2014-05-16 20:37 GMT+02:00 Laszlo Ersek <ler...@redhat.com>:
> 2. The error is in Ludovic's program; I can reproduce it and also fix
> it. Quote:
>
> EFI_STATUS
> EFIAPI
> efi_main (EFI_HANDLE ImageHandle, EFI_SYSTEM_TABLE *SystemTable)
> {
> ...
>
> The problem is "EFIAPI". You *must not* add EFIAPI to efi_main() *in
> gnu-efi*. The compiling/linking magic that gnu-efi does for applications
> automatically takes care of it, and it is thrown off if you specify
> EFIAPI manually for efi_main(). Don't do that.
Not all my sample codes were using EFIAPI.
Examples from [1] are wrong.
Examples from [2] are correct.
Roderick, maybe you should fix your examples at [1].
> This is in fact documented. See README.gnuefi:
>
>> The GNU EFI build environment allows to write EFI applications as
>> described in Intel's EFI documentation, except for two differences:
>>
>> - The EFI application's entry point is always called "efi_main". The
>> declaration of this routine is:
>>
>> EFI_STATUS efi_main (EFI_HANDLE image, EFI_SYSTEM_TABLE *systab);
>>
>> [...]
>
> No EFIAPI. If you check the rest of the applications in apps/ -- no
> EFIAPI on efi_main(), ever.
>
> After removing EFIAPI from Ludovic's app, it works as expected for me
> (invoked from UEFI shell on fresh OVMF).
>
> One strange point is that the D_ERROR output (via DEBUG()) seems to
> overwrite the Print(L"c\n") output on the serial console, except when it
> happens at the bottom of the screen. Not sure why that happens, but it
> doesn't crash.
My debug1.c code is now:
#include <efi.h>
#include <efilib.h>
EFI_STATUS
efi_main (EFI_HANDLE ImageHandle, EFI_SYSTEM_TABLE *SystemTable)
{
InitializeLib(ImageHandle, SystemTable);
Print(L"a\n");
DEBUG(( D_INFO, (CHAR8 *)"D_INFO\r\n"));
Print(L"b\n");
DEBUG(( D_ERROR, (CHAR8 *)"D_ERROR\r\n"));
Print(L"c\n");
return EFI_SUCCESS;
}
On the serial console I get:
FS0:\> debug1.efi
a
b
!!!! X64 Exception Type - 000000000000000E CPU Apic ID - 00000000 !!!!
RIP - 0000000006BA58DE, CS - 0000000000000028, RFLAGS - 0000000000000287
ExceptionData - 0000000000000000
RAX - FFFFFFFFFFFFFFF0, RCX - 0000000000000000, RDX - 0000000000000007
RBX - 0000000080000000, RSP - 0000000007F6A4F0, RBP - 0000000007F6A530
RSI - 000000000000000E, RDI - 0000000006BAA790
R8 - 0000000000000000, R9 - 0000000007F69EA7, R10 - 00000000069A169D
R11 - 0000000000000040, R12 - 0000000000000005, R13 - 0000000000000000
R14 - 0000000000000000, R15 - 0000000000000000
DS - 0000000000000008, ES - 0000000000000008, FS - 0000000000000008
GS - 0000000000000008, SS - 0000000000000008
CR0 - 0000000080000033, CR2 - FFFFFFFFFFFFFFF0, CR3 - 0000000007F09000
CR4 - 0000000000000668, CR8 - 0000000000000000
DR0 - 0000000000000000, DR1 - 0000000000000000, DR2 - 0000000000000000
DR3 - 0000000000000000, DR6 - 00000000FFFF0FF0, DR7 - 0000000000000400
GDTR - 0000000007EF2D98 000000000000003F, LDTR - 0000000000000000
IDTR - 0000000007C1E018 0000000000000FFF, TR - 0000000000000000
FXSAVE_STATE - 0000000007F6A150
!!!! Find PE image
/home/lroussea/Documents/UEFI/edk2/Build/OvmfX64/DEBUG_GCC47/X64/MdeModulePkg/Universal/Console/ConSplitterDxe/ConSplitterDxe/DEBUG/ConSplitterDxe.dll
(ImageBase=0000000006B9F000, EntryPoint=0000000006B9F2AF) !!!!
I get this crash before the debug message is displayed. This is with
OVMF I rebuild myself using edk2 SVN revision 15536
If I use OVMF from OVMF-X64-r15214/OVMF.fd I get on the serial console:
FS0:\> debug1.efi
a
b
D_ERROR
And on the terminal used to start qemu:
qemu: fatal: Trying to execute code outside RAM or ROM at 0x00000000000b0000
RAX=0000000000000000 RBX=0000000080000000 RCX=0000000000000000
RDX=0000000000000018
RSI=0000000000000005 RDI=0000000007dd0c30 RBP=0000000007f94560
RSP=0000000007f94518
R8 =0000000000000000 R9 =0000000007f93f5c R10=00000000000003f8
R11=0000000000000040
R12=0000000000000005 R13=0000000000000000 R14=0000000000000000
R15=0000000000000000
RIP=00000000000affae RFL=00000202 [-------] CPL=0 II=0 A20=1 SMM=0 HLT=0
ES =0008 0000000000000000 ffffffff 00cf9300 DPL=0 DS [-WA]
CS =0028 0000000000000000 ffffffff 00af9b00 DPL=0 CS64 [-RA]
SS =0008 0000000000000000 ffffffff 00cf9300 DPL=0 DS [-WA]
DS =0008 0000000000000000 ffffffff 00cf9300 DPL=0 DS [-WA]
FS =0008 0000000000000000 ffffffff 00cf9300 DPL=0 DS [-WA]
GS =0008 0000000000000000 ffffffff 00cf9300 DPL=0 DS [-WA]
LDT=0000 0000000000000000 0000ffff 00008200 DPL=0 LDT
TR =0000 0000000000000000 0000ffff 00008b00 DPL=0 TSS64-busy
GDT= 0000000007f1ed98 0000003f
IDT= 000000000747f018 00000fff
CR0=80000033 CR2=0000000000000000 CR3=0000000007f33000 CR4=00000668
DR0=0000000000000000 DR1=0000000000000000 DR2=0000000000000000
DR3=0000000000000000
DR6=00000000ffff0ff0 DR7=0000000000000400
CCS=0000000000000000 CCD=0000000000000000 CCO=ADDB
EFER=0000000000000500
FCW=037f FSW=0000 [ST=0] FTW=00 MXCSR=00001f80
FPR0=0000000000000000 0000 FPR1=0000000000000000 0000
FPR2=0000000000000000 0000 FPR3=0000000000000000 0000
FPR4=0000000000000000 0000 FPR5=0000000000000000 0000
FPR6=0000000000000000 0000 FPR7=0000000000000000 0000
XMM00=00000000000000000000000000000000 XMM01=00000000000000000000000000000000
XMM02=00000000000000000000000000000000 XMM03=00000000000000000000000000000000
XMM04=00000000000000000000000000000000 XMM05=00000000000000000000000000000000
XMM06=00000000000000000000000000000000 XMM07=00000000000000000000000000000000
XMM08=00000000000000000000000000000000 XMM09=00000000000000000000000000000000
XMM10=00000000000000000000000000000000 XMM11=00000000000000000000000000000000
XMM12=00000000000000000000000000000000 XMM13=00000000000000000000000000000000
XMM14=00000000000000000000000000000000 XMM15=00000000000000000000000000000000
So I think there is a problem between OVMF and gnu-efi for DEBUG() call.
Since I can use DEBUG() with a code built with edk2 I will continue
with that. I can do more testing and debug if needed. Just tell me.
Thanks a lot
[1] http://www.rodsbooks.com/efi-programming/hello.html
[2] http://mjg59.dreamwidth.org/18773.html
--
Dr. Ludovic Rousseau
------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
edk2-devel mailing list
edk2-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/edk2-devel
