From: "Daniel P. Berrangé" <berrange@redhat.com>
To: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Cc: qemu-devel@nongnu.org, qemu-block@nongnu.org, peterx@redhat.com,
Paolo Bonzini <pbonzini@redhat.com>
Subject: Re: [PATCH 08/10] oslib-posix: add qemu_fds_set_blocking() helper
Date: Wed, 10 Sep 2025 10:52:17 +0100 [thread overview]
Message-ID: <aMFKUcHVxB9q3of6@redhat.com> (raw)
In-Reply-To: <20250903094411.1029449-9-vsementsov@yandex-team.ru>
On Wed, Sep 03, 2025 at 12:44:08PM +0300, Vladimir Sementsov-Ogievskiy wrote:
> And use it in io/channel-socket.c. This simplifies the following
> commit, which will move this functionality from io/channel-socket.c
> to the callers.
Looking at how many callers will still want the current
functionality, I'm inclined to add a flag for QIOChannel
QIO_CHANNEL_READ_FLAG_PRESERVE_BLOCKING
which migration can pass to qio_channel_readv_full and
thus avoid adding this method and duplicated calls to
it in (almost) every caller.
>
> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
> ---
> include/qemu/osdep.h | 7 +++++++
> io/channel-socket.c | 24 +++++++++++++-----------
> util/oslib-posix.c | 12 ++++++++++++
> 3 files changed, 32 insertions(+), 11 deletions(-)
>
> diff --git a/include/qemu/osdep.h b/include/qemu/osdep.h
> index 1b38cb7e45..dde98d588c 100644
> --- a/include/qemu/osdep.h
> +++ b/include/qemu/osdep.h
> @@ -689,6 +689,13 @@ ssize_t qemu_write_full(int fd, const void *buf, size_t count)
> void qemu_set_cloexec(int fd);
> bool qemu_set_blocking(int fd, bool block, Error **errp);
>
> +/*
> + * qemu_fds_set_blockinging:
> + * Call qemu_socket_set_block() on several fds.
> + * When @nfds = 0, does nothing, @fds is not touched.
> + */
> +bool qemu_fds_set_blockinging(int *fds, int nfds, bool block, Error **errp);
> +
> /* Return a dynamically allocated directory path that is appropriate for storing
> * local state.
> *
> diff --git a/io/channel-socket.c b/io/channel-socket.c
> index 4f7e86f72f..96098b5bcc 100644
> --- a/io/channel-socket.c
> +++ b/io/channel-socket.c
> @@ -472,8 +472,11 @@ static void qio_channel_socket_copy_fds(struct msghdr *msg,
> *fds = NULL;
>
> for (cmsg = CMSG_FIRSTHDR(msg); cmsg; cmsg = CMSG_NXTHDR(msg, cmsg)) {
> - int fd_size, i;
> + int fd_size;
> int gotfds;
> +#ifndef MSG_CMSG_CLOEXEC
> + int i;
> +#endif
>
> if (cmsg->cmsg_len < CMSG_LEN(sizeof(int)) ||
> cmsg->cmsg_level != SOL_SOCKET ||
> @@ -491,20 +494,19 @@ static void qio_channel_socket_copy_fds(struct msghdr *msg,
> *fds = g_renew(int, *fds, *nfds + gotfds);
> memcpy(*fds + *nfds, CMSG_DATA(cmsg), fd_size);
>
> + /* O_NONBLOCK is preserved across SCM_RIGHTS so reset it */
> + /* TODO: don't crash on error, just handle it! */
> + qemu_fds_set_blockinging(*fds + *nfds, gotfds, true, &error_abort);
> +
> +#ifndef MSG_CMSG_CLOEXEC
> for (i = 0; i < gotfds; i++) {
> int fd = (*fds)[*nfds + i];
> - if (fd < 0) {
> - continue;
> + if (fd >= 0) {
> + qemu_set_cloexec(fd);
> }
> -
> - /* O_NONBLOCK is preserved across SCM_RIGHTS so reset it */
> - /* TODO: don't crash on error, just handle it! */
> - qemu_set_blocking(fd, true, &error_abort);
> -
> -#ifndef MSG_CMSG_CLOEXEC
> - qemu_set_cloexec(fd);
> -#endif
> }
> +#endif
> +
> *nfds += gotfds;
> }
> }
> diff --git a/util/oslib-posix.c b/util/oslib-posix.c
> index 8891d82db0..8589ff21ec 100644
> --- a/util/oslib-posix.c
> +++ b/util/oslib-posix.c
> @@ -262,6 +262,18 @@ bool qemu_set_blocking(int fd, bool block, Error **errp)
> return true;
> }
>
> +bool qemu_fds_set_blockinging(int *fds, int nfds, bool block, Error **errp)
> +{
> + int i;
> + for (i = 0; i < nfds; i++) {
> + if (fds[i] >= 0 && !qemu_set_blocking(fds[i], block, errp)) {
> + return false;
> + }
> + }
> +
> + return true;
> +}
> +
> int socket_set_fast_reuse(int fd)
> {
> int val = 1, ret;
> --
> 2.48.1
>
With regards,
Daniel
--
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org -o- https://fstop138.berrange.com :|
|: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
next prev parent reply other threads:[~2025-09-10 9:53 UTC|newest]
Thread overview: 44+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-09-03 9:44 [PATCH 00/10] io: deal with blocking/non-blocking fds Vladimir Sementsov-Ogievskiy
2025-09-03 9:44 ` [PATCH 01/10] io/channel: document how qio_channel_readv_full() handles fds Vladimir Sementsov-Ogievskiy
2025-09-09 8:34 ` Daniel P. Berrangé
2025-09-03 9:44 ` [PATCH 02/10] char-socket: rework tcp_chr_recv() Vladimir Sementsov-Ogievskiy
2025-09-08 21:53 ` Peter Xu
2025-09-09 7:49 ` Vladimir Sementsov-Ogievskiy
2025-09-09 8:38 ` Daniel P. Berrangé
2025-09-09 8:46 ` Vladimir Sementsov-Ogievskiy
2025-09-03 9:44 ` [PATCH 03/10] util: add qemu_set_blocking() function Vladimir Sementsov-Ogievskiy
2025-09-08 22:02 ` Peter Xu
2025-09-09 7:51 ` Vladimir Sementsov-Ogievskiy
2025-09-09 8:59 ` Daniel P. Berrangé
2025-09-03 9:44 ` [PATCH 04/10] util: drop qemu_socket_set_nonblock() Vladimir Sementsov-Ogievskiy
2025-09-08 22:16 ` Peter Xu
2025-09-09 8:19 ` Vladimir Sementsov-Ogievskiy
2025-09-10 9:41 ` Daniel P. Berrangé
2025-09-10 10:32 ` Vladimir Sementsov-Ogievskiy
2025-09-10 9:44 ` Daniel P. Berrangé
2025-09-10 10:33 ` Vladimir Sementsov-Ogievskiy
2025-09-10 15:30 ` Vladimir Sementsov-Ogievskiy
2025-09-10 17:55 ` Vladimir Sementsov-Ogievskiy
2025-09-10 18:15 ` Daniel P. Berrangé
2025-09-10 18:52 ` Vladimir Sementsov-Ogievskiy
2025-09-12 14:51 ` Daniel P. Berrangé
2025-09-03 9:44 ` [PATCH 05/10] util: drop qemu_socket_try_set_nonblock() Vladimir Sementsov-Ogievskiy
2025-09-09 0:09 ` Peter Xu
2025-09-03 9:44 ` [PATCH 06/10] util: drop qemu_socket_set_block() Vladimir Sementsov-Ogievskiy
2025-09-03 9:44 ` [PATCH 07/10] use qemu_set_blocking instead of g_unix_set_fd_nonblocking Vladimir Sementsov-Ogievskiy
2025-09-09 9:00 ` Daniel P. Berrangé
2025-09-09 9:11 ` Vladimir Sementsov-Ogievskiy
2025-09-09 9:15 ` Daniel P. Berrangé
2025-09-09 9:50 ` Vladimir Sementsov-Ogievskiy
2025-09-10 9:32 ` Daniel P. Berrangé
2025-09-09 14:24 ` Stefan Hajnoczi
2025-09-03 9:44 ` [PATCH 08/10] oslib-posix: add qemu_fds_set_blocking() helper Vladimir Sementsov-Ogievskiy
2025-09-03 9:48 ` Vladimir Sementsov-Ogievskiy
2025-09-10 9:52 ` Daniel P. Berrangé [this message]
2025-09-10 10:42 ` Vladimir Sementsov-Ogievskiy
2025-09-03 9:44 ` [PATCH 09/10] qio_channel_readv_full(): move setting fd blocking to callers Vladimir Sementsov-Ogievskiy
2025-09-03 9:44 ` [PATCH 10/10] migration/qemu-file: don't make incoming fds blocking again Vladimir Sementsov-Ogievskiy
2025-09-03 9:47 ` Vladimir Sementsov-Ogievskiy
2025-09-04 14:35 ` [PATCH 00/10] io: deal with blocking/non-blocking fds Lei Yang
2025-09-09 8:12 ` Vladimir Sementsov-Ogievskiy
2025-09-09 14:25 ` Stefan Hajnoczi
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=aMFKUcHVxB9q3of6@redhat.com \
--to=berrange@redhat.com \
--cc=pbonzini@redhat.com \
--cc=peterx@redhat.com \
--cc=qemu-block@nongnu.org \
--cc=qemu-devel@nongnu.org \
--cc=vsementsov@yandex-team.ru \
/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).