From: David Gibson <david@gibson.dropbear.id.au>
To: passt-dev@passt.top, Stefano Brivio <sbrivio@redhat.com>
Cc: David Gibson <david@gibson.dropbear.id.au>
Subject: [PATCH v3 10/16] flow,tcp: Generalise TCP epoll_ref to generic flows
Date: Thu, 30 Nov 2023 13:02:16 +1100 [thread overview]
Message-ID: <20231130020222.4056647-11-david@gibson.dropbear.id.au> (raw)
In-Reply-To: <20231130020222.4056647-1-david@gibson.dropbear.id.au>
TCP uses three different epoll object types: one for connected sockets, one
for timers and one for listening sockets. Listening sockets really need
information that's specific to TCP, so need their own epoll_ref field.
Timers and connected sockets, however, only need the connection (flow)
they're associated with. As we expand the use of the flow table, we expect
that to be true for more epoll fds. So, rename the "TCP" epoll_ref field
to be a "flow" epoll_ref field that can be used both for TCP and for other
future cases.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
passt.h | 6 +++---
tcp.c | 10 +++++-----
tcp_splice.c | 4 ++--
3 files changed, 10 insertions(+), 10 deletions(-)
diff --git a/passt.h b/passt.h
index 0fce637..66a819f 100644
--- a/passt.h
+++ b/passt.h
@@ -76,8 +76,8 @@ enum epoll_type {
* union epoll_ref - Breakdown of reference for epoll fd bookkeeping
* @type: Type of fd (tells us what to do with events)
* @fd: File descriptor number (implies < 2^24 total descriptors)
- * @tcp: TCP-specific reference part (connected sockets)
- * @tcp_listen: TCP-specific reference part (listening sockets)
+ * @flow: Index of the flow this fd is linked to
+ * @tcp_listen: TCP-specific reference part for listening sockets
* @udp: UDP-specific reference part
* @icmp: ICMP-specific reference part
* @data: Data handled by protocol handlers
@@ -90,7 +90,7 @@ union epoll_ref {
#define FD_REF_MAX ((int)MAX_FROM_BITS(FD_REF_BITS))
int32_t fd:FD_REF_BITS;
union {
- union tcp_epoll_ref tcp;
+ uint32_t flow;
union tcp_listen_epoll_ref tcp_listen;
union udp_epoll_ref udp;
union icmp_epoll_ref icmp;
diff --git a/tcp.c b/tcp.c
index 512baf3..e80a6aa 100644
--- a/tcp.c
+++ b/tcp.c
@@ -639,7 +639,7 @@ static int tcp_epoll_ctl(const struct ctx *c, struct tcp_tap_conn *conn)
{
int m = conn->in_epoll ? EPOLL_CTL_MOD : EPOLL_CTL_ADD;
union epoll_ref ref = { .type = EPOLL_TYPE_TCP, .fd = conn->sock,
- .tcp.index = FLOW_IDX(conn) };
+ .flow = FLOW_IDX(conn) };
struct epoll_event ev = { .data.u64 = ref.u64 };
if (conn->events == CLOSED) {
@@ -660,7 +660,7 @@ static int tcp_epoll_ctl(const struct ctx *c, struct tcp_tap_conn *conn)
if (conn->timer != -1) {
union epoll_ref ref_t = { .type = EPOLL_TYPE_TCP_TIMER,
.fd = conn->sock,
- .tcp.index = FLOW_IDX(conn) };
+ .flow = FLOW_IDX(conn) };
struct epoll_event ev_t = { .data.u64 = ref_t.u64,
.events = EPOLLIN | EPOLLET };
@@ -688,7 +688,7 @@ static void tcp_timer_ctl(const struct ctx *c, struct tcp_tap_conn *conn)
if (conn->timer == -1) {
union epoll_ref ref = { .type = EPOLL_TYPE_TCP_TIMER,
.fd = conn->sock,
- .tcp.index = FLOW_IDX(conn) };
+ .flow = FLOW_IDX(conn) };
struct epoll_event ev = { .data.u64 = ref.u64,
.events = EPOLLIN | EPOLLET };
int fd;
@@ -2760,7 +2760,7 @@ void tcp_listen_handler(struct ctx *c, union epoll_ref ref,
void tcp_timer_handler(struct ctx *c, union epoll_ref ref)
{
struct itimerspec check_armed = { { 0 }, { 0 } };
- struct tcp_tap_conn *conn = CONN(ref.tcp.index);
+ struct tcp_tap_conn *conn = CONN(ref.flow);
if (c->no_tcp)
return;
@@ -2874,7 +2874,7 @@ static void tcp_tap_sock_handler(struct ctx *c, struct tcp_tap_conn *conn,
*/
void tcp_sock_handler(struct ctx *c, union epoll_ref ref, uint32_t events)
{
- union flow *flow = flowtab + ref.tcp.index;
+ union flow *flow = FLOW(ref.flow);
switch (flow->f.type) {
case FLOW_TCP:
diff --git a/tcp_splice.c b/tcp_splice.c
index 28b91e1..5ebc4e5 100644
--- a/tcp_splice.c
+++ b/tcp_splice.c
@@ -128,8 +128,8 @@ static int tcp_splice_epoll_ctl(const struct ctx *c,
{
int m = conn->in_epoll ? EPOLL_CTL_MOD : EPOLL_CTL_ADD;
union epoll_ref ref[SIDES] = {
- { .type = EPOLL_TYPE_TCP, .fd = conn->s[0], .tcp.index = FLOW_IDX(conn) },
- { .type = EPOLL_TYPE_TCP, .fd = conn->s[1], .tcp.index = FLOW_IDX(conn) }
+ { .type = EPOLL_TYPE_TCP, .fd = conn->s[0], .flow = FLOW_IDX(conn) },
+ { .type = EPOLL_TYPE_TCP, .fd = conn->s[1], .flow = FLOW_IDX(conn) }
};
struct epoll_event ev[SIDES] = { { .data.u64 = ref[0].u64 },
{ .data.u64 = ref[1].u64 } };
--
@@ -128,8 +128,8 @@ static int tcp_splice_epoll_ctl(const struct ctx *c,
{
int m = conn->in_epoll ? EPOLL_CTL_MOD : EPOLL_CTL_ADD;
union epoll_ref ref[SIDES] = {
- { .type = EPOLL_TYPE_TCP, .fd = conn->s[0], .tcp.index = FLOW_IDX(conn) },
- { .type = EPOLL_TYPE_TCP, .fd = conn->s[1], .tcp.index = FLOW_IDX(conn) }
+ { .type = EPOLL_TYPE_TCP, .fd = conn->s[0], .flow = FLOW_IDX(conn) },
+ { .type = EPOLL_TYPE_TCP, .fd = conn->s[1], .flow = FLOW_IDX(conn) }
};
struct epoll_event ev[SIDES] = { { .data.u64 = ref[0].u64 },
{ .data.u64 = ref[1].u64 } };
--
2.43.0
next prev parent reply other threads:[~2023-11-30 2:02 UTC|newest]
Thread overview: 25+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-11-30 2:02 [PATCH v3 00/16] Introduce unified flow table, first steps David Gibson
2023-11-30 2:02 ` [PATCH v3 01/16] treewide: Add messages to static_assert() calls David Gibson
2023-11-30 2:02 ` [PATCH v3 02/16] flow, tcp: Generalise connection types David Gibson
2023-11-30 2:02 ` [PATCH v3 03/16] flow, tcp: Move TCP connection table to unified flow table David Gibson
2023-11-30 2:02 ` [PATCH v3 04/16] flow, tcp: Consolidate flow pointer<->index helpers David Gibson
2023-11-30 2:02 ` [PATCH v3 05/16] util: MAX_FROM_BITS() should be unsigned David Gibson
2023-11-30 2:02 ` [PATCH v3 06/16] flow: Make unified version of flow table compaction David Gibson
2023-11-30 2:02 ` [PATCH v3 07/16] flow, tcp: Add logging helpers for connection related messages David Gibson
2023-11-30 2:02 ` [PATCH v3 08/16] flow: Introduce 'sidx' type to represent one side of one flow David Gibson
2023-12-02 4:35 ` Stefano Brivio
2023-11-30 2:02 ` [PATCH v3 09/16] tcp: Remove unneccessary bounds check in tcp_timer_handler() David Gibson
2023-11-30 2:02 ` David Gibson [this message]
2023-11-30 2:02 ` [PATCH v3 11/16] tcp_splice: Use unsigned to represent side David Gibson
2023-11-30 2:02 ` [PATCH v3 12/16] flow,tcp: Use epoll_ref type including flow and side David Gibson
2023-11-30 2:02 ` [PATCH v3 13/16] test: Avoid hitting guestfish command length limits David Gibson
2023-11-30 2:02 ` [PATCH v3 14/16] pif: Add helpers to get the name of a pif David Gibson
2023-11-30 12:45 ` Stefano Brivio
2023-12-01 0:08 ` David Gibson
2023-11-30 2:02 ` [PATCH v3 15/16] tcp: "TCP" hash secret doesn't need to be TCP specific David Gibson
2023-11-30 2:02 ` [PATCH v3 16/16] tcp: Don't defer hash table removal David Gibson
2023-11-30 12:45 ` Stefano Brivio
2023-12-01 0:07 ` David Gibson
2023-12-02 4:34 ` Stefano Brivio
2023-12-04 0:43 ` David Gibson
2023-12-04 9:54 ` [PATCH v3 00/16] Introduce unified flow table, first steps Stefano Brivio
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=20231130020222.4056647-11-david@gibson.dropbear.id.au \
--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).