Jump to content

Clover General discussion


ErmaC
29,866 posts in this topic

Recommended Posts

@Slice: the embedded Clover theme seems to be broken with the latest Clover, compiled with the latest edk2 source. The rest seems to work as usual.

I think it depends on compiler.

Link to comment
Share on other sites

Oh I mentioned this to others but I forgot to mention it to you, Slice. There is a bug in EDK build system that it doesn't properly rebuild an object if a file it includes changes. So in order for a build that's not messed up in some way you need to completely clean the build and rebuild.

  • Like 1
Link to comment
Share on other sites

The installer uses Boot0082 (I think thats the one) to store boot data and BootNext is a number pointing to which Boot variable to use in booting this one next time. The path in Boot0082 is possibly just to the volume (I'm pretty sure it is)

Yes, I only said it is not used to store the boot file path, because that's done via bless. Though seems like an endless list of boot.efi paths looks nicer than just using bless...

making it not suitable for PC UEFI booting

Except for these solutions that add proper 'Apple compatibility'. :P

Apple EFI inspects the volume for the blessed boot data because bless is filesystem based NOT firmware

'NOT firmware' as in it's not an EFI concept? Not sure what you mean, but the blessed path is deffo stored in the FS header, yeah.

  • Like 1
Link to comment
Share on other sites

Its also there.

Thanks for you confirmation!

 

 

Most interesting...

 

Edit: So sierra supports bootable apfs containers?

 

Well, perhaps, I don't think so though... (I guess they should bring full compatibility at least on 10.13...)

 

Then there's the answer take that driver and put it in the appropriate drivers folder for clover and you can read AFS volumes. Note you won't be able to boot clover from an AFS volume but mac OS, yes.

All right, don't tend to say anything here... Just waiting for Apple for more new stuffs. :)

 

10.12.4 also added firmware updates (which contain the ApfsJumpstart driver). 

 

Yup. We've discussed it before!  :yes:

Link to comment
Share on other sites

apfs.efi is located in /usr/standalone/i386 under 10.12.5 Beta. (Sorry. IDK what about 10.12.4...)

 

Well, but I guess Apple just adds it for experimental purposes.

 

Heh, so does that mean that Hackintosh user will be able to boot from a APFS drive, in contrast to real Apple users? :)

Link to comment
Share on other sites

Heh, so does that mean that Hackintosh user will be able to boot from a APFS drive, in contrast to real Apple users? :)

Well, I don't think so.

After releasing APFS, I suppose that the only thing we should do is just adding APFS.efi like we are now doing for HFSPlus.efi, after that we've given the ability to Clover for booting APFS partitions. ;)

  • Like 1
Link to comment
Share on other sites

Yes, I only said it is not used to store the boot file path, because that's done via bless. Though seems like an endless list of boot.efi paths looks nicer than just using bless...

Except for these solutions that add proper 'Apple compatibility'. :P

'NOT firmware' as in it's not an EFI concept? Not sure what you mean, but the blessed path is deffo stored in the FS header, yeah.

 

Well obviously the firmware reads the bless information, to, you know, boot. But yes bless data is indeed in the filesystem header area. The difference is that with the hardcoded paths there would be firmware updates like mad all the time because they change paths, and they added recovery and booting encrypted volumes. Seems to me it is much easier to just put a special place in the volume filesystem to store a path if the volume is selected to boot. Weirdly, bless used to need an entire directory.

Link to comment
Share on other sites

It will be more easier for Apple do not change the path  ^_^  Or place a link in the old path. 

Will 10.12.4 updater go to bless the hdd? Or just set new boot0080 into NVRAM?

Link to comment
Share on other sites

The difference is that with the hardcoded paths there would be firmware updates like mad all the time because they change paths, and they added recovery and booting encrypted volumes. Seems to me it is much easier to just put a special place in the volume filesystem to store a path if the volume is selected to boot.

That was sarcasm at Clover, not critic at Apple EFI. :)

 

Or place a link in the old path.

Supporting a hundred things (here paths) for no reason at all... I've seen that before...

 

Will 10.12.4 updater go to bless the hdd? Or just set new boot0080 into NVRAM?

Ofc it will bless it, macOS does not store the file path in boot entries (well, except for firmware updates as they are loaded from the ESP).

  • Like 1
Link to comment
Share on other sites

That was sarcasm at Clover, not critic at Apple EFI. :)

 

Dude, I know. I was just making sure that others understood.

 

Supporting a hundred things (here paths) for no reason at all... I've seen that before...

 

Windows?

  • Like 1
Link to comment
Share on other sites

hello 

i want to test other graphics.

 

tested

skylake hd 520 hd530

broadwell hd 5300

haswell hd 4400+rehabman kext(for accel)

 

here is my all cases.

 

 

4:493  0:000  Intel HD Graphics 520 [8086:1916] :: PciRoot(0x0)\Pci(0x2,0x0)

4:493  0:000    Use override Custom EDID
4:493  0:000    AAPL00,DualLink = 1
4:493  0:000    FakeID Intel GFX: not set
4:493  0:000    ig-platform-id: not set
4:493  0:000    Beginning Intel GFX auto-detection with ACPI injection
4:493  0:000    Found FakeID Intel GFX = 0x19168086
4:493  0:000    Found ig-platform-id = 0x19160000
4:493  0:000  Intel GFX revision  = 0x7
 
ig-platform - no
fakeid intel gfx - no
 
 
19:975  0:000  Intel HD Graphics 520 [8086:1916] :: PciRoot(0x0)\Pci(0x2,0x0)
19:975  0:000    Use override Custom EDID
19:975  0:000    AAPL00,DualLink = 1
19:975  0:000    FakeID Intel GFX = 0x19168086
19:975  0:000    ig-platform-id = 0x19160000
19:975  0:000    Beginning ACPI injection
19:975  0:000  Intel GFX revision  = 0x7
 
ig-platform - yes
fakeid intel gfx - yes
 
 
13:421  0:000  Intel HD Graphics 520 [8086:1916] :: PciRoot(0x0)\Pci(0x2,0x0)
13:421  0:000    Use override Custom EDID
13:421  0:000    AAPL00,DualLink = 1
13:421  0:000    FakeID Intel GFX = 0x19168086
13:421  0:000    ig-platform-id: not set
13:421  0:000    Beginning Intel GFX auto-detection with ACPI injection
13:421  0:000    Found ig-platform-id = 0x19160000
13:421  0:000  Intel GFX revision  = 0x7
 
ig-platform - no
fakeid intel gfx - yes
 
 
12:334  0:000  Intel HD Graphics 520 [8086:1916] :: PciRoot(0x0)\Pci(0x2,0x0)
12:334  0:000    Use override Custom EDID
12:334  0:000    AAPL00,DualLink = 1
12:334  0:000    FakeID Intel GFX: not set
12:334  0:000    ig-platform-id = 0x19160000
12:334  0:000    Beginning Intel GFX auto-detection with ACPI injection
12:334  0:000    Found FakeID Intel GFX = 0x19168086
12:335  0:000  Intel GFX revision  = 0x7
 
ig-platform - yes
fakeid intel gfx - no
 

 

 

please test and report clover log like above if you want any cases.

welcome to hard testing.

 

i will make duallink recommend log according to resolution(detect low(1366x768), high(1600x900) for laptop). 

 

to ssdt injection user,

don't worry values in your ssdt

example

you have 1 in "lol" ssdt

clover auto-detection features put 0 in "lol". 

 

first write 0 and next ssdt injection change 0 to 1.

 

result "lol" is 1.

 

add) * Improved the display part of -platform-id in RC2.

post-980913-0-86161500-1492168280_thumb.png 

 

after check other graphic card,  will upload source to slice.

 

thanks in advance

RC2.zip

Link to comment
Share on other sites

DEV prejudice ... because if I do nvram -p out kext dev mode = 1and  rootless = o?( not obsolete?) I also tried to erase the nvram, but on reboot comes out again ..I am using 4049 clover

 

Hellò Api how are u ?

Sergey how are u ?

Link to comment
Share on other sites

@sherlocks -  

happy to help/test. but i dont understand your test here. what are you looking for? how can i help?

 

i have SkyLake HD 520 x1916 

but i use SSDT not clover injection - except when i need to set GFX fakeID  0x12345678 to use generic graphic on Installers.

is this point here to eliminate extra SSDT for GFX?

 

i also have HD3000 and HD4000.

  • Like 1
Link to comment
Share on other sites

 

i have 2 things in config.plist for GFX - so remove these 2 things? what about SSDT-IGPU?

<key>Graphics</key>
<dict>
  <key>Inject</key>
  <dict>
   <key>Intel</key>
   <false/>
  </dict>
  <key>ig-platform-id</key>
  <string>0x19160000</string>
</dict>

<key>AddProperties</key>
  <array>
   <dict>
    <key>Device</key>
    <string>IntelGFX</string>
    <key>Key</key>
    <string>AAPL,GfxYTile</string>
    <key>Value</key>
    <data>
    AQAAAA==
    </data>
   </dict>
  </array>

 

don't need both  

<key>ig-platform-id</key>

<string>0x19160000</string>

<key>AddProperties</key> all values

 

clover now support natively both In RC2 version

 

you can see that clover detect your config status(fakeid, device-id) dynamically and make boot without any configuration of hd 520 etc.

Link to comment
Share on other sites


26:904  0:000  Intel HD Graphics 520 [8086:1916] :: PciRoot(0x0)\Pci(0x2,0x0)
26:904  0:000    AAPL00,DualLink = 1
26:904  0:000    FakeID Intel GFX: not set
26:904  0:000    ig-platform-id: not set
26:904  0:000    Beginning Intel GFX auto-detection with ACPI injection
26:904  0:000    Found FakeID Intel GFX = 0x19168086
26:904  0:000    Found ig-platform-id = 0x19160000
26:904  0:000  Intel GFX revision  = 0x7
26:904  0:000   RCBA disabled; cannot use it
26:904  0:000   setting specified layout-id=28 (0x1C)
26:904  0:000  stringlength = 1948
26:904  0:000  CurrentMode: Width=1920 Height=1080
26:904  0:000  Beginning FSInjection
Link to comment
Share on other sites

 

what about SSDT? all this is defined in the SSDT? should we remove that? 

26:904  0:000  Intel HD Graphics 520 [8086:1916] :: PciRoot(0x0)\Pci(0x2,0x0)
26:904  0:000    AAPL00,DualLink = 1
26:904  0:000    FakeID Intel GFX: not set
26:904  0:000    ig-platform-id: not set
26:904  0:000    Beginning Intel GFX auto-detection with ACPI injection
26:904  0:000    Found FakeID Intel GFX = 0x19168086
26:904  0:000    Found ig-platform-id = 0x19160000
26:904  0:000  Intel GFX revision  = 0x7
26:904  0:000   RCBA disabled; cannot use it
26:904  0:000   setting specified layout-id=28 (0x1C)
26:904  0:000  stringlength = 1948
26:904  0:000  CurrentMode: Width=1920 Height=1080
26:904  0:000  Beginning FSInjection

 

success? please your other machine with clover log

 

thanks in advance.

Link to comment
Share on other sites

Success on HD 520 - works with LCD and External monitor no problem.

- in config.plist, I removed entire Graphics section,

- in config.plist, I removed Add Property for GfxYTile

- and removed SSDT-IGPU.

  • Like 1
Link to comment
Share on other sites

Success on HD 520 - works with LCD and External monitor no problem.

- in config.plist, I removed entire Graphics section,

- in config.plist, I removed Add Property for GfxYTile

- and removed SSDT-IGPU.

 

yes. basic user can boot without any options. just check inject intel.

 

also everytime if you want, you can injection SSDT or Addpropery options. 

 

please test hd 3000, hd 4000. i wonder clover log in these graphics

Link to comment
Share on other sites

On my laptop I removed my SSDT injection for the HD4000 and rebooted with the RC2 files. From second stage onwards my screen is split into 4 as DualLink is enabled.

14:245  0:000  Intel HD Graphics 4000 [8086:0166] :: PciRoot(0x0)\Pci(0x2,0x0)
14:245  0:000    Use override Custom EDID
14:245  0:000    AAPL00,DualLink = 1
14:245  0:000    FakeID Intel GFX: not set
14:245  0:000    ig-platform-id: not set
14:245  0:000    Beginning Intel GFX auto-detection with ACPI injection
14:245  0:000    Found FakeID Intel GFX = 0x01668086
14:245  0:000    Not set ig-platform-id. Use auto-detection or ACPI injection
14:245  0:000    Found ig-platform-id = 0x01660004
14:245  0:000  Intel GFX revision  = 0x9
14:245  0:000   detected codec: 10EC:0269
14:245  0:000  stringlength = 3142
14:245  0:000  CurrentMode: Width=1366 Height=768

Also, I have been injecting 0x01660003 via SSDT while Clover is using 0x01660004.

  • Like 1
Link to comment
Share on other sites

On my laptop I removed my SSDT injection for the HD4000 and rebooted with the RC2 files. From second stage onwards my screen is split into 4 as DualLink is enabled.

14:245  0:000  Intel HD Graphics 4000 [8086:0166] :: PciRoot(0x0)\Pci(0x2,0x0)14:245  0:000    Use override Custom EDID14:245  0:000    AAPL00,DualLink = 114:245  0:000    FakeID Intel GFX: not set14:245  0:000    ig-platform-id: not set14:245  0:000    Beginning Intel GFX auto-detection with ACPI injection14:245  0:000    Found FakeID Intel GFX = 0x0166808614:245  0:000    Not set ig-platform-id. Use auto-detection or ACPI injection14:245  0:000    Found ig-platform-id = 0x0166000414:245  0:000  Intel GFX revision  = 0x914:245  0:000   detected codec: 10EC:026914:245  0:000  stringlength = 314214:245  0:000  CurrentMode: Width=1366 Height=768
Also, I have been injecting 0x01660003 via SSDT while Clover is using 0x01660004.
Set duallink 0 in config.plist. it will be work

 

What is your laptop resolution?

 

Okay i will make dynamic recognize dualink according to resolution.

 

나의 LG-F410S 의 Tapatalk에서 보냄

  • Like 2
Link to comment
Share on other sites

×
×
  • Create New...