On 25/06/2015 05:39, Fam Zheng wrote:
> On Wed, 06/24 18:25, Paolo Bonzini wrote:
>> The next patch will require the BQL to be always taken with
>> qemu_mutex_lock_iothread(), while right now this isn't the case.
>>
>> Outside TCG mode this is not a problem. In TCG mode, we need to be
>> careful and avoid the "prod out of compiled code" step if already
>> in a VCPU thread. This is easily done with a check on current_cpu,
>> i.e. qemu_in_vcpu_thread().
>>
>> Hopefully, multithreaded TCG will get rid of the whole logic to kick
>> VCPUs whenever an I/O event occurs!
>>
>> Cc: Frederic Konrad <fred.kon...@greensocs.com>
>> Message-Id: <1434646046-27150-2-git-send-email-pbonz...@redhat.com>
>
> Why is this "Message-Id:" included in the commit message if it's not final?
>
>> Signed-off-by: Paolo Bonzini <pbonz...@redhat.com>
>> ---
>> cpus.c | 10 +++++++---
>> 1 file changed, 7 insertions(+), 3 deletions(-)
>>
>> diff --git a/cpus.c b/cpus.c
>> index b85fb5f..02cca5d 100644
>> --- a/cpus.c
>> +++ b/cpus.c
>> @@ -953,7 +953,7 @@ static void *qemu_kvm_cpu_thread_fn(void *arg)
>> CPUState *cpu = arg;
>> int r;
>>
>> - qemu_mutex_lock(&qemu_global_mutex);
>> + qemu_mutex_lock_iothread();
>> qemu_thread_get_self(cpu->thread);
>> cpu->thread_id = qemu_get_thread_id();
>> cpu->can_do_io = 1;
>> @@ -1033,10 +1033,10 @@ static void *qemu_tcg_cpu_thread_fn(void *arg)
>> {
>> CPUState *cpu = arg;
>>
>> + qemu_mutex_lock_iothread();
>> qemu_tcg_init_cpu_signals();
>> qemu_thread_get_self(cpu->thread);
>>
>> - qemu_mutex_lock(&qemu_global_mutex);
>> CPU_FOREACH(cpu) {
>> cpu->thread_id = qemu_get_thread_id();
>> cpu->created = true;
>> @@ -1148,7 +1148,11 @@ bool qemu_in_vcpu_thread(void)
>> void qemu_mutex_lock_iothread(void)
>> {
>> atomic_inc(&iothread_requesting_mutex);
>> - if (!tcg_enabled() || !first_cpu || !first_cpu->thread) {
>> + /* In the simple case there is no need to bump the VCPU thread out of
>> + * TCG code execution.
>> + */
>> + if (!tcg_enabled() || qemu_in_vcpu_thread() ||
>> + !first_cpu || !first_cpu->thread) {
>
> This looks like a separate change from the above
> "qemu_mutex_lock(&qemu_global_mutex)" conversion. Why do they belong to the
> same patch?
Previously, a VCPU thread would never call qemu_mutex_lock_iothread().
Now, it can.
While the change is only necessary when a call is introduced with
current_cpu != NULL (that's patch 5), it should be in this patch or
before because this is the one that changes the invariant. I put it in
the same because the patch is already very very smal.
Paolo
> Fam
>
>> qemu_mutex_lock(&qemu_global_mutex);
>> atomic_dec(&iothread_requesting_mutex);
>> } else {
>> --
>> 1.8.3.1
>>
>>
