Re: [Qemu-devel] [PULL 28/31] curl: Remove unnecessary explicit calls to internal event handler

[Matthew Booth <mbooth@redhat.com>]

On 06/05/14 10:56, Kevin Wolf wrote:
> Am 06.05.2014 um 11:44 hat Laurent Desnogues geschrieben:
>> Hello,
>>
>> sorry for another late detection of the use of too recent features.
>>
>> On Wed, Apr 30, 2014 at 8:24 PM, Kevin Wolf <kwolf@redhat.com> wrote:
>>> From: Matthew Booth <mbooth@redhat.com>
>>>
>>> Remove calls to curl_multi_do where the relevant handles are already
>>> registered to the event loop.
>>>
>>> Ensure that we kick off socket handling with CURL_SOCKET_TIMEOUT after
>>> adding a new handle.
>>>
>>> Signed-off-by: Matthew Booth <mbooth@redhat.com>
>>> Tested-by: Richard W.M. Jones <rjones@redhat.com>
>>> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
>>> ---
>>>  block/curl.c | 5 +++--
>>>  1 file changed, 3 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/block/curl.c b/block/curl.c
>>> index 50bd05f..fd2756e 100644
>>> --- a/block/curl.c
>>> +++ b/block/curl.c
>>> @@ -535,7 +535,6 @@ static int curl_open(BlockDriverState *bs, QDict *options, int flags,
>>>      curl_multi_setopt(s->multi, CURLMOPT_TIMERDATA, s);
>>>      curl_multi_setopt(s->multi, CURLMOPT_TIMERFUNCTION, curl_timer_cb);
>>>  #endif
>>> -    curl_multi_do(s);
>>>
>>>      qemu_opts_del(opts);
>>>      return 0;
>>> @@ -564,6 +563,7 @@ static const AIOCBInfo curl_aiocb_info = {
>>>  static void curl_readv_bh_cb(void *p)
>>>  {
>>>      CURLState *state;
>>> +    int running;
>>>
>>>      CURLAIOCB *acb = p;
>>>      BDRVCURLState *s = acb->common.bs->opaque;
>>> @@ -612,8 +612,9 @@ static void curl_readv_bh_cb(void *p)
>>>      curl_easy_setopt(state->curl, CURLOPT_RANGE, state->range);
>>>
>>>      curl_multi_add_handle(s->multi, state->curl);
>>> -    curl_multi_do(s);
>>>
>>> +    /* Tell curl it needs to kick things off */
>>> +    curl_multi_socket_action(s->multi, CURL_SOCKET_TIMEOUT, 0, &running);
>>
>> curl_multi_socket_action isn't available on the CURL lib used
>> by CentOS 5.6.  The obvious workaround is to disable CURL
>> when calling configure.
> 
> Matthew, can you please check whether this call can be avoided, or
> otherwise change configure to check the presence of the function?

I should be able to sort this if I can work out how to add a configure
test. When do you need it by? If I'm too slow it's safe to revert this
change. It's just a cleanup to remove use of a deprecated api.

Matt
-- 
Matthew Booth
Red Hat Engineering, Virtualisation Team

Phone: +442070094448 (UK)
GPG ID:  D33C3490
GPG FPR: 3733 612D 2D05 5458 8A8A 1600 3441 EA19 D33C 3490