Sergey:
So, X64 Thunk16.nasm could pass XCLANG build and ld. What about IA32
Thunk16.nasm code?
Thanks
Liming
From: Sergey Isakov [mailto:isakov...@bk.ru]
Sent: Friday, August 29, 2014 2:15 AM
To: edk2-devel@lists.sourceforge.net
Subject: Re: [edk2] [PATCH 0/8] NASM: Object files; Thunk16; Require for
non-MSVC
That works!
Now I have next same problem (offtop?)
UefiCpuPkg/Library/CpuExceptionHandlerLib/X64/ExceptionHandlerAsm.S
The same using of absolute addresses of CommonInterruptEntry
———
#
Exception0Handle:
.byte 0x6a # push #VectorNum
.byte 0
pushq %rax
.byte 0x48, 0xB8
.quad ASM_PFX(CommonInterruptEntry)
jmp *%rax
————
It looks like more codes should be revised. And I vote to rewrite this part to
nasm too.
On 28 авг. 2014 г., at 21:10, Sergey Isakov
<isakov...@bk.ru<mailto:isakov...@bk.ru>> wrote:
I understand. The value of 16GdtrBase is patched by C-codes in procedure
AsmPrepareThunk16()
So we may set
———
_16GdtrBase:
DQ 0
-----
On 28 авг. 2014 г., at 15:48, Andrew Fish
<af...@apple.com<mailto:af...@apple.com>> wrote:
On Aug 28, 2014, at 4:30 AM, Sergey Isakov
<isakov...@bk.ru<mailto:isakov...@bk.ru>> wrote:
But what can you do with absolute addressing to _NullSeg?
-------
_EntryPoint:
DD _ToUserCode - ASM_PFX(m16Start)
DW CODE16
_16Gdtr:
DW GDT_SIZE - 1
_16GdtrBase:
DQ _NullSeg
It looks like Thunk16.S patches the value.
_EntryPoint: .long ASM_PFX(ToUserCode) - ASM_PFX(m16Start)
.word CODE16
_16Gdtr: .word GDT_SIZE - 1
_16GdtrBase: .quad 0
_16Idtr: .word 0x3ff
.long 0
Thanks,
Andrew Fish
-------
On 28.08.2014, at 15:16, Andrew Fish wrote:
On Aug 28, 2014, at 12:15 AM, Sergey Isakov
<isakov...@bk.ru<mailto:isakov...@bk.ru>> wrote:
Actually I just noticed you are getting an illegal text-relocation error. This
is usually caused by an absolute relocation, and clang wants it to be IP
relative. Basically Xcode/clang only supports a subset of the relocations that
gcc uses for X64.
If I correctly understand you, compilation for Thunk16 is not possible in
XCLANG toolset because of idea of this codes?
No, but this toolchain puts extra restrictions on how things can be coded.
For example Thunk16.asm contains absolute addressing
lea r8, @RetFromRealMode
But in Thunk16.S this is converted to IP relative addressing to support clang.
lea L_RetFromRealMode(%rip), %r8
Some of the .set address math in Thunk16.S is also working around addressing
mode limitations, as it removes relocations from the code.
Thanks,
Andrew Fish
------------------------------------------------------------------------------
Slashdot TV.
Video for Nerds. Stuff that matters.
http://tv.slashdot.org/_______________________________________________
edk2-devel mailing list
edk2-devel@lists.sourceforge.net<mailto:edk2-devel@lists.sourceforge.net>
https://lists.sourceforge.net/lists/listinfo/edk2-devel
------------------------------------------------------------------------------
Slashdot TV.
Video for Nerds. Stuff that matters.
http://tv.slashdot.org/_______________________________________________
edk2-devel mailing list
edk2-devel@lists.sourceforge.net<mailto:edk2-devel@lists.sourceforge.net>
https://lists.sourceforge.net/lists/listinfo/edk2-devel
------------------------------------------------------------------------------
Slashdot TV.
Video for Nerds. Stuff that matters.
http://tv.slashdot.org/_______________________________________________
edk2-devel mailing list
edk2-devel@lists.sourceforge.net<mailto:edk2-devel@lists.sourceforge.net>
https://lists.sourceforge.net/lists/listinfo/edk2-devel
------------------------------------------------------------------------------
Slashdot TV.
Video for Nerds. Stuff that matters.
http://tv.slashdot.org/_______________________________________________
edk2-devel mailing list
edk2-devel@lists.sourceforge.net<mailto:edk2-devel@lists.sourceforge.net>
https://lists.sourceforge.net/lists/listinfo/edk2-devel
------------------------------------------------------------------------------
Slashdot TV.
Video for Nerds. Stuff that matters.
http://tv.slashdot.org/
_______________________________________________
edk2-devel mailing list
edk2-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/edk2-devel
