From: David Gibson <david@gibson.dropbear.id.au>
To: Stefano Brivio <sbrivio@redhat.com>
Cc: passt-dev@passt.top
Subject: Re: [PATCH v2 12/12] icmp: Dedicated functions for starting and closing ping sequences
Date: Sun, 7 Jan 2024 15:30:32 +1100 [thread overview]
Message-ID: <ZZoo6DdqQ7X56egb@zatzit> (raw)
In-Reply-To: <20240106170104.48d4518e@elisabeth>
[-- Attachment #1: Type: text/plain, Size: 6047 bytes --]
On Sat, Jan 06, 2024 at 05:01:04PM +0100, Stefano Brivio wrote:
> On Thu, 21 Dec 2023 17:53:27 +1100
> David Gibson <david@gibson.dropbear.id.au> wrote:
>
> > ICMP sockets are cleaned up on a timeout implemented in icmp_timer_one(),
> > and the logic to do that cleanup is open coded in that function. Similarly
> > new sockets are opened when we discover we don't have an existing one in
> > icmp_tap_handler(), and again the logic is open-coded.
> >
> > That's not the worst thing, but it's a bit cleaner to have dedicated
> > functions for the creation and destruction of ping sockets. This will also
> > make things a bit easier for future changes we have in mind.
> >
> > Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
> > ---
> > icmp.c | 102 +++++++++++++++++++++++++++++++++++++--------------------
> > 1 file changed, 67 insertions(+), 35 deletions(-)
> >
> > diff --git a/icmp.c b/icmp.c
> > index 129a7f1..d669351 100644
> > --- a/icmp.c
> > +++ b/icmp.c
> > @@ -132,6 +132,70 @@ unexpected:
> > warn("%s: Unexpected packet on ping socket", pname);
> > }
> >
> > +/**
> > + * icmp_ping_close() - Close out and cleanup a ping sequence
>
> s/cleanup/clean up/ (verb)
Done.
> The commit title and this comment (referring to a sequence) are a bit
> misleading... this pretty much closes the socket, and as far as I can
> see setting the sequence to -1 doesn't actually have an effect (despite
> the fact it's a good practice).
>
> If you mean "sequence" not as "sequence number" but rather as a
> "sequence of pings", then I think we should rename one of the two...
> either 'seqnum' for the sequence number, or maybe s/sequence/socket/g
> here.
Right, I was basically meaning "flow", but avoiding it because this is
not (yet) linked to the flow table. I've gone with
s/sequence/socket/g, more or less, and I think that works out.
>
> > + * @c: Execution context
> > + * @id_map: id map entry of the sequence to close
> > + */
> > +static void icmp_ping_close(const struct ctx *c, struct icmp_id_sock *id_map)
> > +{
> > + epoll_ctl(c->epollfd, EPOLL_CTL_DEL, id_map->sock, NULL);
> > + close(id_map->sock);
> > + id_map->sock = -1;
> > + id_map->seq = -1;
> > +}
> > +
> > +/**
> > + * icmp_ping_new() - Prepare a new ping socket for a new id
> > + * @c: Execution context
> > + * @id_map: id map entry of the sequence to open
> > + * @af: Address family, AF_INET or AF_INET6
> > + * @id: ICMP id for the new sequence
> > + *
> > + * Return: Newly opened ping socket fd, or -1 on failure
> > + */
> > +static int icmp_ping_new(const struct ctx *c, struct icmp_id_sock *id_map,
> > + int af, uint16_t id)
> > +{
> > + uint8_t proto = af == AF_INET ? IPPROTO_ICMP : IPPROTO_ICMPV6;
> > + const char *const pname = af == AF_INET ? "ICMP" : "ICMPv6";
> > + union icmp_epoll_ref iref = { .id = id };
> > + const void *bind_addr;
> > + const char *bind_if;
> > + int s;
> > +
> > + if (af == AF_INET) {
> > + bind_addr = &c->ip4.addr_out;
> > + bind_if = c->ip4.ifname_out;
> > + } else {
> > + bind_addr = &c->ip6.addr_out;
> > + bind_if = c->ip6.ifname_out;
> > + }
> > +
> > + s = sock_l4(c, af, proto, bind_addr, bind_if, 0, iref.u32);
> > +
> > + if (s < 0) {
> > + warn("Cannot open \"ping\" socket. You might need to:");
> > + warn(" sysctl -w net.ipv4.ping_group_range=\"0 2147483647\"");
> > + warn("...echo requests/replies will fail.");
> > + goto cancel;
> > + }
> > +
> > + if (s > FD_REF_MAX)
> > + goto cancel;
> > +
> > + id_map->sock = s;
> > +
> > + debug("%s: new socket %i for echo ID %"PRIu16, pname, s, id);
> > +
> > + return s;
> > +
> > +cancel:
> > + if (s >= 0)
> > + close(s);
> > + return -1;
> > +}
> > +
> > /**
> > * icmp_tap_handler() - Handle packets from tap
> > * @c: Execution context
> > @@ -148,7 +212,6 @@ int icmp_tap_handler(const struct ctx *c, uint8_t pif, int af,
> > const void *saddr, const void *daddr,
> > const struct pool *p, const struct timespec *now)
> > {
> > - uint8_t proto = af == AF_INET ? IPPROTO_ICMP : IPPROTO_ICMPV6;
> > const char *const pname = af == AF_INET ? "ICMP" : "ICMPv6";
> > union {
> > struct sockaddr sa;
> > @@ -200,37 +263,9 @@ int icmp_tap_handler(const struct ctx *c, uint8_t pif, int af,
> > ASSERT(0);
> > }
> >
> > - if ((s = id_map->sock) < 0) {
> > - union icmp_epoll_ref iref = { .id = id };
> > - const void *bind_addr;
> > - const char *bind_if;
> > -
> > - if (af == AF_INET) {
> > - bind_addr = &c->ip4.addr_out;
> > - bind_if = c->ip4.ifname_out;
> > - } else {
> > - bind_addr = &c->ip6.addr_out;
> > - bind_if = c->ip6.ifname_out;
> > - }
> > -
> > - s = sock_l4(c, af, proto, bind_addr, bind_if, 0, iref.u32);
> > -
> > - if (s < 0) {
> > - warn("Cannot open \"ping\" socket. You might need to:");
> > - warn(" sysctl -w net.ipv4.ping_group_range=\"0 2147483647\"");
> > - warn("...echo requests/replies will fail.");
> > - return 1;
> > - }
> > -
> > - if (s > FD_REF_MAX) {
> > - close(s);
> > + if ((s = id_map->sock) < 0)
> > + if ((s = icmp_ping_new(c, id_map, af, id)) < 0)
> > return 1;
> > - }
> > -
> > - id_map->sock = s;
> > -
> > - debug("%s: new socket %i for echo ID %"PRIu16, pname, s, id);
> > - }
> >
> > id_map->ts = now->tv_sec;
> >
> > @@ -257,10 +292,7 @@ static void icmp_timer_one(const struct ctx *c, struct icmp_id_sock *id_map,
> > if (id_map->sock < 0 || now->tv_sec - id_map->ts <= ICMP_ECHO_TIMEOUT)
> > return;
> >
> > - epoll_ctl(c->epollfd, EPOLL_CTL_DEL, id_map->sock, NULL);
> > - close(id_map->sock);
> > - id_map->sock = -1;
> > - id_map->seq = -1;
> > + icmp_ping_close(c, id_map);
> > }
> >
> > /**
>
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]
prev parent reply other threads:[~2024-01-07 4:30 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-12-21 6:53 [PATCH v2 00/12] RFC: ICMP reworks preliminary to flow table integration David Gibson
2023-12-21 6:53 ` [PATCH v2 01/12] checksum: Don't use linux/icmp.h when netinet/ip_icmp.h will do David Gibson
2023-12-21 6:53 ` [PATCH v2 02/12] icmp: Don't set "port" on destination sockaddr for ping sockets David Gibson
2023-12-21 6:53 ` [PATCH v2 03/12] icmp: Remove redundant initialisation of sendto() address David Gibson
2023-12-21 6:53 ` [PATCH v2 04/12] icmp: Don't attempt to handle "wrong direction" ping socket traffic David Gibson
2024-01-06 15:59 ` Stefano Brivio
2024-01-07 0:37 ` David Gibson
2024-01-07 14:59 ` Stefano Brivio
2023-12-21 6:53 ` [PATCH v2 05/12] icmp: Don't attempt to match host IDs to guest IDs David Gibson
2023-12-21 6:53 ` [PATCH v2 06/12] icmp: Use -1 to represent "missing" sockets David Gibson
2024-01-06 15:59 ` Stefano Brivio
2024-01-07 0:38 ` David Gibson
2023-12-21 6:53 ` [PATCH v2 07/12] icmp: Simplify socket expiry scanning David Gibson
2024-01-06 15:59 ` Stefano Brivio
2024-01-07 0:41 ` David Gibson
2023-12-21 6:53 ` [PATCH v2 08/12] icmp: Share more between IPv4 and IPv6 paths in icmp_tap_handler() David Gibson
2024-01-06 16:00 ` Stefano Brivio
2024-01-07 4:41 ` David Gibson
2023-12-21 6:53 ` [PATCH v2 09/12] icmp: Consolidate icmp_sock_handler() with icmpv6_sock_handler() David Gibson
2024-01-06 16:00 ` Stefano Brivio
2024-01-07 0:45 ` David Gibson
2023-12-21 6:53 ` [PATCH v2 10/12] icmp: Warn on receive errors from ping sockets David Gibson
2023-12-21 6:53 ` [PATCH v2 11/12] icmp: Validate packets received on " David Gibson
2023-12-21 6:53 ` [PATCH v2 12/12] icmp: Dedicated functions for starting and closing ping sequences David Gibson
2024-01-06 16:01 ` Stefano Brivio
2024-01-07 4:30 ` David Gibson [this message]
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=ZZoo6DdqQ7X56egb@zatzit \
--to=david@gibson.dropbear.id.au \
--cc=passt-dev@passt.top \
--cc=sbrivio@redhat.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.
Code repositories for project(s) associated with this public inbox
https://passt.top/passt
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 IMAP folder(s).