Jump to content

Clover Problems and Solutions

ErmaC

By ErmaC,
in Clover Development

 Share
3205 posts in this topic

Recommended Posts

Sherlocks
Sherlocks
Posted
Posted

Because of wrong shifting Cho, Jong, Joong in the char square.

 

embedded theme looks like here

attachicon.gifScreen Shot 2016-09-12 at 16.10.10.png

 

I think you should just draw new korean font.

Okay i try and report

 

나의 LG-F410S 의 Tapatalk에서 보냄

Sherlocks
Sherlocks
Posted
Posted

It will be good also to create FontKorean for dark themes.

attachicon.gifScreen Shot 2016-09-12 at 16.31.00.png

 

r3735 is good. if i dont have korean font png, clover use english font.

 

0:301  0:000  Loading font from ThemeDir: Error

0:301  0:000  font image not loaded, use english
0:301  0:000  load font image type 1
0:301  0:000  Using embedded font
0:302  0:000  Font loaded: ImageWidth=176 ImageHeight=288
0:302  0:000  Font 1 prepared WxH=11x18 CharWidth=9

 

i will try to optimized korean font in GUI. and report 

 

thank in advance. Slice. have a nice day

Sherlocks
Sherlocks
Posted
Posted

i report one.

 

hibernate wake is not working.

 

here is screenshot.

select option

post-980913-0-80877600-1474099486_thumb.png

 

and hold this screen.

post-980913-0-64647700-1474099470_thumb.jpeg

 

 

add. i'm working korean font optimizing now. but some still have issue. i will report more detail.

 

thanks in advance.

Riley Freeman
Riley Freeman
Posted
Posted

There seems to be a bug with the EDID Fake ID injection. The Apple DisplayVendorID is supposed to be 0x0610 but it's showing up in the IOReg as 0x1006.

 

In the Clover Graphics Injector submenu it's displayed correctly as 0x0610.

 

MBP9,1 IOReg:

 

 

    | | | |   | |   |       "DisplayProductID" = 40099

    | | | |   | |   |       "IODisplayPrefsKey" = "Alias:0/AppleBacklightDisplay-610-9ca3"

    | | | |   | |   |       "DisplayVendorID" = 1552

 

 

 

My IOReg:

 

 

    | |   | | |       "DisplayProductID" = 40099

    | |   | | |       "IODisplayPrefsKey" = "IOService:/AppleACPIPlatformExpert/PCI0@0/AppleACPIPCI/IGPU@2/AppleIntelFramebuffer@0/display0/AppleBacklightDisplay-1006-9ca3"

    | |   | | |       "DisplayVendorID" = 4102

 

 

mhaeuser
mhaeuser
Posted
Posted

There seems to be a bug with the EDID Fake ID injection. The Apple DisplayVendorID is supposed to be 0x0610 but it's showing up in the IOReg as 0x1006.

 

In the Clover Graphics Injector submenu it's displayed correctly as 0x0610.

 

MBP9,1 IOReg:

 

 

| | | | | | | "DisplayProductID" = 40099

| | | | | | | "IODisplayPrefsKey" = "Alias:0/AppleBacklightDisplay-610-9ca3"

| | | | | | | "DisplayVendorID" = 1552

 

 

 

My IOReg:

 

 

| | | | | "DisplayProductID" = 40099

| | | | | "IODisplayPrefsKey" = "IOService:/AppleACPIPlatformExpert/PCI0@0/AppleACPIPCI/IGPU@2/AppleIntelFramebuffer@0/display0/AppleBacklightDisplay-1006-9ca3"

| | | | | "DisplayVendorID" = 4102

 

 

Little vs big endian, just switch the two bytes for now.

LockDown
LockDown
Posted
Posted

 

Enabling Nvidia web in clover GUI works, but not in config.plist. GUI doesn't detect that it is enabled in config

	<key>SystemParameters</key>
	<dict>
		<key>NvidiaWeb</key>
		<true/>
	</dict>

+1 running on ML it doesn't work. I cannot test Maverick orYosemite as i don't have it.. Works well on EL Capitan.

Maybe because ML Web driver doesn't have "WEB" in its filename? it works in GUI though, so I'm guessing :D

Slice
Slice
Posted
Posted

There seems to be a bug with the EDID Fake ID injection. The Apple DisplayVendorID is supposed to be 0x0610 but it's showing up in the IOReg as 0x1006.

 

In the Clover Graphics Injector submenu it's displayed correctly as 0x0610.

 

MBP9,1 IOReg:

 

 

    | | | |   | |   |       "DisplayProductID" = 40099

    | | | |   | |   |       "IODisplayPrefsKey" = "Alias:0/AppleBacklightDisplay-610-9ca3"

    | | | |   | |   |       "DisplayVendorID" = 1552

 

 

 

My IOReg:

 

 

    | |   | | |       "DisplayProductID" = 40099

    | |   | | |       "IODisplayPrefsKey" = "IOService:/AppleACPIPlatformExpert/PCI0@0/AppleACPIPCI/IGPU@2/AppleIntelFramebuffer@0/display0/AppleBacklightDisplay-1006-9ca3"

    | |   | | |       "DisplayVendorID" = 4102

 

 

MacBookPro5,4 ioreg

    | |   |     | |   "EDID" =
So byte 8 = 0x06, byte 9 = 0x10
word = 0x1006 - LittleEndian low byte 0x06 high byte 0x10
 
    | |   |     |   |       "DisplayProductID" = 40099
    | |   |     |   |       "DisplayVendorID" = 1552
 
 What did you set in your config?

 

Enabling Nvidia web in clover GUI works, but not in config.plist. GUI doesn't detect that it is enabled in config

	<key>SystemParameters</key>
	<dict>
		<key>NvidiaWeb</key>
		<true/>
	</dict>

It appears to work only in Sierra.

For other systems use 

	<key>Boot</key>
	<dict>
		<key>Arguments</key>
		<string>nvda_drv=1</string>

But I don't think it should be set statically in config.plist.

Let macOS set this argument by using pref panel.

  • Like 1
Mork vom Ork
Mork vom Ork
Posted
Posted

build CLOVER Rev. 3763 with XCode8 and Mickeys BuildCloverScript V4: everything works ok:

0:100  0:100  MemLog inited, TSC freq: 4008550260
0:100  0:000  MemLog was calibrated without ACPI PM Timer: ACPI I/O space is not enabled.
0:100  0:000  
0:100  0:000  Now is 21.9.2016,  9:54:51 (GMT)
0:100  0:000  Starting Clover revision: 3763 on American Megatrends EFI
0:100  0:000  Build with: [Args: -x64 -D NO_GRUB_DRIVERS_EMBEDDED -D CHECK_FLAGS -t XCODE5 | -D NO_GRUB_DRIVERS_EMBEDDED -D CHECK_FLAGS -D USE_LOW_EBDA -a X64 -b RELEASE -t XCODE5 -n 9 | OS: 10.12 | XCODE: 8.0]
0:100  0:000  SelfDevicePath=PciRoot(0x0)\Pci(0x1,0x2)\Pci(0x0,0x0)\Sata(0x0,0xFFFF,0x0)\HD(1,GPT,AB53964E-4A3A-454A-A09B-6E00B6117415,0x28,0x64000) @634F7198
0:100  0:000  SelfDirPath = \EFI\BOOT
Riley Freeman
Riley Freeman
Posted
Posted

@Slice: I set 1552 in config for Vendor.

 

I did some more digging and now I see that with EDID the Vendor ID is big endian, while the Product ID is little endian. I assumed wrongly they would both be the same.

 

So there's nothing wrong, except with me :)

  • Like 1
Micky1979
Micky1979
Posted
Posted

Rev 3761 at sf.net is bad, use boot file from other revisions

 

attachicon.gifScreen Shot 2016-09-21 at 12.12.50.png

This file is good. rev 3763 compiled by Xcode4

Please someone compile and test with Xcode8.

attachicon.gifboot6.zip

Not sure what happened, but just some seconds ago I do an svn up: edk2 has received some new files but the fact is that is the same revision 22628 as since last week. How is that possible??

 

EDIT

the same in Linux:

BaseTools:
checking..available, continuing..
Updating '.':

Fetching external item into 'Bin/Win32':
U    Bin/Win32/GenFds.exe
U    Bin/Win32/BPDG.exe
U    Bin/Win32/Readme.txt
U    Bin/Win32/build.exe
U    Bin/Win32/GenDepex.exe
U    Bin/Win32/Trim.exe
U    Bin/Win32/PatchPcdValue.exe
U    Bin/Win32/TargetTool.exe
U    Bin/Win32/GenPatchPcdTable.exe
Updated external to revision 190.

maybe does nothing for us?

Slice
Slice
Posted
Posted

 

build CLOVER Rev. 3763 with XCode8 and Mickeys BuildCloverScript V4: everything works ok:

0:100  0:100  MemLog inited, TSC freq: 4008550260
0:100  0:000  MemLog was calibrated without ACPI PM Timer: ACPI I/O space is not enabled.
0:100  0:000  
0:100  0:000  Now is 21.9.2016,  9:54:51 (GMT)
0:100  0:000  Starting Clover revision: 3763 on American Megatrends EFI
0:100  0:000  Build with: [Args: -x64 -D NO_GRUB_DRIVERS_EMBEDDED -D CHECK_FLAGS -t XCODE5 | -D NO_GRUB_DRIVERS_EMBEDDED -D CHECK_FLAGS -D USE_LOW_EBDA -a X64 -b RELEASE -t XCODE5 -n 9 | OS: 10.12 | XCODE: 8.0]
0:100  0:000  SelfDevicePath=PciRoot(0x0)\Pci(0x1,0x2)\Pci(0x0,0x0)\Sata(0x0,0xFFFF,0x0)\HD(1,GPT,AB53964E-4A3A-454A-A09B-6E00B6117415,0x28,0x64000) @634F7198
0:100  0:000  SelfDirPath = \EFI\BOOT

No problem with UEFI boot.

The problem is legacy boot file (boot6 == CloverEFI SATA) rev.3761 uploaded to sf.net is bad leading to red panic.

Slice
Slice
Posted
Posted

Have you warnings with RIP-relative addressing with optimization disabled?

(I've read that this is a possible cause of cpu exceptions)

/Users/slice/src/edk2/Build/Clover/RELEASE_XCODE5/X64/Clover/OsxAptioFixDrv/OsxAptioFix2Drv/OUTPUT/X64/AsmFuncsX64.iii:221: warning: absolute address can not be RIP-relative
/Users/slice/src/edk2/Build/Clover/RELEASE_XCODE5/X64/Clover/OsxAptioFixDrv/OsxAptioFix2Drv/OUTPUT/X64/AsmFuncsX64.iii:225: warning: absolute address can not be RIP-relative
/Users/slice/src/edk2/Build/Clover/RELEASE_XCODE5/X64/Clover/OsxAptioFixDrv/OsxAptioFix2Drv/OUTPUT/X64/AsmFuncsX64.iii:226: warning: absolute address can not be RIP-relative
[NASM] AsmFuncsX64
/Users/slice/src/edk2/Build/Clover/RELEASE_XCODE5/X64/Clover/OsxAptioFixDrv/OsxAptioFixDrv/OUTPUT/X64/AsmFuncsX64.iii:221: warning: absolute address can not be RIP-relative
/Users/slice/src/edk2/Build/Clover/RELEASE_XCODE5/X64/Clover/OsxAptioFixDrv/OsxAptioFixDrv/OUTPUT/X64/AsmFuncsX64.iii:225: warning: absolute address can not be RIP-relative
/Users/slice/src/edk2/Build/Clover/RELEASE_XCODE5/X64/Clover/OsxAptioFixDrv/OsxAptioFixDrv/OUTPUT/X64/AsmFuncsX64.iii:226: warning: absolute address can not be RIP-relative

I see numerous messages that OsxAptioFix cause reboot. I am using OsxAptioFix2Drv and it works. I will think about.

  • Like 1
Slice
Slice
Posted
Posted

This warning appears in NASM 2.12.02 and was absent in 2.12.01 so why dmazar didn't see it.

The subroutine is very tricky making self-modification and NASM warning us about this.

Anyway I think all works as expected.

LockDown
LockDown
Posted
Posted

about the NvidiaWeb in SystemParameters

It appears to work only in Sierra.

For other systems use

 

<key>Boot</key>	<dict>		<key>Arguments</key>		<string>nvda_drv=1</string>
But I don't think it should be set statically in config.plist.

Let macOS set this argument by using pref panel.

wouldnt be nice if it can work with lower builds too?
 Share
Go to topic listing
×
×
  • Create New...