Jump to content

Clover General discussion

ErmaC

By ErmaC,
in Clover

 Share
30960 posts in this topic

Recommended Posts

Sherlocks
Sherlocks
Posted
Posted

IntelGraphicsDVMTFixup does not work with 10.12.4 installer.Still panic.

Just placing it and lilu in kexts/10.12,is it wrong?

Update:Thing does not work in installer is lilu.

Lilu.kext 1.1.1

IntelGraphicsDVMTFixup 1.1.3

 

Tested with other user. Redownload kext and Try test again

 

나의 LG-F410S 의 Tapatalk에서 보냄

  • Like 1
Sherlocks
Sherlocks
Posted
Posted

Thanks, here you go. This is with using InjectNvidia=true.

 

i checked all.

in ioreg, vram.totalsize shown 0x00002000(8192 MB (8 GB) )

 

log too. also i checked clover code. there is no problem. maybe need to change something in config.plist

 

how about this?

Inject Nvidia -> uncheck

NvidiaGeneric -> uncheck

 

NvidiaWeb -> only check

gills83
gills83
Posted
Posted

Yeah... I got that from when you said it before. These attachments do not help. Need debug.log from failed clover boot to determine if it is indeed because CPU detection is failing or if it's for another reason.

 

euh ?? 

 

There can not be a debug.log clover, it does not start.

Ritz Carltn
Ritz Carltn
Posted
Posted

Should. Set config.plist -> CPU -> QPI to 100 then test if it works.

 

http://www.insanelymac.com/forum/topic/284656-clover-general-discussion/?p=2409594

 

It still did not fix it.

I attached DarwinDumper.

DarwinDumper_3.0.3_14.05_09.26.00_MacPro6,1_Clover_X64_4061_Yosemite_14F2315_admin.zip

apianti
apianti
Posted
Posted

euh ?? 

 

There can not be a debug.log clover, it does not start.

 

You enabled Boot/Debug and there was no debug.log? Or you just assume?

 

It still did not fix it.

I attached DarwinDumper.

 

I doubt that's the setting you want to change, mostly superficial. You probably want CPU/BusSpeedkHz.

  • Like 1
Zenith432
Zenith432
Posted
Posted

@apianti:

Why is this needed (r4071)?

 

--- a/Library/OpensslLib/openssl-1.0.1e/OpensslLib.inf
+++ b/Library/OpensslLib/openssl-1.0.1e/OpensslLib.inf
@@ -641,7 +641,7 @@
    INTEL:*_*_X64_CC_FLAGS                 = -U_WIN32 -U_WIN64 -U_MSC_VER -U__ICC $(OPENSSL_FLAGS) $(OPENSSL_EXFLAGS) /w -DSIXTY_FOUR_BIT
    INTEL:*_*_IPF_CC_FLAGS                 = -U_WIN32 -U_WIN64 -U_MSC_VER -U__ICC $(OPENSSL_FLAGS) $(OPENSSL_EXFLAGS) /w -DSIXTY_FOUR_BIT
    GCC:*_*_IA32_CC_FLAGS                  = -U_WIN32 -U_WIN64 $(OPENSSL_FLAGS) $(OPENSSL_EXFLAGS) -w -DTHIRTY_TWO_BIT
-   GCC:*_*_X64_CC_FLAGS                   = -U_WIN32 -U_WIN64 $(OPENSSL_FLAGS) $(OPENSSL_EXFLAGS) -w -DSIXTY_FOUR_BIT
+   GCC:*_*_X64_CC_FLAGS                   = -U_WIN32 -U_WIN64 $(OPENSSL_FLAGS) $(OPENSSL_EXFLAGS) -w -DSIXTY_FOUR_BIT -DNO_MSABI_VA_FUNCS
    GCC:*_*_IPF_CC_FLAGS                   = -U_WIN32 -U_WIN64 $(OPENSSL_FLAGS) $(OPENSSL_EXFLAGS) -w -DSIXTY_FOUR_BIT
    RVCT:*_*_IA32_CC_FLAGS                 = $(OPENSSL_FLAGS) $(OPENSSL_EXFLAGS) --library_interface=aeabi_clib99 --fpu=vfpv3 -DTHIRTY_TWO_BIT
    RVCT:*_*_X64_CC_FLAGS                  = $(OPENSSL_FLAGS) $(OPENSSL_EXFLAGS) --library_interface=aeabi_clib99 --fpu=vfpv3 -DSIXTY_FOUR_BIT

 

This problem was supposed to be fixed in r4050 (see diff for that rev under /Library/OpensslLib). Does the code not compile without this? Or generate faulty machine code?

apianti
apianti
Posted
Posted

It's probably not, you can revert it if you want. It was because someone said that it was needed over in the build_clover.command thread, but I think that's actually a problem with that script.

 

EDIT: Fixed the script name from clover_build.command to build_clover.command.

Philip Petev
Philip Petev
Posted
Posted

The Build_Clover.command script has nothing to do with the compiler flags/settings (except choosing the build toolkit: GCCxx or XCODE5/8). Besides, Build_Clover.command only passes options, that are already present in ebuild.sh, based on the user input (selection).

  • Like 1
apianti
apianti
Posted
Posted

The Build_Clover.command script has nothing to do with the compiler flags/settings (except choosing the build toolkit: GCCxx or XCODE5/8). Besides, Build_Clover.command only passes options, that are already present in ebuild.sh, based on the user input (selection).

 

Nah, I know what it does. I didn't mean there was a bug in the script, I meant that it was a problem with using the script.

 

After some more digging, I see that this is actually because of CLANG38 and GCC44 compilation that is included in the tools_def.txt(.template). So the reason appears to be the tool definitions. I'm not sure why it's quoted... or only used for these two versions?

DEFINE GCC44_X64_CC_FLAGS            = DEF(GCC44_ALL_CC_FLAGS) -m64 -fno-stack-protector "-DEFIAPI=__attribute__((ms_abi))" -maccumulate-outgoing-args -mno-red-zone -Wno-address -mcmodel=small -fpie -fno-asynchronous-unwind-tables

DEBUG_CLANG38_X64_CC_FLAGS         = DEF(CLANG38_ALL_CC_FLAGS) -m64 "-DEFIAPI=__attribute__((ms_abi))" -mno-red-zone -mcmodel=small -fpie -Oz -flto DEF(CLANG38_X64_TARGET) -g
RELEASE_CLANG38_X64_CC_FLAGS       = DEF(CLANG38_ALL_CC_FLAGS) -m64 "-DEFIAPI=__attribute__((ms_abi))" -mno-red-zone -mcmodel=small -fpie -Oz -flto DEF(CLANG38_X64_TARGET)

EDIT: These should be escaped differently, like

-D'EFIAPI=__attribute__((ms_abi))'

EDIT2: Or maybe not at all? I think that the EDK2 build system might automatically escape parentheses with a backslash...

  • Like 1
gills83
gills83
Posted
Posted

 

You enabled Boot/Debug and there was no debug.log? Or you just assume?

 

 

No, there was nothing, I will follow your advice to get better information.
 
thank you

 

 

@apianti,

 

no files debug.log 

 

thanks

 

"Debug

<key>Debug</key><false/>

Formerly known as "Log" before revision 3064

If you are unable to boot into the Clover you can use this setting to produces a debug output to /EFI/CLOVER/misc/debug.log. This log then can be assessed to find out what the problem is. !WARNING turning on this log will dramatically increase loading time as it has to write the log to the disk as it goes. So please be patient and give plenty of time before reseting your computer."

 

Debug

<key>Debug</key><false/>

Formerly known as "Log" before revision 3064

If you are unable to boot into the Clover you can use this setting to produces a debug output to /EFI/CLOVER/misc/debug.log. This log then can be assessed to find out what the problem is. !WARNING turning on this log will dramatically increase loading time as it has to write the log to the disk as it goes. So please be patient and give plenty of time before reseting your computer.

Zenith432
Zenith432
Posted
Posted

@apianti:

 

FYI

  • Variadic EFIAPI should use VA_LIST macros.
  • Variadic non-EFIAPI should use __builtin_va_list (and corresponding builtins for MS/Intel compilers.)
  • This is set in include file.  I reverted r4071.  Don't use NO_MSABI_VA_FUNC as everything is set up right in MdePkg/Include/Base.h
  • In, r4072, the settings for XCODE were missing -DSIXTY_FOUR_BIT, etc.
  • It is not always necessary to make XCODE, since XCODE uses GCC if specific XCODE is not available.
  • No need for -mmmx, -msse as x64 always has this, and even for ia32 (which is deprecated), minimum target architecture supports these.
  • == vs = in INF means "replace" instead of "append".  Whoever used == (Andy V??) must have been thinking of something.
  • So I reverted r4072 as well.
  • OTOH there was a duplicate symbol _fltused when compiling -D ENABLE_SECURE_BOOT, so eliminated dup with #ifndef.
 Share
Go to topic listing
×
×
  • Create New...