Re: svn commit: r198845 - head/sys/netinet/ipfw
On Tue, Nov 03, 2009 at 08:41:14AM +, Oleg Bulyzhin wrote: > Author: oleg > Date: Tue Nov 3 08:41:14 2009 > New Revision: 198845 > URL: http://svn.freebsd.org/changeset/base/198845 > > Log: > Fix two issues that can lead to exceeding configured pipe bandwidth: > - do not expire queues which are not ready to be expired. > - properly calculate available burst size. > > MFC after: 3 days > > Modified: > head/sys/netinet/ipfw/ip_dummynet.c Reviews are welcome, since i'm going to push this into 8.0-RELEASE. Thanks. -- Oleg. ==== === Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru === ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r198845 - head/sys/netinet/ipfw
Author: oleg
Date: Tue Nov 3 08:41:14 2009
New Revision: 198845
URL: http://svn.freebsd.org/changeset/base/198845
Log:
Fix two issues that can lead to exceeding configured pipe bandwidth:
- do not expire queues which are not ready to be expired.
- properly calculate available burst size.
MFC after:3 days
Modified:
head/sys/netinet/ipfw/ip_dummynet.c
Modified: head/sys/netinet/ipfw/ip_dummynet.c
==
--- head/sys/netinet/ipfw/ip_dummynet.c Tue Nov 3 07:29:58 2009
(r198844)
+++ head/sys/netinet/ipfw/ip_dummynet.c Tue Nov 3 08:41:14 2009
(r198845)
@@ -244,6 +244,17 @@ void dummynet_drain(void);
static int dummynet_io(struct mbuf **, int , struct ip_fw_args *);
/*
+ * Flow queue is idle if:
+ * 1) it's empty for at least 1 tick
+ * 2) it has invalid timestamp (WF2Q case)
+ * 3) parent pipe has no 'exhausted' burst.
+ */
+#define QUEUE_IS_IDLE(q) ((q)->head == NULL && (q)->S == (q)->F + 1 && \
+ curr_time > (q)->idle_time + 1 && \
+ ((q)->numbytes + (curr_time - (q)->idle_time - 1) * \
+ (q)->fs->pipe->bandwidth >= q->fs->pipe->burst))
+
+/*
* Heap management functions.
*
* In the heap, first node is element 0. Children of i are 2i+1 and 2i+2.
@@ -1004,7 +1015,7 @@ expire_queues(struct dn_flow_set *fs)
fs->last_expired = time_uptime ;
for (i = 0 ; i <= fs->rq_size ; i++) /* last one is overflow */
for (prev=NULL, q = fs->rq[i] ; q != NULL ; )
- if (q->head != NULL || q->S != q->F+1) {
+ if (!QUEUE_IS_IDLE(q)) {
prev = q ;
q = q->next ;
} else { /* entry is idle, expire it */
@@ -1134,7 +1145,7 @@ find_queue(struct dn_flow_set *fs, struc
break ; /* found */
/* No match. Check if we can expire the entry */
- if (pipe_expire && q->head == NULL && q->S == q->F+1 ) {
+ if (pipe_expire && QUEUE_IS_IDLE(q)) {
/* entry is idle and not in any heap, expire it */
struct dn_flow_queue *old_q = q ;
@@ -1408,7 +1419,7 @@ dummynet_io(struct mbuf **m0, int dir, s
if (q->idle_time < curr_time) {
/* Calculate available burst size. */
q->numbytes +=
- (curr_time - q->idle_time) * pipe->bandwidth;
+ (curr_time - q->idle_time - 1) * pipe->bandwidth;
if (q->numbytes > pipe->burst)
q->numbytes = pipe->burst;
if (io_fast)
@@ -1418,8 +1429,8 @@ dummynet_io(struct mbuf **m0, int dir, s
if (pipe->idle_time < curr_time) {
/* Calculate available burst size. */
pipe->numbytes +=
- (curr_time - pipe->idle_time) * pipe->bandwidth;
- if (pipe->numbytes > pipe->burst)
+ (curr_time - pipe->idle_time - 1) * pipe->bandwidth;
+ if (pipe->numbytes > 0 && pipe->numbytes > pipe->burst)
pipe->numbytes = pipe->burst;
if (io_fast)
pipe->numbytes += pipe->bandwidth;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r199073 - head/sys/netinet/ipfw
Author: oleg Date: Mon Nov 9 09:12:45 2009 New Revision: 199073 URL: http://svn.freebsd.org/changeset/base/199073 Log: style(9): add missing parentheses Modified: head/sys/netinet/ipfw/ip_dummynet.c Modified: head/sys/netinet/ipfw/ip_dummynet.c == --- head/sys/netinet/ipfw/ip_dummynet.c Mon Nov 9 08:54:47 2009 (r199072) +++ head/sys/netinet/ipfw/ip_dummynet.c Mon Nov 9 09:12:45 2009 (r199073) @@ -252,7 +252,7 @@ static int dummynet_io(struct mbuf **, i #define QUEUE_IS_IDLE(q) ((q)->head == NULL && (q)->S == (q)->F + 1 && \ curr_time > (q)->idle_time + 1 && \ ((q)->numbytes + (curr_time - (q)->idle_time - 1) * \ - (q)->fs->pipe->bandwidth >= q->fs->pipe->burst)) + (q)->fs->pipe->bandwidth >= (q)->fs->pipe->burst)) /* * Heap management functions. ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r199075 - stable/8/sys/netinet/ipfw
Author: oleg
Date: Mon Nov 9 10:13:24 2009
New Revision: 199075
URL: http://svn.freebsd.org/changeset/base/199075
Log:
MFC r198845:
Fix two issues that can lead to exceeding configured pipe bandwidth:
- do not expire queues which are not ready to be expired.
- properly calculate available burst size.
MFC r199073:
style(9): add missing parentheses
Modified:
stable/8/sys/netinet/ipfw/ip_dummynet.c
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/amd64/include/xen/ (props changed)
stable/8/sys/cddl/contrib/opensolaris/ (props changed)
stable/8/sys/contrib/dev/acpica/ (props changed)
stable/8/sys/contrib/pf/ (props changed)
stable/8/sys/dev/xen/xenpci/ (props changed)
Modified: stable/8/sys/netinet/ipfw/ip_dummynet.c
==
--- stable/8/sys/netinet/ipfw/ip_dummynet.c Mon Nov 9 09:27:09 2009
(r199074)
+++ stable/8/sys/netinet/ipfw/ip_dummynet.c Mon Nov 9 10:13:24 2009
(r199075)
@@ -244,6 +244,17 @@ void dummynet_drain(void);
static int dummynet_io(struct mbuf **, int , struct ip_fw_args *);
/*
+ * Flow queue is idle if:
+ * 1) it's empty for at least 1 tick
+ * 2) it has invalid timestamp (WF2Q case)
+ * 3) parent pipe has no 'exhausted' burst.
+ */
+#define QUEUE_IS_IDLE(q) ((q)->head == NULL && (q)->S == (q)->F + 1 && \
+ curr_time > (q)->idle_time + 1 && \
+ ((q)->numbytes + (curr_time - (q)->idle_time - 1) * \
+ (q)->fs->pipe->bandwidth >= (q)->fs->pipe->burst))
+
+/*
* Heap management functions.
*
* In the heap, first node is element 0. Children of i are 2i+1 and 2i+2.
@@ -1004,7 +1015,7 @@ expire_queues(struct dn_flow_set *fs)
fs->last_expired = time_uptime ;
for (i = 0 ; i <= fs->rq_size ; i++) /* last one is overflow */
for (prev=NULL, q = fs->rq[i] ; q != NULL ; )
- if (q->head != NULL || q->S != q->F+1) {
+ if (!QUEUE_IS_IDLE(q)) {
prev = q ;
q = q->next ;
} else { /* entry is idle, expire it */
@@ -1134,7 +1145,7 @@ find_queue(struct dn_flow_set *fs, struc
break ; /* found */
/* No match. Check if we can expire the entry */
- if (pipe_expire && q->head == NULL && q->S == q->F+1 ) {
+ if (pipe_expire && QUEUE_IS_IDLE(q)) {
/* entry is idle and not in any heap, expire it */
struct dn_flow_queue *old_q = q ;
@@ -1408,7 +1419,7 @@ dummynet_io(struct mbuf **m0, int dir, s
if (q->idle_time < curr_time) {
/* Calculate available burst size. */
q->numbytes +=
- (curr_time - q->idle_time) * pipe->bandwidth;
+ (curr_time - q->idle_time - 1) * pipe->bandwidth;
if (q->numbytes > pipe->burst)
q->numbytes = pipe->burst;
if (io_fast)
@@ -1418,8 +1429,8 @@ dummynet_io(struct mbuf **m0, int dir, s
if (pipe->idle_time < curr_time) {
/* Calculate available burst size. */
pipe->numbytes +=
- (curr_time - pipe->idle_time) * pipe->bandwidth;
- if (pipe->numbytes > pipe->burst)
+ (curr_time - pipe->idle_time - 1) * pipe->bandwidth;
+ if (pipe->numbytes > 0 && pipe->numbytes > pipe->burst)
pipe->numbytes = pipe->burst;
if (io_fast)
pipe->numbytes += pipe->bandwidth;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r216730 - stable/8/sys/net
Author: oleg
Date: Mon Dec 27 06:59:59 2010
New Revision: 216730
URL: http://svn.freebsd.org/changeset/base/216730
Log:
MFC r203548:
Propagate the vlan events to the underlying interfaces/members so they
can do initialization of hw related features.
PR: kern/141646
Modified:
stable/8/sys/net/if_lagg.c
stable/8/sys/net/if_lagg.h
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/amd64/include/xen/ (props changed)
stable/8/sys/cddl/contrib/opensolaris/ (props changed)
stable/8/sys/contrib/dev/acpica/ (props changed)
stable/8/sys/contrib/pf/ (props changed)
Modified: stable/8/sys/net/if_lagg.c
==
--- stable/8/sys/net/if_lagg.c Mon Dec 27 05:47:24 2010(r216729)
+++ stable/8/sys/net/if_lagg.c Mon Dec 27 06:59:59 2010(r216730)
@@ -39,6 +39,7 @@ __FBSDID("$FreeBSD$");
#include
#include
#include
+#include
#include
#include
@@ -204,6 +205,50 @@ static moduledata_t lagg_mod = {
DECLARE_MODULE(if_lagg, lagg_mod, SI_SUB_PSEUDO, SI_ORDER_ANY);
+#if __FreeBSD_version >= 80
+/*
+ * This routine is run via an vlan
+ * config EVENT
+ */
+static void
+lagg_register_vlan(void *arg, struct ifnet *ifp, u_int16_t vtag)
+{
+struct lagg_softc *sc = ifp->if_softc;
+struct lagg_port*lp;
+
+if (ifp->if_softc != arg) /* Not our event */
+return;
+
+LAGG_RLOCK(sc);
+if (!SLIST_EMPTY(&sc->sc_ports)) {
+SLIST_FOREACH(lp, &sc->sc_ports, lp_entries)
+EVENTHANDLER_INVOKE(vlan_config, lp->lp_ifp, vtag);
+}
+LAGG_RUNLOCK(sc);
+}
+
+/*
+ * This routine is run via an vlan
+ * unconfig EVENT
+ */
+static void
+lagg_unregister_vlan(void *arg, struct ifnet *ifp, u_int16_t vtag)
+{
+struct lagg_softc *sc = ifp->if_softc;
+struct lagg_port*lp;
+
+if (ifp->if_softc != arg) /* Not our event */
+return;
+
+LAGG_RLOCK(sc);
+if (!SLIST_EMPTY(&sc->sc_ports)) {
+SLIST_FOREACH(lp, &sc->sc_ports, lp_entries)
+EVENTHANDLER_INVOKE(vlan_unconfig, lp->lp_ifp, vtag);
+}
+LAGG_RUNLOCK(sc);
+}
+#endif
+
static int
lagg_clone_create(struct if_clone *ifc, int unit, caddr_t params)
{
@@ -259,6 +304,13 @@ lagg_clone_create(struct if_clone *ifc,
*/
ether_ifattach(ifp, eaddr);
+#if __FreeBSD_version >= 80
+ sc->vlan_attach = EVENTHANDLER_REGISTER(vlan_config,
+ lagg_register_vlan, sc, EVENTHANDLER_PRI_FIRST);
+ sc->vlan_detach = EVENTHANDLER_REGISTER(vlan_unconfig,
+ lagg_unregister_vlan, sc, EVENTHANDLER_PRI_FIRST);
+#endif
+
/* Insert into the global list of laggs */
mtx_lock(&lagg_list_mtx);
SLIST_INSERT_HEAD(&lagg_list, sc, sc_entries);
@@ -278,6 +330,11 @@ lagg_clone_destroy(struct ifnet *ifp)
lagg_stop(sc);
ifp->if_flags &= ~IFF_UP;
+#if __FreeBSD_version >= 80
+ EVENTHANDLER_DEREGISTER(vlan_config, sc->vlan_attach);
+ EVENTHANDLER_DEREGISTER(vlan_unconfig, sc->vlan_detach);
+#endif
+
/* Shutdown and remove lagg ports */
while ((lp = SLIST_FIRST(&sc->sc_ports)) != NULL)
lagg_port_destroy(lp, 1);
Modified: stable/8/sys/net/if_lagg.h
==
--- stable/8/sys/net/if_lagg.h Mon Dec 27 05:47:24 2010(r216729)
+++ stable/8/sys/net/if_lagg.h Mon Dec 27 06:59:59 2010(r216730)
@@ -198,6 +198,10 @@ struct lagg_softc {
void(*sc_lladdr)(struct lagg_softc *);
void(*sc_req)(struct lagg_softc *, caddr_t);
void(*sc_portreq)(struct lagg_port *, caddr_t);
+#if __FreeBSD_version >= 80
+ eventhandler_tag vlan_attach;
+ eventhandler_tag vlan_detach;
+#endif
};
struct lagg_port {
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r216741 - releng/8.2/sys/net
Author: oleg
Date: Mon Dec 27 18:55:16 2010
New Revision: 216741
URL: http://svn.freebsd.org/changeset/base/216741
Log:
MFC r203548:
Propagate the vlan events to the underlying interfaces/members so they
can do initialization of hw related features.
PR: kern/141646
Approved by: re (bz)
Modified:
releng/8.2/sys/net/if_lagg.c
releng/8.2/sys/net/if_lagg.h
Directory Properties:
releng/8.2/sys/ (props changed)
releng/8.2/sys/amd64/include/xen/ (props changed)
releng/8.2/sys/cddl/contrib/opensolaris/ (props changed)
releng/8.2/sys/contrib/dev/acpica/ (props changed)
releng/8.2/sys/contrib/pf/ (props changed)
Modified: releng/8.2/sys/net/if_lagg.c
==
--- releng/8.2/sys/net/if_lagg.cMon Dec 27 16:02:14 2010
(r216740)
+++ releng/8.2/sys/net/if_lagg.cMon Dec 27 18:55:16 2010
(r216741)
@@ -39,6 +39,7 @@ __FBSDID("$FreeBSD$");
#include
#include
#include
+#include
#include
#include
@@ -204,6 +205,50 @@ static moduledata_t lagg_mod = {
DECLARE_MODULE(if_lagg, lagg_mod, SI_SUB_PSEUDO, SI_ORDER_ANY);
+#if __FreeBSD_version >= 80
+/*
+ * This routine is run via an vlan
+ * config EVENT
+ */
+static void
+lagg_register_vlan(void *arg, struct ifnet *ifp, u_int16_t vtag)
+{
+struct lagg_softc *sc = ifp->if_softc;
+struct lagg_port*lp;
+
+if (ifp->if_softc != arg) /* Not our event */
+return;
+
+LAGG_RLOCK(sc);
+if (!SLIST_EMPTY(&sc->sc_ports)) {
+SLIST_FOREACH(lp, &sc->sc_ports, lp_entries)
+EVENTHANDLER_INVOKE(vlan_config, lp->lp_ifp, vtag);
+}
+LAGG_RUNLOCK(sc);
+}
+
+/*
+ * This routine is run via an vlan
+ * unconfig EVENT
+ */
+static void
+lagg_unregister_vlan(void *arg, struct ifnet *ifp, u_int16_t vtag)
+{
+struct lagg_softc *sc = ifp->if_softc;
+struct lagg_port*lp;
+
+if (ifp->if_softc != arg) /* Not our event */
+return;
+
+LAGG_RLOCK(sc);
+if (!SLIST_EMPTY(&sc->sc_ports)) {
+SLIST_FOREACH(lp, &sc->sc_ports, lp_entries)
+EVENTHANDLER_INVOKE(vlan_unconfig, lp->lp_ifp, vtag);
+}
+LAGG_RUNLOCK(sc);
+}
+#endif
+
static int
lagg_clone_create(struct if_clone *ifc, int unit, caddr_t params)
{
@@ -259,6 +304,13 @@ lagg_clone_create(struct if_clone *ifc,
*/
ether_ifattach(ifp, eaddr);
+#if __FreeBSD_version >= 80
+ sc->vlan_attach = EVENTHANDLER_REGISTER(vlan_config,
+ lagg_register_vlan, sc, EVENTHANDLER_PRI_FIRST);
+ sc->vlan_detach = EVENTHANDLER_REGISTER(vlan_unconfig,
+ lagg_unregister_vlan, sc, EVENTHANDLER_PRI_FIRST);
+#endif
+
/* Insert into the global list of laggs */
mtx_lock(&lagg_list_mtx);
SLIST_INSERT_HEAD(&lagg_list, sc, sc_entries);
@@ -278,6 +330,11 @@ lagg_clone_destroy(struct ifnet *ifp)
lagg_stop(sc);
ifp->if_flags &= ~IFF_UP;
+#if __FreeBSD_version >= 80
+ EVENTHANDLER_DEREGISTER(vlan_config, sc->vlan_attach);
+ EVENTHANDLER_DEREGISTER(vlan_unconfig, sc->vlan_detach);
+#endif
+
/* Shutdown and remove lagg ports */
while ((lp = SLIST_FIRST(&sc->sc_ports)) != NULL)
lagg_port_destroy(lp, 1);
Modified: releng/8.2/sys/net/if_lagg.h
==
--- releng/8.2/sys/net/if_lagg.hMon Dec 27 16:02:14 2010
(r216740)
+++ releng/8.2/sys/net/if_lagg.hMon Dec 27 18:55:16 2010
(r216741)
@@ -198,6 +198,10 @@ struct lagg_softc {
void(*sc_lladdr)(struct lagg_softc *);
void(*sc_req)(struct lagg_softc *, caddr_t);
void(*sc_portreq)(struct lagg_port *, caddr_t);
+#if __FreeBSD_version >= 80
+ eventhandler_tag vlan_attach;
+ eventhandler_tag vlan_detach;
+#endif
};
struct lagg_port {
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r216757 - stable/8/sys/netinet/ipfw
Author: oleg
Date: Tue Dec 28 11:34:59 2010
New Revision: 216757
URL: http://svn.freebsd.org/changeset/base/216757
Log:
MFC r213265:
Fix handling of initial credit for an idle pipe. This fixes the bug where
setting bw > 1 MTU/tick resulted in infinite bandwidth if io_fast=1
PR: kern/147245, kern/148429
Obtained from:Riccardo Panicucci
Modified:
stable/8/sys/netinet/ipfw/ip_dn_io.c
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/amd64/include/xen/ (props changed)
stable/8/sys/cddl/contrib/opensolaris/ (props changed)
stable/8/sys/contrib/dev/acpica/ (props changed)
stable/8/sys/contrib/pf/ (props changed)
Modified: stable/8/sys/netinet/ipfw/ip_dn_io.c
==
--- stable/8/sys/netinet/ipfw/ip_dn_io.cTue Dec 28 10:08:50 2010
(r216756)
+++ stable/8/sys/netinet/ipfw/ip_dn_io.cTue Dec 28 11:34:59 2010
(r216757)
@@ -742,8 +742,11 @@ dummynet_io(struct mbuf **m0, int dir, s
}
/* compute the initial allowance */
- {
+ if (si->idle_time < dn_cfg.curr_time) {
+ /* Do this only on the first packet on an idle pipe */
struct dn_link *p = &fs->sched->link;
+
+ si->sched_time = dn_cfg.curr_time;
si->credit = dn_cfg.io_fast ? p->bandwidth : 0;
if (p->burst) {
uint64_t burst = (dn_cfg.curr_time - si->idle_time) *
p->bandwidth;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r216759 - releng/8.2/sys/netinet/ipfw
Author: oleg
Date: Tue Dec 28 12:18:46 2010
New Revision: 216759
URL: http://svn.freebsd.org/changeset/base/216759
Log:
MFC r213265:
Fix handling of initial credit for an idle pipe. This fixes the bug where
setting bw > 1 MTU/tick resulted in infinite bandwidth if io_fast=1
PR: kern/147245, kern/148429
Obtained from:Riccardo Panicucci
Approved by: re (bz)
Modified:
releng/8.2/sys/netinet/ipfw/ip_dn_io.c
Directory Properties:
releng/8.2/sys/ (props changed)
releng/8.2/sys/amd64/include/xen/ (props changed)
releng/8.2/sys/cddl/contrib/opensolaris/ (props changed)
releng/8.2/sys/contrib/dev/acpica/ (props changed)
releng/8.2/sys/contrib/pf/ (props changed)
Modified: releng/8.2/sys/netinet/ipfw/ip_dn_io.c
==
--- releng/8.2/sys/netinet/ipfw/ip_dn_io.c Tue Dec 28 12:13:30 2010
(r216758)
+++ releng/8.2/sys/netinet/ipfw/ip_dn_io.c Tue Dec 28 12:18:46 2010
(r216759)
@@ -742,8 +742,11 @@ dummynet_io(struct mbuf **m0, int dir, s
}
/* compute the initial allowance */
- {
+ if (si->idle_time < dn_cfg.curr_time) {
+ /* Do this only on the first packet on an idle pipe */
struct dn_link *p = &fs->sched->link;
+
+ si->sched_time = dn_cfg.curr_time;
si->credit = dn_cfg.io_fast ? p->bandwidth : 0;
if (p->burst) {
uint64_t burst = (dn_cfg.curr_time - si->idle_time) *
p->bandwidth;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r190865 - in head: sbin/ipfw sys/netinet
Converting numbytes (struct dn_pipe) to dn_key (which is uint64_t) may break WFQ. ready_event_wfq() code depends on signed 'numbytes'. P.S. i vote for modifing ready_event_wfq() to deal with unsigned numbytes. -- Oleg. ==== === Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru === ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r191570 - head/sys/netinet
Author: oleg
Date: Mon Apr 27 17:37:36 2009
New Revision: 191570
URL: http://svn.freebsd.org/changeset/base/191570
Log:
Optimize packet flow: if net.inet.ip.fw.one_pass != 0 and packet was
processed by ipfw once - avoid second ipfw_chk() call.
This saves us from unnecessary IPFW_RLOCK(), m_tag_find() calls and
ip/tcp/udp header parsing.
MFC after:2 month
Modified:
head/sys/netinet/ip_fw2.c
head/sys/netinet/ip_fw_pfil.c
Modified: head/sys/netinet/ip_fw2.c
==
--- head/sys/netinet/ip_fw2.c Mon Apr 27 17:36:41 2009(r191569)
+++ head/sys/netinet/ip_fw2.c Mon Apr 27 17:37:36 2009(r191570)
@@ -2515,16 +2515,7 @@ do {
\
/*
* Packet has already been tagged. Look for the next rule
* to restart processing.
-*
-* If fw_one_pass != 0 then just accept it.
-* XXX should not happen here, but optimized out in
-* the caller.
*/
- if (V_fw_one_pass) {
- IPFW_RUNLOCK(chain);
- return (IP_FW_PASS);
- }
-
f = args->rule->next_rule;
if (f == NULL)
f = lookup_next_rule(args->rule, 0);
Modified: head/sys/netinet/ip_fw_pfil.c
==
--- head/sys/netinet/ip_fw_pfil.c Mon Apr 27 17:36:41 2009
(r191569)
+++ head/sys/netinet/ip_fw_pfil.c Mon Apr 27 17:37:36 2009
(r191570)
@@ -51,7 +51,6 @@ __FBSDID("$FreeBSD$");
#include
#include
-#define _NET_IF_VAR_H_ /* we don't want if_var.h, only if.h */
#include
#include
#include
@@ -63,6 +62,7 @@ __FBSDID("$FreeBSD$");
#include
#include
#include
+#include
#include
@@ -131,10 +131,14 @@ again:
args.m = *m0;
args.inp = inp;
- ipfw = ipfw_chk(&args);
- *m0 = args.m;
tee = 0;
+ if (V_fw_one_pass == 0 || args.rule == NULL) {
+ ipfw = ipfw_chk(&args);
+ *m0 = args.m;
+ } else
+ ipfw = IP_FW_PASS;
+
KASSERT(*m0 != NULL || ipfw == IP_FW_DENY, ("%s: m0 is NULL",
__func__));
@@ -257,10 +261,14 @@ again:
args.m = *m0;
args.oif = ifp;
args.inp = inp;
- ipfw = ipfw_chk(&args);
- *m0 = args.m;
tee = 0;
+ if (V_fw_one_pass == 0 || args.rule == NULL) {
+ ipfw = ipfw_chk(&args);
+ *m0 = args.m;
+ } else
+ ipfw = IP_FW_PASS;
+
KASSERT(*m0 != NULL || ipfw == IP_FW_DENY, ("%s: m0 is NULL",
__func__));
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r253256 - head/sbin/recoverdisk
Author: oleg Date: Fri Jul 12 09:20:55 2013 New Revision: 253256 URL: http://svnweb.freebsd.org/changeset/base/253256 Log: Fix 'SEE ALSO' list. Modified: head/sbin/recoverdisk/recoverdisk.1 Modified: head/sbin/recoverdisk/recoverdisk.1 == --- head/sbin/recoverdisk/recoverdisk.1 Fri Jul 12 08:03:10 2013 (r253255) +++ head/sbin/recoverdisk/recoverdisk.1 Fri Jul 12 09:20:55 2013 (r253256) @@ -125,9 +125,9 @@ recoverdisk -b 0 /dev/ad3 /somewhere .Ed .Sh SEE ALSO .Xr dd 1 , -.Xr ada 4, -.Xr cam 4, -.Xr cd 4, +.Xr ada 4 , +.Xr cam 4 , +.Xr cd 4 , .Xr da 4 .Sh HISTORY The ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r249628 - head/sys/net
Author: oleg
Date: Thu Apr 18 20:13:33 2013
New Revision: 249628
URL: http://svnweb.freebsd.org/changeset/base/249628
Log:
Recover missing arp_ifinit() call.
MFC after:2 weeks
Modified:
head/sys/net/if_vlan.c
Modified: head/sys/net/if_vlan.c
==
--- head/sys/net/if_vlan.c Thu Apr 18 19:52:11 2013(r249627)
+++ head/sys/net/if_vlan.c Thu Apr 18 20:13:33 2013(r249628)
@@ -41,6 +41,7 @@
#include
__FBSDID("$FreeBSD$");
+#include "opt_inet.h"
#include "opt_vlan.h"
#include
@@ -66,6 +67,11 @@ __FBSDID("$FreeBSD$");
#include
#include
+#ifdef INET
+#include
+#include
+#endif
+
#defineVLAN_DEF_HWIDTH 4
#defineVLAN_IFFLAGS(IFF_BROADCAST | IFF_MULTICAST)
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r249742 - in head/sys: netinet netinet6
Author: oleg
Date: Sun Apr 21 21:28:38 2013
New Revision: 249742
URL: http://svnweb.freebsd.org/changeset/base/249742
Log:
Plug static llentry leak (ipv4 & ipv6 were affected).
PR: kern/172985
MFC after:1 month
Modified:
head/sys/netinet/in.c
head/sys/netinet6/in6.c
Modified: head/sys/netinet/in.c
==
--- head/sys/netinet/in.c Sun Apr 21 20:51:53 2013(r249741)
+++ head/sys/netinet/in.c Sun Apr 21 21:28:38 2013(r249742)
@@ -1467,10 +1467,14 @@ in_lltable_lookup(struct lltable *llt, u
LLE_WLOCK(lle);
lle->la_flags |= LLE_DELETED;
EVENTHANDLER_INVOKE(lle_event, lle, LLENTRY_DELETED);
- LLE_WUNLOCK(lle);
#ifdef DIAGNOSTIC
- log(LOG_INFO, "ifaddr cache = %p is deleted\n", lle);
+ log(LOG_INFO, "ifaddr cache = %p is deleted\n", lle);
#endif
+ if ((lle->la_flags &
+ (LLE_STATIC | LLE_IFADDR)) == LLE_STATIC)
+ llentry_free(lle);
+ else
+ LLE_WUNLOCK(lle);
}
lle = (void *)-1;
Modified: head/sys/netinet6/in6.c
==
--- head/sys/netinet6/in6.c Sun Apr 21 20:51:53 2013(r249741)
+++ head/sys/netinet6/in6.c Sun Apr 21 21:28:38 2013(r249742)
@@ -195,10 +195,10 @@ in6_ifremloop(struct ifaddr *ifa)
ia = ifa2ia6(ifa);
ifp = ifa->ifa_ifp;
- IF_AFDATA_LOCK(ifp);
- lla_lookup(LLTABLE6(ifp), (LLE_DELETE | LLE_IFADDR),
- (struct sockaddr *)&ia->ia_addr);
- IF_AFDATA_UNLOCK(ifp);
+ memcpy(&addr, &ia->ia_addr, sizeof(ia->ia_addr));
+ memcpy(&mask, &ia->ia_prefixmask, sizeof(ia->ia_prefixmask));
+ lltable_prefix_free(AF_INET6, (struct sockaddr *)&addr,
+ (struct sockaddr *)&mask, LLE_STATIC);
/*
* initialize for rtmsg generation
@@ -210,8 +210,6 @@ in6_ifremloop(struct ifaddr *ifa)
gateway.sdl_alen = ifp->if_addrlen;
bzero(&rt0, sizeof(rt0));
rt0.rt_gateway = (struct sockaddr *)&gateway;
- memcpy(&mask, &ia->ia_prefixmask, sizeof(ia->ia_prefixmask));
- memcpy(&addr, &ia->ia_addr, sizeof(ia->ia_addr));
rt_mask(&rt0) = (struct sockaddr *)&mask;
rt_key(&rt0) = (struct sockaddr *)&addr;
rt0.rt_flags = RTF_HOST | RTF_STATIC;
@@ -2604,10 +2602,14 @@ in6_lltable_lookup(struct lltable *llt,
if (!(lle->la_flags & LLE_IFADDR) || (flags & LLE_IFADDR)) {
LLE_WLOCK(lle);
lle->la_flags |= LLE_DELETED;
- LLE_WUNLOCK(lle);
#ifdef DIAGNOSTIC
- log(LOG_INFO, "ifaddr cache = %p is deleted\n", lle);
+ log(LOG_INFO, "ifaddr cache = %p is deleted\n", lle);
#endif
+ if ((lle->la_flags &
+ (LLE_STATIC | LLE_IFADDR)) == LLE_STATIC)
+ llentry_free(lle);
+ else
+ LLE_WUNLOCK(lle);
}
lle = (void *)-1;
}
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r250927 - in stable/9/sys: net netinet netinet6
Author: oleg
Date: Thu May 23 11:29:28 2013
New Revision: 250927
URL: http://svnweb.freebsd.org/changeset/base/250927
Log:
MFC: r249628, r249742
- recover missing arp_ifinit() call.
- plug static llentry leak (ipv4 & ipv6 were affected).
PR: kern/172985
Modified:
stable/9/sys/net/if_vlan.c
stable/9/sys/netinet/in.c
stable/9/sys/netinet6/in6.c
Directory Properties:
stable/9/sys/ (props changed)
Modified: stable/9/sys/net/if_vlan.c
==
--- stable/9/sys/net/if_vlan.c Thu May 23 05:42:35 2013(r250926)
+++ stable/9/sys/net/if_vlan.c Thu May 23 11:29:28 2013(r250927)
@@ -41,6 +41,7 @@
#include
__FBSDID("$FreeBSD$");
+#include "opt_inet.h"
#include "opt_vlan.h"
#include
@@ -66,6 +67,11 @@ __FBSDID("$FreeBSD$");
#include
#include
+#ifdef INET
+#include
+#include
+#endif
+
#define VLANNAME "vlan"
#defineVLAN_DEF_HWIDTH 4
#defineVLAN_IFFLAGS(IFF_BROADCAST | IFF_MULTICAST)
Modified: stable/9/sys/netinet/in.c
==
--- stable/9/sys/netinet/in.c Thu May 23 05:42:35 2013(r250926)
+++ stable/9/sys/netinet/in.c Thu May 23 11:29:28 2013(r250927)
@@ -1528,10 +1528,14 @@ in_lltable_lookup(struct lltable *llt, u
LLE_WLOCK(lle);
lle->la_flags |= LLE_DELETED;
EVENTHANDLER_INVOKE(arp_update_event, lle);
- LLE_WUNLOCK(lle);
#ifdef DIAGNOSTIC
- log(LOG_INFO, "ifaddr cache = %p is deleted\n", lle);
+ log(LOG_INFO, "ifaddr cache = %p is deleted\n", lle);
#endif
+ if ((lle->la_flags &
+ (LLE_STATIC | LLE_IFADDR)) == LLE_STATIC)
+ llentry_free(lle);
+ else
+ LLE_WUNLOCK(lle);
}
lle = (void *)-1;
Modified: stable/9/sys/netinet6/in6.c
==
--- stable/9/sys/netinet6/in6.c Thu May 23 05:42:35 2013(r250926)
+++ stable/9/sys/netinet6/in6.c Thu May 23 11:29:28 2013(r250927)
@@ -194,10 +194,10 @@ in6_ifremloop(struct ifaddr *ifa)
ia = ifa2ia6(ifa);
ifp = ifa->ifa_ifp;
- IF_AFDATA_LOCK(ifp);
- lla_lookup(LLTABLE6(ifp), (LLE_DELETE | LLE_IFADDR),
- (struct sockaddr *)&ia->ia_addr);
- IF_AFDATA_UNLOCK(ifp);
+ memcpy(&addr, &ia->ia_addr, sizeof(ia->ia_addr));
+ memcpy(&mask, &ia->ia_prefixmask, sizeof(ia->ia_prefixmask));
+ lltable_prefix_free(AF_INET6, (struct sockaddr *)&addr,
+ (struct sockaddr *)&mask, LLE_STATIC);
/*
* initialize for rtmsg generation
@@ -209,8 +209,6 @@ in6_ifremloop(struct ifaddr *ifa)
gateway.sdl_alen = ifp->if_addrlen;
bzero(&rt0, sizeof(rt0));
rt0.rt_gateway = (struct sockaddr *)&gateway;
- memcpy(&mask, &ia->ia_prefixmask, sizeof(ia->ia_prefixmask));
- memcpy(&addr, &ia->ia_addr, sizeof(ia->ia_addr));
rt_mask(&rt0) = (struct sockaddr *)&mask;
rt_key(&rt0) = (struct sockaddr *)&addr;
rt0.rt_flags = RTF_HOST | RTF_STATIC;
@@ -2581,10 +2579,14 @@ in6_lltable_lookup(struct lltable *llt,
if (!(lle->la_flags & LLE_IFADDR) || (flags & LLE_IFADDR)) {
LLE_WLOCK(lle);
lle->la_flags |= LLE_DELETED;
- LLE_WUNLOCK(lle);
#ifdef DIAGNOSTIC
- log(LOG_INFO, "ifaddr cache = %p is deleted\n", lle);
+ log(LOG_INFO, "ifaddr cache = %p is deleted\n", lle);
#endif
+ if ((lle->la_flags &
+ (LLE_STATIC | LLE_IFADDR)) == LLE_STATIC)
+ llentry_free(lle);
+ else
+ LLE_WUNLOCK(lle);
}
lle = (void *)-1;
}
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r258346 - stable/10/sys/dev/ixgbe
Author: oleg
Date: Tue Nov 19 14:24:25 2013
New Revision: 258346
URL: http://svnweb.freebsd.org/changeset/base/258346
Log:
MFC: 257695
- Fix link loss on vlan reconfiguration.
- Fix issues with 'vlanhwfilter'.
Approved by: re (glebius)
Modified:
stable/10/sys/dev/ixgbe/ixgbe.c
Directory Properties:
stable/10/sys/ (props changed)
Modified: stable/10/sys/dev/ixgbe/ixgbe.c
==
--- stable/10/sys/dev/ixgbe/ixgbe.c Tue Nov 19 13:32:24 2013
(r258345)
+++ stable/10/sys/dev/ixgbe/ixgbe.c Tue Nov 19 14:24:25 2013
(r258346)
@@ -1253,9 +1253,6 @@ ixgbe_init_locked(struct adapter *adapte
IXGBE_WRITE_REG(hw, IXGBE_RDT(i), adapter->num_rx_desc - 1);
}
- /* Set up VLAN support and filter */
- ixgbe_setup_vlan_hw_support(adapter);
-
/* Enable Receive engine */
rxctrl = IXGBE_READ_REG(hw, IXGBE_RXCTRL);
if (hw->mac.type == ixgbe_mac_82598EB)
@@ -1339,6 +1336,9 @@ ixgbe_init_locked(struct adapter *adapte
/* Initialize the FC settings */
ixgbe_start_hw(hw);
+ /* Set up VLAN support and filter */
+ ixgbe_setup_vlan_hw_support(adapter);
+
/* And now turn on interrupts */
ixgbe_enable_intr(adapter);
@@ -4691,7 +4691,7 @@ ixgbe_register_vlan(void *arg, struct if
bit = vtag & 0x1F;
adapter->shadow_vfta[index] |= (1 << bit);
++adapter->num_vlans;
- ixgbe_init_locked(adapter);
+ ixgbe_setup_vlan_hw_support(adapter);
IXGBE_CORE_UNLOCK(adapter);
}
@@ -4718,7 +4718,7 @@ ixgbe_unregister_vlan(void *arg, struct
adapter->shadow_vfta[index] &= ~(1 << bit);
--adapter->num_vlans;
/* Re-init to load the changes */
- ixgbe_init_locked(adapter);
+ ixgbe_setup_vlan_hw_support(adapter);
IXGBE_CORE_UNLOCK(adapter);
}
@@ -4740,6 +4740,20 @@ ixgbe_setup_vlan_hw_support(struct adapt
if (adapter->num_vlans == 0)
return;
+ /* Setup the queues for vlans */
+ for (int i = 0; i < adapter->num_queues; i++) {
+ rxr = &adapter->rx_rings[i];
+ /* On 82599 the VLAN enable is per/queue in RXDCTL */
+ if (hw->mac.type != ixgbe_mac_82598EB) {
+ ctrl = IXGBE_READ_REG(hw, IXGBE_RXDCTL(i));
+ ctrl |= IXGBE_RXDCTL_VME;
+ IXGBE_WRITE_REG(hw, IXGBE_RXDCTL(i), ctrl);
+ }
+ rxr->vtag_strip = TRUE;
+ }
+
+ if ((ifp->if_capenable & IFCAP_VLAN_HWFILTER) == 0)
+ return;
/*
** A soft reset zero's out the VFTA, so
** we need to repopulate it now.
@@ -4758,18 +4772,6 @@ ixgbe_setup_vlan_hw_support(struct adapt
if (hw->mac.type == ixgbe_mac_82598EB)
ctrl |= IXGBE_VLNCTRL_VME;
IXGBE_WRITE_REG(hw, IXGBE_VLNCTRL, ctrl);
-
- /* Setup the queues for vlans */
- for (int i = 0; i < adapter->num_queues; i++) {
- rxr = &adapter->rx_rings[i];
- /* On 82599 the VLAN enable is per/queue in RXDCTL */
- if (hw->mac.type != ixgbe_mac_82598EB) {
- ctrl = IXGBE_READ_REG(hw, IXGBE_RXDCTL(i));
- ctrl |= IXGBE_RXDCTL_VME;
- IXGBE_WRITE_REG(hw, IXGBE_RXDCTL(i), ctrl);
- }
- rxr->vtag_strip = TRUE;
- }
}
static void
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r258346 - stable/10/sys/dev/ixgbe
On Mon, Dec 16, 2013 at 08:53:21PM +0400, Andrey V. Elsukov wrote: > On 19.11.2013 18:24, Oleg Bulyzhin wrote: > > Author: oleg > > Date: Tue Nov 19 14:24:25 2013 > > New Revision: 258346 > > URL: http://svnweb.freebsd.org/changeset/base/258346 > > > > Log: > > MFC: 257695 > > > > - Fix link loss on vlan reconfiguration. > > - Fix issues with 'vlanhwfilter'. > > Hi, Oleg, > > do you plan to merge this in stable/9 too? > > -- > WBR, Andrey V. Elsukov I didn't plan it (i'm still hoping jfv@ will fix it). -- Oleg. === Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru === ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r257695 - head/sys/dev/ixgbe
Author: oleg
Date: Tue Nov 5 09:46:01 2013
New Revision: 257695
URL: http://svnweb.freebsd.org/changeset/base/257695
Log:
- Fix link loss on vlan reconfiguration.
- Fix issues with 'vlanhwfilter'.
MFC after:1 week
Silence from: jfv 5 weeks
Modified:
head/sys/dev/ixgbe/ixgbe.c
Modified: head/sys/dev/ixgbe/ixgbe.c
==
--- head/sys/dev/ixgbe/ixgbe.c Tue Nov 5 09:30:06 2013(r257694)
+++ head/sys/dev/ixgbe/ixgbe.c Tue Nov 5 09:46:01 2013(r257695)
@@ -1253,9 +1253,6 @@ ixgbe_init_locked(struct adapter *adapte
IXGBE_WRITE_REG(hw, IXGBE_RDT(i), adapter->num_rx_desc - 1);
}
- /* Set up VLAN support and filter */
- ixgbe_setup_vlan_hw_support(adapter);
-
/* Enable Receive engine */
rxctrl = IXGBE_READ_REG(hw, IXGBE_RXCTRL);
if (hw->mac.type == ixgbe_mac_82598EB)
@@ -1339,6 +1336,9 @@ ixgbe_init_locked(struct adapter *adapte
/* Initialize the FC settings */
ixgbe_start_hw(hw);
+ /* Set up VLAN support and filter */
+ ixgbe_setup_vlan_hw_support(adapter);
+
/* And now turn on interrupts */
ixgbe_enable_intr(adapter);
@@ -4688,7 +4688,7 @@ ixgbe_register_vlan(void *arg, struct if
bit = vtag & 0x1F;
adapter->shadow_vfta[index] |= (1 << bit);
++adapter->num_vlans;
- ixgbe_init_locked(adapter);
+ ixgbe_setup_vlan_hw_support(adapter);
IXGBE_CORE_UNLOCK(adapter);
}
@@ -4715,7 +4715,7 @@ ixgbe_unregister_vlan(void *arg, struct
adapter->shadow_vfta[index] &= ~(1 << bit);
--adapter->num_vlans;
/* Re-init to load the changes */
- ixgbe_init_locked(adapter);
+ ixgbe_setup_vlan_hw_support(adapter);
IXGBE_CORE_UNLOCK(adapter);
}
@@ -4737,6 +4737,20 @@ ixgbe_setup_vlan_hw_support(struct adapt
if (adapter->num_vlans == 0)
return;
+ /* Setup the queues for vlans */
+ for (int i = 0; i < adapter->num_queues; i++) {
+ rxr = &adapter->rx_rings[i];
+ /* On 82599 the VLAN enable is per/queue in RXDCTL */
+ if (hw->mac.type != ixgbe_mac_82598EB) {
+ ctrl = IXGBE_READ_REG(hw, IXGBE_RXDCTL(i));
+ ctrl |= IXGBE_RXDCTL_VME;
+ IXGBE_WRITE_REG(hw, IXGBE_RXDCTL(i), ctrl);
+ }
+ rxr->vtag_strip = TRUE;
+ }
+
+ if ((ifp->if_capenable & IFCAP_VLAN_HWFILTER) == 0)
+ return;
/*
** A soft reset zero's out the VFTA, so
** we need to repopulate it now.
@@ -4755,18 +4769,6 @@ ixgbe_setup_vlan_hw_support(struct adapt
if (hw->mac.type == ixgbe_mac_82598EB)
ctrl |= IXGBE_VLNCTRL_VME;
IXGBE_WRITE_REG(hw, IXGBE_VLNCTRL, ctrl);
-
- /* Setup the queues for vlans */
- for (int i = 0; i < adapter->num_queues; i++) {
- rxr = &adapter->rx_rings[i];
- /* On 82599 the VLAN enable is per/queue in RXDCTL */
- if (hw->mac.type != ixgbe_mac_82598EB) {
- ctrl = IXGBE_READ_REG(hw, IXGBE_RXDCTL(i));
- ctrl |= IXGBE_RXDCTL_VME;
- IXGBE_WRITE_REG(hw, IXGBE_RXDCTL(i), ctrl);
- }
- rxr->vtag_strip = TRUE;
- }
}
static void
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r257695 - head/sys/dev/ixgbe
Discussion is here: http://lists.freebsd.org/pipermail/freebsd-stable/2013-September/075280.html -- Oleg. === Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru === ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r232272 - head/sys/netinet/ipfw
Author: oleg
Date: Tue Feb 28 21:53:39 2012
New Revision: 232272
URL: http://svn.freebsd.org/changeset/base/232272
Log:
lookup_dyn_rule_locked(): style(9) cleanup
MFC after:1 month
Modified:
head/sys/netinet/ipfw/ip_fw_dynamic.c
Modified: head/sys/netinet/ipfw/ip_fw_dynamic.c
==
--- head/sys/netinet/ipfw/ip_fw_dynamic.c Tue Feb 28 21:45:21 2012
(r232271)
+++ head/sys/netinet/ipfw/ip_fw_dynamic.c Tue Feb 28 21:53:39 2012
(r232272)
@@ -390,72 +390,68 @@ ipfw_remove_dyn_children(struct ip_fw *r
IPFW_DYN_UNLOCK();
}
-/**
- * lookup a dynamic rule, locked version
+/*
+ * Lookup a dynamic rule, locked version.
*/
static ipfw_dyn_rule *
lookup_dyn_rule_locked(struct ipfw_flow_id *pkt, int *match_direction,
struct tcphdr *tcp)
{
/*
-* stateful ipfw extensions.
-* Lookup into dynamic session queue
+* Stateful ipfw extensions.
+* Lookup into dynamic session queue.
*/
#define MATCH_REVERSE 0
#define MATCH_FORWARD 1
#define MATCH_NONE 2
#define MATCH_UNKNOWN 3
int i, dir = MATCH_NONE;
- ipfw_dyn_rule *prev, *q=NULL;
+ ipfw_dyn_rule *prev, *q = NULL;
IPFW_DYN_LOCK_ASSERT();
if (V_ipfw_dyn_v == NULL)
- goto done; /* not found */
- i = hash_packet( pkt );
- for (prev=NULL, q = V_ipfw_dyn_v[i] ; q != NULL ; ) {
+ goto done; /* not found */
+ i = hash_packet(pkt);
+ for (prev = NULL, q = V_ipfw_dyn_v[i]; q != NULL;) {
if (q->dyn_type == O_LIMIT_PARENT && q->count)
goto next;
- if (TIME_LEQ( q->expire, time_uptime)) { /* expire entry */
+ if (TIME_LEQ(q->expire, time_uptime)) { /* expire entry */
UNLINK_DYN_RULE(prev, V_ipfw_dyn_v[i], q);
continue;
}
- if (pkt->proto == q->id.proto &&
- q->dyn_type != O_LIMIT_PARENT) {
- if (IS_IP6_FLOW_ID(pkt)) {
- if (IN6_ARE_ADDR_EQUAL(&(pkt->src_ip6),
- &(q->id.src_ip6)) &&
- IN6_ARE_ADDR_EQUAL(&(pkt->dst_ip6),
- &(q->id.dst_ip6)) &&
+ if (pkt->proto != q->id.proto || q->dyn_type == O_LIMIT_PARENT)
+ goto next;
+
+ if (IS_IP6_FLOW_ID(pkt)) {
+ if (IN6_ARE_ADDR_EQUAL(&pkt->src_ip6, &q->id.src_ip6) &&
+ IN6_ARE_ADDR_EQUAL(&pkt->dst_ip6, &q->id.dst_ip6) &&
pkt->src_port == q->id.src_port &&
- pkt->dst_port == q->id.dst_port ) {
+ pkt->dst_port == q->id.dst_port) {
dir = MATCH_FORWARD;
break;
- }
- if (IN6_ARE_ADDR_EQUAL(&(pkt->src_ip6),
- &(q->id.dst_ip6)) &&
- IN6_ARE_ADDR_EQUAL(&(pkt->dst_ip6),
- &(q->id.src_ip6)) &&
- pkt->src_port == q->id.dst_port &&
- pkt->dst_port == q->id.src_port ) {
- dir = MATCH_REVERSE;
- break;
- }
- } else {
- if (pkt->src_ip == q->id.src_ip &&
- pkt->dst_ip == q->id.dst_ip &&
- pkt->src_port == q->id.src_port &&
- pkt->dst_port == q->id.dst_port ) {
- dir = MATCH_FORWARD;
- break;
- }
- if (pkt->src_ip == q->id.dst_ip &&
- pkt->dst_ip == q->id.src_ip &&
- pkt->src_port == q->id.dst_port &&
- pkt->dst_port == q->id.src_port ) {
- dir = MATCH_REVERSE;
- break;
- }
+ }
+ if (IN6_ARE_ADDR_EQUAL(&pkt->src_ip6, &q->id.dst_ip6) &&
+ IN6_ARE_ADDR_EQUAL(&pkt->dst_ip6, &q->id.src_ip6) &&
+ pkt->src_port == q->id.dst_port &&
+ pkt->dst_port == q->id.src_port) {
+ dir = MATCH_REVERSE;
+ break;
+ }
+ } else {
+ if (pkt->src_ip == q->id.src_ip &&
+ pkt->dst_ip == q->id.dst_ip &&
+ pkt->src_port == q->id.
svn commit: r232273 - head/sys/netinet/ipfw
Author: oleg
Date: Tue Feb 28 22:00:41 2012
New Revision: 232273
URL: http://svn.freebsd.org/changeset/base/232273
Log:
- Refresh dynamic tcp rule only if both sides answered keepalive packets.
- Remove some useless assignments.
MFC after:1 month
Modified:
head/sys/netinet/ipfw/ip_fw_dynamic.c
Modified: head/sys/netinet/ipfw/ip_fw_dynamic.c
==
--- head/sys/netinet/ipfw/ip_fw_dynamic.c Tue Feb 28 21:53:39 2012
(r232272)
+++ head/sys/netinet/ipfw/ip_fw_dynamic.c Tue Feb 28 22:00:41 2012
(r232273)
@@ -472,8 +472,12 @@ next:
#define BOTH_SYN (TH_SYN | (TH_SYN << 8))
#define BOTH_FIN (TH_FIN | (TH_FIN << 8))
+#defineTCP_FLAGS (TH_FLAGS | (TH_FLAGS << 8))
+#defineACK_FWD 0x1 /* fwd ack seen */
+#defineACK_REV 0x2 /* rev ack seen */
+
q->state |= (dir == MATCH_FORWARD) ? flags : (flags << 8);
- switch (q->state) {
+ switch (q->state & TCP_FLAGS) {
case TH_SYN:/* opening */
q->expire = time_uptime + V_dyn_syn_lifetime;
break;
@@ -482,24 +486,28 @@ next:
case BOTH_SYN | TH_FIN: /* one side tries to close */
case BOTH_SYN | (TH_FIN << 8):
#define _SEQ_GE(a,b) ((int)(a) - (int)(b) >= 0)
- if (tcp == NULL) {
- q->expire = time_uptime + V_dyn_ack_lifetime;
+ if (tcp == NULL)
break;
- }
ack = ntohl(tcp->th_ack);
if (dir == MATCH_FORWARD) {
- if (q->ack_fwd == 0 || _SEQ_GE(ack, q->ack_fwd))
+ if (q->ack_fwd == 0 ||
+ _SEQ_GE(ack, q->ack_fwd)) {
q->ack_fwd = ack;
- else/* ignore out-of-sequence */
- break;
+ q->state |= ACK_FWD;
+ }
} else {
- if (q->ack_rev == 0 || _SEQ_GE(ack, q->ack_rev))
+ if (q->ack_rev == 0 ||
+ _SEQ_GE(ack, q->ack_rev)) {
q->ack_rev = ack;
- else/* ignore out-of-sequence */
- break;
+ q->state |= ACK_REV;
+ }
+ }
+ if ((q->state & (ACK_FWD | ACK_REV)) ==
+ (ACK_FWD | ACK_REV)) {
+ q->expire = time_uptime + V_dyn_ack_lifetime;
+ q->state &= ~(ACK_FWD | ACK_REV);
}
- q->expire = time_uptime + V_dyn_ack_lifetime;
break;
case BOTH_SYN | BOTH_FIN: /* both sides closed */
@@ -1074,10 +1082,12 @@ ipfw_tick(void * vnetx)
if (TIME_LEQ(q->expire, time_uptime))
continue; /* too late, rule expired */
- m = ipfw_send_pkt(NULL, &(q->id), q->ack_rev - 1,
- q->ack_fwd, TH_SYN);
- mnext = ipfw_send_pkt(NULL, &(q->id), q->ack_fwd - 1,
- q->ack_rev, 0);
+ m = (q->state & ACK_REV) ? NULL :
+ ipfw_send_pkt(NULL, &(q->id), q->ack_rev - 1,
+ q->ack_fwd, TH_SYN);
+ mnext = (q->state & ACK_FWD) ? NULL :
+ ipfw_send_pkt(NULL, &(q->id), q->ack_fwd - 1,
+ q->ack_rev, 0);
switch (q->id.addr_type) {
case 4:
@@ -1103,18 +1113,16 @@ ipfw_tick(void * vnetx)
break;
#endif
}
-
- m = mnext = NULL;
}
}
IPFW_DYN_UNLOCK();
- for (m = mnext = m0; m != NULL; m = mnext) {
+ for (m = m0; m != NULL; m = mnext) {
mnext = m->m_nextpkt;
m->m_nextpkt = NULL;
ip_output(m, NULL, NULL, 0, NULL, NULL);
}
#ifdef INET6
- for (m = mnext = m6; m != NULL; m = mnext) {
+ for (m = m6; m != NULL; m = mnext) {
mnext = m->m_nextpkt;
m->m_nextpkt = NULL;
ip6_output(m, NULL, NULL, 0, NULL, NULL, NULL);
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listin
Re: svn commit: r232272 - head/sys/netinet/ipfw
On Tue, Feb 28, 2012 at 11:22:52PM +0100, Luigi Rizzo wrote: > On Tue, Feb 28, 2012 at 09:53:39PM +0000, Oleg Bulyzhin wrote: > > Author: oleg > > Date: Tue Feb 28 21:53:39 2012 > > New Revision: 232272 > > URL: http://svn.freebsd.org/changeset/base/232272 > > > > Log: > > lookup_dyn_rule_locked(): style(9) cleanup > > > > MFC after:1 month > > is there a reason for delaying the MFC, such as this is > preparatory work for further changes ? > Otherwise i'd suggest that whitespace changes like this > are quickly pushed to the stable branch so that the code > versions do not diverge. > > cheers > luigi Actual commit is r232273, this one used for reduce diff of meaningful commit. -- Oleg. === Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru === ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r236692 - stable/9/sys/netinet/ipfw
Author: oleg
Date: Wed Jun 6 18:00:19 2012
New Revision: 236692
URL: http://svn.freebsd.org/changeset/base/236692
Log:
MFC: r232272, r232273
- lookup_dyn_rule_locked(): style(9) cleanup
- Refresh dynamic tcp rule only if both sides answered keepalive packets.
- Remove some useless assignments.
Modified:
stable/9/sys/netinet/ipfw/ip_fw_dynamic.c
Directory Properties:
stable/9/sys/ (props changed)
Modified: stable/9/sys/netinet/ipfw/ip_fw_dynamic.c
==
--- stable/9/sys/netinet/ipfw/ip_fw_dynamic.c Wed Jun 6 17:28:46 2012
(r236691)
+++ stable/9/sys/netinet/ipfw/ip_fw_dynamic.c Wed Jun 6 18:00:19 2012
(r236692)
@@ -390,72 +390,68 @@ ipfw_remove_dyn_children(struct ip_fw *r
IPFW_DYN_UNLOCK();
}
-/**
- * lookup a dynamic rule, locked version
+/*
+ * Lookup a dynamic rule, locked version.
*/
static ipfw_dyn_rule *
lookup_dyn_rule_locked(struct ipfw_flow_id *pkt, int *match_direction,
struct tcphdr *tcp)
{
/*
-* stateful ipfw extensions.
-* Lookup into dynamic session queue
+* Stateful ipfw extensions.
+* Lookup into dynamic session queue.
*/
#define MATCH_REVERSE 0
#define MATCH_FORWARD 1
#define MATCH_NONE 2
#define MATCH_UNKNOWN 3
int i, dir = MATCH_NONE;
- ipfw_dyn_rule *prev, *q=NULL;
+ ipfw_dyn_rule *prev, *q = NULL;
IPFW_DYN_LOCK_ASSERT();
if (V_ipfw_dyn_v == NULL)
- goto done; /* not found */
- i = hash_packet( pkt );
- for (prev=NULL, q = V_ipfw_dyn_v[i] ; q != NULL ; ) {
+ goto done; /* not found */
+ i = hash_packet(pkt);
+ for (prev = NULL, q = V_ipfw_dyn_v[i]; q != NULL;) {
if (q->dyn_type == O_LIMIT_PARENT && q->count)
goto next;
- if (TIME_LEQ( q->expire, time_uptime)) { /* expire entry */
+ if (TIME_LEQ(q->expire, time_uptime)) { /* expire entry */
UNLINK_DYN_RULE(prev, V_ipfw_dyn_v[i], q);
continue;
}
- if (pkt->proto == q->id.proto &&
- q->dyn_type != O_LIMIT_PARENT) {
- if (IS_IP6_FLOW_ID(pkt)) {
- if (IN6_ARE_ADDR_EQUAL(&(pkt->src_ip6),
- &(q->id.src_ip6)) &&
- IN6_ARE_ADDR_EQUAL(&(pkt->dst_ip6),
- &(q->id.dst_ip6)) &&
+ if (pkt->proto != q->id.proto || q->dyn_type == O_LIMIT_PARENT)
+ goto next;
+
+ if (IS_IP6_FLOW_ID(pkt)) {
+ if (IN6_ARE_ADDR_EQUAL(&pkt->src_ip6, &q->id.src_ip6) &&
+ IN6_ARE_ADDR_EQUAL(&pkt->dst_ip6, &q->id.dst_ip6) &&
pkt->src_port == q->id.src_port &&
- pkt->dst_port == q->id.dst_port ) {
+ pkt->dst_port == q->id.dst_port) {
+ dir = MATCH_FORWARD;
+ break;
+ }
+ if (IN6_ARE_ADDR_EQUAL(&pkt->src_ip6, &q->id.dst_ip6) &&
+ IN6_ARE_ADDR_EQUAL(&pkt->dst_ip6, &q->id.src_ip6) &&
+ pkt->src_port == q->id.dst_port &&
+ pkt->dst_port == q->id.src_port) {
+ dir = MATCH_REVERSE;
+ break;
+ }
+ } else {
+ if (pkt->src_ip == q->id.src_ip &&
+ pkt->dst_ip == q->id.dst_ip &&
+ pkt->src_port == q->id.src_port &&
+ pkt->dst_port == q->id.dst_port) {
dir = MATCH_FORWARD;
break;
- }
- if (IN6_ARE_ADDR_EQUAL(&(pkt->src_ip6),
- &(q->id.dst_ip6)) &&
- IN6_ARE_ADDR_EQUAL(&(pkt->dst_ip6),
- &(q->id.src_ip6)) &&
- pkt->src_port == q->id.dst_port &&
- pkt->dst_port == q->id.src_port ) {
- dir = MATCH_REVERSE;
- break;
- }
- } else {
- if (pkt->src_ip == q->id.src_ip &&
- pkt->dst_ip == q->id.dst_ip &&
- pkt->src_port == q->id.src_port &&
- pkt->dst_port == q->id.dst_port ) {
- dir = MATCH_FORWARD;
- break;
- }
- if (pkt->src_ip =
svn commit: r236694 - stable/8/sys/netinet/ipfw
Author: oleg
Date: Wed Jun 6 18:04:49 2012
New Revision: 236694
URL: http://svn.freebsd.org/changeset/base/236694
Log:
MFC: r232272, r232273
- lookup_dyn_rule_locked(): style(9) cleanup
- Refresh dynamic tcp rule only if both sides answered keepalive packets.
- Remove some useless assignments.
Modified:
stable/8/sys/netinet/ipfw/ip_fw_dynamic.c
Directory Properties:
stable/8/sys/ (props changed)
Modified: stable/8/sys/netinet/ipfw/ip_fw_dynamic.c
==
--- stable/8/sys/netinet/ipfw/ip_fw_dynamic.c Wed Jun 6 18:00:38 2012
(r236693)
+++ stable/8/sys/netinet/ipfw/ip_fw_dynamic.c Wed Jun 6 18:04:49 2012
(r236694)
@@ -390,72 +390,68 @@ ipfw_remove_dyn_children(struct ip_fw *r
IPFW_DYN_UNLOCK();
}
-/**
- * lookup a dynamic rule, locked version
+/*
+ * Lookup a dynamic rule, locked version.
*/
static ipfw_dyn_rule *
lookup_dyn_rule_locked(struct ipfw_flow_id *pkt, int *match_direction,
struct tcphdr *tcp)
{
/*
-* stateful ipfw extensions.
-* Lookup into dynamic session queue
+* Stateful ipfw extensions.
+* Lookup into dynamic session queue.
*/
#define MATCH_REVERSE 0
#define MATCH_FORWARD 1
#define MATCH_NONE 2
#define MATCH_UNKNOWN 3
int i, dir = MATCH_NONE;
- ipfw_dyn_rule *prev, *q=NULL;
+ ipfw_dyn_rule *prev, *q = NULL;
IPFW_DYN_LOCK_ASSERT();
if (V_ipfw_dyn_v == NULL)
- goto done; /* not found */
- i = hash_packet( pkt );
- for (prev=NULL, q = V_ipfw_dyn_v[i] ; q != NULL ; ) {
+ goto done; /* not found */
+ i = hash_packet(pkt);
+ for (prev = NULL, q = V_ipfw_dyn_v[i]; q != NULL;) {
if (q->dyn_type == O_LIMIT_PARENT && q->count)
goto next;
- if (TIME_LEQ( q->expire, time_uptime)) { /* expire entry */
+ if (TIME_LEQ(q->expire, time_uptime)) { /* expire entry */
UNLINK_DYN_RULE(prev, V_ipfw_dyn_v[i], q);
continue;
}
- if (pkt->proto == q->id.proto &&
- q->dyn_type != O_LIMIT_PARENT) {
- if (IS_IP6_FLOW_ID(pkt)) {
- if (IN6_ARE_ADDR_EQUAL(&(pkt->src_ip6),
- &(q->id.src_ip6)) &&
- IN6_ARE_ADDR_EQUAL(&(pkt->dst_ip6),
- &(q->id.dst_ip6)) &&
+ if (pkt->proto != q->id.proto || q->dyn_type == O_LIMIT_PARENT)
+ goto next;
+
+ if (IS_IP6_FLOW_ID(pkt)) {
+ if (IN6_ARE_ADDR_EQUAL(&pkt->src_ip6, &q->id.src_ip6) &&
+ IN6_ARE_ADDR_EQUAL(&pkt->dst_ip6, &q->id.dst_ip6) &&
pkt->src_port == q->id.src_port &&
- pkt->dst_port == q->id.dst_port ) {
+ pkt->dst_port == q->id.dst_port) {
+ dir = MATCH_FORWARD;
+ break;
+ }
+ if (IN6_ARE_ADDR_EQUAL(&pkt->src_ip6, &q->id.dst_ip6) &&
+ IN6_ARE_ADDR_EQUAL(&pkt->dst_ip6, &q->id.src_ip6) &&
+ pkt->src_port == q->id.dst_port &&
+ pkt->dst_port == q->id.src_port) {
+ dir = MATCH_REVERSE;
+ break;
+ }
+ } else {
+ if (pkt->src_ip == q->id.src_ip &&
+ pkt->dst_ip == q->id.dst_ip &&
+ pkt->src_port == q->id.src_port &&
+ pkt->dst_port == q->id.dst_port) {
dir = MATCH_FORWARD;
break;
- }
- if (IN6_ARE_ADDR_EQUAL(&(pkt->src_ip6),
- &(q->id.dst_ip6)) &&
- IN6_ARE_ADDR_EQUAL(&(pkt->dst_ip6),
- &(q->id.src_ip6)) &&
- pkt->src_port == q->id.dst_port &&
- pkt->dst_port == q->id.src_port ) {
- dir = MATCH_REVERSE;
- break;
- }
- } else {
- if (pkt->src_ip == q->id.src_ip &&
- pkt->dst_ip == q->id.dst_ip &&
- pkt->src_port == q->id.src_port &&
- pkt->dst_port == q->id.dst_port ) {
- dir = MATCH_FORWARD;
- break;
- }
- if (pkt->src_ip =
svn commit: r200170 - head/sys/netinet/ipfw
Author: oleg
Date: Sat Dec 5 23:27:21 2009
New Revision: 200170
URL: http://svn.freebsd.org/changeset/base/200170
Log:
Fix burst processing for WF2Q pipes - do not increase available burst size
unless pipe is idle. This should fix follwing issues:
- 'dummynet: OUCH! pipe should have been idle!' log messages.
- exceeding configured pipe bandwidth.
MFC after:1 week
Modified:
head/sys/netinet/ipfw/ip_dummynet.c
Modified: head/sys/netinet/ipfw/ip_dummynet.c
==
--- head/sys/netinet/ipfw/ip_dummynet.c Sat Dec 5 23:23:46 2009
(r200169)
+++ head/sys/netinet/ipfw/ip_dummynet.c Sat Dec 5 23:27:21 2009
(r200170)
@@ -1447,7 +1447,9 @@ dummynet_io(struct mbuf **m0, int dir, s
q->numbytes += pipe->bandwidth;
}
} else {/* WF2Q. */
- if (pipe->idle_time < curr_time) {
+ if (pipe->idle_time < curr_time &&
+ pipe->scheduler_heap.elements == 0 &&
+ pipe->not_eligible_heap.elements == 0) {
/* Calculate available burst size. */
pipe->numbytes +=
(curr_time - pipe->idle_time - 1) * pipe->bandwidth;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r194245 - head/sys/netinet/ipfw
Author: oleg
Date: Mon Jun 15 17:14:47 2009
New Revision: 194245
URL: http://svn.freebsd.org/changeset/base/194245
Log:
Since dn_pipe.numbytes is int64_t now - remove unnecessary overflow detection
code in ready_event_wfq().
Modified:
head/sys/netinet/ipfw/ip_dummynet.c
Modified: head/sys/netinet/ipfw/ip_dummynet.c
==
--- head/sys/netinet/ipfw/ip_dummynet.c Mon Jun 15 16:51:07 2009
(r194244)
+++ head/sys/netinet/ipfw/ip_dummynet.c Mon Jun 15 17:14:47 2009
(r194245)
@@ -56,7 +56,6 @@ __FBSDID("$FreeBSD$");
* include files marked with XXX are probably not needed
*/
-#include
#include
#include
#include
@@ -686,16 +685,11 @@ ready_event_wfq(struct dn_pipe *p, struc
int p_was_empty = (p->head == NULL);
struct dn_heap *sch = &(p->scheduler_heap);
struct dn_heap *neh = &(p->not_eligible_heap);
- int64_t p_numbytes = p->numbytes;
DUMMYNET_LOCK_ASSERT();
if (p->if_name[0] == 0) /* tx clock is simulated */
- /*
-* Since result may not fit into p->numbytes (32bit) we
-* are using 64bit var here.
-*/
- p_numbytes += (curr_time - p->sched_time) * p->bandwidth;
+ p->numbytes += (curr_time - p->sched_time) * p->bandwidth;
else { /*
* tx clock is for real,
* the ifq must be empty or this is a NOP.
@@ -712,7 +706,7 @@ ready_event_wfq(struct dn_pipe *p, struc
* While we have backlogged traffic AND credit, we need to do
* something on the queue.
*/
- while (p_numbytes >= 0 && (sch->elements > 0 || neh->elements > 0)) {
+ while (p->numbytes >= 0 && (sch->elements > 0 || neh->elements > 0)) {
if (sch->elements > 0) {
/* Have some eligible pkts to send out. */
struct dn_flow_queue *q = sch->p[0].object;
@@ -722,7 +716,7 @@ ready_event_wfq(struct dn_pipe *p, struc
int len_scaled = p->bandwidth ? len * 8 * hz : 0;
heap_extract(sch, NULL); /* Remove queue from heap. */
- p_numbytes -= len_scaled;
+ p->numbytes -= len_scaled;
move_pkt(pkt, q, p, len);
p->V += (len << MY_M) / p->sum; /* Update V. */
@@ -763,11 +757,11 @@ ready_event_wfq(struct dn_pipe *p, struc
}
if (p->if_name[0] != '\0') { /* Tx clock is from a real thing */
- p_numbytes = -1;/* Mark not ready for I/O. */
+ p->numbytes = -1; /* Mark not ready for I/O. */
break;
}
}
- if (sch->elements == 0 && neh->elements == 0 && p_numbytes >= 0 &&
+ if (sch->elements == 0 && neh->elements == 0 && p->numbytes >= 0 &&
p->idle_heap.elements > 0) {
/*
* No traffic and no events scheduled.
@@ -790,11 +784,11 @@ ready_event_wfq(struct dn_pipe *p, struc
* If we are under credit, schedule the next ready event.
* Also fix the delivery time of the last packet.
*/
- if (p->if_name[0]==0 && p_numbytes < 0) { /* This implies bw > 0. */
+ if (p->if_name[0]==0 && p->numbytes < 0) { /* This implies bw > 0. */
dn_key t = 0; /* Number of ticks i have to wait. */
if (p->bandwidth > 0)
- t = (p->bandwidth - 1 - p_numbytes) / p->bandwidth;
+ t = (p->bandwidth - 1 - p->numbytes) / p->bandwidth;
dn_tag_get(p->tail)->output_time += t;
p->sched_time = curr_time;
heap_insert(&wfq_ready_heap, curr_time + t, (void *)p);
@@ -804,14 +798,6 @@ ready_event_wfq(struct dn_pipe *p, struc
*/
}
- /* Fit (adjust if necessary) 64bit result into 32bit variable. */
- if (p_numbytes > INT_MAX)
- p->numbytes = INT_MAX;
- else if (p_numbytes < INT_MIN)
- p->numbytes = INT_MIN;
- else
- p->numbytes = p_numbytes;
-
/*
* If the delay line was empty call transmit_event() now.
* Otherwise, the scheduler will take care of it.
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r194930 - in head: sbin/ipfw sys/netinet sys/netinet/ipfw
Author: oleg
Date: Wed Jun 24 22:57:07 2009
New Revision: 194930
URL: http://svn.freebsd.org/changeset/base/194930
Log:
- fix dummynet 'fast' mode for WF2Q case.
- fix printing of pipe profile data.
- introduce new pipe parameter: 'burst' - how much data can be sent through
pipe bypassing bandwidth limit.
Modified:
head/sbin/ipfw/Makefile
head/sbin/ipfw/dummynet.c
head/sbin/ipfw/ipfw.8
head/sbin/ipfw/ipfw2.h
head/sys/netinet/ip_dummynet.h
head/sys/netinet/ipfw/ip_dummynet.c
Modified: head/sbin/ipfw/Makefile
==
--- head/sbin/ipfw/Makefile Wed Jun 24 22:42:52 2009(r194929)
+++ head/sbin/ipfw/Makefile Wed Jun 24 22:57:07 2009(r194930)
@@ -3,6 +3,7 @@
PROG= ipfw
SRCS= ipfw2.c dummynet.c ipv6.c main.c nat.c altq.c
WARNS?=2
+LDADD= -lutil
MAN= ipfw.8
.include
Modified: head/sbin/ipfw/dummynet.c
==
--- head/sbin/ipfw/dummynet.c Wed Jun 24 22:42:52 2009(r194929)
+++ head/sbin/ipfw/dummynet.c Wed Jun 24 22:57:07 2009(r194930)
@@ -32,6 +32,8 @@
#include
#include
+#include
+#include
#include
#include
#include
@@ -70,6 +72,7 @@ static struct _s_x dummynet_params[] = {
{ "src-ipv6", TOK_SRCIP6},
{ "src-ip6",TOK_SRCIP6},
{ "profile",TOK_PIPE_PROFILE},
+ { "burst", TOK_BURST},
{ "dummynet-params",TOK_NULL },
{ NULL, 0 } /* terminator */
};
@@ -236,7 +239,7 @@ print_flowset_parms(struct dn_flow_set *
plr[0] = '\0';
if (fs->flags_fs & DN_IS_RED) /* RED parameters */
sprintf(red,
- "\n\t %cRED w_q %f min_th %d max_th %d max_p %f",
+ "\n\t %cRED w_q %f min_th %d max_th %d max_p %f",
(fs->flags_fs & DN_IS_GENTLE_RED) ? 'G' : ' ',
1.0 * fs->w_q / (double)(1 << SCALE_RED),
SCALE_VAL(fs->min_th),
@@ -250,7 +253,7 @@ print_flowset_parms(struct dn_flow_set *
}
static void
-print_extra_delay_parms(struct dn_pipe *p, char *prefix)
+print_extra_delay_parms(struct dn_pipe *p)
{
double loss;
if (p->samples_no <= 0)
@@ -258,8 +261,8 @@ print_extra_delay_parms(struct dn_pipe *
loss = p->loss_level;
loss /= p->samples_no;
- printf("%s profile: name \"%s\" loss %f samples %d\n",
- prefix, p->name, loss, p->samples_no);
+ printf("\t profile: name \"%s\" loss %f samples %d\n",
+ p->name, loss, p->samples_no);
}
void
@@ -280,6 +283,7 @@ ipfw_list_pipes(void *data, uint nbytes,
double b = p->bandwidth;
char buf[30];
char prefix[80];
+ char burst[5 + 7];
if (SLIST_NEXT(p, next) != (struct dn_pipe *)DN_IS_PIPE)
break; /* done with pipes, now queues */
@@ -311,10 +315,16 @@ ipfw_list_pipes(void *data, uint nbytes,
sprintf(prefix, "%05d: %s %4d ms ",
p->pipe_nr, buf, p->delay);
- print_extra_delay_parms(p, prefix);
-
print_flowset_parms(&(p->fs), prefix);
+ if (humanize_number(burst, sizeof(burst), p->burst,
+ "Byte", HN_AUTOSCALE, 0) < 0 || co.verbose)
+ printf("\t burst: %ju Byte\n", p->burst);
+ else
+ printf("\t burst: %s\n", burst);
+
+ print_extra_delay_parms(p);
+
q = (struct dn_flow_queue *)(p+1);
list_queues(&(p->fs), q);
}
@@ -933,6 +943,21 @@ end_mask:
--ac; ++av;
break;
+ case TOK_BURST:
+ if (co.do_pipe != 1)
+ errx(EX_DATAERR, "burst only valid for pipes");
+ NEED1("burst needs argument\n");
+ errno = 0;
+ if (expand_number(av[0], &p.burst) < 0)
+ if (errno != ERANGE)
+ errx(EX_DATAERR,
+ "burst: invalid argument");
+ if (errno || p.burst > (1ULL << 48) - 1)
+ errx(EX_DATAERR,
+ "burst: out of range (0..2^48-1)");
+ ac--; av++;
+ break;
+
default:
errx(EX_DATAERR, "unrecognised option ``%s''", av[-1]);
}
Modified: head/sbin/ipfw/ipfw.8
==
--- head/sbin/ipfw/ipfw.8 Wed Jun 24 22:42:52 2009(r194929)
+++ head/sbin/ipfw/ipfw.8 Wed Jun 24 22:57:07 2009(r194930)
@@ -1,7 +1,7 @@
.\"
.\
Re: svn commit: r194930 - in head: sbin/ipfw sys/netinet sys/netinet/ipfw
On Wed, Jun 24, 2009 at 07:44:04PM -0400, Ben Kaduk wrote:
> There's a grammar error and a style error, here. I'm actually not
> entirely sure what
> the intended meaning is, so it's a bit hard to fix the grammar error.
It's no wonder - my english writing skill is poor.
> Looking at the code, it seems that in burst mode, extra data will be
> allowed to be sent, though
> it is capped to not exceed the pipe bandwidth.
> Perhaps "... bytes of data is allowed to bypass the dummynet scheduler
> (...), though
> the transmission rate will still be capped so as to not exceed the
> pipe's bandwidth."?
Let me explain. For example, we have pipe with bw 1Mbit/s and burst size is
5GByte. If we try to download 10Gbyte through this pipe following will happen:
1) 1st 5Gbyte of data will go with 'wire speed'.
2) last 5Gbyte will be shaped to 1Mbit/s
Could you please mail me whole 'burst' part (as it should be), and i will fix
it.
>
> The style error is that the new sentence ("Effective burst size ...")
> should start
> on a new line.
>
> I would also prefer to see the new sentence be an actual complete sentence
> (i.e., "The effective burst size is calculated as follows"), though
> there appears
> to be at least one other bug of this form in the file already (see,
> for example, the
> quoted text at the beginning of this hunk: "Default value is no delay.", which
> would benefit from a "the".)
>
>
> Unrelated to this commit, there is a grammar error early in the file:
>
> 312 Once
> 313 .Fl p
> 314 has been specified, any additional arguments as passed on to the
> preproc
> essor
> 315 for interpretation.
>
> The 'as' in line 314 should be 'are'.
> (This is from CVS r1.220, so the line numbers may not be current.)
>
>
> Thanks,
>
> Ben Kaduk
>
>
> > +MAX(
> > +.Ar size
> > +,
> > +.Nm bw
> > +* pipe_idle_time).
> > +.Pp
> > Â .It Cm profile Ar filename
> > Â A file specifying the additional overhead incurred in the transmission
> > Â of a packet on the link.
> >
--
Oleg.
=== Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru ===
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r195075 - head/sbin/ipfw
Author: oleg Date: Fri Jun 26 19:49:06 2009 New Revision: 195075 URL: http://svn.freebsd.org/changeset/base/195075 Log: - 'burst' description rewritten. Submitted by: Ben Kaduk Approved by: re (kib) Modified: head/sbin/ipfw/ipfw.8 Modified: head/sbin/ipfw/ipfw.8 == --- head/sbin/ipfw/ipfw.8 Fri Jun 26 19:39:33 2009(r195074) +++ head/sbin/ipfw/ipfw.8 Fri Jun 26 19:49:06 2009(r195075) @@ -1944,13 +1944,18 @@ the granularity to 1ms or less). The default value is 0, meaning no delay. .Pp .It Cm burst Ar size -If the data rate exceeds the pipe bandwith limit -(and pipe was idle long enough), +If the data to be sent exceeds the pipe's bandwidth limit +(and the pipe was previously idle), up to .Ar size -bytes of data is allowed to bypass the +bytes of data are allowed to bypass the .Nm dummynet -scheduler (i.e. it will be sent without shaping), then transmission rate -will not exceed pipe bandwidth. Effective burst size calculated as follows: +scheduler, and will be sent as fast as the physical link allows. +Any additional data will be transmitted at the rate specified +by the +.Nm pipe +bandwidth. +The burst size depends on how long the pipe has been idle; +the effective burst size is calculated as follows: MAX( .Ar size , ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r184414 - in head: . sys/netinet
Author: oleg
Date: Tue Oct 28 14:14:57 2008
New Revision: 184414
URL: http://svn.freebsd.org/changeset/base/184414
Log:
Type of q_time (start of queue idle time) has changed: uint32_t -> uint64_t.
This should fix q_time overflow, which happens after 2^32/(86400*hz) days of
uptime (~50days for hz = 1000).
q_time overflow cause following:
- traffic shaping may not work in 'fast' mode (not enabled by default).
- incorrect average queue length calculation in RED/GRED algorithm.
NB: due to ABI change this change is not applicable to stable.
PR: kern/128401
Modified:
head/UPDATING
head/sys/netinet/ip_dummynet.h
Modified: head/UPDATING
==
--- head/UPDATING Tue Oct 28 13:44:11 2008(r184413)
+++ head/UPDATING Tue Oct 28 14:14:57 2008(r184414)
@@ -22,6 +22,9 @@ NOTE TO PEOPLE WHO THINK THAT FreeBSD 8.
to maximize performance. (To disable malloc debugging, run
ln -s aj /etc/malloc.conf.)
+20081028:
+ dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
+
20081009:
The uhci, ohci, ehci and slhci USB Host controller drivers have
been put into separate modules. If you load the usb module
Modified: head/sys/netinet/ip_dummynet.h
==
--- head/sys/netinet/ip_dummynet.h Tue Oct 28 13:44:11 2008
(r184413)
+++ head/sys/netinet/ip_dummynet.h Tue Oct 28 14:14:57 2008
(r184414)
@@ -216,7 +216,7 @@ struct dn_flow_queue {
int avg ; /* average queue length est. (scaled) */
int count ; /* arrivals since last RED drop */
int random ;/* random value (scaled) */
-u_int32_t q_time ; /* start of queue idle time */
+dn_key q_time; /* start of queue idle time */
/* WF2Q+ support */
struct dn_flow_set *fs ; /* parent flow set */
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "[EMAIL PROTECTED]"
svn commit: r184504 - stable/7/sys/netinet
Author: oleg
Date: Fri Oct 31 12:58:12 2008
New Revision: 184504
URL: http://svn.freebsd.org/changeset/base/184504
Log:
Direct commit (r184414 is not applicable to stable due to ABI change):
Workaround possible q_time overflow (will happen after 2^32/(86400*hz)
days of uptime (~50days for hz = 1000)), which may lead to:
- broken shaping in 'fast' io mode.
- incorrect average queue length calculation in RED/GRED algorithm.
PR: kern/128401
Approved by: re (kensmith)
Modified:
stable/7/sys/netinet/ip_dummynet.c
Modified: stable/7/sys/netinet/ip_dummynet.c
==
--- stable/7/sys/netinet/ip_dummynet.c Fri Oct 31 11:47:51 2008
(r184503)
+++ stable/7/sys/netinet/ip_dummynet.c Fri Oct 31 12:58:12 2008
(r184504)
@@ -1155,7 +1155,8 @@ red_drops(struct dn_flow_set *fs, struct
* XXX check wraps...
*/
if (q->avg) {
- u_int t = (curr_time - q->q_time) / fs->lookup_step;
+ u_int t = ((uint32_t)curr_time - q->q_time) /
+ fs->lookup_step;
q->avg = (t < fs->lookup_depth) ?
SCALE_MUL(q->avg, fs->w_q_lookup[t]) : 0;
@@ -1350,7 +1351,7 @@ dummynet_io(struct mbuf **m0, int dir, s
if (q->head != m) /* Flow was not idle, we are done. */
goto done;
- if (q->q_time < curr_time)
+ if (q->q_time < (uint32_t)curr_time)
q->numbytes = io_fast ? fs->pipe->bandwidth : 0;
q->q_time = curr_time;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "[EMAIL PROTECTED]"
svn commit: r184505 - stable/6/sys/netinet
Author: oleg
Date: Fri Oct 31 13:00:34 2008
New Revision: 184505
URL: http://svn.freebsd.org/changeset/base/184505
Log:
Direct commit (r184414 is not applicable to stable due to ABI change):
Workaround possible q_time overflow (will happen after 2^32/(86400*hz)
days of uptime (~50days for hz = 1000)), which may lead to:
- broken shaping in 'fast' io mode.
- incorrect average queue length calculation in RED/GRED algorithm.
PR: kern/128401
Approved by: re (kensmith)
Modified:
stable/6/sys/netinet/ip_dummynet.c
Modified: stable/6/sys/netinet/ip_dummynet.c
==
--- stable/6/sys/netinet/ip_dummynet.c Fri Oct 31 12:58:12 2008
(r184504)
+++ stable/6/sys/netinet/ip_dummynet.c Fri Oct 31 13:00:34 2008
(r184505)
@@ -1186,7 +1186,8 @@ red_drops(struct dn_flow_set *fs, struct
* XXX check wraps...
*/
if (q->avg) {
- u_int t = (curr_time - q->q_time) / fs->lookup_step;
+ u_int t = ((uint32_t)curr_time - q->q_time) /
+ fs->lookup_step;
q->avg = (t < fs->lookup_depth) ?
SCALE_MUL(q->avg, fs->w_q_lookup[t]) : 0;
@@ -1382,7 +1383,7 @@ dummynet_io(struct mbuf **m0, int dir, s
if (q->head != m) /* Flow was not idle, we are done. */
goto done;
- if (q->q_time < curr_time)
+ if (q->q_time < (uint32_t)curr_time)
q->numbytes = io_fast ? fs->pipe->bandwidth : 0;
q->q_time = curr_time;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "[EMAIL PROTECTED]"
svn commit: r184506 - releng/6.4/sys/netinet
Author: oleg
Date: Fri Oct 31 13:01:31 2008
New Revision: 184506
URL: http://svn.freebsd.org/changeset/base/184506
Log:
Direct commit (r184414 is not applicable to stable due to ABI change):
Workaround possible q_time overflow (will happen after 2^32/(86400*hz)
days of uptime (~50days for hz = 1000)), which may lead to:
- broken shaping in 'fast' io mode.
- incorrect average queue length calculation in RED/GRED algorithm.
PR: kern/128401
Approved by: re (kensmith)
Modified:
releng/6.4/sys/netinet/ip_dummynet.c
Modified: releng/6.4/sys/netinet/ip_dummynet.c
==
--- releng/6.4/sys/netinet/ip_dummynet.cFri Oct 31 13:00:34 2008
(r184505)
+++ releng/6.4/sys/netinet/ip_dummynet.cFri Oct 31 13:01:31 2008
(r184506)
@@ -1186,7 +1186,8 @@ red_drops(struct dn_flow_set *fs, struct
* XXX check wraps...
*/
if (q->avg) {
- u_int t = (curr_time - q->q_time) / fs->lookup_step;
+ u_int t = ((uint32_t)curr_time - q->q_time) /
+ fs->lookup_step;
q->avg = (t < fs->lookup_depth) ?
SCALE_MUL(q->avg, fs->w_q_lookup[t]) : 0;
@@ -1382,7 +1383,7 @@ dummynet_io(struct mbuf **m0, int dir, s
if (q->head != m) /* Flow was not idle, we are done. */
goto done;
- if (q->q_time < curr_time)
+ if (q->q_time < (uint32_t)curr_time)
q->numbytes = io_fast ? fs->pipe->bandwidth : 0;
q->q_time = curr_time;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "[EMAIL PROTECTED]"
svn commit: r193859 - in head/sys: net netgraph netinet netinet/ipfw
Author: oleg
Date: Tue Jun 9 21:27:11 2009
New Revision: 193859
URL: http://svn.freebsd.org/changeset/base/193859
Log:
Close long existed race with net.inet.ip.fw.one_pass = 0:
If packet leaves ipfw to other kernel subsystem (dummynet, netgraph, etc)
it carries pointer to matching ipfw rule. If this packet then reinjected back
to ipfw, ruleset processing starts from that rule. If rule was deleted
meanwhile, due to existed race condition panic was possible (as well as
other odd effects like parsing rules in 'reap list').
P.S. this commit changes ABI so userland ipfw related binaries should be
recompiled.
MFC after:1 month
Tested by:Mikolaj Golub
Modified:
head/sys/net/if_bridge.c
head/sys/net/if_ethersubr.c
head/sys/netgraph/ng_ipfw.c
head/sys/netgraph/ng_ipfw.h
head/sys/netinet/ip_dummynet.h
head/sys/netinet/ip_fw.h
head/sys/netinet/ipfw/ip_dummynet.c
head/sys/netinet/ipfw/ip_fw2.c
head/sys/netinet/ipfw/ip_fw_pfil.c
Modified: head/sys/net/if_bridge.c
==
--- head/sys/net/if_bridge.cTue Jun 9 21:17:57 2009(r193858)
+++ head/sys/net/if_bridge.cTue Jun 9 21:27:11 2009(r193859)
@@ -3041,11 +3041,19 @@ bridge_pfil(struct mbuf **mp, struct ifn
if (ip_fw_chk_ptr && pfil_ipfw != 0 && dir == PFIL_OUT && ifp != NULL) {
INIT_VNET_INET(curvnet);
+ struct dn_pkt_tag *dn_tag;
error = -1;
- args.rule = ip_dn_claim_rule(*mp);
- if (args.rule != NULL && V_fw_one_pass)
- goto ipfwpass; /* packet already partially processed */
+ dn_tag = ip_dn_claim_tag(*mp);
+ if (dn_tag != NULL) {
+ if (dn_tag->rule != NULL && V_fw_one_pass)
+ /* packet already partially processed */
+ goto ipfwpass;
+ args.rule = dn_tag->rule; /* matching rule to restart */
+ args.rule_id = dn_tag->rule_id;
+ args.chain_id = dn_tag->chain_id;
+ } else
+ args.rule = NULL;
args.m = *mp;
args.oif = ifp;
Modified: head/sys/net/if_ethersubr.c
==
--- head/sys/net/if_ethersubr.c Tue Jun 9 21:17:57 2009(r193858)
+++ head/sys/net/if_ethersubr.c Tue Jun 9 21:27:11 2009(r193859)
@@ -145,8 +145,7 @@ MALLOC_DEFINE(M_ARPCOM, "arpcom", "802.*
#if defined(INET) || defined(INET6)
int
-ether_ipfw_chk(struct mbuf **m0, struct ifnet *dst,
- struct ip_fw **rule, int shared);
+ether_ipfw_chk(struct mbuf **m0, struct ifnet *dst, int shared);
#ifdef VIMAGE_GLOBALS
static int ether_ipfw;
#endif
@@ -428,10 +427,9 @@ ether_output_frame(struct ifnet *ifp, st
{
#if defined(INET) || defined(INET6)
INIT_VNET_NET(ifp->if_vnet);
- struct ip_fw *rule = ip_dn_claim_rule(m);
if (ip_fw_chk_ptr && V_ether_ipfw != 0) {
- if (ether_ipfw_chk(&m, ifp, &rule, 0) == 0) {
+ if (ether_ipfw_chk(&m, ifp, 0) == 0) {
if (m) {
m_freem(m);
return EACCES; /* pkt dropped */
@@ -455,8 +453,7 @@ ether_output_frame(struct ifnet *ifp, st
* ether_output_frame.
*/
int
-ether_ipfw_chk(struct mbuf **m0, struct ifnet *dst,
- struct ip_fw **rule, int shared)
+ether_ipfw_chk(struct mbuf **m0, struct ifnet *dst, int shared)
{
INIT_VNET_INET(dst->if_vnet);
struct ether_header *eh;
@@ -464,9 +461,19 @@ ether_ipfw_chk(struct mbuf **m0, struct
struct mbuf *m;
int i;
struct ip_fw_args args;
+ struct dn_pkt_tag *dn_tag;
- if (*rule != NULL && V_fw_one_pass)
- return 1; /* dummynet packet, already partially processed */
+ dn_tag = ip_dn_claim_tag(*m0);
+
+ if (dn_tag != NULL) {
+ if (dn_tag->rule != NULL && V_fw_one_pass)
+ /* dummynet packet, already partially processed */
+ return (1);
+ args.rule = dn_tag->rule; /* matching rule to restart */
+ args.rule_id = dn_tag->rule_id;
+ args.chain_id = dn_tag->chain_id;
+ } else
+ args.rule = NULL;
/*
* I need some amt of data to be contiguous, and in case others need
@@ -487,7 +494,6 @@ ether_ipfw_chk(struct mbuf **m0, struct
args.m = m; /* the packet we are looking at */
args.oif = dst; /* destination, if any */
- args.rule = *rule; /* matching rule to restart */
args.next_hop = NULL; /* we do not support forward yet*/
args.eh = &save_eh; /* MAC header for bridged/MAC packets */
svn commit: r193861 - head/share/man/man4
Author: oleg
Date: Tue Jun 9 21:35:32 2009
New Revision: 193861
URL: http://svn.freebsd.org/changeset/base/193861
Log:
Sync struct ng_ipfw_tag description with sources.
Submitted by: Mikolaj Golub
Modified:
head/share/man/man4/ng_ipfw.4
Modified: head/share/man/man4/ng_ipfw.4
==
--- head/share/man/man4/ng_ipfw.4 Tue Jun 9 21:29:16 2009
(r193860)
+++ head/share/man/man4/ng_ipfw.4 Tue Jun 9 21:35:32 2009
(r193861)
@@ -24,7 +24,7 @@
.\"
.\" $FreeBSD$
.\"
-.Dd February 5, 2005
+.Dd June 10, 2009
.Dt NG_IPFW 4
.Os
.Sh NAME
@@ -84,11 +84,12 @@ to
struct ng_ipfw_tag {
struct m_tagmt; /* tag header */
struct ip_fw*rule; /* matching rule */
+ uint32_trule_id;/* matching rule id */
+ uint32_tchain_id; /* ruleset id */
struct ifnet *ifp; /* interface, for ip_output */
int dir;/* packet direction */
#defineNG_IPFW_OUT 0
#defineNG_IPFW_IN 1
- int flags; /* flags, for ip_output() */
};
.Ed
.Pp
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r340724 - head/sys/net
Author: oleg
Date: Wed Nov 21 13:34:21 2018
New Revision: 340724
URL: https://svnweb.freebsd.org/changeset/base/340724
Log:
Unbreak kernel build with VLAN_ARRAY defined.
MFC after:1 week
Modified:
head/sys/net/if_vlan.c
Modified: head/sys/net/if_vlan.c
==
--- head/sys/net/if_vlan.c Wed Nov 21 12:46:28 2018(r340723)
+++ head/sys/net/if_vlan.c Wed Nov 21 13:34:21 2018(r340724)
@@ -314,15 +314,15 @@ VNET_DEFINE_STATIC(struct if_clone *, vlan_cloner);
#defineV_vlan_cloner VNET(vlan_cloner)
#endif
-#ifndef VLAN_ARRAY
-#define HASH(n, m) n) >> 8) ^ ((n) >> 4) ^ (n)) & (m))
-
static void
vlan_mc_free(struct epoch_context *ctx)
{
struct vlan_mc_entry *mc = __containerof(ctx, struct vlan_mc_entry,
mc_epoch_ctx);
free(mc, M_VLAN);
}
+
+#ifndef VLAN_ARRAY
+#define HASH(n, m) n) >> 8) ^ ((n) >> 4) ^ (n)) & (m))
static void
vlan_inithash(struct ifvlantrunk *trunk)
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r341139 - stable/12/sys/net
Author: oleg
Date: Wed Nov 28 11:58:40 2018
New Revision: 341139
URL: https://svnweb.freebsd.org/changeset/base/341139
Log:
MFC r340724:
Unbreak kernel build with VLAN_ARRAY defined.
Modified:
stable/12/sys/net/if_vlan.c
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/sys/net/if_vlan.c
==
--- stable/12/sys/net/if_vlan.c Wed Nov 28 09:25:43 2018(r341138)
+++ stable/12/sys/net/if_vlan.c Wed Nov 28 11:58:40 2018(r341139)
@@ -314,15 +314,15 @@ VNET_DEFINE_STATIC(struct if_clone *, vlan_cloner);
#defineV_vlan_cloner VNET(vlan_cloner)
#endif
-#ifndef VLAN_ARRAY
-#define HASH(n, m) n) >> 8) ^ ((n) >> 4) ^ (n)) & (m))
-
static void
vlan_mc_free(struct epoch_context *ctx)
{
struct vlan_mc_entry *mc = __containerof(ctx, struct vlan_mc_entry,
mc_epoch_ctx);
free(mc, M_VLAN);
}
+
+#ifndef VLAN_ARRAY
+#define HASH(n, m) n) >> 8) ^ ((n) >> 4) ^ (n)) & (m))
static void
vlan_inithash(struct ifvlantrunk *trunk)
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r341236 - releng/12.0/sys/net
Author: oleg
Date: Thu Nov 29 15:26:07 2018
New Revision: 341236
URL: https://svnweb.freebsd.org/changeset/base/341236
Log:
MFS r341139:
MFC r340724:
Unbreak kernel build with VLAN_ARRAY defined.
Approved by: re (gjb)
Modified:
releng/12.0/sys/net/if_vlan.c
Directory Properties:
releng/12.0/ (props changed)
Modified: releng/12.0/sys/net/if_vlan.c
==
--- releng/12.0/sys/net/if_vlan.c Thu Nov 29 15:07:59 2018
(r341235)
+++ releng/12.0/sys/net/if_vlan.c Thu Nov 29 15:26:07 2018
(r341236)
@@ -314,15 +314,15 @@ VNET_DEFINE_STATIC(struct if_clone *, vlan_cloner);
#defineV_vlan_cloner VNET(vlan_cloner)
#endif
-#ifndef VLAN_ARRAY
-#define HASH(n, m) n) >> 8) ^ ((n) >> 4) ^ (n)) & (m))
-
static void
vlan_mc_free(struct epoch_context *ctx)
{
struct vlan_mc_entry *mc = __containerof(ctx, struct vlan_mc_entry,
mc_epoch_ctx);
free(mc, M_VLAN);
}
+
+#ifndef VLAN_ARRAY
+#define HASH(n, m) n) >> 8) ^ ((n) >> 4) ^ (n)) & (m))
static void
vlan_inithash(struct ifvlantrunk *trunk)
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r304154 - head/sys/netpfil/ipfw
Author: oleg Date: Mon Aug 15 13:06:29 2016 New Revision: 304154 URL: https://svnweb.freebsd.org/changeset/base/304154 Log: Fix command: ipfw set (enable|disable) N (where N > 4). enable_sets() expects set bitmasks, not set numbers. MFC after:3 days Modified: head/sys/netpfil/ipfw/ip_fw_sockopt.c Modified: head/sys/netpfil/ipfw/ip_fw_sockopt.c == --- head/sys/netpfil/ipfw/ip_fw_sockopt.c Mon Aug 15 12:56:45 2016 (r304153) +++ head/sys/netpfil/ipfw/ip_fw_sockopt.c Mon Aug 15 13:06:29 2016 (r304154) @@ -1420,8 +1420,10 @@ manage_sets(struct ip_fw_chain *chain, i if (rh->range.head.length != sizeof(ipfw_range_tlv)) return (1); - if (rh->range.set >= IPFW_MAX_SETS || - rh->range.new_set >= IPFW_MAX_SETS) + /* enable_sets() expects bitmasks. */ + if (op3->opcode != IP_FW_SET_ENABLE && + (rh->range.set >= IPFW_MAX_SETS || + rh->range.new_set >= IPFW_MAX_SETS)) return (EINVAL); ret = 0; ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r304415 - stable/11/sys/netpfil/ipfw
Author: oleg Date: Thu Aug 18 11:23:02 2016 New Revision: 304415 URL: https://svnweb.freebsd.org/changeset/base/304415 Log: MFC r304154 Fix command: ipfw set (enable|disable) N (where N > 4). Modified: stable/11/sys/netpfil/ipfw/ip_fw_sockopt.c Modified: stable/11/sys/netpfil/ipfw/ip_fw_sockopt.c == --- stable/11/sys/netpfil/ipfw/ip_fw_sockopt.c Thu Aug 18 11:17:36 2016 (r304414) +++ stable/11/sys/netpfil/ipfw/ip_fw_sockopt.c Thu Aug 18 11:23:02 2016 (r304415) @@ -1414,8 +1414,10 @@ manage_sets(struct ip_fw_chain *chain, i if (rh->range.head.length != sizeof(ipfw_range_tlv)) return (1); - if (rh->range.set >= IPFW_MAX_SETS || - rh->range.new_set >= IPFW_MAX_SETS) + /* enable_sets() expects bitmasks. */ + if (op3->opcode != IP_FW_SET_ENABLE && + (rh->range.set >= IPFW_MAX_SETS || + rh->range.new_set >= IPFW_MAX_SETS)) return (EINVAL); ret = 0; ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r304419 - stable/11
Author: oleg Date: Thu Aug 18 11:41:58 2016 New Revision: 304419 URL: https://svnweb.freebsd.org/changeset/base/304419 Log: Fix directory properties missing in previous (r304415) commit. Modified: Directory Properties: stable/11/ (props changed) ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r339358 - head/sys/net
btw, this breaks kernel build with VLAN_ARRAY option defined. -- Oleg. === Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru === ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r309143 - head/sys/libkern
On Fri, Nov 25, 2016 at 01:49:33PM +, Fabien Thomas wrote:
> + if (reseed || atomic_cmpset_int(&arc4rand_iniseed_state,
> + ARC4_ENTR_HAVE, ARC4_ENTR_SEED)) {
> + ARC4_FOREACH(arc4)
> + arc4_randomstir(arc4);
> + }
How about eliminating this atomic? Converting it to:
if (reseed || arc4rand_iniseed_state == ARC4_ENTR_HAVE) {
arc4rand_iniseed_state = ARC4_ENTR_SEED;
ARC4_FOREACH(arc4);
arc4_randomstir(arc4);
}
should give you free performance boost. Having locked cmpxchg in main path is
not cheap (and as i can see it's really used only once,
when randomdev unblocks).
P.S. quick and dirty test gives me 10-30% of extra perfomance.
--
Oleg.
====
=== Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru ===
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r309281 - head/sbin/ipfw
Author: oleg
Date: Tue Nov 29 10:43:58 2016
New Revision: 309281
URL: https://svnweb.freebsd.org/changeset/base/309281
Log:
Fix 'ipfw delete set N':
do not emit meaningless 'rule 0 not found' warning if set was already empty.
MFC after:1 week
Modified:
head/sbin/ipfw/ipfw2.c
Modified: head/sbin/ipfw/ipfw2.c
==
--- head/sbin/ipfw/ipfw2.c Tue Nov 29 08:20:55 2016(r309280)
+++ head/sbin/ipfw/ipfw2.c Tue Nov 29 10:43:58 2016(r309281)
@@ -3221,7 +3221,7 @@ ipfw_delete(char *av[])
exitval = EX_UNAVAILABLE;
warn("rule %u: setsockopt(IP_FW_XDEL)",
rt.start_rule);
- } else if (rt.new_set == 0) {
+ } else if (rt.new_set == 0 && do_set == 0) {
exitval = EX_UNAVAILABLE;
if (rt.start_rule != rt.end_rule)
warnx("no rules rules in %u-%u range",
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r309372 - head/sys/sys
On Fri, Dec 02, 2016 at 04:11:32PM +0800, Sepherosa Ziehau wrote:
> peek_clear_sc is added to address the issue you mentioned. IMHO, this
> commit weakens the proper assertion.
I would say this check:
#ifdef DEBUG_BUFRING
int i;
for (i = br->br_cons_head; i != br->br_prod_head;
i = ((i + 1) & br->br_cons_mask))
if(br->br_ring[i] == buf)
panic("buf=%p already enqueue at %d prod=%d cons=%d",
buf, i, br->br_prod_tail, br->br_cons_tail);
#endif
can't be relied upon:
1) it does not synchronize with anything, neither consumers nor producers.
So you can read inconsistent data. For example, you are storing
NULL in buf_ring_peek_clear_sc() but there is no guarantee you will see
it here.
2) it's not covered by critical section so thread running this check can be
preempted. Consider the following scenario:
a) thread is running this check and gets preempted by other producer
before i != br->br_prod_head check.
b) other producer moves br->br_prod_head forward.
c) if we are unlucky we can spin forever.
Current buf_ring implementation has insufficient memory ordering constraints.
I've tried to fix acq/rel usage here:
https://reviews.freebsd.org/D8637
but didn't get any review yet.
>
> On Fri, Dec 2, 2016 at 5:08 AM, Ryan Stone wrote:
> > Author: rstone
> > Date: Thu Dec 1 21:08:42 2016
> > New Revision: 309372
> > URL: https://svnweb.freebsd.org/changeset/base/309372
> >
> > Log:
> > Fix a false positive in a buf_ring assert
> >
> > buf_ring contains an assert that checks whether an item being
> > enqueued already exists on the ring. There is a subtle bug in
> > this assert. An item can be returned by a peek() function and
> > freed, and then the consumer thread can be preempted before
> > calling advance(). If this happens the item appears to still be
> > on the queue, but another thread may allocate the item from the
> > free pool and wind up trying to enqueue it again, causing the
> > assert to trigger incorrectly.
> >
> > Fix this by skipping the head of the consumer's portion of the
> > ring, as this index is what will be returned by peek().
> >
> > Sponsored by: Dell EMC Isilon
> > MFC After:1 week
> > Differential Revision:https://reviews.freebsd.org/D8685
> > Reviewed by: hselasky
> >
> > Modified:
> > head/sys/sys/buf_ring.h
> >
> > Modified: head/sys/sys/buf_ring.h
> > ==
> > --- head/sys/sys/buf_ring.h Thu Dec 1 20:36:48 2016(r309371)
> > +++ head/sys/sys/buf_ring.h Thu Dec 1 21:08:42 2016(r309372)
> > @@ -67,11 +67,13 @@ buf_ring_enqueue(struct buf_ring *br, vo
> > uint32_t prod_head, prod_next, cons_tail;
> > #ifdef DEBUG_BUFRING
> > int i;
> > - for (i = br->br_cons_head; i != br->br_prod_head;
> > -i = ((i + 1) & br->br_cons_mask))
> > - if(br->br_ring[i] == buf)
> > - panic("buf=%p already enqueue at %d prod=%d
> > cons=%d",
> > - buf, i, br->br_prod_tail, br->br_cons_tail);
> > + if (br->br_cons_head != br->br_prod_head) {
> > + for (i = (br->br_cons_head + 1) & br->br_cons_mask; i !=
> > br->br_prod_head;
> > + i = ((i + 1) & br->br_cons_mask))
> > + if(br->br_ring[i] == buf)
> > + panic("buf=%p already enqueue at %d prod=%d
> > cons=%d",
> > + buf, i, br->br_prod_tail,
> > br->br_cons_tail);
> > + }
> > #endif
> > critical_enter();
> > do {
> > ___
> > svn-src-all@freebsd.org mailing list
> > https://lists.freebsd.org/mailman/listinfo/svn-src-all
> > To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
>
>
>
> --
> Tomorrow Will Never Die
--
Oleg.
=== Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru ===
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r305738 - releng/11.0/sys/netpfil/ipfw
On Mon, Sep 12, 2016 at 03:57:35PM +, Andrey V. Elsukov wrote: > Author: ae > Date: Mon Sep 12 15:57:35 2016 > New Revision: 305738 > URL: https://svnweb.freebsd.org/changeset/base/305738 > > Log: > Merge from stable/11 r304415,304419 (by oleg): > Fix command: ipfw set (enable|disable) N (where N > 4). > > PR: 212595 > Approved by:re (kib) Thank you. -- Oleg. ==== === Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru === ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r332402 - head/sys/netpfil/ipfw
Author: oleg
Date: Wed Apr 11 11:12:20 2018
New Revision: 332402
URL: https://svnweb.freebsd.org/changeset/base/332402
Log:
Fix ipfw table creation when net.inet.ip.fw.tables_sets = 0 and non zero set
specified on table creation. This fixes following:
# sysctl net.inet.ip.fw.tables_sets
net.inet.ip.fw.tables_sets: 0
# ipfw table all info
# ipfw set 1 table 1 create type addr
# ipfw set 1 table 1 create type addr
# ipfw add 10 set 1 count ip from table\(1\) to any
00010 count ip from table(1) to any
# ipfw add 10 set 1 count ip from table\(1\) to any
00010 count ip from table(1) to any
# ipfw table all info
--- table(1), set(1) ---
kindex: 4, type: addr
references: 1, valtype: legacy
algorithm: addr:radix
items: 0, size: 296
--- table(1), set(1) ---
kindex: 3, type: addr
references: 1, valtype: legacy
algorithm: addr:radix
items: 0, size: 296
--- table(1), set(1) ---
kindex: 2, type: addr
references: 0, valtype: legacy
algorithm: addr:radix
items: 0, size: 296
--- table(1), set(1) ---
kindex: 1, type: addr
references: 0, valtype: legacy
algorithm: addr:radix
items: 0, size: 296
#
MFC after:1 week
Modified:
head/sys/netpfil/ipfw/ip_fw_table.c
Modified: head/sys/netpfil/ipfw/ip_fw_table.c
==
--- head/sys/netpfil/ipfw/ip_fw_table.c Wed Apr 11 10:36:20 2018
(r332401)
+++ head/sys/netpfil/ipfw/ip_fw_table.c Wed Apr 11 11:12:20 2018
(r332402)
@@ -3171,7 +3171,7 @@ alloc_table_config(struct ip_fw_chain *ch, struct tid_
if (ntlv == NULL)
return (NULL);
name = ntlv->name;
- set = ntlv->set;
+ set = (V_fw_tables_sets == 0) ? 0 : ntlv->set;
} else {
/* Compat part: convert number to string representation */
snprintf(bname, sizeof(bname), "%d", ti->uidx);
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r332403 - head/sbin/ipfw
Author: oleg Date: Wed Apr 11 11:17:57 2018 New Revision: 332403 URL: https://svnweb.freebsd.org/changeset/base/332403 Log: Fix typo. MFC after:1 week Modified: head/sbin/ipfw/ipfw.8 Modified: head/sbin/ipfw/ipfw.8 == --- head/sbin/ipfw/ipfw.8 Wed Apr 11 11:12:20 2018(r332402) +++ head/sbin/ipfw/ipfw.8 Wed Apr 11 11:17:57 2018(r332403) @@ -2233,7 +2233,7 @@ of the firewall and quickly (and atomically) switch be By default, tables from set 0 are referenced when adding rule with table opcodes regardless of rule set. This behavior can be changed by setting -.Va net.inet.ip.fw.tables_set +.Va net.inet.ip.fw.tables_sets variable to 1. Rule's set will then be used for table references. .Pp ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r332772 - stable/11/sys/netpfil/ipfw
Author: oleg
Date: Thu Apr 19 15:02:42 2018
New Revision: 332772
URL: https://svnweb.freebsd.org/changeset/base/332772
Log:
Fix ipfw table creation when net.inet.ip.fw.tables_sets = 0 and non zero set
specified on table creation. This fixes following:
# sysctl net.inet.ip.fw.tables_sets
net.inet.ip.fw.tables_sets: 0
# ipfw table all info
# ipfw set 1 table 1 create type addr
# ipfw set 1 table 1 create type addr
# ipfw add 10 set 1 count ip from table\(1\) to any
00010 count ip from table(1) to any
# ipfw add 10 set 1 count ip from table\(1\) to any
00010 count ip from table(1) to any
# ipfw table all info
--- table(1), set(1) ---
kindex: 4, type: addr
references: 1, valtype: legacy
algorithm: addr:radix
items: 0, size: 296
--- table(1), set(1) ---
kindex: 3, type: addr
references: 1, valtype: legacy
algorithm: addr:radix
items: 0, size: 296
--- table(1), set(1) ---
kindex: 2, type: addr
references: 0, valtype: legacy
algorithm: addr:radix
items: 0, size: 296
--- table(1), set(1) ---
kindex: 1, type: addr
references: 0, valtype: legacy
algorithm: addr:radix
items: 0, size: 296
#
Modified:
stable/11/sys/netpfil/ipfw/ip_fw_table.c
Directory Properties:
stable/11/ (props changed)
Modified: stable/11/sys/netpfil/ipfw/ip_fw_table.c
==
--- stable/11/sys/netpfil/ipfw/ip_fw_table.cThu Apr 19 14:09:44 2018
(r332771)
+++ stable/11/sys/netpfil/ipfw/ip_fw_table.cThu Apr 19 15:02:42 2018
(r332772)
@@ -3169,7 +3169,7 @@ alloc_table_config(struct ip_fw_chain *ch, struct tid_
if (ntlv == NULL)
return (NULL);
name = ntlv->name;
- set = ntlv->set;
+ set = (V_fw_tables_sets == 0) ? 0 : ntlv->set;
} else {
/* Compat part: convert number to string representation */
snprintf(bname, sizeof(bname), "%d", ti->uidx);
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r332775 - stable/11/sbin/ipfw
Author: oleg Date: Thu Apr 19 15:07:53 2018 New Revision: 332775 URL: https://svnweb.freebsd.org/changeset/base/332775 Log: MFC r332403: Fix a typo. Modified: stable/11/sbin/ipfw/ipfw.8 Directory Properties: stable/11/ (props changed) Modified: stable/11/sbin/ipfw/ipfw.8 == --- stable/11/sbin/ipfw/ipfw.8 Thu Apr 19 15:03:48 2018(r332774) +++ stable/11/sbin/ipfw/ipfw.8 Thu Apr 19 15:07:53 2018(r332775) @@ -2233,7 +2233,7 @@ of the firewall and quickly (and atomically) switch be By default, tables from set 0 are referenced when adding rule with table opcodes regardless of rule set. This behavior can be changed by setting -.Va net.inet.ip.fw.tables_set +.Va net.inet.ip.fw.tables_sets variable to 1. Rule's set will then be used for table references. .Pp ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r327013 - head/sys/net
On Wed, Dec 20, 2017 at 01:03:34AM +, Stephen Hurd wrote: > Author: shurd > Date: Wed Dec 20 01:03:34 2017 > New Revision: 327013 > URL: https://svnweb.freebsd.org/changeset/base/327013 > > Log: > Support attaching tx queues to cpus > > This will attempt to use a different thread/core on the same L2 > cache when possible, or use the same cpu as the rx thread when not. > If SMP isn't enabled, don't go looking for cores to use. This is mostly > useful when using shared TX/RX queues. > > Reviewed by:sbruno > Sponsored by: Limelight Networks > Differential Revision: https://reviews.freebsd.org/D12446 i'm getting this panic: [skipped] ix0: port 0xcc00-0xcc1f mem 0xfde8-0xfdef,0xfde7c000-0xfde7 irq 18 at device 0.0 on pci2 ix0: using 2048 tx descriptors and 2048 rx descriptors ix0: msix_init qsets capped at 16 ix0: pxm cpus: 4 queue msgs: 63 admincnt: 1 ix0: using 4 rx queues 4 tx queues ix0: attempting to allocate 5 MSI-X vectors (64 supported) msi: routing MSI-X IRQ 259 to local APIC 0 vector 54 msi: routing MSI-X IRQ 260 to local APIC 0 vector 55 msi: routing MSI-X IRQ 261 to local APIC 0 vector 56 msi: routing MSI-X IRQ 262 to local APIC 0 vector 57 msi: routing MSI-X IRQ 263 to local APIC 0 vector 58 ix0: using IRQs 259-263 for MSI-X ix0: Using MSIX interrupts with 5 vectors ix0: allocated for 4 queues ix0: allocated for 4 rx queues panic: Assertion gtask->gt_taskqueue != NULL failed at /usr/src/sys/net/iflib.c:5177 cpuid = 0 time = 1 KDB: enter: panic [ thread pid 0 tid 10 ] Stopped at kdb_enter+0x3b: movq$0,kdb_why db> bt Tracing pid 0 tid 10 td 0x80bf63c0 kdb_enter() at kdb_enter+0x3b/frame 0x81004130 vpanic() at vpanic+0x1b9/frame 0x810041b0 kassert_panic() at kassert_panic+0x173/frame 0x81004220 iflib_irq_set_affinity() at iflib_irq_set_affinity+0xae2/frame 0x810042e0 iflib_irq_alloc_generic() at iflib_irq_alloc_generic+0x1f7/frame 0x81004350 ixgbe_if_msix_intr_assign() at ixgbe_if_msix_intr_assign+0xa2/frame 0x810043d0 iflib_device_register() at iflib_device_register+0x2a6a/frame 0x810046f0 iflib_device_attach() at iflib_device_attach+0xb7/frame 0x81004720 device_attach() at device_attach+0x3f7/frame 0x81004770 bus_generic_attach() at bus_generic_attach+0x5a/frame 0x81004790 pci_attach() at pci_attach+0xd5/frame 0x810047d0 device_attach() at device_attach+0x3f7/frame 0x81004820 bus_generic_attach() at bus_generic_attach+0x5a/frame 0x81004840 acpi_pcib_pci_attach() at acpi_pcib_pci_attach+0xa1/frame 0x81004880 device_attach() at device_attach+0x3f7/frame 0x810048d0 bus_generic_attach() at bus_generic_attach+0x5a/frame 0x810048f0 pci_attach() at pci_attach+0xd5/frame 0x81004930 device_attach() at device_attach+0x3f7/frame 0x81004980 bus_generic_attach() at bus_generic_attach+0x5a/frame 0x810049a0 acpi_pcib_acpi_attach() at acpi_pcib_acpi_attach+0x3bc/frame 0x81004a10 device_attach() at device_attach+0x3f7/frame 0x81004a60 bus_generic_attach() at bus_generic_attach+0x5a/frame 0x81004a80 acpi_attach() at acpi_attach+0xe85/frame 0x81004b30 device_attach() at device_attach+0x3f7/frame 0x81004b80 bus_generic_attach() at bus_generic_attach+0x5a/frame 0x81004ba0 nexus_acpi_attach() at nexus_acpi_attach+0x73/frame 0x81004bd0 device_attach() at device_attach+0x3f7/frame 0x81004c20 bus_generic_new_pass() at bus_generic_new_pass+0x116/frame 0x81004c50 bus_set_pass() at bus_set_pass+0x8c/frame 0x81004c80 configure() at configure+0x9/frame 0x81004c90 mi_startup() at mi_startup+0x9c/frame 0x81004cb0 btext() at btext+0x2c -- Oleg. ==== === Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru === ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r327013 - head/sys/net
On Fri, Dec 22, 2017 at 01:22:20PM -0500, Stephen Hurd wrote: > Can you provide me with the output of ``sysctl kern.sched.topology_spec''? # sysctl kern.sched.topology_spec kern.sched.topology_spec: 0, 1, 2, 3 0 1 2 3 > And provide the new dmesg with this added output. ix0: port 0xcc00-0xcc1f mem 0xfde8-0xfdef,0xfde7c000-0xfde7 irq 18 at device 0.0 on pci2 ix0: using 2048 tx descriptors and 2048 rx descriptors ix0: msix_init qsets capped at 16 ix0: pxm cpus: 4 queue msgs: 63 admincnt: 1 ix0: using 4 rx queues 4 tx queues ix0: attempting to allocate 5 MSI-X vectors (64 supported) msi: routing MSI-X IRQ 259 to local APIC 0 vector 54 msi: routing MSI-X IRQ 260 to local APIC 0 vector 55 msi: routing MSI-X IRQ 261 to local APIC 0 vector 56 msi: routing MSI-X IRQ 262 to local APIC 0 vector 57 msi: routing MSI-X IRQ 263 to local APIC 0 vector 58 ix0: using IRQs 259-263 for MSI-X ix0: Using MSIX interrupts with 5 vectors ix0: allocated for 4 queues ix0: allocated for 4 rx queues ix0: irq=259, type=0, qid=0, name=rxq0 panic: Assertion gtask->gt_taskqueue != NULL failed at /usr/src/sys/net/iflib.c:5178 cpuid = 0 time = 1 KDB: enter: panic -- Oleg. ==== === Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru === ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r327013 - head/sys/net
On Fri, Dec 22, 2017 at 07:11:36PM -0500, Stephen Hurd wrote: > Interesting... is this a standard GENERIC kernel, or is this a custom > config? If a custom config, can you send it to me? > > Also, does this system have any other iflib drivers? igb, em, or bnxt? > Can you include the log info (if any) for them as well? It's custom kernel on my testing machine. It has em too but ix is probed first.. If i comment out if_ix.ko load - it boots just fine (looks like if_em is not calling iflib_irq_set_affinity()): em0: port 0xdc00-0xdc1f mem 0xfeae-0x feaf,0xfeac-0xfead irq 17 at device 0.0 on pci3 em0: attach_pre capping queues at 1 em0: using 1024 tx descriptors and 1024 rx descriptors em0: msix_init qsets capped at 1 em0: PCIY_MSIX capability not found; or rid 0 == 0. em0: Using a Legacy interrupt em0: allocated for 1 tx_queues em0: allocated for 1 rx_queues ioapic0: routing intpin 17 (PCI IRQ 17) to lapic 0 vector 54 em0: bpf attached em0: Ethernet address: 00:15:17:3a:0c:7a em0: netmap queues/slots: TX 1/1024, RX 1/1024 P.S. verbose dmesg & kernel conf attached -- Oleg. ==== === Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru === /boot/config: 0:ad(0,a)/boot/loader -Dh -S115200 FreeBSD/x86 boot Default: 0:ad(0,a)/boot/loader boot: Consoles: internal video/keyboard serial port BIOS drive C: is disk0 BIOS 619kB/3111360kB available memory FreeBSD/x86 bootstrap loader, Revision 1.1 (Thu Dec 21 19:31:04 MSK 2017 r...@slave.rinet.ru) Loading /boot/defaults/loader.conf /boot/kernel/kernel text=0x5826b8 data=0x7d500+0x2b3378 syms=[0x8+0x9e880+0x8+0xa0fbf] /boot/entropy size=0x1000 /boot/kernel/nullfs.ko size 0xbd30 at 0xdf4000 /boot/kernel/geom_mirror.ko size 0x2c968 at 0xe0 /boot/kernel/ipfw.ko size 0x47820 at 0xe2d000 /boot/kernel/carp.ko size 0x13420 at 0xe75000 /boot/kernel/if_bge.ko size 0x247c8 at 0xe89000 loading required module 'miibus' /boot/kernel/miibus.ko size 0x414b8 at 0xeae000 /boot/kernel/if_em.ko size 0x6ca88 at 0xef /boot/kernel/if_nfe.ko size 0x13480 at 0xf5d000 /boot/kernel/if_re.ko size 0x13e98 at 0xf71000 /boot/kernel/if_ix.ko size 0x52bc0 at 0xf85000 /boot/kernel/libalias.ko size 0x19dd8 at 0xfd8000 Hit [Enter] to boot immediately, or any other key for command prompt. Booting [/boot/kernel/kernel]... GDB: no debug ports present KDB: debugger backends: ddb KDB: current backend: ddb Table 'FACP' at 0xbdf70290 Table 'APIC' at 0xbdf70390 APIC: Found table at 0xbdf70390 APIC: Using the MADT enumerator. Copyright (c) 1992-2017 The FreeBSD Project. Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994 The Regents of the University of California. All rights reserved. FreeBSD is a registered trademark of The FreeBSD Foundation. FreeBSD 12.0-CURRENT #5 r327061M: Sat Dec 23 01:53:58 MSK 2017 r...@slave.rinet.ru:/usr/obj/usr/src/amd64.amd64/sys/slave-amd64-smp-debug amd64 FreeBSD clang version 5.0.1 (tags/RELEASE_501/final 320880) (based on LLVM 5.0.1) WARNING: WITNESS option enabled, expect reduced performance. WARNING: DIAGNOSTIC option enabled, expect reduced performance. Table 'FACP' at 0xbdf70290 Table 'APIC' at 0xbdf70390 Table 'MCFG' at 0xbdf70420 Table 'OEMB' at 0xbdf88040 Table 'HPET' at 0xbdf7f8b0 Table 'SSDT' at 0xbdf7f8f0 ACPI: No SRAT table found Preloaded elf kernel "/boot/kernel/kernel" at 0x80ff2000. Preloaded boot_entropy_cache "/boot/entropy" at 0x80ffacd8. Preloaded elf obj module "/boot/kernel/nullfs.ko" at 0x80ffad30. Preloaded elf obj module "/boot/kernel/geom_mirror.ko" at 0x80ffb358. Preloaded elf obj module "/boot/kernel/ipfw.ko" at 0x80ffba48. Preloaded elf obj module "/boot/kernel/carp.ko" at 0x80ffc0f0. Preloaded elf obj module "/boot/kernel/if_bge.ko" at 0x80ffc7d8. Preloaded elf obj module "/boot/kernel/miibus.ko" at 0x80ffce40. Preloaded elf obj module "/boot/kernel/if_em.ko" at 0x80ffd428. Preloaded elf obj module "/boot/kernel/if_nfe.ko" at 0x80ffda90. Preloaded elf obj module "/boot/kernel/if_re.ko" at 0x80ffe038. Preloaded elf obj module "/boot/kernel/if_ix.ko" at 0x80ffe5e0. Preloaded elf obj module "/boot/kernel/libalias.ko" at 0x80ffed08. link_elf_obj: symbol sctp_calculate_cksum undefined KLD file ipfw.ko - could not finalize loading Calibrating TSC clock ... TSC clock: 3515725584 Hz CPU: AMD Phenom(tm) II X4 970 Processor (3515.73-MHz K8-class CPU) Origin="AuthenticAMD" Id=0x100f43 Family=0x10 Model=0x4 Stepping=3 Features=0x178bfbff Features2=0x802009 AMD Features=0xee500800 AMD Fea
Re: svn commit: r327013 - head/sys/net
On Sat, Dec 23, 2017 at 11:13:37PM -0500, Stephen Hurd wrote: > Hrm, can you try it with: > optionsEARLY_AP_STARTUP > > In slave-amd64-smp-debug? yes, it does boot now: ix0: port 0xcc00-0xcc1f mem 0xfd e8-0xfdef,0xfde7c000-0xfde7 irq 18 at device 0.0 on pci2 ix0: using 2048 tx descriptors and 2048 rx descriptors ix0: msix_init qsets capped at 16 ix0: pxm cpus: 4 queue msgs: 63 admincnt: 1 ix0: using 4 rx queues 4 tx queues ix0: attempting to allocate 5 MSI-X vectors (64 supported) msi: routing MSI-X IRQ 259 to local APIC 2 vector 48 msi: routing MSI-X IRQ 260 to local APIC 3 vector 48 msi: routing MSI-X IRQ 261 to local APIC 0 vector 53 msi: routing MSI-X IRQ 262 to local APIC 1 vector 49 msi: routing MSI-X IRQ 263 to local APIC 2 vector 49 ix0: using IRQs 259-263 for MSI-X ix0: Using MSIX interrupts with 5 vectors ix0: allocated for 4 queues ix0: allocated for 4 rx queues ix0: irq=259, type=0, qid=0, name=rxq0 msi: Assigning MSI-X IRQ 259 to local APIC 0 vector 54 ix0: irq=260, type=0, qid=1, name=rxq1 msi: Assigning MSI-X IRQ 260 to local APIC 1 vector 50 ix0: irq=261, type=0, qid=2, name=rxq2 msi: Assigning MSI-X IRQ 261 to local APIC 2 vector 48 ix0: irq=262, type=0, qid=3, name=rxq3 msi: Assigning MSI-X IRQ 262 to local APIC 3 vector 48 ix0: irq=259, type=1, qid=0, name=txq0 ix0: irq=260, type=1, qid=1, name=txq1 ix0: irq=261, type=1, qid=2, name=txq2 ix0: irq=262, type=1, qid=3, name=txq3 ix0: bpf attached ix0: Ethernet address: 90:e2:ba:49:ab:70 ix0: PCI Express Bus: Speed 5.0GT/s Width x8 ix0: netmap queues/slots: TX 4/2048, RX 4/2048 random: harvesting attach, 8 bytes (4 bits) from ix0 ix1: port 0xc880-0xc89f mem 0xfdd8-0xfddf,0xfde78000-0xfde7bfff irq 19 at device 0.1 on pci2 ix1: using 2048 tx descriptors and 2048 rx descriptors ix1: msix_init qsets capped at 16 ix1: pxm cpus: 4 queue msgs: 63 admincnt: 1 ix1: using 4 rx queues 4 tx queues ix1: attempting to allocate 5 MSI-X vectors (64 supported) msi: routing MSI-X IRQ 264 to local APIC 3 vector 49 msi: routing MSI-X IRQ 265 to local APIC 0 vector 53 msi: routing MSI-X IRQ 266 to local APIC 1 vector 49 msi: routing MSI-X IRQ 267 to local APIC 2 vector 50 msi: routing MSI-X IRQ 268 to local APIC 3 vector 50 ix1: using IRQs 264-268 for MSI-X ix1: Using MSIX interrupts with 5 vectors ix1: allocated for 4 queues ix1: allocated for 4 rx queues ix1: irq=264, type=0, qid=0, name=rxq0 msi: Assigning MSI-X IRQ 264 to local APIC 0 vector 55 ix1: irq=265, type=0, qid=1, name=rxq1 msi: Assigning MSI-X IRQ 265 to local APIC 1 vector 51 ix1: irq=266, type=0, qid=2, name=rxq2 msi: Assigning MSI-X IRQ 266 to local APIC 2 vector 51 ix1: irq=267, type=0, qid=3, name=rxq3 msi: Assigning MSI-X IRQ 267 to local APIC 3 vector 49 ix1: irq=264, type=1, qid=0, name=txq0 ix1: irq=265, type=1, qid=1, name=txq1 ix1: irq=266, type=1, qid=2, name=txq2 ix1: irq=267, type=1, qid=3, name=txq3 ix1: bpf attached ix1: Ethernet address: 90:e2:ba:49:ab:71 ix1: PCI Express Bus: Speed 5.0GT/s Width x8 ix1: netmap queues/slots: TX 4/2048, RX 4/2048 random: harvesting attach, 8 bytes (4 bits) from ix1 random: harvesting attach, 8 bytes (4 bits) from pci2 random: harvesting attach, 8 bytes (4 bits) from pcib2 -- Oleg. ==== === Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru === ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r309713 - stable/11/sbin/ipfw
Author: oleg
Date: Thu Dec 8 12:42:36 2016
New Revision: 309713
URL: https://svnweb.freebsd.org/changeset/base/309713
Log:
MFC r309281:
Fix 'ipfw delete set N':
do not emit meaningless 'rule 0 not found' warning if set was already empty.
Modified:
stable/11/sbin/ipfw/ipfw2.c
Directory Properties:
stable/11/ (props changed)
Modified: stable/11/sbin/ipfw/ipfw2.c
==
--- stable/11/sbin/ipfw/ipfw2.c Thu Dec 8 11:39:08 2016(r309712)
+++ stable/11/sbin/ipfw/ipfw2.c Thu Dec 8 12:42:36 2016(r309713)
@@ -3194,7 +3194,7 @@ ipfw_delete(char *av[])
exitval = EX_UNAVAILABLE;
warn("rule %u: setsockopt(IP_FW_XDEL)",
rt.start_rule);
- } else if (rt.new_set == 0) {
+ } else if (rt.new_set == 0 && do_set == 0) {
exitval = EX_UNAVAILABLE;
if (rt.start_rule != rt.end_rule)
warnx("no rules rules in %u-%u range",
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r322307 - head/sys/netinet
Author: oleg Date: Wed Aug 9 10:46:34 2017 New Revision: 322307 URL: https://svnweb.freebsd.org/changeset/base/322307 Log: Fix comment typo. Modified: head/sys/netinet/if_ether.c Modified: head/sys/netinet/if_ether.c == --- head/sys/netinet/if_ether.c Wed Aug 9 10:32:51 2017(r322306) +++ head/sys/netinet/if_ether.c Wed Aug 9 10:46:34 2017(r322307) @@ -1279,7 +1279,7 @@ arp_mark_lle_reachable(struct llentry *la) } /* - * Add pernament link-layer record for given interface address. + * Add permanent link-layer record for given interface address. */ static __noinline void arp_add_ifa_lle(struct ifnet *ifp, const struct sockaddr *dst) ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r322353 - head/usr.bin/calendar/calendars
Author: oleg Date: Thu Aug 10 12:31:55 2017 New Revision: 322353 URL: https://svnweb.freebsd.org/changeset/base/322353 Log: Add myself. Reported by: mckusick Modified: head/usr.bin/calendar/calendars/calendar.freebsd Modified: head/usr.bin/calendar/calendars/calendar.freebsd == --- head/usr.bin/calendar/calendars/calendar.freebsdThu Aug 10 12:30:34 2017(r322352) +++ head/usr.bin/calendar/calendars/calendar.freebsdThu Aug 10 12:31:55 2017(r322353) @@ -155,6 +155,7 @@ 04/22 Jakub Klama born in Blachownia, Silesia, Poland, 1989 04/25 Richard Gallamore born in Kissimmee, Florida, United States, 1987 04/26 Rene Ladan born in Geldrop, the Netherlands, 1980 +04/28 Oleg Bulyzhin born in Kharkov, USSR, 1976 04/28 Andriy Voskoboinyk born in Bila Tserkva, Ukraine, 1992 04/29 Adam Weinberger born in Berkeley, California, United States, 1980 04/29 Eric Anholt born in Portland, Oregon, United States, 1983 ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r322628 - head/sys/boot/common
Author: oleg Date: Thu Aug 17 19:39:42 2017 New Revision: 322628 URL: https://svnweb.freebsd.org/changeset/base/322628 Log: Fix BSD label partition end sector calculation. Reviewed by: ae MFC after:2 weeks Differential Revision:https://reviews.freebsd.org/D12066 Modified: head/sys/boot/common/part.c Modified: head/sys/boot/common/part.c == --- head/sys/boot/common/part.c Thu Aug 17 19:16:23 2017(r322627) +++ head/sys/boot/common/part.c Thu Aug 17 19:39:42 2017(r322628) @@ -506,7 +506,7 @@ ptable_bsdread(struct ptable *table, void *dev, diskre break; entry->part.start = le32toh(part->p_offset) - raw_offset; entry->part.end = entry->part.start + - le32toh(part->p_size) + 1; + le32toh(part->p_size) - 1; entry->part.type = bsd_parttype(part->p_fstype); entry->part.index = i; /* starts from zero */ entry->type.bsd = part->p_fstype; ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r323027 - in head/sys: dev/e1000 net
On Wed, Aug 30, 2017 at 06:56:24PM +, Sean Bruno wrote: > Author: sbruno > Date: Wed Aug 30 18:56:24 2017 > New Revision: 323027 > URL: https://svnweb.freebsd.org/changeset/base/323027 > > Log: > Revert r323008 and its conversion of e1000/iflib to using SX locks. > > This seems to be missing something on the 82574L causing NFS root mounts > to hang. JFYI: with r323008 my test machine keeps crashing during boot: [skipped] pcib3: allocated I/O port range (0xdc00-0xdc1f) for rid 18 of pci0:3:0:0 pcib3: matched entry for 3.0.INTA pcib3: slot 0 INTA hardwired to IRQ 17 em0: port 0xdc00-0xdc1f mem 0xfeae-0xfeaf,0xfeac-0xfead irq 17 at device 0.0 on pci3 em0: attach_pre capping queues at 1 panic: unknown mac type e cpuid = 0 time = 1 KDB: enter: panic [ thread pid 0 tid 10 ] Stopped at kdb_enter+0x3b: movq$0,kdb_why db> bt Tracing pid 0 tid 10 td 0x80bd5c80 kdb_enter() at kdb_enter+0x3b/frame 0x80ffc2b0 vpanic() at vpanic+0x1b9/frame 0x80ffc330 panic() at panic+0x43/frame 0x80ffc390 e1000_reset_hw_82571() at e1000_reset_hw_82571+0x626/frame 0x80ffc3d0 em_if_attach_pre() at em_if_attach_pre+0xe49/frame 0x80ffc440 iflib_device_register() at iflib_device_register+0x74b/frame 0x80ffc760 iflib_device_attach() at iflib_device_attach+0xb7/frame 0x80ffc790 device_attach() at device_attach+0x3ee/frame 0x80ffc7d0 bus_generic_attach() at bus_generic_attach+0x5a/frame 0x80ffc7f0 pci_attach() at pci_attach+0xd5/frame 0x80ffc830 device_attach() at device_attach+0x3ee/frame 0x80ffc870 bus_generic_attach() at bus_generic_attach+0x5a/frame 0x80ffc890 acpi_pcib_pci_attach() at acpi_pcib_pci_attach+0xa1/frame 0x80ffc8d0 device_attach() at device_attach+0x3ee/frame 0x80ffc910 bus_generic_attach() at bus_generic_attach+0x5a/frame 0x80ffc930 pci_attach() at pci_attach+0xd5/frame 0x80ffc970 device_attach() at device_attach+0x3ee/frame 0x80ffc9b0 bus_generic_attach() at bus_generic_attach+0x5a/frame 0x80ffc9d0 acpi_pcib_acpi_attach() at acpi_pcib_acpi_attach+0x3bc/frame 0x80ffca40 --More--^M^Mdevice_attach() at device_attach+0x3ee/frame 0x80ffca80 bus_generic_attach() at bus_generic_attach+0x5a/frame 0x80ffcaa0 acpi_attach() at acpi_attach+0xe85/frame 0x80ffcb50 device_attach() at device_attach+0x3ee/frame 0x80ffcb90 bus_generic_attach() at bus_generic_attach+0x5a/frame 0x80ffcbb0 nexus_acpi_attach() at nexus_acpi_attach+0x73/frame 0x80ffcbe0 device_attach() at device_attach+0x3ee/frame 0x80ffcc20 bus_generic_new_pass() at bus_generic_new_pass+0x116/frame 0x80ffcc50 bus_set_pass() at bus_set_pass+0x8c/frame 0x80ffcc80 configure() at configure+0x9/frame 0x80ffcc90 mi_startup() at mi_startup+0x9c/frame 0x80ffccb0 btext() at btext+0x2c db> Hardware is: # pciconf -lv|grep -A4 em0 em0@pci0:3:0:0: class=0x02 card=0x10828086 chip=0x107d8086 rev=0x06 hdr=0x00 vendor = 'Intel Corporation' device = '82572EI Gigabit Ethernet Controller (Copper)' class = network subclass = ethernet -- Oleg. ==== === Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru === ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323050 - stable/11/sys/boot/common
Author: oleg Date: Thu Aug 31 12:36:09 2017 New Revision: 323050 URL: https://svnweb.freebsd.org/changeset/base/323050 Log: MFC r322628: Fix BSD label partition end sector calculation. Differential Revision:https://reviews.freebsd.org/D12066 Modified: stable/11/sys/boot/common/part.c Directory Properties: stable/11/ (props changed) Modified: stable/11/sys/boot/common/part.c == --- stable/11/sys/boot/common/part.cThu Aug 31 12:02:14 2017 (r323049) +++ stable/11/sys/boot/common/part.cThu Aug 31 12:36:09 2017 (r323050) @@ -486,7 +486,7 @@ ptable_bsdread(struct ptable *table, void *dev, diskre break; entry->part.start = le32toh(part->p_offset) - raw_offset; entry->part.end = entry->part.start + - le32toh(part->p_size) + 1; + le32toh(part->p_size) - 1; entry->part.type = bsd_parttype(part->p_fstype); entry->part.index = i; /* starts from zero */ entry->type.bsd = part->p_fstype; ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r323052 - stable/10/sys/boot/common
Author: oleg Date: Thu Aug 31 13:32:01 2017 New Revision: 323052 URL: https://svnweb.freebsd.org/changeset/base/323052 Log: MFC r322628: Fix BSD label partition end sector calculation. Approved by: re (marius) Differential Revision:https://reviews.freebsd.org/D12066 Modified: stable/10/sys/boot/common/part.c Directory Properties: stable/10/ (props changed) Modified: stable/10/sys/boot/common/part.c == --- stable/10/sys/boot/common/part.cThu Aug 31 13:02:17 2017 (r323051) +++ stable/10/sys/boot/common/part.cThu Aug 31 13:32:01 2017 (r323052) @@ -483,7 +483,7 @@ ptable_bsdread(struct ptable *table, void *dev, diskre break; entry->part.start = le32toh(part->p_offset) - raw_offset; entry->part.end = entry->part.start + - le32toh(part->p_size) + 1; + le32toh(part->p_size) - 1; entry->part.type = bsd_parttype(part->p_fstype); entry->part.index = i; /* starts from zero */ entry->type.bsd = part->p_fstype; ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r305177 - head/sys/net
On Thu, Sep 01, 2016 at 06:32:35AM +, Sepherosa Ziehau wrote:
> Author: sephe
> Date: Thu Sep 1 06:32:35 2016
> New Revision: 305177
> URL: https://svnweb.freebsd.org/changeset/base/305177
>
> Log:
> net/vlan: Shift for pri is 13 (pri mask 0xe000) not 1.
>
> Reviewed by:araujo, hps
> MFC after: 1 week
> Sponsored by: Microsoft
> Differential Revision: https://reviews.freebsd.org/D7710
>
> Modified:
> head/sys/net/ethernet.h
>
> Modified: head/sys/net/ethernet.h
> ==
> --- head/sys/net/ethernet.h Thu Sep 1 06:05:08 2016(r305176)
> +++ head/sys/net/ethernet.h Thu Sep 1 06:32:35 2016(r305177)
> @@ -92,7 +92,7 @@ struct ether_vlan_header {
> #define EVL_PRIOFTAG(tag) (((tag) >> 13) & 7)
> #define EVL_CFIOFTAG(tag) (((tag) >> 12) & 1)
> #define EVL_MAKETAG(vlid, pri, cfi)
> \
> - ((pri) & 7) << 1) | ((cfi) & 1)) << 12) | ((vlid) & EVL_VLID_MASK))
> + ((pri) & 7) << 13) | ((cfi) & 1)) << 12) | ((vlid) & EVL_VLID_MASK))
>
> /*
> * NOTE: 0x-0x05DC (0..1500) are generally IEEE 802.3 length fields.
Please revert this one. It's just plain wrong and previous one was ok.
If you care about readability it should be:
((((pri) & 7) << 13) | (((cfi) & 1) << 12) | ((vlid) & EVL_VLID_MASK))
--
Oleg.
=== Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru ===
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r305177 - head/sys/net
On Thu, May 18, 2017 at 04:25:01PM +0200, Hans Petter Selasky wrote:
> On 05/18/17 16:04, Oleg Bulyzhin wrote:
> > On Thu, Sep 01, 2016 at 06:32:35AM +, Sepherosa Ziehau wrote:
> >> Author: sephe
> >> Date: Thu Sep 1 06:32:35 2016
> >> New Revision: 305177
> >> URL: https://svnweb.freebsd.org/changeset/base/305177
> >>
> >> Log:
> >>net/vlan: Shift for pri is 13 (pri mask 0xe000) not 1.
> >>
> >>Reviewed by:araujo, hps
> >>MFC after: 1 week
> >>Sponsored by: Microsoft
> >>Differential Revision: https://reviews.freebsd.org/D7710
> >>
> >> Modified:
> >>head/sys/net/ethernet.h
> >>
> >> Modified: head/sys/net/ethernet.h
> >> ==
> >> --- head/sys/net/ethernet.hThu Sep 1 06:05:08 2016
> >> (r305176)
> >> +++ head/sys/net/ethernet.hThu Sep 1 06:32:35 2016
> >> (r305177)
> >> @@ -92,7 +92,7 @@ struct ether_vlan_header {
> >> #define EVL_PRIOFTAG(tag) (((tag) >> 13) & 7)
> >> #define EVL_CFIOFTAG(tag) (((tag) >> 12) & 1)
> >> #define EVL_MAKETAG(vlid, pri, cfi)
> >> \
> >> - ((pri) & 7) << 1) | ((cfi) & 1)) << 12) | ((vlid) & EVL_VLID_MASK))
> >> + ((pri) & 7) << 13) | ((cfi) & 1)) << 12) | ((vlid) & EVL_VLID_MASK))
> >>
> >> /*
> >>* NOTE: 0x-0x05DC (0..1500) are generally IEEE 802.3 length fields.
> >
> > Please revert this one. It's just plain wrong and previous one was ok.
> >
>
> Hi,
>
> Can you explain a bit more what is wrong?
>
> > If you care about readability it should be:
> > pri) & 7) << 13) | (((cfi) & 1) << 12) | ((vlid) & EVL_VLID_MASK))
>
> Isn't this exactly what the patch is doing? -R ???
Current version is shifting pri out of uint16. If you examine parentheses:
pri is shifted left 13, then 12.
Original version did it right (shift 1, then 12 (total 13)).
--
Oleg.
=== Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru ===
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r287856 - head/sys/net
Author: oleg Date: Wed Sep 16 10:07:45 2015 New Revision: 287856 URL: https://svnweb.freebsd.org/changeset/base/287856 Log: Remove superfluous m_freem(). MFC after:1 month Modified: head/sys/net/if_ethersubr.c Modified: head/sys/net/if_ethersubr.c == --- head/sys/net/if_ethersubr.c Wed Sep 16 09:59:05 2015(r287855) +++ head/sys/net/if_ethersubr.c Wed Sep 16 10:07:45 2015(r287856) @@ -499,7 +499,6 @@ ether_input_internal(struct ifnet *ifp, if_printf(ifp, "cannot pullup VLAN header\n"); #endif if_inc_counter(ifp, IFCOUNTER_IERRORS, 1); - m_freem(m); CURVNET_RESTORE(); return; } ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r297481 - head/sys/kern
On Fri, Apr 01, 2016 at 08:12:50AM +0200, Mateusz Guzik wrote: > Author: mjg > Date: Wed Apr 1 08:10:00 2016 > New Revision: 280963 > URL: https://svnweb.freebsd.org/changeset/base/297481 Something is wrong with this commit. Revision 297481 is about head/sys/dev/hyperv/vmbus/hv_hv.c -- Oleg. ==== === Oleg Bulyzhin -- OBUL-RIPN -- OBUL-RIPE -- o...@rinet.ru === ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
