Re: [Qemu-devel] [PATCH] SVM support

["Blue Swirl" <blauwirbel@gmail.com>]

On 9/18/07, J. Mayer <l_indien@magic.fr> wrote:
> Going this way, I see no reason why the interruption code in the cpu_exec loop
> could not finally become:
>                 interrupt_request = env->interrupt_request;
>                 if (__builtin_expect(interrupt_request, 0)) {
>                     if (interrupt_request & CPU_INTERRUPT_DEBUG) {
>                         env->interrupt_request &= ~CPU_INTERRUPT_DEBUG;
>                         env->exception_index = EXCP_DEBUG;
>                         cpu_loop_exit();
>                     }
>                     if (interrupt_request & CPU_INTERRUPT_HALT) {
>                         env->interrupt_request &= ~CPU_INTERRUPT_HALT;
>                         env->halted = 1;
>                         env->exception_index = EXCP_HLT;
>                         cpu_loop_exit();
>                     }
>                     if (interrupt_request & CPU_INTERRUPT_HARD) {
>                         hw_interrupt(env);
>                         if (env->pending_interrupts == 0)
>                             env->interrupt_request &= ~CPU_INTERRUPT_HARD;
> #if defined(__sparc__) && !defined(HOST_SOLARIS)
>                         tmp_T0 = 0;
> #else
>                         T0 = 0;
> #endif
>                     }
>                    /* Don't use the cached interupt_request value,
>                       do_interrupt may have updated the EXITTB flag. */
>                     if (interrupt_request & CPU_INTERRUPT_EXITTB) {
>                         env->interrupt_request &= ~CPU_INTERRUPT_EXITTB;
>                         /* ensure that no TB jump will be modified as
>                            the program flow was changed */
> #if defined(__sparc__) && !defined(HOST_SOLARIS)
>                         tmp_T0 = 0;
> #else
>                         T0 = 0;
> #endif
>                     }
>                     if (interrupt_request & CPU_INTERRUPT_EXIT) {
>                         env->interrupt_request &= ~CPU_INTERRUPT_EXIT;
>                         env->exception_index = EXCP_INTERRUPT;
>                         cpu_loop_exit();
>                     }
>                 }
> All the targets specific tricks could be done in the hw_interrupt routine. And the generic code
> would become much more readable. But this needs some works (not so much) and intensive tests...
> And I guess nobody feels like taking this risk right now ;-)
> But I think this will have to be done someday...

I vote for this. There are some other places where we could use the same API:
cpu_dump_state()
cpu_list()
cpu_register()