qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: John Snow <jsnow@redhat.com>
To: Max Reitz <mreitz@redhat.com>, qemu-block@nongnu.org
Cc: Kevin Wolf <kwolf@redhat.com>, qemu-devel@nongnu.org
Subject: Re: [Qemu-devel] [Qemu-block] [PATCH 3/6] curl: Pass CURLSocket to curl_multi_{do, read}()
Date: Mon, 9 Sep 2019 16:10:36 -0400	[thread overview]
Message-ID: <8fea3235-3f4c-86b1-f88e-b469081f0a6a@redhat.com> (raw)
In-Reply-To: <20190827163439.16686-4-mreitz@redhat.com>



On 8/27/19 12:34 PM, Max Reitz wrote:
> curl_multi_do_locked() currently marks all sockets as ready.  That is
> not only inefficient, but in fact unsafe (the loop is).  A follow-up
> patch will change that, but to do so, curl_multi_do_locked() needs to
> know exactly which socket is ready; and that is accomplished by this
> patch here.
> 
> Cc: qemu-stable@nongnu.org
> Signed-off-by: Max Reitz <mreitz@redhat.com>
> ---
>  block/curl.c | 29 ++++++++++++++++-------------
>  1 file changed, 16 insertions(+), 13 deletions(-)
> 
> diff --git a/block/curl.c b/block/curl.c
> index 8a45b371cc..05f77a38c2 100644
> --- a/block/curl.c
> +++ b/block/curl.c
> @@ -189,15 +189,15 @@ static int curl_sock_cb(CURL *curl, curl_socket_t fd, int action,
>      switch (action) {
>          case CURL_POLL_IN:
>              aio_set_fd_handler(s->aio_context, fd, false,
> -                               curl_multi_read, NULL, NULL, state);
> +                               curl_multi_read, NULL, NULL, socket);
>              break;
>          case CURL_POLL_OUT:
>              aio_set_fd_handler(s->aio_context, fd, false,
> -                               NULL, curl_multi_do, NULL, state);
> +                               NULL, curl_multi_do, NULL, socket);
>              break;
>          case CURL_POLL_INOUT:
>              aio_set_fd_handler(s->aio_context, fd, false,
> -                               curl_multi_read, curl_multi_do, NULL, state);
> +                               curl_multi_read, curl_multi_do, NULL, socket);
>              break;
>          case CURL_POLL_REMOVE:
>              aio_set_fd_handler(s->aio_context, fd, false,
> @@ -394,9 +394,10 @@ static void curl_multi_check_completion(BDRVCURLState *s)
>  }
>  
>  /* Called with s->mutex held.  */
> -static void curl_multi_do_locked(CURLState *s)
> +static void curl_multi_do_locked(CURLSocket *ready_socket)
>  {
>      CURLSocket *socket, *next_socket;
> +    CURLState *s = socket->state;

Did you mean to use ready_socket here instead?

>      int running;
>      int r;
>  
> @@ -415,21 +416,23 @@ static void curl_multi_do_locked(CURLState *s)
>  
>  static void curl_multi_do(void *arg)
>  {
> -    CURLState *s = (CURLState *)arg;
> +    CURLSocket *socket = arg;
> +    BDRVCURLState *s = socket->state->s;
>  
> -    qemu_mutex_lock(&s->s->mutex);
> -    curl_multi_do_locked(s);
> -    qemu_mutex_unlock(&s->s->mutex);
> +    qemu_mutex_lock(&s->mutex);
> +    curl_multi_do_locked(socket);
> +    qemu_mutex_unlock(&s->mutex);
>  }
>  
>  static void curl_multi_read(void *arg)
>  {
> -    CURLState *s = (CURLState *)arg;
> +    CURLSocket *socket = arg;
> +    BDRVCURLState *s = socket->state->s;
>  
> -    qemu_mutex_lock(&s->s->mutex);
> -    curl_multi_do_locked(s);
> -    curl_multi_check_completion(s->s);
> -    qemu_mutex_unlock(&s->s->mutex);

bye bye &s->s->mutex ! you're very nasty !!

> +    qemu_mutex_lock(&s->mutex);
> +    curl_multi_do_locked(socket);
> +    curl_multi_check_completion(s);
> +    qemu_mutex_unlock(&s->mutex);
>  }
>  
>  static void curl_multi_timeout_do(void *arg)
> 


  reply	other threads:[~2019-09-09 20:11 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-08-27 16:34 [Qemu-devel] [PATCH 0/6] block/curl: Fix hang and potential crash Max Reitz
2019-08-27 16:34 ` [Qemu-devel] [PATCH 1/6] curl: Keep pointer to the CURLState in CURLSocket Max Reitz
2019-09-09 20:05   ` [Qemu-devel] [Qemu-block] " John Snow
2019-08-27 16:34 ` [Qemu-devel] [PATCH 2/6] curl: Keep *socket until the end of curl_sock_cb() Max Reitz
2019-09-09 20:09   ` [Qemu-devel] [Qemu-block] " John Snow
2019-09-10  7:50     ` Max Reitz
2019-08-27 16:34 ` [Qemu-devel] [PATCH 3/6] curl: Pass CURLSocket to curl_multi_{do, read}() Max Reitz
2019-09-09 20:10   ` John Snow [this message]
2019-09-10  7:52     ` [Qemu-devel] [Qemu-block] " Max Reitz
2019-08-27 16:34 ` [Qemu-devel] [PATCH 4/6] curl: Report only ready sockets Max Reitz
2019-09-09 20:16   ` [Qemu-devel] [Qemu-block] " John Snow
2019-09-10  7:53     ` Max Reitz
2019-08-27 16:34 ` [Qemu-devel] [PATCH 5/6] curl: Handle success in multi_check_completion Max Reitz
2019-09-09 20:30   ` [Qemu-devel] [Qemu-block] " John Snow
2019-09-10  8:17     ` Max Reitz
2019-08-27 16:34 ` [Qemu-devel] [PATCH 6/6] curl: Check curl_multi_add_handle()'s return code Max Reitz
2019-09-09 20:32   ` [Qemu-devel] [Qemu-block] " John Snow

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=8fea3235-3f4c-86b1-f88e-b469081f0a6a@redhat.com \
    --to=jsnow@redhat.com \
    --cc=kwolf@redhat.com \
    --cc=mreitz@redhat.com \
    --cc=qemu-block@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).