You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@httpd.apache.org by Yann Ylavic <yl...@gmail.com> on 2018/02/08 19:37:20 UTC

Re: svn commit: r1807876 - /httpd/httpd/trunk/modules/proxy/mod_proxy_fcgi.c

Hi Luca,

On Sat, Sep 9, 2017 at 12:24 PM,  <el...@apache.org> wrote:
>
> +            if (((conn->worker->s->flush_packets == flush_on) ||
> +                 ((conn->worker->s->flush_packets == flush_auto) &&
> +                  (apr_poll(flushpoll, 1, &flushpoll_fd,
> +                            conn->worker->s->flush_wait) == APR_TIMEUP))) && mayflush) {

Don't we want to test "mayflush" before anything here, why would we
poll() otherwise?

Thanks,
Yann.

Re: svn commit: r1807876 - /httpd/httpd/trunk/modules/proxy/mod_proxy_fcgi.c

Posted by Yann Ylavic <yl...@gmail.com>.

On Thu, Feb 8, 2018 at 8:50 PM, Luca Toscano <to...@gmail.com> wrote:
>
> 2018-02-08 20:37 GMT+01:00 Yann Ylavic <yl...@gmail.com>:
>>
>> Hi Luca,
>>
>> On Sat, Sep 9, 2017 at 12:24 PM,  <el...@apache.org> wrote:
>> >
>> > +            if (((conn->worker->s->flush_packets == flush_on) ||
>> > +                 ((conn->worker->s->flush_packets == flush_auto) &&
>> > +                  (apr_poll(flushpoll, 1, &flushpoll_fd,
>> > +                            conn->worker->s->flush_wait) ==
>> > APR_TIMEUP))) && mayflush) {
>>
>> Don't we want to test "mayflush" before anything here, why would we
>> poll() otherwise?
>
>
> it makes a lot of sense, +1
>
> If you want to commit please go ahead :)

Not really, since it's part of a patch series you may want to follow up on it ;)

Re: svn commit: r1807876 - /httpd/httpd/trunk/modules/proxy/mod_proxy_fcgi.c

Posted by Luca Toscano <to...@gmail.com>.

Hi Yann,

2018-02-08 20:37 GMT+01:00 Yann Ylavic <yl...@gmail.com>:

> Hi Luca,
>
> On Sat, Sep 9, 2017 at 12:24 PM,  <el...@apache.org> wrote:
> >
> > +            if (((conn->worker->s->flush_packets == flush_on) ||
> > +                 ((conn->worker->s->flush_packets == flush_auto) &&
> > +                  (apr_poll(flushpoll, 1, &flushpoll_fd,
> > +                            conn->worker->s->flush_wait) ==
> APR_TIMEUP))) && mayflush) {
>
> Don't we want to test "mayflush" before anything here, why would we
> poll() otherwise?
>

it makes a lot of sense, +1

If you want to commit please go ahead :)

Luca