Bernhard Kaindl wrote:
> Hi,
>
> I found that on kvm-61 the cpuid in the guest was reported incorrectly
> when qemu-kvm was compiled with gcc-4.1 or 4.3.
>
> This resulted in linux-64bit not booting, complaining that it is not
> running on a 64-bit machine.
>
> Symptom: Unexpected behaviour after the assembly snippet.
>
> Solution: New assembly which is simpler and leaves optimizations to gcc,
> resulting in much shorter and maintainable code.
>
> Comments are welcome,
>
There were two reasons for the obfuscated code: first, qemu-kvm.c used
to be compiled with ebx assigned to a global register, which can't be
clobbered, and second not to cause excessive register pressure on i386.
The first reason is gone, but can you test the second by compiling on i386?
> Bernhard
>
> PS: Thanks a lot to Alex Graf for the fix.
>
> --- qemu/qemu-kvm-x86.c
> +++ qemu/qemu-kvm-x86.c
> @@ -428,35 +428,7 @@ static void host_cpuid(uint32_t function
> uint32_t vec[4];
>
> vec[0] = function;
> - asm volatile (
> -#ifdef __x86_64__
> - "sub $128, %%rsp \n\t" /* skip red zone */
> - "push %0; push %%rsi \n\t"
> - "push %%rax; push %%rbx; push %%rcx; push %%rdx \n\t"
> - "mov 8*5(%%rsp), %%rsi \n\t"
> - "mov (%%rsi), %%eax \n\t"
> - "cpuid \n\t"
> - "mov %%eax, (%%rsi) \n\t"
> - "mov %%ebx, 4(%%rsi) \n\t"
> - "mov %%ecx, 8(%%rsi) \n\t"
> - "mov %%edx, 12(%%rsi) \n\t"
> - "pop %%rdx; pop %%rcx; pop %%rbx; pop %%rax \n\t"
> - "pop %%rsi; pop %0 \n\t"
> - "add $128, %%rsp"
> -#else
> - "push %0; push %%esi \n\t"
> - "push %%eax; push %%ebx; push %%ecx; push %%edx \n\t"
> - "mov 4*5(%%esp), %%esi \n\t"
> - "mov (%%esi), %%eax \n\t"
> - "cpuid \n\t"
> - "mov %%eax, (%%esi) \n\t"
> - "mov %%ebx, 4(%%esi) \n\t"
> - "mov %%ecx, 8(%%esi) \n\t"
> - "mov %%edx, 12(%%esi) \n\t"
> - "pop %%edx; pop %%ecx; pop %%ebx; pop %%eax \n\t"
> - "pop %%esi; pop %0 \n\t"
> -#endif
> - : : "rm"(vec) : "memory");
> + asm volatile("cpuid" : "+a" (vec[0]), "=b" (vec[1]),"=c" (vec[2]), "=d"
> (vec[3]));
> if (eax)
> *eax = vec[0];
> if (ebx)
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> kvm-devel mailing list
> kvm-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/kvm-devel
>
--
error compiling committee.c: too many arguments to function
-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
kvm-devel mailing list
kvm-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/kvm-devel
