From: Eric Blake <eblake@redhat.com>
To: Markus Armbruster <armbru@redhat.com>, qemu-devel@nongnu.org
Cc: kwolf@redhat.com, berrange@redhat.com, zxq_yx_007@163.com,
kraxel@redhat.com, pbonzini@redhat.com,
marcandre.lureau@redhat.com
Subject: Re: [PATCH v2 11/11] sockets: Make abstract UnixSocketAddress depend on CONFIG_LINUX
Date: Mon, 2 Nov 2020 08:12:57 -0600 [thread overview]
Message-ID: <e98fe4e1-b641-97ef-e697-11281d085c04@redhat.com> (raw)
In-Reply-To: <20201102094422.173975-12-armbru@redhat.com>
On 11/2/20 3:44 AM, Markus Armbruster wrote:
> The abstract socket namespace is a non-portable Linux extension. An
> attempt to use it elsewhere should fail with ENOENT (the abstract
> address looks like a "" pathname, which does not resolve). We report
> this failure like
>
> Failed to connect socket abc: No such file or directory
>
> Tolerable, although ENOTSUP would be better.
>
> However, introspection lies: it has @abstract regardless of host
> support. Easy enough to fix: since Linux provides them since 2.2,
> 'if': 'defined(CONFIG_LINUX)' should do.
>
> The above failure becomes
>
> Parameter 'backend.data.addr.data.abstract' is unexpected
>
> I consider this an improvement.
>
Commit message lacks mention of the fact that we are now explicitly not
outputting 'strict' for non-abstract sockets (in fact, that change could
be squashed in 9/11 if you wanted to do it there). But as this cleans
up the code I mentioned in 9/11, I'll leave it up to Dan if the commit
message needs a tweak; the end result is fine if we don't feel like a v3
spin just for moving hunks around.
Reviewed-by: Eric Blake <eblake@redhat.com>
> +++ b/chardev/char-socket.c
> @@ -444,14 +444,20 @@ static char *qemu_chr_socket_address(SocketChardev *s, const char *prefix)
> break;
> case SOCKET_ADDRESS_TYPE_UNIX:
> {
> + const char *tight = "", *abstract = "";
> UnixSocketAddress *sa = &s->addr->u.q_unix;
>
> - return g_strdup_printf("%sunix:%s%s%s%s", prefix,
> - s->addr->u.q_unix.path,
> - sa->has_abstract && sa->abstract
> - ? ",abstract" : "",
> - sa->has_tight && sa->tight
> - ? ",tight" : "",
Unconditional output if tight is true (which is its stated default)...
> +#ifdef CONFIG_LINUX
> + if (sa->has_abstract && sa->abstract) {
> + abstract = ",abstract";
> + if (sa->has_tight && sa->tight) {
> + tight = ",tight";
> + }
> + }
> +#endif
> +
> + return g_strdup_printf("%sunix:%s%s%s%s", prefix, sa->path,
> + abstract, tight,
...vs. the now-nicer conditional where tight is only present if abstract.
--
Eric Blake, Principal Software Engineer
Red Hat, Inc. +1-919-301-3226
Virtualization: qemu.org | libvirt.org
next prev parent reply other threads:[~2020-11-02 14:14 UTC|newest]
Thread overview: 20+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-11-02 9:44 [PATCH v2 00/11] sockets: Attempt to drain the abstract socket swamp Markus Armbruster
2020-11-02 9:44 ` [PATCH v2 01/11] test-util-sockets: Plug file descriptor leak Markus Armbruster
2020-11-02 14:10 ` Philippe Mathieu-Daudé
2020-11-02 9:44 ` [PATCH v2 02/11] test-util-sockets: Correct to set has_abstract, has_tight Markus Armbruster
2020-11-02 9:44 ` [PATCH v2 03/11] test-util-sockets: Clean up SocketAddress construction Markus Armbruster
2020-11-02 9:44 ` [PATCH v2 04/11] test-util-sockets: Factor out test_socket_unix_abstract_one() Markus Armbruster
2020-11-02 9:44 ` [PATCH v2 05/11] test-util-sockets: Synchronize properly, don't sleep(1) Markus Armbruster
2020-11-02 9:44 ` [PATCH v2 06/11] test-util-sockets: Test the complete abstract socket matrix Markus Armbruster
2020-11-02 9:44 ` [PATCH v2 07/11] sockets: Fix default of UnixSocketAddress member @tight Markus Armbruster
2020-11-02 9:44 ` [PATCH v2 08/11] sockets: Fix socket_sockaddr_to_address_unix() for abstract sockets Markus Armbruster
2020-11-02 14:04 ` Eric Blake
2020-11-02 9:44 ` [PATCH v2 09/11] char-socket: Fix qemu_chr_socket_address() " Markus Armbruster
2020-11-02 14:08 ` Eric Blake
2020-11-03 6:28 ` Markus Armbruster
2020-11-03 13:17 ` Daniel P. Berrangé
2020-11-03 15:21 ` Markus Armbruster
2020-11-02 9:44 ` [PATCH v2 10/11] sockets: Bypass "replace empty @path" for abstract unix sockets Markus Armbruster
2020-11-02 9:44 ` [PATCH v2 11/11] sockets: Make abstract UnixSocketAddress depend on CONFIG_LINUX Markus Armbruster
2020-11-02 14:12 ` Eric Blake [this message]
2020-11-03 6:35 ` Markus Armbruster
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=e98fe4e1-b641-97ef-e697-11281d085c04@redhat.com \
--to=eblake@redhat.com \
--cc=armbru@redhat.com \
--cc=berrange@redhat.com \
--cc=kraxel@redhat.com \
--cc=kwolf@redhat.com \
--cc=marcandre.lureau@redhat.com \
--cc=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=zxq_yx_007@163.com \
/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).