Re: [Qemu-devel] [PATCH 3/3] slirp: fill mainloop timeout with more precise value

[liu ping fan]

On Tue, Aug 13, 2013 at 4:19 PM, Jan Kiszka <address@hidden> wrote:
> On 2013-08-13 05:15, Liu Ping Fan wrote:
>> If slirp needs to emulate tcp timeout, then the timeout value
>> for mainloop should be more precise, which is determined by
>> slirp's fasttimo or slowtimo.
>>
>> Signed-off-by: Liu Ping Fan <address@hidden>
>> ---
>>  main-loop.c   |  2 +-
>>  slirp/slirp.c | 18 ++++++++++++++++++
>>  2 files changed, 19 insertions(+), 1 deletion(-)
>>
>> diff --git a/main-loop.c b/main-loop.c
>> index a44fff6..04120d2 100644
>> --- a/main-loop.c
>> +++ b/main-loop.c
>> @@ -458,8 +458,8 @@ int main_loop_wait(int nonblocking)
>>      g_array_set_size(gpollfds, 0); /* reset for new iteration */
>>      /* XXX: separate device handlers from system ones */
>>  #ifdef CONFIG_SLIRP
>> -    slirp_update_timeout(&timeout);
>>      slirp_pollfds_fill(gpollfds);
>> +    slirp_update_timeout(&timeout);
>
> Why this reordering?
>
In order to give timeout more precise value, which is based on the
result of fasttimo or slowtimo after slirp_pollfds_fill()
>>  #endif
>>      qemu_iohandler_fill(gpollfds);
>>      ret = os_host_main_loop_wait(timeout);
>> diff --git a/slirp/slirp.c b/slirp/slirp.c
>> index 1deaad9..af66006 100644
>> --- a/slirp/slirp.c
>> +++ b/slirp/slirp.c
>> @@ -262,9 +262,27 @@ void slirp_cleanup(Slirp *slirp)
>>
>>  void slirp_update_timeout(uint32_t *timeout)
>>  {
>> +    Slirp *slirp;
>> +
>>      if (!QTAILQ_EMPTY(&slirp_instances)) {
>
> if (QTAILQ_EMPTY(&slirp_instances) || *timeout <= TIMEOUT_FAST) {
>     return;
> }
>
> *timeout = MIN(1000, *timeout);
> ...
>
> would be nicer.
>
Ok.
>>          *timeout = MIN(1000, *timeout);
>>      }
>> +    if (*timeout <= TIMEOUT_FAST) {
>> +        return;
>> +    }
>> +
>> +    /* If we have tcp timeout with slirp, then we will fill @timeout with
>> +     * more precise value.
>> +     */
>> +    QTAILQ_FOREACH(slirp, &slirp_instances, entry) {
>> +        if (slirp->time_fasttimo) {
>> +            *timeout = TIMEOUT_FAST;
>> +            break;
>> +        }
>> +        if (slirp->do_slowtimo) {
>> +            *timeout = MIN(TIMEOUT_SLOW, *timeout);
>> +        }
>> +    }
>
> Not sure if the compiler is smart enough, but I would suggest to keep
> timeout local until returning and only write it back by then.
>
Ok, got it.

Thx,
Pingfan