Re: [PULL 07/37] cpus: extract out hax-specific code to target/i386/

From: Claudio Fontana <cfontana@suse.de>
To: "Volker Rümelin" <vr_qemu@t-online.de>
Cc: Paolo Bonzini <pbonzini@redhat.com>,
	Richard Henderson <richard.henderson@linaro.org>,
	qemu-devel@nongnu.org
Subject: Re: [PULL 07/37] cpus: extract out hax-specific code to target/i386/
Date: Fri, 16 Oct 2020 10:00:20 +0200	[thread overview]
Message-ID: <39069adb-ee6d-ddf9-cc70-f069bcab5843@suse.de> (raw)
In-Reply-To: <a412667e-c486-b002-1e9c-5128345cad75@t-online.de>

On 10/16/20 8:48 AM, Volker Rümelin wrote:
>> From: Claudio Fontana <cfontana@suse.de>
>>
>> register a "CpusAccel" interface for HAX as well.
>>
> 
>> diff --git a/softmmu/cpus.c b/softmmu/cpus.c
>> index 9fa73735a2..900fff827a 100644
>> --- a/softmmu/cpus.c
>> +++ b/softmmu/cpus.c
>> @@ -416,35 +403,6 @@ void qemu_wait_io_event(CPUState *cpu)
>>      qemu_wait_io_event_common(cpu);
>>  }
>>  
>> -static void *qemu_hax_cpu_thread_fn(void *arg)
>> -{
>> -    CPUState *cpu = arg;
>> -    int r;
>> -
>> -    rcu_register_thread();
>> -    qemu_mutex_lock_iothread();
>> -    qemu_thread_get_self(cpu->thread);
>> -
>> -    cpu->thread_id = qemu_get_thread_id();
>> -    current_cpu = cpu;
> 
> Hi Claudio,
> 
> is there a reason why you removed current_cpu = cpu; from hax_cpu_thread_fn() when you moved that function to target/i386/hax-cpus.c? This change broke HAX on Windows. Adding back that line makes it work again.

Hello Volker, I see the change in the history and it was clearly an ugly mistake on my part.
There was no reason or intention to remove the current_cpu = cpu assignment

The fix seems indeed to just + current_cpu = cpu;
and I will send a patch momentarily that does just that,

but I don't know of any CI coverage for Windows + hax currently,
so it would be good if you could spin the change to verify that it fixes the problem.

Ciao,

Claudio

> 
> The simplest reproducer is:
> $ ./qemu-system-x86_64.exe -machine pc,accel=hax -smp 2 -display gtk
> HAX is working and emulator runs in fast virt mode.
> 
> Then the QEMU window opens and shows 'Guest has not initialized the display (yet).' forever.
> 
> A look at the Windows Task Manager suggests one thread is busy looping.
> 
> With best regards,
> Volker
> 
>> -    hax_init_vcpu(cpu);
>> -    cpu_thread_signal_created(cpu);
>> -    qemu_guest_random_seed_thread_part2(cpu->random_seed);
>> -
>> -    do {
>> -        if (cpu_can_run(cpu)) {
>> -            r = hax_smp_cpu_exec(cpu);
>> -            if (r == EXCP_DEBUG) {
>> -                cpu_handle_guest_debug(cpu);
>> -            }
>> -        }
>> -
>> -        qemu_wait_io_event(cpu);
>> -    } while (!cpu->unplug || cpu_can_run(cpu));
>> -    rcu_unregister_thread();
>> -    return NULL;
>> -}
>> -
>>  /* The HVF-specific vCPU thread function. This one should only run when the host
>>   * CPU supports the VMX "unrestricted guest" feature. */
>>  static void *qemu_hvf_cpu_thread_fn(void *arg)
>>