or criteria going forward so go away'. I have noticed a steady decline in membership and resignations, is it a wonder as to why? I am neither a Moralist or a Holier than Thou person but I like to treat others with respect and courtesy no matter how much I am irked by the person or their query, surely we all can employ better diplomacy when dealing with others. Mods and Admins since this post is off topic, please edit or remove as you see fit - Thank you. 

[mhaeuser]

9 minutes ago, eSaF said:

'No your idea or suggestion is not part of our planning or criteria going forward so go away'

What is wrong with us not supporting ideas we think are wrong? It is not like literally any project head does this, not in this entire community and not anywhere else, they allow bad ideas to exist but do not endorse them (just as we do). The difference is we take the time to comment why it is a bad idea before we ignore any future development.

[Henry.M.]

20 hours ago, Andrey1970 said:

 

It boot.efi verbose.

@Andrey1970 Can you explain, what this reports mean and how to get rid of them, because I have no idea where to start?

[MacPato]

@Download-Fritz i figured as much only i spoke to some dev on the MacRumors discord server that has a fully functional Acer Laptop using Opencore but won't share how they managed to get it booted and installed. He's afraid apple will patch it, I also have seen on unsupported real macs they are using open core and some hax.dylib file to get it installed. do you know anything about this?

[mhaeuser]

@MacPato AvoidRuntimeDefrag needs to be off (causes early crash), kexts will not be injected unless prelinkedkernel is enforced over KC

[eSaF]

2 minutes ago, Download-Fritz said:

What is wrong with us not supporting ideas we think are wrong? It is not like literally any project head does this, not in this entire community and not anywhere else, they allow bad ideas to exist but do not endorse them (just as we do). The difference is we take the time to comment why it is a bad idea before we ignore any future development.

Than you Mr. Fritz, your point is duly noted and taken on board. Can I say as I did not call any names or point to any particular individual, I am not guilty as such but I do agree as it is a project you're heavily involved with, you're entitled to accept or reject ideas, all that my post is highlighting is how responses are handled. As an avid user of Opencore I am grateful for the work done by all the developers and contributors and as I have nothing meaningful to offer to further it's development I am sorry if you feel I have over step the mark with my post.

[pkdesign]

8 hours ago, blackosx said:

EDIT2:

I'm attaching an initial version of the icons for testing. I've also included an Apple.icns (Thanks @crusher). Note: these don't include any with labels at the moment..

 

Alt_Icons_Resources_v0.1.zip

 

Oh man! I've been thinking about this too. Of course the master beat me too it! Might make me switch to OpenCore

[Andrey1970]

10 hours ago, blackosx said:

I know a lot of users like to keep their hacks looking as close to a mac as possible and therefore the use of OpenCanopy with Apple's own icons makes perfect sense.

Please create a separate topic.

[pkdesign]

There should be an OpenCanppy topic. I asked about this in May:

 

[MrTrip]

9 hours ago, Download-Fritz said:

@MacPato It's a bloody mess, expect nothing this week.

 

Hackintosh X86 Discord just reported you said later today or tomorrow for OC, WEG and Lilu patches for Big Sur? Got some awesome update?

[MacNB]

On 6/22/2020 at 2:31 AM, vit9696 said:

That means you have OpenRuntime (you should not use it on legacy systems) and RequestBootVarRouting (it makes no sense on legacy systems as well).

 

@vit9696 Thx.

 

OK so I removed OpenRuntime and set RequestBootVarRouting=False.

 

The log looks better but no cigar. Default boot is not set. See attached Log & config.

 

On 1st boot I ran ResetNVRAM from OC Boot picker which resets & reboots the system.

On 2nd boot, the OC BootPicker display a '*' next to the 1st Entry (Windows) so I arrow down to my EVO250 Entry and press CTRL+ENTER (to set the Default). macOS boots OK, I login and I shut it down.

On 3rd boot,  the OC BootPicker display a '*' next to the 1st entry (Windows) instead of EVO250 (that I set as default previously).

 

Selecting my EVO250 disk from System Preferences->Startup Disk makes not difference.

Setting ShowPicker=False makes no difference as OC still boots the 1st Entry (Windows) which of course crashes with ACPI_ERROR.

OC Boot Timeout = 10s.

 

The Log states "OCB: Setting default - Success" but on next boot it is not boot the Default.

 

Any suggestions what could be wrong and what else I can do to debug further ?

 

BTW, May I make a suggestion that the Configuration.pdf mentions NOT to use OpenRuntime with Legacy BIOS systems.

 

MacNB-opencore-2020-06-23-202849.txt

MacNB-config-2.plist

[MacPato]

1 hour ago, MrTrip said:

 

Hackintosh X86 Discord just reported you said later today or tomorrow for OC, WEG and Lilu patches for Big Sur? Got some awesome update?

We were told by other people an update was coming tomorrow,  @Download-Fritz  had written this somewhere else. 

You should read the other thread.

Edited June 23, 2020 by MacPato

[mhaeuser]

The stuff that is to be fixed soon is the early kernel crash. Kext Injection will only work with forcing prelinkedkernel for a good while, and that is not (yet?) possible with the installer. However it seems things besides the new kernel stuff are not as terrible as feared.

[MacPato]

8 minutes ago, Download-Fritz said:

The stuff that is to be fixed soon is the early kernel crash. Kext Injection will only work with forcing prelinkedkernel for a good while, and that is not (yet?) possible with the installer. However it seems things besides the new kernel stuff are not as terrible as feared.

Always good to hear, I'll make do with my MacPro for now! 

[vit9696]

@MacNB, do you have the latest version of DuetPkg? Your BootOrder looks really weird. Also, there is an error here:

0 -> Boot0080 - failed to read

This looks like your Boot0080 variable data is not valid. Could you post it here?

 

03:296 00:010 OCB: MacNB: InternalGetBootOptionData: FilePathListSize (154) > LoadOptionSize (92) so Returning NULL
03:307 00:010 OCB: 0 -> Boot0080 - failed to read

 

CC @Download-Fritz

[MacNB]

59 minutes ago, vit9696 said:

@MacNB, do you have the latest version of DuetPkg? Your BootOrder looks really weird. Also, there is an error here:

0 -> Boot0080 - failed to read

This looks like your Boot0080 variable data is not valid. Could you post it here?

 

03:296 00:010 OCB: MacNB: InternalGetBootOptionData: FilePathListSize (154) > LoadOptionSize (92) so Returning NULL
03:307 00:010 OCB: 0 -> Boot0080 - failed to read

 

CC @Download-Fritz

 

@vit9696 I assume Duet is the latest as I assumed it is now included with OC. I just did a git pull and ran the build_oc.tool on the 19th June.

 

I noticed the 0 -> Boot0080 - failed to read so I added the extra DEBUG_INFO to see why and it shows FilePathListSize (154) > LoadOptionSize (92) so Returning NULL. 

 

Attached below the nvram.plist file from EFI root.

Also attached output from the following commands:

SR-Mac-530:~ MacNB$ nvram -x 8BE4DF61-93CA-11D2-AA0D-00E098032B8C:BootOrder > nvram-BootOrder.plist
SR-Mac-530:~ MacNB$ nvram -x 8BE4DF61-93CA-11D2-AA0D-00E098032B8C:Boot0080 > nvram-Boot0080.plist

 

You will notice that the value for Boot0080 in nvram-Boot0080.plist differs from the value in the nvram.plist.

I think the value in the macOS nvram (nvram-Boot0080.plist) may have been generated AFTER I ran System Pref->Startup disk.

 

Thanks for the great support all you guys provide.

nvram.plist

nvram-Boot0080.plist

nvram-BootOrder.plist

[koder]

imo don't rush on release patch.. don't forget this is beta only and if you release a patch now then they will patch it on alpha.

[sudoali]

On 6/15/2020 at 8:45 PM, Andrey1970 said:

Don't consider comments // _CRS

 

Yeah, I removed all comments in the file, also I searched only for `(_CRS` to find only where they declared. the problem is OC will search in the .aml file not .dsl. so there should be a tool to search in the .aml file and match them to the .dsl file.

 

In my case, when the `_CRS` in the .dsl is used in other functions, it is not exactly `_CRS` in the .aml file. so I have to ignore them. that's solved by searching for `(_CRS`. However there was something like `^^^.PCI0._CRS` in the .dsl file and it was repeated in the .aml file. so when OC is counting to skip them. it counts this one too. 

 

One also can search the device's name in the .aml file and find where it is declared. and count the `_CRS` till that point. 

[REKTIMU2]

9 minutes ago, sudoali said:

Yeah, I removed all comments in the file, also I searched only for `(_CRS` to find only where they declared. the problem is OC will search in the .aml file not .dsl. so there should be a tool to search in the .aml file and match them to the .dsl file.

 

In my case, when the `_CRS` in the .dsl is used in other functions, it is not exactly `_CRS` in the .aml file. so I have to ignore them. that's solved by searching for `(_CRS`. However there was something like `^^^.PCI0._CRS` in the .dsl file and it was repeated in the .aml file. so when OC is counting to skip them. it counts this one too. 

 

One also can search the device's name in the .aml file and find where it is declared. and count the `_CRS` till that point. 

 

there is no need to compare against dsl, you don't even need to count _CRS occurrences. using iasl -l DSDT.aml will provide a file with hex of methods/devices so you can rename a specific _CRS to XCRS (this is how SSDT-HPET and patch works).

 

you can also disable the device (_.STA = Zero) and add a new device with its modified implementation

[sudoali]

6 hours ago, REKTIMU2 said:

 

there is no need to compare against dsl, you don't even need to count _CRS occurrences. using iasl -l DSDT.aml will provide a file with hex of methods/devices so you can rename a specific _CRS to XCRS (this is how SSDT-HPET and patch works).

 

you can also disable the device (_.STA = Zero) and add a new device with its modified implementation

Thanks, this is another method for it (if you mean to extend the find and replace to make it unique). back then I didn't want to use this method, I wanted a more clean way. now I think both ways are OK. 

I have one question about using `-l` flag, would the bytecode for a specific method be changed if I recompile the .aml file?

[btwise]

In the new EDK version, the address of the submodule boringssl.googlesource.com  seems to be unable to be properly accessed in many places. It is suggested to change it to an easily accessible mirror address!