svn commit: r361355 - head/share/man/man4
Author: rgrimes Date: Fri May 22 03:13:29 2020 New Revision: 361355 URL: https://svnweb.freebsd.org/changeset/base/361355 Log: Include all currently present kernel options for IPFW Also fix igor complaint about manpage/s/man page Reported by: rgri...@freebsd.org PR: 219075 Submitted by: Dries Michiels driesm.michiels_gmail.com Reported by: rgrimes Reviewed by: bcr (manpages), 0mp MFC after:3 days Differential Revision:https://reviews.freebsd.org/D24541 Modified: head/share/man/man4/ipfirewall.4 Modified: head/share/man/man4/ipfirewall.4 == --- head/share/man/man4/ipfirewall.4Fri May 22 03:11:33 2020 (r361354) +++ head/share/man/man4/ipfirewall.4Fri May 22 03:13:29 2020 (r361355) @@ -1,7 +1,7 @@ .\" .\" $FreeBSD$ .\" -.Dd October 25, 2012 +.Dd May 21, 2020 .Dt IPFW 4 .Os .Sh NAME @@ -20,8 +20,14 @@ Other related kernel options which may also be useful are: .Bd -ragged -offset indent .Cd "options IPFIREWALL_DEFAULT_TO_ACCEPT" +.Cd "options IPDIVERT" +.Cd "options IPFIREWALL_NAT" +.Cd "options IPFIREWALL_NAT64" +.Cd "options IPFIREWALL_NPTV6" +.Cd "options IPFIREWALL_PMOD" .Cd "options IPFIREWALL_VERBOSE" .Cd "options IPFIREWALL_VERBOSE_LIMIT=100" +.Cd "options LIBALIAS" .Ed .Pp To load @@ -57,6 +63,54 @@ If the default behavior is to allow everything, it is easier to cope with firewall-tuning mistakes which may accidentally block all traffic. .Pp +When using +.Xr natd 8 +in conjunction with +.Nm +as +.Tn NAT +facility, the kernel option +.Dv IPDIVERT +enables diverting packets to +.Xr natd 8 +for translation. +.Pp +When using the in-kernel +.Tn NAT +facility of +.Nm , +the kernel option +.Dv IPFIREWALL_NAT +enables basic +.Xr libalias 3 +functionality in the kernel. +.Pp +When using any of the +.Tn IPv4 +to +.Tn IPv6 +transition mechanisms in +.Nm , +the kernel option +.Dv IPFIREWALL_NAT64 +enables all of these +.Tn NAT64 +methods in the kernel. +.Pp +When using the +.Tn IPv6 +network prefix translation facility of +.Nm , +the kernel option +.Dv IPFIREWALL_NPTV6 +enables this functionality in the kernel. +.Pp +When using the packet modification facility of +.Nm , +the kernel option +.Dv IPFIREWALL_PMOD +enables this functionality in the kernel. +.Pp To enable logging of packets passing through .Nm , enable the @@ -70,20 +124,39 @@ from flooding system logs or causing local Denial of S This option may be set to the number of packets which will be logged on a per-entry basis before the entry is rate-limited. .Pp +When using the in-kernel +.Tn NAT +facility of +.Nm , +the kernel option +.Dv LIBALIAS +enables full +.Xr libalias 3 +functionality in the kernel. +Full functionality refers to included support for cuseeme, ftp, bbt, +skinny, irc, pptp and smedia packets, which are missing in the basic +.Xr libalias 3 +functionality accomplished with the +.Dv IPFIREWALL_NAT +kernel option. +.Pp The user interface for .Nm is implemented by the .Xr ipfw 8 utility, so please refer to the .Xr ipfw 8 -manpage for a complete description of the +man page for a complete description of the .Nm capabilities and how to use it. .Sh SEE ALSO .Xr setsockopt 2 , .Xr divert 4 , .Xr ip 4 , +.Xr ip6 4 , .Xr ipfw 8 , +.Xr libalias 3 , +.Xr natd 8 , .Xr sysctl 8 , .Xr syslogd 8 , .Xr pfil 9 ___ 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: r361354 - stable/11/etc/ntp
Author: delphij Date: Fri May 22 03:11:33 2020 New Revision: 361354 URL: https://svnweb.freebsd.org/changeset/base/361354 Log: MFC r361260: Update leap-seconds to leap-seconds.3676924800. Modified: stable/11/etc/ntp/leap-seconds Directory Properties: stable/11/ (props changed) Modified: stable/11/etc/ntp/leap-seconds == --- stable/11/etc/ntp/leap-seconds Fri May 22 02:06:56 2020 (r361353) +++ stable/11/etc/ntp/leap-seconds Fri May 22 03:11:33 2020 (r361354) @@ -204,10 +204,10 @@ # current -- the update time stamp, the data and the name of the file # will not change. # -# Updated through IERS Bulletin C58 -# File expires on: 28 June 2020 +# Updated through IERS Bulletin C59 +# File expires on: 28 December 2020 # -#@ 3802291200 +#@ 3818102400 # 2272060800 10 # 1 Jan 1972 2287785600 11 # 1 Jul 1972 @@ -252,4 +252,4 @@ # the hash line is also ignored in the # computation. # -#h f28827d2 f263b6c3 ec0f19eb a3e0dbf0 97f3fa30 +#h a1c168ae 27c79a7d 9dddcfc3 bcfe616b 2e2c44ea ___ 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: r361353 - stable/12/usr.sbin/ntp/ntpd
Author: delphij Date: Fri May 22 02:06:56 2020 New Revision: 361353 URL: https://svnweb.freebsd.org/changeset/base/361353 Log: MFC r361260: Update leap-seconds to leap-seconds.3676924800. Modified: stable/12/usr.sbin/ntp/ntpd/leap-seconds Directory Properties: stable/12/ (props changed) Modified: stable/12/usr.sbin/ntp/ntpd/leap-seconds == --- stable/12/usr.sbin/ntp/ntpd/leap-secondsFri May 22 01:18:55 2020 (r361352) +++ stable/12/usr.sbin/ntp/ntpd/leap-secondsFri May 22 02:06:56 2020 (r361353) @@ -204,10 +204,10 @@ # current -- the update time stamp, the data and the name of the file # will not change. # -# Updated through IERS Bulletin C58 -# File expires on: 28 June 2020 +# Updated through IERS Bulletin C59 +# File expires on: 28 December 2020 # -#@ 3802291200 +#@ 3818102400 # 2272060800 10 # 1 Jan 1972 2287785600 11 # 1 Jul 1972 @@ -252,4 +252,4 @@ # the hash line is also ignored in the # computation. # -#h f28827d2 f263b6c3 ec0f19eb a3e0dbf0 97f3fa30 +#h a1c168ae 27c79a7d 9dddcfc3 bcfe616b 2e2c44ea ___ 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: r361334 - in stable/12/sys: amd64/amd64 arm64/arm64 dev/acpica i386/i386 x86/acpica
On Fri, May 22, 2020 at 12:45:03AM +0200, Herbert J. Skuhra wrote: > On Thu, 21 May 2020 17:28:35 +0200, Mark Johnston wrote: > > > > Author: markj > > Date: Thu May 21 15:28:35 2020 > > New Revision: 361334 > > URL: https://svnweb.freebsd.org/changeset/base/361334 > > > > Log: > > MFC r361033: > > Call acpi_pxm_set_proximity_info() slightly earlier on x86. > > > > Modified: > > stable/12/sys/amd64/amd64/mp_machdep.c > > stable/12/sys/arm64/arm64/mp_machdep.c > > stable/12/sys/dev/acpica/acpi_pxm.c > > stable/12/sys/dev/acpica/acpivar.h > > stable/12/sys/i386/i386/mp_machdep.c > > stable/12/sys/x86/acpica/srat.c > > Directory Properties: > > stable/12/ (props changed) > > > > Modified: stable/12/sys/amd64/amd64/mp_machdep.c > > == > > --- stable/12/sys/amd64/amd64/mp_machdep.c Thu May 21 15:18:59 2020 > > (r361333) > > +++ stable/12/sys/amd64/amd64/mp_machdep.c Thu May 21 15:28:35 2020 > > (r361334) > > @@ -265,8 +265,9 @@ cpu_mp_start(void) > > init_ops.start_all_aps(); > > > > set_interrupt_apic_ids(); > > -} > > > > + acpi_pxm_set_cpu_locality(); > > +} > > > > /* > > * AP CPU's call this to initialize themselves. > > Until now it was possible to build a kernel (amd64) without 'device > acpi'. After this commit it fails with this error: > > --- kernel.full --- > linking kernel.full > ld: error: undefined symbol: acpi_pxm_set_cpu_locality > >>> referenced by mp_machdep.c:269 (/usr/src/sys/amd64/amd64/mp_machdep.c:269) > >>> mp_machdep.o:(cpu_mp_start) > *** [kernel.full] Error code 1 > > Was that intended? It was unintentional. I committed a fix to head in r361352 and will merge to stable/12 shortly. ___ 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: r361352 - in head/sys: amd64/amd64 i386/i386
Author: markj
Date: Fri May 22 01:18:55 2020
New Revision: 361352
URL: https://svnweb.freebsd.org/changeset/base/361352
Log:
Fix the build after r361033 when ACPI is disabled.
Reported by: Herbert J. Skuhra
Modified:
head/sys/amd64/amd64/mp_machdep.c
head/sys/i386/i386/mp_machdep.c
Modified: head/sys/amd64/amd64/mp_machdep.c
==
--- head/sys/amd64/amd64/mp_machdep.c Fri May 22 00:00:55 2020
(r361351)
+++ head/sys/amd64/amd64/mp_machdep.c Fri May 22 01:18:55 2020
(r361352)
@@ -29,6 +29,7 @@
#include
__FBSDID("$FreeBSD$");
+#include "opt_acpi.h"
#include "opt_cpu.h"
#include "opt_ddb.h"
#include "opt_kstack_pages.h"
@@ -78,8 +79,10 @@ __FBSDID("$FreeBSD$");
#include
#include
+#ifdef DEV_ACPI
#include
#include
+#endif
#define WARMBOOT_TARGET0
#define WARMBOOT_OFF (KERNBASE + 0x0467)
@@ -265,7 +268,9 @@ cpu_mp_start(void)
set_interrupt_apic_ids();
+#if defined(DEV_ACPI) && MAXMEMDOM > 1
acpi_pxm_set_cpu_locality();
+#endif
}
/*
Modified: head/sys/i386/i386/mp_machdep.c
==
--- head/sys/i386/i386/mp_machdep.c Fri May 22 00:00:55 2020
(r361351)
+++ head/sys/i386/i386/mp_machdep.c Fri May 22 01:18:55 2020
(r361352)
@@ -28,6 +28,7 @@
#include
__FBSDID("$FreeBSD$");
+#include "opt_acpi.h"
#include "opt_apic.h"
#include "opt_cpu.h"
#include "opt_kstack_pages.h"
@@ -83,8 +84,10 @@ __FBSDID("$FreeBSD$");
#include
#include
+#ifdef DEV_ACPI
#include
#include
+#endif
#define WARMBOOT_TARGET0
#define WARMBOOT_OFF (PMAP_MAP_LOW + 0x0467)
@@ -202,7 +205,7 @@ cpu_mp_start(void)
set_interrupt_apic_ids();
-#if MAXMEMDOM > 1
+#if defined(DEV_ACPI) && MAXMEMDOM > 1
acpi_pxm_set_cpu_locality();
#endif
}
___
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: r361351 - releng/11.4/sys/conf
Author: gjb
Date: Fri May 22 00:00:55 2020
New Revision: 361351
URL: https://svnweb.freebsd.org/changeset/base/361351
Log:
Update releng/11.4 to RC1 as part of the 11.4-RELEASE cycle.
Approved by: re (implicit)
Sponsored by: Rubicon Communications, LLC (netgate.com)
Modified:
releng/11.4/sys/conf/newvers.sh
Modified: releng/11.4/sys/conf/newvers.sh
==
--- releng/11.4/sys/conf/newvers.sh Thu May 21 22:47:39 2020
(r361350)
+++ releng/11.4/sys/conf/newvers.sh Fri May 22 00:00:55 2020
(r361351)
@@ -44,7 +44,7 @@
TYPE="FreeBSD"
REVISION="11.4"
-BRANCH="BETA2"
+BRANCH="RC1"
if [ -n "${BRANCH_OVERRIDE}" ]; then
BRANCH=${BRANCH_OVERRIDE}
fi
___
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: r361342 - in stable/12/sys/netinet: . tcp_stacks
> Author: rscheff
> Date: Thu May 21 19:46:11 2020
> New Revision: 361342
> URL: https://svnweb.freebsd.org/changeset/base/361342
>
> Log:
Partial merge of r360477:
> MFC r360477: Correctly set up the initial TCP congestion window in all cases
>
> by not including the SYN bit sequence space in cwnd related calculations.
> Snd_und is adjusted explicitly in all cases, outside the cwnd update,
> instead.
>
> This fixes an off-by-one conformance issue with regular TCP sessions
> not using Appropriate Byte Counting (RFC3465), sending one more
> packet during the initial window than expected.
There should of been a note here:
This does NOT contain the merge of the change to bbrv1 since at this
time that code does not exist in stable/12, and there is no plan to
merge that code to stable/12.
>
> PR: 235256
> Reviewed by:tuexen (mentor), rgrimes (mentor, blanket)
> Approved by:tuexen (mentor), rgrimes (mentor, blanket)
> MFC after: 3 weeks
> Sponsored by: NetApp, Inc.
> Differential Revision: https://reviews.freebsd.org/D19000
>
> Modified:
> stable/12/sys/netinet/tcp_input.c
> stable/12/sys/netinet/tcp_stacks/rack.c
> Directory Properties:
> stable/12/ (props changed)
>
> Modified: stable/12/sys/netinet/tcp_input.c
> ==
> --- stable/12/sys/netinet/tcp_input.c Thu May 21 19:45:14 2020
> (r361341)
> +++ stable/12/sys/netinet/tcp_input.c Thu May 21 19:46:11 2020
> (r361342)
> @@ -1519,7 +1519,7 @@ tcp_do_segment(struct mbuf *m, struct tcphdr *th, stru
> struct tcpcb *tp, int drop_hdrlen, int tlen, uint8_t iptos)
> {
> int thflags, acked, ourfinisacked, needoutput = 0, sack_changed;
> - int rstreason, todrop, win;
> + int rstreason, todrop, win, incforsyn = 0;
> uint32_t tiwin;
> uint16_t nsegs;
> char *s;
> @@ -2432,12 +2432,6 @@ tcp_do_segment(struct mbuf *m, struct tcphdr *th, stru
> if (IS_FASTOPEN(tp->t_flags) && tp->t_tfo_pending) {
> tcp_fastopen_decrement_counter(tp->t_tfo_pending);
> tp->t_tfo_pending = NULL;
> -
> - /*
> - * Account for the ACK of our SYN prior to
> - * regular ACK processing below.
> - */
> - tp->snd_una++;
> }
> if (tp->t_flags & TF_NEEDFIN) {
> tcp_state_change(tp, TCPS_FIN_WAIT_1);
> @@ -2458,6 +2452,13 @@ tcp_do_segment(struct mbuf *m, struct tcphdr *th, stru
> tcp_timer_activate(tp, TT_KEEP, TP_KEEPIDLE(tp));
> }
> /*
> + * Account for the ACK of our SYN prior to
> + * regular ACK processing below, except for
> + * simultaneous SYN, which is handled later.
> + */
> + if (SEQ_GT(th->th_ack, tp->snd_una) && !(tp->t_flags &
> TF_NEEDSYN))
> + incforsyn = 1;
> + /*
>* If segment contains data or ACK, will call tcp_reass()
>* later; if not, do so now to pass queued data to user.
>*/
> @@ -2751,6 +2752,15 @@ tcp_do_segment(struct mbuf *m, struct tcphdr *th, stru
> process_ACK:
> INP_WLOCK_ASSERT(tp->t_inpcb);
>
> + /*
> + * Adjust for the SYN bit in sequence space,
> + * but don't account for it in cwnd calculations.
> + * This is for the SYN_RECEIVED, non-simultaneous
> + * SYN case. SYN_SENT and simultaneous SYN are
> + * treated elsewhere.
> + */
> + if (incforsyn)
> + tp->snd_una++;
> acked = BYTES_THIS_ACK(tp, th);
> KASSERT(acked >= 0, ("%s: acked unexepectedly negative "
> "(tp->snd_una=%u, th->th_ack=%u, tp=%p, m=%p)", __func__,
>
> Modified: stable/12/sys/netinet/tcp_stacks/rack.c
> ==
> --- stable/12/sys/netinet/tcp_stacks/rack.c Thu May 21 19:45:14 2020
> (r361341)
> +++ stable/12/sys/netinet/tcp_stacks/rack.c Thu May 21 19:46:11 2020
> (r361342)
> @@ -5580,12 +5580,6 @@ rack_do_syn_recv(struct mbuf *m, struct tcphdr *th, st
> if (IS_FASTOPEN(tp->t_flags) && tp->t_tfo_pending) {
> tcp_fastopen_decrement_counter(tp->t_tfo_pending);
> tp->t_tfo_pending = NULL;
> -
> - /*
> - * Account for the ACK of our SYN prior to
> - * regular ACK processing below.
> - */
> - tp->snd_una++;
> }
> if (tp->t_flags & TF_NEEDFIN) {
> tcp_state_change(tp, TCPS_FIN_WAIT_1);
> @@ -5603,6 +5597,13 @@ rack_do_syn_recv(struct mbuf *m, struct tcphdr *th, st
> if
Re: svn commit: r361340 - in stable/12/sys/netinet: . tcp_stacks
> Author: rscheff
> Date: Thu May 21 19:41:25 2020
> New Revision: 361340
> URL: https://svnweb.freebsd.org/changeset/base/361340
>
> Log:
> MFC r360479: Prevent premature shrinking of the scaled receive window
>
> which can cause a TCP client to use invalid or stale TCP sequence numbers
> for ACK packets.
>
> Packets with old sequence numbers are ignored and not used to update the
> send window size.
> This might cause the TCP session to hang indefinitely under some
> circumstances.
There should of been a note here:
This does NOT contain the merge of the change to bbrv1 since at this
time that code does not exist in stable/12, and there is no plan to
merge that code to stable/12.
> Reported by:Cui Cheng
> Reviewed by:tuexen (mentor), rgrimes (mentor, blanket)
> Approved by:tuexen (mentor), rgrimes (mentor, blanket)
> MFC after: 3 weeks
> Sponsored by: NetApp, Inc.
> Differential Revision: https://reviews.freebsd.org/D24515
>
> Modified:
> stable/12/sys/netinet/tcp_output.c
> stable/12/sys/netinet/tcp_stacks/rack.c
> Directory Properties:
> stable/12/ (props changed)
>
> Modified: stable/12/sys/netinet/tcp_output.c
> ==
> --- stable/12/sys/netinet/tcp_output.cThu May 21 18:50:05 2020
> (r361339)
> +++ stable/12/sys/netinet/tcp_output.cThu May 21 19:41:25 2020
> (r361340)
> @@ -1206,8 +1206,11 @@ send:
> if (flags & TH_SYN)
> th->th_win = htons((u_short)
> (min(sbspace(>so_rcv), TCP_MAXWIN)));
> - else
> + else {
> + /* Avoid shrinking window with window scaling. */
> + recwin = roundup2(recwin, 1 << tp->rcv_scale);
> th->th_win = htons((u_short)(recwin >> tp->rcv_scale));
> + }
>
> /*
>* Adjust the RXWIN0SENT flag - indicate that we have advertised
>
> Modified: stable/12/sys/netinet/tcp_stacks/rack.c
> ==
> --- stable/12/sys/netinet/tcp_stacks/rack.c Thu May 21 18:50:05 2020
> (r361339)
> +++ stable/12/sys/netinet/tcp_stacks/rack.c Thu May 21 19:41:25 2020
> (r361340)
> @@ -8355,8 +8355,11 @@ send:
> if (flags & TH_SYN)
> th->th_win = htons((u_short)
> (min(sbspace(>so_rcv), TCP_MAXWIN)));
> - else
> + else {
> + /* Avoid shrinking window with window scaling. */
> + recwin = roundup2(recwin, 1 << tp->rcv_scale);
> th->th_win = htons((u_short)(recwin >> tp->rcv_scale));
> + }
> /*
>* Adjust the RXWIN0SENT flag - indicate that we have advertised a 0
>* window. This may cause the remote transmitter to stall. This
>
--
Rod Grimes rgri...@freebsd.org
___
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: r361350 - in stable/12: contrib/netbsd-tests/lib/libc/sys lib/libc/sys lib/libc/tests/sys sys/amd64/vmm sys/sys sys/vm usr.bin/vmstat
Author: markj
Date: Thu May 21 22:47:39 2020
New Revision: 361350
URL: https://svnweb.freebsd.org/changeset/base/361350
Log:
MFC r347532, r347541
Provide separate accounting for user-wired pages.
Modified:
stable/12/contrib/netbsd-tests/lib/libc/sys/t_mlock.c
stable/12/lib/libc/sys/mlock.2
stable/12/lib/libc/sys/mlockall.2
stable/12/lib/libc/tests/sys/mlock_helper.c
stable/12/sys/amd64/vmm/vmm.c
stable/12/sys/sys/param.h
stable/12/sys/sys/vmmeter.h
stable/12/sys/vm/vm_glue.c
stable/12/sys/vm/vm_map.c
stable/12/sys/vm/vm_map.h
stable/12/sys/vm/vm_meter.c
stable/12/sys/vm/vm_mmap.c
stable/12/sys/vm/vm_pageout.c
stable/12/sys/vm/vm_pageout.h
stable/12/sys/vm/vm_unix.c
stable/12/usr.bin/vmstat/vmstat.c
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/contrib/netbsd-tests/lib/libc/sys/t_mlock.c
==
--- stable/12/contrib/netbsd-tests/lib/libc/sys/t_mlock.c Thu May 21
22:24:23 2020(r361349)
+++ stable/12/contrib/netbsd-tests/lib/libc/sys/t_mlock.c Thu May 21
22:47:39 2020(r361350)
@@ -51,7 +51,7 @@ __RCSID("$NetBSD: t_mlock.c,v 1.6 2016/08/09 12:02:44
#define _KMEMUSER
#include
-void set_vm_max_wired(int);
+void set_vm_max_wired(u_long);
void restore_vm_max_wired(void);
#endif
Modified: stable/12/lib/libc/sys/mlock.2
==
--- stable/12/lib/libc/sys/mlock.2 Thu May 21 22:24:23 2020
(r361349)
+++ stable/12/lib/libc/sys/mlock.2 Thu May 21 22:47:39 2020
(r361350)
@@ -28,7 +28,7 @@
.\"@(#)mlock.2 8.2 (Berkeley) 12/11/93
.\" $FreeBSD$
.\"
-.Dd March 20, 2018
+.Dd May 13, 2019
.Dt MLOCK 2
.Os
.Sh NAME
@@ -97,13 +97,13 @@ resource limit and the
system-wide
.Dq wired pages
limit
-.Va vm.max_wired .
-.Va vm.max_wired
+.Va vm.max_user_wired .
+.Va vm.max_user_wired
applies to the system as a whole, so the amount available to a single
process at any given time is the difference between
-.Va vm.max_wired
+.Va vm.max_user_wired
and
-.Va vm.stats.vm.v_wire_count .
+.Va vm.stats.vm.v_user_wire_count .
.Pp
If
.Va security.bsd.unprivileged_mlock
@@ -124,13 +124,11 @@ will fail if:
is set to 0 and the caller is not the super-user.
.It Bq Er EINVAL
The address range given wraps around zero.
-.It Bq Er EAGAIN
-Locking the indicated range would exceed the system limit for locked memory.
.It Bq Er ENOMEM
Some portion of the indicated address range is not allocated.
There was an error faulting/mapping a page.
-Locking the indicated range would exceed the per-process limit for locked
-memory.
+Locking the indicated range would exceed the per-process or system-wide limits
+for locked memory.
.El
The
.Fn munlock
@@ -171,11 +169,11 @@ system calls first appeared in
Allocating too much wired memory can lead to a memory-allocation deadlock
which requires a reboot to recover from.
.Pp
-The per-process resource limit is a limit on the amount of virtual
-memory locked, while the system-wide limit is for the number of locked
-physical pages.
-Hence a process with two distinct locked mappings of the same physical page
-counts as 2 pages against the per-process limit and as only a single page
-in the system limit.
+The per-process and system-wide resource limits of locked memory apply
+to the amount of virtual memory locked, not the amount of locked physical
+pages.
+Hence two distinct locked mappings of the same physical page counts as
+2 pages aginst the system limit, and also against the per-process limit
+if both mappings belong to the same physical map.
.Pp
The per-process resource limit is not currently supported.
Modified: stable/12/lib/libc/sys/mlockall.2
==
--- stable/12/lib/libc/sys/mlockall.2 Thu May 21 22:24:23 2020
(r361349)
+++ stable/12/lib/libc/sys/mlockall.2 Thu May 21 22:47:39 2020
(r361350)
@@ -30,7 +30,7 @@
.\"
.\" $FreeBSD$
.\"
-.Dd December 25, 2012
+.Dd May 13, 2019
.Dt MLOCKALL 2
.Os
.Sh NAME
@@ -69,7 +69,7 @@ limited in how much they can lock down.
A single process can lock the minimum of a system-wide
.Dq wired pages
limit
-.Va vm.max_wired
+.Va vm.max_user_wired
and the per-process
.Dv RLIMIT_MEMLOCK
resource limit.
@@ -138,9 +138,9 @@ and
functions first appeared in
.Fx 5.1 .
.Sh BUGS
-The per-process resource limit is a limit on the amount of virtual
-memory locked, while the system-wide limit is for the number of locked
-physical pages.
-Hence a process with two distinct locked mappings of the same physical page
-counts as 2 pages against the per-process limit and as only a single page
-in the system limit.
+The per-process and system-wide resource limits of locked memory apply
+to the amount of virtual memory locked, not the amount of locked physical
+pages.
+Hence two distinct locked mappings of
Re: svn commit: r361334 - in stable/12/sys: amd64/amd64 arm64/arm64 dev/acpica i386/i386 x86/acpica
On Thu, 21 May 2020 17:28:35 +0200, Mark Johnston wrote: > > Author: markj > Date: Thu May 21 15:28:35 2020 > New Revision: 361334 > URL: https://svnweb.freebsd.org/changeset/base/361334 > > Log: > MFC r361033: > Call acpi_pxm_set_proximity_info() slightly earlier on x86. > > Modified: > stable/12/sys/amd64/amd64/mp_machdep.c > stable/12/sys/arm64/arm64/mp_machdep.c > stable/12/sys/dev/acpica/acpi_pxm.c > stable/12/sys/dev/acpica/acpivar.h > stable/12/sys/i386/i386/mp_machdep.c > stable/12/sys/x86/acpica/srat.c > Directory Properties: > stable/12/ (props changed) > > Modified: stable/12/sys/amd64/amd64/mp_machdep.c > == > --- stable/12/sys/amd64/amd64/mp_machdep.cThu May 21 15:18:59 2020 > (r361333) > +++ stable/12/sys/amd64/amd64/mp_machdep.cThu May 21 15:28:35 2020 > (r361334) > @@ -265,8 +265,9 @@ cpu_mp_start(void) > init_ops.start_all_aps(); > > set_interrupt_apic_ids(); > -} > > + acpi_pxm_set_cpu_locality(); > +} > > /* > * AP CPU's call this to initialize themselves. Until now it was possible to build a kernel (amd64) without 'device acpi'. After this commit it fails with this error: --- kernel.full --- linking kernel.full ld: error: undefined symbol: acpi_pxm_set_cpu_locality >>> referenced by mp_machdep.c:269 (/usr/src/sys/amd64/amd64/mp_machdep.c:269) >>> mp_machdep.o:(cpu_mp_start) *** [kernel.full] Error code 1 Was that intended? -- Herbert ___ 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: r361349 - in head: cddl/contrib/opensolaris/lib/libdtrace/common lib/libc/gen libexec/rtld-elf sys/sys
Author: kib
Date: Thu May 21 22:24:23 2020
New Revision: 361349
URL: https://svnweb.freebsd.org/changeset/base/361349
Log:
Restore the binary compatibility for link_map l_addr.
Keep link_map l_addr binary layout compatible, rename l_addr to l_base
where rtld returns map base. Provide relocbase in newly added l_addr.
This effectively reverts the patch to the initial version of D24918.
Reported by: antoine (portmgr)
Reviewed by: jhb, markj
Tested by:markj
Sponsored by: The FreeBSD Foundation
MFC after:1 week
Differential revision:https://reviews.freebsd.org/D24946
Modified:
head/cddl/contrib/opensolaris/lib/libdtrace/common/drti.c
head/lib/libc/gen/dlinfo.3
head/libexec/rtld-elf/rtld.c
head/sys/sys/link_elf.h
Modified: head/cddl/contrib/opensolaris/lib/libdtrace/common/drti.c
==
--- head/cddl/contrib/opensolaris/lib/libdtrace/common/drti.c Thu May 21
21:42:49 2020(r361348)
+++ head/cddl/contrib/opensolaris/lib/libdtrace/common/drti.c Thu May 21
22:24:23 2020(r361349)
@@ -143,12 +143,18 @@ dtrace_dof_init(void)
return;
}
+#ifdef __FreeBSD__
+ elf = (void *)lmp->l_base;
+#else
elf = (void *)lmp->l_addr;
+#endif
dh.dofhp_dof = (uintptr_t)dof;
- dh.dofhp_addr = elf->e_type == ET_DYN ? (uintptr_t) lmp->l_addr : 0;
#ifdef __FreeBSD__
+ dh.dofhp_addr = elf->e_type == ET_DYN ? (uintptr_t) lmp->l_base : 0;
dh.dofhp_pid = getpid();
+#else
+ dh.dofhp_addr = elf->e_type == ET_DYN ? (uintptr_t) lmp->l_addr : 0;
#endif
if (lmid == 0) {
Modified: head/lib/libc/gen/dlinfo.3
==
--- head/lib/libc/gen/dlinfo.3 Thu May 21 21:42:49 2020(r361348)
+++ head/lib/libc/gen/dlinfo.3 Thu May 21 22:24:23 2020(r361349)
@@ -105,17 +105,16 @@ structure is defined in
.In link.h
and has the following members:
.Bd -literal -offset indent
-caddr_t l_addr;/* Load Offset of library */
+caddr_t l_base;/* Base Address of library */
const char *l_name; /* Absolute Path to Library */
const void *l_ld; /* Pointer to .dynamic in memory */
struct link_map *l_next, /* linked list of mapped libs */
*l_prev;
+caddr_t l_addr; /* Load Offset of library */
.Ed
.Bl -tag -width ".Va l_addr"
-.It Va l_addr
-The load offset of the object, that is, the difference between
-the actual load address and the base virtual address the object
-was linked at.
+.It Va l_base
+The base address of the object loaded into memory.
.It Va l_name
The full name of the loaded shared object.
.It Va l_ld
@@ -130,6 +129,10 @@ structure on the link-map list.
The previous
.Vt Link_map
structure on the link-map list.
+.It Va l_addr
+The load offset of the object, that is, the difference between
+the actual load address and the base virtual address the object
+was linked at.
.El
.It Dv RTLD_DI_SERINFO
Retrieve the library search paths associated with the given
Modified: head/libexec/rtld-elf/rtld.c
==
--- head/libexec/rtld-elf/rtld.cThu May 21 21:42:49 2020
(r361348)
+++ head/libexec/rtld-elf/rtld.cThu May 21 22:24:23 2020
(r361349)
@@ -4032,8 +4032,9 @@ linkmap_add(Obj_Entry *obj)
struct link_map *prev;
obj->linkmap.l_name = obj->path;
-obj->linkmap.l_addr = obj->relocbase;
+obj->linkmap.l_base = obj->mapbase;
obj->linkmap.l_ld = obj->dynamic;
+obj->linkmap.l_addr = obj->relocbase;
if (r_debug.r_map == NULL) {
r_debug.r_map = l;
Modified: head/sys/sys/link_elf.h
==
--- head/sys/sys/link_elf.h Thu May 21 21:42:49 2020(r361348)
+++ head/sys/sys/link_elf.h Thu May 21 22:24:23 2020(r361349)
@@ -57,13 +57,14 @@
#defineLA_SER_SECURE 0x80/* default (secure) path prepended */
typedef struct link_map {
- caddr_t l_addr; /* Load Offset of library */
+ caddr_t l_base; /* Base Address of library */
#ifdef __mips__
caddr_t l_xxx; /* unused */
#endif
const char *l_name;/* Absolute Path to Library */
const void *l_ld; /* Pointer to .dynamic in
memory */
struct link_map *l_next, *l_prev; /* linked list of of mapped
libs */
+ caddr_t l_addr; /* Load Offset of library */
} Link_map;
struct r_debug {
___
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: r361348 - head/sys/netinet/cc
Author: rscheff Date: Thu May 21 21:42:49 2020 New Revision: 361348 URL: https://svnweb.freebsd.org/changeset/base/361348 Log: DCTCP: update alpha only once after loss recovery. In mixed ECN marking and loss scenarios it was found, that the alpha value of DCTCP is updated two times. The second update happens with freshly initialized counters indicating to ECN loss. Overall this leads to alpha not adjusting as quickly as expected to ECN markings, and therefore lead to excessive loss. Reported by: Cheng Cui Reviewed by: chengc_netapp.com, rrs, tuexen (mentor) Approved by: tuexen (mentor) MFC after:2 weeks Sponsored by: NetApp, Inc. Differential Revision:https://reviews.freebsd.org/D24817 Modified: head/sys/netinet/cc/cc_dctcp.c Modified: head/sys/netinet/cc/cc_dctcp.c == --- head/sys/netinet/cc/cc_dctcp.c Thu May 21 21:33:15 2020 (r361347) +++ head/sys/netinet/cc/cc_dctcp.c Thu May 21 21:42:49 2020 (r361348) @@ -154,10 +154,8 @@ dctcp_ack_received(struct cc_var *ccv, uint16_t type) * Update the fraction of marked bytes at the end of * current window size. */ - if ((IN_FASTRECOVERY(CCV(ccv, t_flags)) && - SEQ_GEQ(ccv->curack, CCV(ccv, snd_recover))) || - (!IN_FASTRECOVERY(CCV(ccv, t_flags)) && - SEQ_GT(ccv->curack, dctcp_data->save_sndnxt))) + if (!IN_FASTRECOVERY(CCV(ccv, t_flags)) && + SEQ_GT(ccv->curack, dctcp_data->save_sndnxt)) dctcp_update_alpha(ccv); } else newreno_cc_algo.ack_received(ccv, type); ___ 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: r361347 - in head/sys/netinet: . tcp_stacks
Author: rscheff
Date: Thu May 21 21:33:15 2020
New Revision: 361347
URL: https://svnweb.freebsd.org/changeset/base/361347
Log:
With RFC3168 ECN, CWR SHOULD only be sent with new data
Overly conservative data receivers may ignore the CWR flag
on other packets, and keep ECE latched. This can result in
continous reduction of the congestion window, and very poor
performance when ECN is enabled.
Reviewed by: rgrimes (mentor), rrs
Approved by: rgrimes (mentor), tuexen (mentor)
MFC after:3 days
Sponsored by: NetApp, Inc.
Differential Revision:https://reviews.freebsd.org/D23364
Modified:
head/sys/netinet/tcp_input.c
head/sys/netinet/tcp_output.c
head/sys/netinet/tcp_stacks/rack.c
Modified: head/sys/netinet/tcp_input.c
==
--- head/sys/netinet/tcp_input.cThu May 21 21:26:21 2020
(r361346)
+++ head/sys/netinet/tcp_input.cThu May 21 21:33:15 2020
(r361347)
@@ -447,9 +447,15 @@ cc_cong_signal(struct tcpcb *tp, struct tcphdr *th, ui
}
break;
case CC_ECN:
- if (!IN_CONGRECOVERY(tp->t_flags)) {
+ if (!IN_CONGRECOVERY(tp->t_flags) ||
+ /*
+* Allow ECN reaction on ACK to CWR, if
+* that data segment was also CE marked.
+*/
+ SEQ_GEQ(th->th_ack, tp->snd_recover)) {
+ EXIT_CONGRECOVERY(tp->t_flags);
TCPSTAT_INC(tcps_ecn_rcwnd);
- tp->snd_recover = tp->snd_max;
+ tp->snd_recover = tp->snd_max + 1;
if (tp->t_flags2 & TF2_ECN_PERMIT)
tp->t_flags2 |= TF2_ECN_SND_CWR;
}
Modified: head/sys/netinet/tcp_output.c
==
--- head/sys/netinet/tcp_output.c Thu May 21 21:26:21 2020
(r361346)
+++ head/sys/netinet/tcp_output.c Thu May 21 21:33:15 2020
(r361347)
@@ -1170,7 +1170,8 @@ send:
*/
if (len > 0 && SEQ_GEQ(tp->snd_nxt, tp->snd_max) &&
(sack_rxmit == 0) &&
- !((tp->t_flags & TF_FORCEDATA) && len == 1)) {
+ !((tp->t_flags & TF_FORCEDATA) && len == 1 &&
+ SEQ_LT(tp->snd_una, tp->snd_max))) {
#ifdef INET6
if (isipv6)
ip6->ip6_flow |= htonl(IPTOS_ECN_ECT0 << 20);
@@ -1178,14 +1179,14 @@ send:
#endif
ip->ip_tos |= IPTOS_ECN_ECT0;
TCPSTAT_INC(tcps_ecn_ect0);
- }
-
- /*
-* Reply with proper ECN notifications.
-*/
- if (tp->t_flags2 & TF2_ECN_SND_CWR) {
- flags |= TH_CWR;
- tp->t_flags2 &= ~TF2_ECN_SND_CWR;
+ /*
+* Reply with proper ECN notifications.
+* Only set CWR on new data segments.
+*/
+ if (tp->t_flags2 & TF2_ECN_SND_CWR) {
+ flags |= TH_CWR;
+ tp->t_flags2 &= ~TF2_ECN_SND_CWR;
+ }
}
if (tp->t_flags2 & TF2_ECN_SND_ECE)
flags |= TH_ECE;
Modified: head/sys/netinet/tcp_stacks/rack.c
==
--- head/sys/netinet/tcp_stacks/rack.c Thu May 21 21:26:21 2020
(r361346)
+++ head/sys/netinet/tcp_stacks/rack.c Thu May 21 21:33:15 2020
(r361347)
@@ -4095,9 +4095,15 @@ rack_cong_signal(struct tcpcb *tp, struct tcphdr *th,
}
break;
case CC_ECN:
- if (!IN_CONGRECOVERY(tp->t_flags)) {
+ if (!IN_CONGRECOVERY(tp->t_flags) ||
+ /*
+* Allow ECN reaction on ACK to CWR, if
+* that data segment was also CE marked.
+*/
+ SEQ_GEQ(th->th_ack, tp->snd_recover)) {
+ EXIT_CONGRECOVERY(tp->t_flags);
KMOD_TCPSTAT_INC(tcps_ecn_rcwnd);
- tp->snd_recover = tp->snd_max;
+ tp->snd_recover = tp->snd_max + 1;
if (tp->t_flags2 & TF2_ECN_PERMIT)
tp->t_flags2 |= TF2_ECN_SND_CWR;
}
@@ -13556,13 +13562,14 @@ send:
#endif
ip->ip_tos |= IPTOS_ECN_ECT0;
KMOD_TCPSTAT_INC(tcps_ecn_ect0);
- }
- /*
-* Reply with proper ECN notifications.
-*/
- if (tp->t_flags2 & TF2_ECN_SND_CWR) {
-
svn commit: r361346 - in head/sys/netinet: . tcp_stacks
Author: rscheff
Date: Thu May 21 21:26:21 2020
New Revision: 361346
URL: https://svnweb.freebsd.org/changeset/base/361346
Log:
Retain only mutually supported TCP options after simultaneous SYN
When receiving a parallel SYN in SYN-SENT state, remove all the
options only we supported locally before sending the SYN,ACK.
This addresses a consistency issue on parallel opens.
Also, on such a parallel open, the stack could be coaxed into
running with timestamps enabled, even if administratively disabled.
Reviewed by: tuexen (mentor)
Approved by: tuexen (mentor)
MFC after:2 weeks
Sponsored by: NetApp, Inc.
Differential Revision:https://reviews.freebsd.org/D23371
Modified:
head/sys/netinet/tcp_input.c
head/sys/netinet/tcp_stacks/bbr.c
head/sys/netinet/tcp_stacks/rack.c
Modified: head/sys/netinet/tcp_input.c
==
--- head/sys/netinet/tcp_input.cThu May 21 21:15:25 2020
(r361345)
+++ head/sys/netinet/tcp_input.cThu May 21 21:26:21 2020
(r361346)
@@ -1623,17 +1623,20 @@ tcp_do_segment(struct mbuf *m, struct tcphdr *th, stru
(tp->t_flags & TF_REQ_SCALE)) {
tp->t_flags |= TF_RCVD_SCALE;
tp->snd_scale = to.to_wscale;
- }
+ } else
+ tp->t_flags &= ~TF_REQ_SCALE;
/*
* Initial send window. It will be updated with
* the next incoming segment to the scaled value.
*/
tp->snd_wnd = th->th_win;
- if (to.to_flags & TOF_TS) {
+ if ((to.to_flags & TOF_TS) &&
+ (tp->t_flags & TF_REQ_TSTMP)) {
tp->t_flags |= TF_RCVD_TSTMP;
tp->ts_recent = to.to_tsval;
tp->ts_recent_age = tcp_ts_getticks();
- }
+ } else
+ tp->t_flags &= ~TF_REQ_TSTMP;
if (to.to_flags & TOF_MSS)
tcp_mss(tp, to.to_mss);
if ((tp->t_flags & TF_SACK_PERMIT) &&
Modified: head/sys/netinet/tcp_stacks/bbr.c
==
--- head/sys/netinet/tcp_stacks/bbr.c Thu May 21 21:15:25 2020
(r361345)
+++ head/sys/netinet/tcp_stacks/bbr.c Thu May 21 21:26:21 2020
(r361346)
@@ -11595,17 +11595,20 @@ bbr_do_segment_nounlock(struct mbuf *m, struct tcphdr
(tp->t_flags & TF_REQ_SCALE)) {
tp->t_flags |= TF_RCVD_SCALE;
tp->snd_scale = to.to_wscale;
- }
+ } else
+ tp->t_flags &= ~TF_REQ_SCALE;
/*
* Initial send window. It will be updated with the
* next incoming segment to the scaled value.
*/
tp->snd_wnd = th->th_win;
- if (to.to_flags & TOF_TS) {
+ if ((to.to_flags & TOF_TS) &&
+ (tp->t_flags & TF_REQ_TSTMP)) {
tp->t_flags |= TF_RCVD_TSTMP;
tp->ts_recent = to.to_tsval;
tp->ts_recent_age =
tcp_tv_to_mssectick(>rc_tv);
- }
+ } else
+ tp->t_flags &= ~TF_REQ_TSTMP;
if (to.to_flags & TOF_MSS)
tcp_mss(tp, to.to_mss);
if ((tp->t_flags & TF_SACK_PERMIT) &&
Modified: head/sys/netinet/tcp_stacks/rack.c
==
--- head/sys/netinet/tcp_stacks/rack.c Thu May 21 21:15:25 2020
(r361345)
+++ head/sys/netinet/tcp_stacks/rack.c Thu May 21 21:26:21 2020
(r361346)
@@ -11082,17 +11082,20 @@ rack_do_segment_nounlock(struct mbuf *m, struct tcphdr
(tp->t_flags & TF_REQ_SCALE)) {
tp->t_flags |= TF_RCVD_SCALE;
tp->snd_scale = to.to_wscale;
- }
+ } else
+ tp->t_flags &= ~TF_REQ_SCALE;
/*
* Initial send window. It will be updated with the
* next incoming segment to the scaled value.
*/
tp->snd_wnd = th->th_win;
- if (to.to_flags & TOF_TS) {
+ if ((to.to_flags & TOF_TS) &&
+ (tp->t_flags & TF_REQ_TSTMP)) {
tp->t_flags |= TF_RCVD_TSTMP;
tp->ts_recent =
svn commit: r361345 - in head/sys/netinet: . tcp_stacks
Author: rscheff
Date: Thu May 21 21:15:25 2020
New Revision: 361345
URL: https://svnweb.freebsd.org/changeset/base/361345
Log:
Handle ECN handshake in simultaneous open
While testing simultaneous open TCP with ECN, found that
negotiation fails to arrive at the expected final state.
Reviewed by: tuexen (mentor)
Approved by: tuexen (mentor), rgrimes (mentor)
MFC after:2 weeks
Sponsored by: NetApp, Inc.
Differential Revision:https://reviews.freebsd.org/D23373
Modified:
head/sys/netinet/tcp_input.c
head/sys/netinet/tcp_output.c
head/sys/netinet/tcp_stacks/rack.c
Modified: head/sys/netinet/tcp_input.c
==
--- head/sys/netinet/tcp_input.cThu May 21 21:00:46 2020
(r361344)
+++ head/sys/netinet/tcp_input.cThu May 21 21:15:25 2020
(r361345)
@@ -1611,6 +1611,14 @@ tcp_do_segment(struct mbuf *m, struct tcphdr *th, stru
* XXX this is traditional behavior, may need to be cleaned up.
*/
if (tp->t_state == TCPS_SYN_SENT && (thflags & TH_SYN)) {
+ /* Handle parallel SYN for ECN */
+ if (!(thflags & TH_ACK) &&
+ ((thflags & (TH_CWR | TH_ECE)) == (TH_CWR | TH_ECE)) &&
+ ((V_tcp_do_ecn == 1) || (V_tcp_do_ecn == 2))) {
+ tp->t_flags2 |= TF2_ECN_PERMIT;
+ tp->t_flags2 |= TF2_ECN_SND_ECE;
+ TCPSTAT_INC(tcps_ecn_shs);
+ }
if ((to.to_flags & TOF_SCALE) &&
(tp->t_flags & TF_REQ_SCALE)) {
tp->t_flags |= TF_RCVD_SCALE;
Modified: head/sys/netinet/tcp_output.c
==
--- head/sys/netinet/tcp_output.c Thu May 21 21:00:46 2020
(r361344)
+++ head/sys/netinet/tcp_output.c Thu May 21 21:15:25 2020
(r361345)
@@ -1154,6 +1154,12 @@ send:
} else
flags |= TH_ECE|TH_CWR;
}
+ /* Handle parallel SYN for ECN */
+ if ((tp->t_state == TCPS_SYN_RECEIVED) &&
+ (tp->t_flags2 & TF2_ECN_SND_ECE)) {
+ flags |= TH_ECE;
+ tp->t_flags2 &= ~TF2_ECN_SND_ECE;
+ }
if (tp->t_state == TCPS_ESTABLISHED &&
(tp->t_flags2 & TF2_ECN_PERMIT)) {
Modified: head/sys/netinet/tcp_stacks/rack.c
==
--- head/sys/netinet/tcp_stacks/rack.c Thu May 21 21:00:46 2020
(r361344)
+++ head/sys/netinet/tcp_stacks/rack.c Thu May 21 21:15:25 2020
(r361345)
@@ -11070,6 +11070,14 @@ rack_do_segment_nounlock(struct mbuf *m, struct tcphdr
* this is traditional behavior, may need to be cleaned up.
*/
if (tp->t_state == TCPS_SYN_SENT && (thflags & TH_SYN)) {
+ /* Handle parallel SYN for ECN */
+ if (!(thflags & TH_ACK) &&
+ ((thflags & (TH_CWR | TH_ECE)) == (TH_CWR |
TH_ECE)) &&
+ ((V_tcp_do_ecn == 1) || (V_tcp_do_ecn == 2))) {
+ tp->t_flags2 |= TF2_ECN_PERMIT;
+ tp->t_flags2 |= TF2_ECN_SND_ECE;
+ TCPSTAT_INC(tcps_ecn_shs);
+ }
if ((to.to_flags & TOF_SCALE) &&
(tp->t_flags & TF_REQ_SCALE)) {
tp->t_flags |= TF_RCVD_SCALE;
@@ -13522,6 +13530,12 @@ send:
flags |= TH_ECE | TH_CWR;
} else
flags |= TH_ECE | TH_CWR;
+ }
+ /* Handle parallel SYN for ECN */
+ if ((tp->t_state == TCPS_SYN_RECEIVED) &&
+ (tp->t_flags2 & TF2_ECN_SND_ECE)) {
+ flags |= TH_ECE;
+ tp->t_flags2 &= ~TF2_ECN_SND_ECE;
}
if (tp->t_state == TCPS_ESTABLISHED &&
(tp->t_flags2 & TF2_ECN_PERMIT)) {
___
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: r361344 - in head: contrib/unbound contrib/unbound/.github contrib/unbound/cachedb contrib/unbound/compat contrib/unbound/contrib contrib/unbound/daemon contrib/unbound/doc contrib/unbo...
Author: cy Date: Thu May 21 21:00:46 2020 New Revision: 361344 URL: https://svnweb.freebsd.org/changeset/base/361344 Log: MFV r361322: Update unbound 1.9.6 --> 1.10.1. Bug Fixes: - CVE-2020-12662 Unbound can be tricked into amplifying an incoming query into a large number of queries directed to a target. - CVE-2020-12663 Malformed answers from upstream name servers can be used to make Unbound unresponsive. Reported by: emaste MFC after:3 days Relnotes: yes Security: CVE-2020-12662, CVE-2020-12663 Added: head/contrib/unbound/.github/ - copied from r361322, vendor/unbound/dist/.github/ head/contrib/unbound/contrib/drop2rpz - copied unchanged from r361322, vendor/unbound/dist/contrib/drop2rpz head/contrib/unbound/contrib/unbound_portable.service.in - copied unchanged from r361322, vendor/unbound/dist/contrib/unbound_portable.service.in head/contrib/unbound/contrib/unbound_smf23.tar.gz - copied unchanged from r361322, vendor/unbound/dist/contrib/unbound_smf23.tar.gz head/contrib/unbound/services/rpz.c - copied unchanged from r361322, vendor/unbound/dist/services/rpz.c head/contrib/unbound/services/rpz.h - copied unchanged from r361322, vendor/unbound/dist/services/rpz.h Deleted: head/contrib/unbound/contrib/unbound_smf22.tar.gz Modified: head/contrib/unbound/Makefile.in head/contrib/unbound/README.md head/contrib/unbound/aclocal.m4 head/contrib/unbound/cachedb/cachedb.c head/contrib/unbound/compat/getentropy_solaris.c head/contrib/unbound/config.guess head/contrib/unbound/config.h.in head/contrib/unbound/config.sub head/contrib/unbound/configure head/contrib/unbound/configure.ac head/contrib/unbound/contrib/README head/contrib/unbound/contrib/fastrpz.patch head/contrib/unbound/contrib/libunbound.pc.in head/contrib/unbound/contrib/unbound.service.in head/contrib/unbound/contrib/unbound_munin_ head/contrib/unbound/daemon/daemon.c head/contrib/unbound/daemon/daemon.h head/contrib/unbound/daemon/remote.c head/contrib/unbound/daemon/stats.c head/contrib/unbound/daemon/unbound.c head/contrib/unbound/daemon/worker.c head/contrib/unbound/doc/Changelog head/contrib/unbound/doc/README head/contrib/unbound/doc/example.conf.in head/contrib/unbound/doc/libunbound.3.in head/contrib/unbound/doc/unbound-anchor.8.in head/contrib/unbound/doc/unbound-checkconf.8.in head/contrib/unbound/doc/unbound-control.8.in head/contrib/unbound/doc/unbound-host.1.in head/contrib/unbound/doc/unbound.8.in head/contrib/unbound/doc/unbound.conf.5.in head/contrib/unbound/edns-subnet/subnetmod.c head/contrib/unbound/install-sh head/contrib/unbound/iterator/iter_delegpt.c head/contrib/unbound/iterator/iter_delegpt.h head/contrib/unbound/iterator/iter_scrub.c head/contrib/unbound/iterator/iter_utils.c head/contrib/unbound/iterator/iterator.c head/contrib/unbound/iterator/iterator.h head/contrib/unbound/libunbound/context.c head/contrib/unbound/libunbound/libworker.c head/contrib/unbound/libunbound/unbound.h head/contrib/unbound/respip/respip.c head/contrib/unbound/respip/respip.h head/contrib/unbound/services/authzone.c head/contrib/unbound/services/authzone.h head/contrib/unbound/services/cache/dns.c head/contrib/unbound/services/cache/dns.h head/contrib/unbound/services/localzone.c head/contrib/unbound/services/localzone.h head/contrib/unbound/services/mesh.c head/contrib/unbound/services/mesh.h head/contrib/unbound/services/outside_network.c head/contrib/unbound/services/view.c head/contrib/unbound/sldns/parse.c head/contrib/unbound/sldns/str2wire.c head/contrib/unbound/smallapp/unbound-checkconf.c head/contrib/unbound/smallapp/unbound-control.c head/contrib/unbound/util/config_file.c head/contrib/unbound/util/config_file.h head/contrib/unbound/util/configlexer.lex head/contrib/unbound/util/configparser.y head/contrib/unbound/util/data/dname.c head/contrib/unbound/util/data/dname.h head/contrib/unbound/util/data/msgencode.c head/contrib/unbound/util/data/msgparse.c head/contrib/unbound/util/data/msgparse.h head/contrib/unbound/util/data/msgreply.c head/contrib/unbound/util/data/packed_rrset.c head/contrib/unbound/util/data/packed_rrset.h head/contrib/unbound/util/fptr_wlist.c head/contrib/unbound/util/fptr_wlist.h head/contrib/unbound/util/iana_ports.inc head/contrib/unbound/util/log.c head/contrib/unbound/util/log.h head/contrib/unbound/util/module.h head/contrib/unbound/util/net_help.c head/contrib/unbound/util/net_help.h head/contrib/unbound/util/netevent.c head/contrib/unbound/util/random.c head/contrib/unbound/util/storage/dnstree.c head/contrib/unbound/util/storage/dnstree.h head/contrib/unbound/validator/val_secalgo.c head/contrib/unbound/validator/validator.c head/lib/libunbound/Makefile Directory Properties: head/contrib/unbound/ (props changed) Modified:
svn commit: r361343 - in head/sys/compat/linuxkpi/common: include/linux src
Author: manu
Date: Thu May 21 20:18:38 2020
New Revision: 361343
URL: https://svnweb.freebsd.org/changeset/base/361343
Log:
linuxkpi: Add rcu_work functions
The rcu_work function helps to queue some work after waiting for a grace
period.
This is needed by DRM drivers.
Sponsored-by: The FreeBSD Foundation
Reviewed by: hselasky
Differential Revision:https://reviews.freebsd.org/D24942
Modified:
head/sys/compat/linuxkpi/common/include/linux/workqueue.h
head/sys/compat/linuxkpi/common/src/linux_work.c
Modified: head/sys/compat/linuxkpi/common/include/linux/workqueue.h
==
--- head/sys/compat/linuxkpi/common/include/linux/workqueue.h Thu May 21
19:46:11 2020(r361342)
+++ head/sys/compat/linuxkpi/common/include/linux/workqueue.h Thu May 21
20:18:38 2020(r361343)
@@ -72,6 +72,13 @@ struct work_struct {
atomic_t state;
};
+struct rcu_work {
+ struct work_struct work;
+ struct rcu_head rcu;
+
+ struct workqueue_struct *wq;
+};
+
#defineDECLARE_WORK(name, fn)
\
struct work_struct name;\
static void name##_init(void *arg) \
@@ -112,6 +119,9 @@ do {
\
TASK_INIT(&(work)->work_task, 0, linux_work_fn, (work));\
} while (0)
+#define INIT_RCU_WORK(_work, _fn) \
+ INIT_WORK(&(_work)->work, (_fn))
+
#defineINIT_WORK_ONSTACK(work, fn) \
INIT_WORK(work, fn)
@@ -192,6 +202,12 @@ do {
\
#defineflush_work(work) \
linux_flush_work(work)
+#definequeue_rcu_work(wq, rwork) \
+ linux_queue_rcu_work(wq, rwork)
+
+#defineflush_rcu_work(rwork) \
+ linux_flush_rcu_work(rwork)
+
#defineflush_delayed_work(dwork) \
linux_flush_delayed_work(dwork)
@@ -237,5 +253,7 @@ extern bool linux_flush_delayed_work(struct delayed_wo
extern bool linux_work_pending(struct work_struct *);
extern bool linux_work_busy(struct work_struct *);
extern struct work_struct *linux_current_work(void);
+extern bool linux_queue_rcu_work(struct workqueue_struct *wq, struct rcu_work
*rwork);
+extern bool linux_flush_rcu_work(struct rcu_work *rwork);
#endif /* _LINUX_WORKQUEUE_H_ */
Modified: head/sys/compat/linuxkpi/common/src/linux_work.c
==
--- head/sys/compat/linuxkpi/common/src/linux_work.cThu May 21 19:46:11
2020(r361342)
+++ head/sys/compat/linuxkpi/common/src/linux_work.cThu May 21 20:18:38
2020(r361343)
@@ -31,6 +31,7 @@ __FBSDID("$FreeBSD$");
#include
#include
#include
+#include
#include
@@ -153,6 +154,53 @@ linux_queue_work_on(int cpu __unused, struct workqueue
default:
return (false); /* already on a queue */
}
+}
+
+/*
+ * Callback func for linux_queue_rcu_work
+ */
+static void
+rcu_work_func(struct rcu_head *rcu)
+{
+ struct rcu_work *rwork;
+
+ rwork = container_of(rcu, struct rcu_work, rcu);
+ linux_queue_work_on(WORK_CPU_UNBOUND, rwork->wq, >work);
+}
+
+/*
+ * This function queue a work after a grace period
+ * If the work was already pending it returns false,
+ * if not it calls call_rcu and returns true.
+ */
+bool
+linux_queue_rcu_work(struct workqueue_struct *wq, struct rcu_work *rwork)
+{
+
+ if (!linux_work_pending(>work)) {
+ rwork->wq = wq;
+ linux_call_rcu(RCU_TYPE_REGULAR, >rcu, rcu_work_func);
+ return (true);
+ }
+ return (false);
+}
+
+/*
+ * This function waits for the last execution of a work and then
+ * flush the work.
+ * It returns true if the work was pending and we waited, it returns
+ * false otherwise.
+ */
+bool
+linux_flush_rcu_work(struct rcu_work *rwork)
+{
+
+ if (linux_work_pending(>work)) {
+ linux_rcu_barrier(RCU_TYPE_REGULAR);
+ linux_flush_work(>work);
+ return (true);
+ }
+ return (linux_flush_work(>work));
}
/*
___
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: r361303 - in head: lib/libc/gen libexec/rtld-elf sys/sys
On Thu, May 21, 2020 at 12:30:47PM -0700, John Baldwin wrote: > On 5/21/20 9:56 AM, Konstantin Belousov wrote: > > On Thu, May 21, 2020 at 09:03:44AM -0700, John Baldwin wrote: > >> On 5/21/20 8:12 AM, Mark Johnston wrote: > >>> On Thu, May 21, 2020 at 04:41:52PM +0300, Konstantin Belousov wrote: > On Thu, May 21, 2020 at 03:02:07PM +0200, Antoine Brodin wrote: > > On Thu, May 21, 2020 at 12:08 AM Konstantin Belousov > > wrote: > >> > >> Author: kib > >> Date: Wed May 20 22:08:26 2020 > >> New Revision: 361303 > >> URL: https://svnweb.freebsd.org/changeset/base/361303 > >> > >> Log: > >> Change the samantic of struct link_map l_addr member. > >> > >> It previously returned the object map base address, while all other > >> ELF operating systems return load offset, i.e. the difference between > >> map base and the link base. > >> > >> Explain the meaning of the field in the man page. > >> > >> Stop filling the mips-only l_offs member, which is apparently unused. > >> > >> PR: 246561 > >> Requested by: Damjan Jovanovic > >> Reviewed by: emaste, jhb, cem (previous version) > >> Sponsored by: The FreeBSD Foundation > >> MFC after:1 week > >> Differential revision:https://reviews.freebsd.org/D24918 > >> > >> Modified: > >> head/lib/libc/gen/dlinfo.3 > >> head/libexec/rtld-elf/rtld.c > >> head/sys/sys/link_elf.h > > > > Hi, > > > > After this commit, some ports fail to build with signal 11. > > For instance lang/perl5.30 fails to build with default options (DTRACE > > on) > > Disabling the DTRACE option makes it able to build again. > > > I see, thank you for reporting. > > So drti.c:dtrace_dof_init() does read l_addr, and the dtrace code assumes > that l_addr is the base, not relocbase. > > Mark, was dofhp_addr initialization changed comparing to Solaris ? > >>> > >>> It appears it has been the same since DTrace was imported. illumos > >>> still has similar code. > >>> > >>> Note that drti.o is linked into any executable and shlib that defines > >>> static probes, so the ABI change affects more than just dtrace(1). > >>> Would it be possible to define a new value for RTLD_DI_LINKMAP, and > >>> preserve the old behaviour for the old value? > >> > >> I think a bigger question is if Solaris/illumos treat l_addr as mapbase > >> (absolute address) or relocbase (relative address). In the discussion > >> in the phabricator I had assumed that all other OS's treated l_addr as > >> the relative offset (relocbase). Does the code for illumos assume an > >> absolute address or does it assume a relative address in l_addr? > > > > It is rather clear, since the dtrace code was pristine, that Solaris > > provides the mapbase. I do not have Solaris/Illumos box anymore > > (for quite some time), so I cannot check directly. > > > > My current PoV is that l_addr semantic must be restored, and relocbase > > provided by newly added member. > > I am fine with reverting the l_addr semantic. I'm still not sure how to > resolve the original PR, though perhaps Wine just has to carry a local > patch forever? GDB will work via the current accident so long as we > never pre-link libraries. As long as PIE binaries have a starting VA of > 0 like our shared libraries then I think GDB will be ok with our PIE > binaries as well. Wine should work without patch now, and after the D24918 is applied, too. I do not intend to revert l_addr to the 'load address' semantic. I am actually trying to find a solaris box to compile the test program. ___ 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: r361342 - in stable/12/sys/netinet: . tcp_stacks
Author: rscheff
Date: Thu May 21 19:46:11 2020
New Revision: 361342
URL: https://svnweb.freebsd.org/changeset/base/361342
Log:
MFC r360477: Correctly set up the initial TCP congestion window in all cases
by not including the SYN bit sequence space in cwnd related calculations.
Snd_und is adjusted explicitly in all cases, outside the cwnd update, instead.
This fixes an off-by-one conformance issue with regular TCP sessions
not using Appropriate Byte Counting (RFC3465), sending one more
packet during the initial window than expected.
PR: 235256
Reviewed by: tuexen (mentor), rgrimes (mentor, blanket)
Approved by: tuexen (mentor), rgrimes (mentor, blanket)
MFC after:3 weeks
Sponsored by: NetApp, Inc.
Differential Revision:https://reviews.freebsd.org/D19000
Modified:
stable/12/sys/netinet/tcp_input.c
stable/12/sys/netinet/tcp_stacks/rack.c
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/sys/netinet/tcp_input.c
==
--- stable/12/sys/netinet/tcp_input.c Thu May 21 19:45:14 2020
(r361341)
+++ stable/12/sys/netinet/tcp_input.c Thu May 21 19:46:11 2020
(r361342)
@@ -1519,7 +1519,7 @@ tcp_do_segment(struct mbuf *m, struct tcphdr *th, stru
struct tcpcb *tp, int drop_hdrlen, int tlen, uint8_t iptos)
{
int thflags, acked, ourfinisacked, needoutput = 0, sack_changed;
- int rstreason, todrop, win;
+ int rstreason, todrop, win, incforsyn = 0;
uint32_t tiwin;
uint16_t nsegs;
char *s;
@@ -2432,12 +2432,6 @@ tcp_do_segment(struct mbuf *m, struct tcphdr *th, stru
if (IS_FASTOPEN(tp->t_flags) && tp->t_tfo_pending) {
tcp_fastopen_decrement_counter(tp->t_tfo_pending);
tp->t_tfo_pending = NULL;
-
- /*
-* Account for the ACK of our SYN prior to
-* regular ACK processing below.
-*/
- tp->snd_una++;
}
if (tp->t_flags & TF_NEEDFIN) {
tcp_state_change(tp, TCPS_FIN_WAIT_1);
@@ -2458,6 +2452,13 @@ tcp_do_segment(struct mbuf *m, struct tcphdr *th, stru
tcp_timer_activate(tp, TT_KEEP, TP_KEEPIDLE(tp));
}
/*
+* Account for the ACK of our SYN prior to
+* regular ACK processing below, except for
+* simultaneous SYN, which is handled later.
+*/
+ if (SEQ_GT(th->th_ack, tp->snd_una) && !(tp->t_flags &
TF_NEEDSYN))
+ incforsyn = 1;
+ /*
* If segment contains data or ACK, will call tcp_reass()
* later; if not, do so now to pass queued data to user.
*/
@@ -2751,6 +2752,15 @@ tcp_do_segment(struct mbuf *m, struct tcphdr *th, stru
process_ACK:
INP_WLOCK_ASSERT(tp->t_inpcb);
+ /*
+* Adjust for the SYN bit in sequence space,
+* but don't account for it in cwnd calculations.
+* This is for the SYN_RECEIVED, non-simultaneous
+* SYN case. SYN_SENT and simultaneous SYN are
+* treated elsewhere.
+*/
+ if (incforsyn)
+ tp->snd_una++;
acked = BYTES_THIS_ACK(tp, th);
KASSERT(acked >= 0, ("%s: acked unexepectedly negative "
"(tp->snd_una=%u, th->th_ack=%u, tp=%p, m=%p)", __func__,
Modified: stable/12/sys/netinet/tcp_stacks/rack.c
==
--- stable/12/sys/netinet/tcp_stacks/rack.c Thu May 21 19:45:14 2020
(r361341)
+++ stable/12/sys/netinet/tcp_stacks/rack.c Thu May 21 19:46:11 2020
(r361342)
@@ -5580,12 +5580,6 @@ rack_do_syn_recv(struct mbuf *m, struct tcphdr *th, st
if (IS_FASTOPEN(tp->t_flags) && tp->t_tfo_pending) {
tcp_fastopen_decrement_counter(tp->t_tfo_pending);
tp->t_tfo_pending = NULL;
-
- /*
-* Account for the ACK of our SYN prior to
-* regular ACK processing below.
-*/
- tp->snd_una++;
}
if (tp->t_flags & TF_NEEDFIN) {
tcp_state_change(tp, TCPS_FIN_WAIT_1);
@@ -5603,6 +5597,13 @@ rack_do_syn_recv(struct mbuf *m, struct tcphdr *th, st
if (!IS_FASTOPEN(tp->t_flags))
cc_conn_init(tp);
}
+ /*
+* Account for the ACK of our SYN prior to
+* regular ACK processing below, except for
+* simultaneous SYN, which is handled later.
+*/
+ if (SEQ_GT(th->th_ack, tp->snd_una) && !(tp->t_flags & TF_NEEDSYN))
+ tp->snd_una++;
svn commit: r361341 - releng/11.4/secure/caroot/trusted
Author: kevans Date: Thu May 21 19:45:14 2020 New Revision: 361341 URL: https://svnweb.freebsd.org/changeset/base/361341 Log: Revert r360395: MFC r353095, r355376: add root bundle certctl(8) demonstrably has some logistics issues that still need to be worked out, as pointed out in at least PR 228913, 246614. I do not feel comfortable with proceeding with my original plan for the impending release without resolving at least PR 246614 and being able to get the release(7) scripts into a state where they're producing VM images that more closesly resemble a bsdinstall-produced install. 11.4 will still maintain the current version of certctl(8) that works well for most cases, and it still includes the caroot infrastructure. I do not currently intend to revert this in stable/11, as I would still like folks along stable/11 to be able to participate in Q/A'ing this feature. Approved by: re (gjb) Deleted: releng/11.4/secure/caroot/trusted/ACCVRAIZ1.pem releng/11.4/secure/caroot/trusted/AC_RAIZ_FNMT-RCM.pem releng/11.4/secure/caroot/trusted/Actalis_Authentication_Root_CA.pem releng/11.4/secure/caroot/trusted/AddTrust_External_Root.pem releng/11.4/secure/caroot/trusted/AddTrust_Low-Value_Services_Root.pem releng/11.4/secure/caroot/trusted/AffirmTrust_Commercial.pem releng/11.4/secure/caroot/trusted/AffirmTrust_Networking.pem releng/11.4/secure/caroot/trusted/AffirmTrust_Premium.pem releng/11.4/secure/caroot/trusted/AffirmTrust_Premium_ECC.pem releng/11.4/secure/caroot/trusted/Amazon_Root_CA_1.pem releng/11.4/secure/caroot/trusted/Amazon_Root_CA_2.pem releng/11.4/secure/caroot/trusted/Amazon_Root_CA_3.pem releng/11.4/secure/caroot/trusted/Amazon_Root_CA_4.pem releng/11.4/secure/caroot/trusted/Atos_TrustedRoot_2011.pem releng/11.4/secure/caroot/trusted/Autoridad_de_Certificacion_Firmaprofesional_CIF_A62634068.pem releng/11.4/secure/caroot/trusted/Baltimore_CyberTrust_Root.pem releng/11.4/secure/caroot/trusted/Buypass_Class_2_Root_CA.pem releng/11.4/secure/caroot/trusted/Buypass_Class_3_Root_CA.pem releng/11.4/secure/caroot/trusted/CA_Disig_Root_R2.pem releng/11.4/secure/caroot/trusted/CFCA_EV_ROOT.pem releng/11.4/secure/caroot/trusted/COMODO_Certification_Authority.pem releng/11.4/secure/caroot/trusted/COMODO_ECC_Certification_Authority.pem releng/11.4/secure/caroot/trusted/COMODO_RSA_Certification_Authority.pem releng/11.4/secure/caroot/trusted/Camerfirma_Chambers_of_Commerce_Root.pem releng/11.4/secure/caroot/trusted/Camerfirma_Global_Chambersign_Root.pem releng/11.4/secure/caroot/trusted/Certigna.pem releng/11.4/secure/caroot/trusted/Certigna_Root_CA.pem releng/11.4/secure/caroot/trusted/Certum_Root_CA.pem releng/11.4/secure/caroot/trusted/Certum_Trusted_Network_CA.pem releng/11.4/secure/caroot/trusted/Certum_Trusted_Network_CA_2.pem releng/11.4/secure/caroot/trusted/Chambers_of_Commerce_Root_-_2008.pem releng/11.4/secure/caroot/trusted/Comodo_AAA_Services_root.pem releng/11.4/secure/caroot/trusted/Cybertrust_Global_Root.pem releng/11.4/secure/caroot/trusted/D-TRUST_Root_CA_3_2013.pem releng/11.4/secure/caroot/trusted/D-TRUST_Root_Class_3_CA_2_2009.pem releng/11.4/secure/caroot/trusted/D-TRUST_Root_Class_3_CA_2_EV_2009.pem releng/11.4/secure/caroot/trusted/DST_Root_CA_X3.pem releng/11.4/secure/caroot/trusted/DigiCert_Assured_ID_Root_CA.pem releng/11.4/secure/caroot/trusted/DigiCert_Assured_ID_Root_G2.pem releng/11.4/secure/caroot/trusted/DigiCert_Assured_ID_Root_G3.pem releng/11.4/secure/caroot/trusted/DigiCert_Global_Root_CA.pem releng/11.4/secure/caroot/trusted/DigiCert_Global_Root_G2.pem releng/11.4/secure/caroot/trusted/DigiCert_Global_Root_G3.pem releng/11.4/secure/caroot/trusted/DigiCert_High_Assurance_EV_Root_CA.pem releng/11.4/secure/caroot/trusted/DigiCert_Trusted_Root_G4.pem releng/11.4/secure/caroot/trusted/E-Tugra_Certification_Authority.pem releng/11.4/secure/caroot/trusted/EC-ACC.pem releng/11.4/secure/caroot/trusted/EE_Certification_Centre_Root_CA.pem releng/11.4/secure/caroot/trusted/Entrust_Root_Certification_Authority.pem releng/11.4/secure/caroot/trusted/Entrust_Root_Certification_Authority_-_EC1.pem releng/11.4/secure/caroot/trusted/Entrust_Root_Certification_Authority_-_G2.pem releng/11.4/secure/caroot/trusted/Entrust_Root_Certification_Authority_-_G4.pem releng/11.4/secure/caroot/trusted/Entrust_net_Premium_2048_Secure_Server_CA.pem releng/11.4/secure/caroot/trusted/GDCA_TrustAUTH_R5_ROOT.pem releng/11.4/secure/caroot/trusted/GTS_Root_R1.pem releng/11.4/secure/caroot/trusted/GTS_Root_R2.pem releng/11.4/secure/caroot/trusted/GTS_Root_R3.pem releng/11.4/secure/caroot/trusted/GTS_Root_R4.pem releng/11.4/secure/caroot/trusted/GeoTrust_Global_CA.pem releng/11.4/secure/caroot/trusted/GeoTrust_Primary_Certification_Authority.pem releng/11.4/secure/caroot/trusted/GeoTrust_Primary_Certification_Authority_-_G2.pem
svn commit: r361340 - in stable/12/sys/netinet: . tcp_stacks
Author: rscheff
Date: Thu May 21 19:41:25 2020
New Revision: 361340
URL: https://svnweb.freebsd.org/changeset/base/361340
Log:
MFC r360479: Prevent premature shrinking of the scaled receive window
which can cause a TCP client to use invalid or stale TCP sequence numbers for
ACK packets.
Packets with old sequence numbers are ignored and not used to update the send
window size.
This might cause the TCP session to hang indefinitely under some
circumstances.
Reported by: Cui Cheng
Reviewed by: tuexen (mentor), rgrimes (mentor, blanket)
Approved by: tuexen (mentor), rgrimes (mentor, blanket)
MFC after:3 weeks
Sponsored by: NetApp, Inc.
Differential Revision:https://reviews.freebsd.org/D24515
Modified:
stable/12/sys/netinet/tcp_output.c
stable/12/sys/netinet/tcp_stacks/rack.c
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/sys/netinet/tcp_output.c
==
--- stable/12/sys/netinet/tcp_output.c Thu May 21 18:50:05 2020
(r361339)
+++ stable/12/sys/netinet/tcp_output.c Thu May 21 19:41:25 2020
(r361340)
@@ -1206,8 +1206,11 @@ send:
if (flags & TH_SYN)
th->th_win = htons((u_short)
(min(sbspace(>so_rcv), TCP_MAXWIN)));
- else
+ else {
+ /* Avoid shrinking window with window scaling. */
+ recwin = roundup2(recwin, 1 << tp->rcv_scale);
th->th_win = htons((u_short)(recwin >> tp->rcv_scale));
+ }
/*
* Adjust the RXWIN0SENT flag - indicate that we have advertised
Modified: stable/12/sys/netinet/tcp_stacks/rack.c
==
--- stable/12/sys/netinet/tcp_stacks/rack.c Thu May 21 18:50:05 2020
(r361339)
+++ stable/12/sys/netinet/tcp_stacks/rack.c Thu May 21 19:41:25 2020
(r361340)
@@ -8355,8 +8355,11 @@ send:
if (flags & TH_SYN)
th->th_win = htons((u_short)
(min(sbspace(>so_rcv), TCP_MAXWIN)));
- else
+ else {
+ /* Avoid shrinking window with window scaling. */
+ recwin = roundup2(recwin, 1 << tp->rcv_scale);
th->th_win = htons((u_short)(recwin >> tp->rcv_scale));
+ }
/*
* Adjust the RXWIN0SENT flag - indicate that we have advertised a 0
* window. This may cause the remote transmitter to stall. This
___
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: r361303 - in head: lib/libc/gen libexec/rtld-elf sys/sys
On 5/21/20 9:56 AM, Konstantin Belousov wrote: > On Thu, May 21, 2020 at 09:03:44AM -0700, John Baldwin wrote: >> On 5/21/20 8:12 AM, Mark Johnston wrote: >>> On Thu, May 21, 2020 at 04:41:52PM +0300, Konstantin Belousov wrote: On Thu, May 21, 2020 at 03:02:07PM +0200, Antoine Brodin wrote: > On Thu, May 21, 2020 at 12:08 AM Konstantin Belousov > wrote: >> >> Author: kib >> Date: Wed May 20 22:08:26 2020 >> New Revision: 361303 >> URL: https://svnweb.freebsd.org/changeset/base/361303 >> >> Log: >> Change the samantic of struct link_map l_addr member. >> >> It previously returned the object map base address, while all other >> ELF operating systems return load offset, i.e. the difference between >> map base and the link base. >> >> Explain the meaning of the field in the man page. >> >> Stop filling the mips-only l_offs member, which is apparently unused. >> >> PR: 246561 >> Requested by: Damjan Jovanovic >> Reviewed by: emaste, jhb, cem (previous version) >> Sponsored by: The FreeBSD Foundation >> MFC after:1 week >> Differential revision:https://reviews.freebsd.org/D24918 >> >> Modified: >> head/lib/libc/gen/dlinfo.3 >> head/libexec/rtld-elf/rtld.c >> head/sys/sys/link_elf.h > > Hi, > > After this commit, some ports fail to build with signal 11. > For instance lang/perl5.30 fails to build with default options (DTRACE on) > Disabling the DTRACE option makes it able to build again. > I see, thank you for reporting. So drti.c:dtrace_dof_init() does read l_addr, and the dtrace code assumes that l_addr is the base, not relocbase. Mark, was dofhp_addr initialization changed comparing to Solaris ? >>> >>> It appears it has been the same since DTrace was imported. illumos >>> still has similar code. >>> >>> Note that drti.o is linked into any executable and shlib that defines >>> static probes, so the ABI change affects more than just dtrace(1). >>> Would it be possible to define a new value for RTLD_DI_LINKMAP, and >>> preserve the old behaviour for the old value? >> >> I think a bigger question is if Solaris/illumos treat l_addr as mapbase >> (absolute address) or relocbase (relative address). In the discussion >> in the phabricator I had assumed that all other OS's treated l_addr as >> the relative offset (relocbase). Does the code for illumos assume an >> absolute address or does it assume a relative address in l_addr? > > It is rather clear, since the dtrace code was pristine, that Solaris > provides the mapbase. I do not have Solaris/Illumos box anymore > (for quite some time), so I cannot check directly. > > My current PoV is that l_addr semantic must be restored, and relocbase > provided by newly added member. I am fine with reverting the l_addr semantic. I'm still not sure how to resolve the original PR, though perhaps Wine just has to carry a local patch forever? GDB will work via the current accident so long as we never pre-link libraries. As long as PIE binaries have a starting VA of 0 like our shared libraries then I think GDB will be ok with our PIE binaries as well. -- John Baldwin ___ 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: r361339 - releng/11.4/usr.sbin/certctl
Author: kevans
Date: Thu May 21 18:50:05 2020
New Revision: 361339
URL: https://svnweb.freebsd.org/changeset/base/361339
Log:
MFS r361310: MFC r361022-361023, 361148: certctl(8) fixes
r361022: certctl(8): don't completely nuke $CERTDESTDIR
It's been reported/noted that a well-timed `certctl rehash` will completely
obliterate $CERTDESTDIR, which may get used by ports or system
administrators. While we can't guarantee the certctl semantics when other
non-certctl-controlled bits live here, we should make some amount of effort
to play nice.
Pruning all existing links, which we'll subsequently rebuild as needed, is
sufficient for our needs. This can still be destructive, but it's perhaps
less likely to cause issues.
I also note that we should probably be pruning /etc/ssl/blacklisted upon
rehash as well.
r361023: certctl: follow-up to r361022, prune blacklist as well
Otherwise, removals from the blacklist may not get processed as they should.
While we're here, restructure these to not bother with mkdir(1) if we've
already tested them to exist.
r361148: certctl: don't fall over flat with relative DESTDIR
Up until now, all of our DESTDIR use has been with absolute paths. It turned
out that the cd in/out dance we do here breaks us down later on, as the
relative path no longer resolves.
Convert EXTENSIONS to an ERE that we'll use to grep ls -1 of the dir we're
inspecting, rather than cd'ing into it and globbing it up.
Approved by: re (kib)
Modified:
releng/11.4/usr.sbin/certctl/certctl.sh
Directory Properties:
releng/11.4/ (props changed)
Modified: releng/11.4/usr.sbin/certctl/certctl.sh
==
--- releng/11.4/usr.sbin/certctl/certctl.sh Thu May 21 18:38:41 2020
(r361338)
+++ releng/11.4/usr.sbin/certctl/certctl.sh Thu May 21 18:50:05 2020
(r361339)
@@ -34,7 +34,7 @@
:
${BLACKLISTPATH:=${DESTDIR}/usr/share/certs/blacklisted:${DESTDIR}/usr/local/etc/ssl/blacklisted}
: ${CERTDESTDIR:=${DESTDIR}/etc/ssl/certs}
: ${BLACKLISTDESTDIR:=${DESTDIR}/etc/ssl/blacklisted}
-: ${EXTENSIONS:="*.pem *.crt *.cer *.crl *.0"}
+: ${FILEPAT:="\.pem$|\.crt$|\.cer$|\.crl$|\.0$"}
: ${VERBOSE:=0}
GLOBALS
@@ -104,13 +104,11 @@ do_scan()
for CPATH in "$@"; do
[ -d "$CPATH" ] || continue
echo "Scanning $CPATH for certificates..."
- cd "$CPATH"
- for CFILE in $EXTENSIONS; do
- [ -e "$CFILE" ] || continue
+ for CFILE in $(ls -1 "${CPATH}" | grep -Ee "${FILEPAT}"); do
+ [ -e "$CPATH/$CFILE" ] || continue
[ $VERBOSE -gt 0 ] && echo "Reading $CFILE"
"$CFUNC" "$CPATH/$CFILE"
done
- cd -
done
}
@@ -142,9 +140,18 @@ do_list()
cmd_rehash()
{
- [ $NOOP -eq 0 ] && rm -rf "$CERTDESTDIR"
- [ $NOOP -eq 0 ] && mkdir -p "$CERTDESTDIR"
- [ $NOOP -eq 0 ] && mkdir -p "$BLACKLISTDESTDIR"
+ if [ $NOOP -eq 0 ]; then
+ if [ -e "$CERTDESTDIR" ]; then
+ find "$CERTDESTDIR" -type link -delete
+ else
+ mkdir -p "$CERTDESTDIR"
+ fi
+ if [ -e "$BLACKLISTDESTDIR" ]; then
+ find "$BLACKLISTDESTDIR" -type link -delete
+ else
+ mkdir -p "$BLACKLISTDESTDIR"
+ fi
+ fi
do_scan create_blacklisted "$BLACKLISTPATH"
do_scan create_trusted_link "$TRUSTPATH"
___
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: r361338 - head/sys/sys
Author: markj
Date: Thu May 21 18:38:41 2020
New Revision: 361338
URL: https://svnweb.freebsd.org/changeset/base/361338
Log:
Fix ACCEPT_FILTER_DEFINE to pass the version to MODULE_VERSION.
MFC with: r361263
Modified:
head/sys/sys/socketvar.h
Modified: head/sys/sys/socketvar.h
==
--- head/sys/sys/socketvar.hThu May 21 17:34:31 2020(r361337)
+++ head/sys/sys/socketvar.hThu May 21 18:38:41 2020(r361338)
@@ -349,7 +349,7 @@ struct accept_filter {
}; \
DECLARE_MODULE(modname, modname##_mod, SI_SUB_DRIVERS, \
SI_ORDER_MIDDLE); \
- MODULE_VERSION(modname, 1)
+ MODULE_VERSION(modname, ver)
#ifdef MALLOC_DECLARE
MALLOC_DECLARE(M_ACCF);
___
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: r361337 - head/usr.bin/indent
Author: pstef
Date: Thu May 21 17:34:31 2020
New Revision: 361337
URL: https://svnweb.freebsd.org/changeset/base/361337
Log:
indent(1): add fallthrough markers
This silences -Wimplicit-fallthrough warnings.
Submitted by: Michael Paquier
Obtained from:postgresql.org
MFC after:3 days
Modified:
head/usr.bin/indent/indent.c
head/usr.bin/indent/parse.c
Modified: head/usr.bin/indent/indent.c
==
--- head/usr.bin/indent/indent.cThu May 21 15:53:16 2020
(r361336)
+++ head/usr.bin/indent/indent.cThu May 21 17:34:31 2020
(r361337)
@@ -967,6 +967,7 @@ check_type:
case structure:
if (ps.p_l_follow > 0)
goto copy_id;
+ /* FALLTHROUGH */
case decl: /* we have a declaration type (int, etc.) */
parse(decl);/* let parser worry about indentation */
if (ps.last_token == rparen && ps.tos <= 1) {
Modified: head/usr.bin/indent/parse.c
==
--- head/usr.bin/indent/parse.c Thu May 21 15:53:16 2020(r361336)
+++ head/usr.bin/indent/parse.c Thu May 21 17:34:31 2020(r361337)
@@ -107,6 +107,7 @@ parse(int tk) /* tk: the code for the construct scanne
*/
ps.i_l_follow = ps.il[ps.tos--];
/* the rest is the same as for dolit and forstmt */
+ /* FALLTHROUGH */
case dolit:/* 'do' */
case forstmt: /* for (...) */
ps.p_stack[++ps.tos] = tk;
___
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: r361303 - in head: lib/libc/gen libexec/rtld-elf sys/sys
On Thu, May 21, 2020 at 08:09:58PM +0300, Konstantin Belousov wrote: > On Thu, May 21, 2020 at 01:01:24PM -0400, Mark Johnston wrote: > > On Thu, May 21, 2020 at 07:56:46PM +0300, Konstantin Belousov wrote: > > > On Thu, May 21, 2020 at 09:03:44AM -0700, John Baldwin wrote: > > > > On 5/21/20 8:12 AM, Mark Johnston wrote: > > > > > On Thu, May 21, 2020 at 04:41:52PM +0300, Konstantin Belousov wrote: > > > > >> On Thu, May 21, 2020 at 03:02:07PM +0200, Antoine Brodin wrote: > > > > >>> On Thu, May 21, 2020 at 12:08 AM Konstantin Belousov > > > > >>> wrote: > > > > > > > > Author: kib > > > > Date: Wed May 20 22:08:26 2020 > > > > New Revision: 361303 > > > > URL: https://svnweb.freebsd.org/changeset/base/361303 > > > > > > > > Log: > > > > Change the samantic of struct link_map l_addr member. > > > > > > > > It previously returned the object map base address, while all > > > > other > > > > ELF operating systems return load offset, i.e. the difference > > > > between > > > > map base and the link base. > > > > > > > > Explain the meaning of the field in the man page. > > > > > > > > Stop filling the mips-only l_offs member, which is apparently > > > > unused. > > > > > > > > PR: 246561 > > > > Requested by: Damjan Jovanovic > > > > Reviewed by: emaste, jhb, cem (previous version) > > > > Sponsored by: The FreeBSD Foundation > > > > MFC after:1 week > > > > Differential revision:https://reviews.freebsd.org/D24918 > > > > > > > > Modified: > > > > head/lib/libc/gen/dlinfo.3 > > > > head/libexec/rtld-elf/rtld.c > > > > head/sys/sys/link_elf.h > > > > >>> > > > > >>> Hi, > > > > >>> > > > > >>> After this commit, some ports fail to build with signal 11. > > > > >>> For instance lang/perl5.30 fails to build with default options > > > > >>> (DTRACE on) > > > > >>> Disabling the DTRACE option makes it able to build again. > > > > >>> > > > > >> I see, thank you for reporting. > > > > >> > > > > >> So drti.c:dtrace_dof_init() does read l_addr, and the dtrace code > > > > >> assumes > > > > >> that l_addr is the base, not relocbase. > > > > >> > > > > >> Mark, was dofhp_addr initialization changed comparing to Solaris ? > > > > > > > > > > It appears it has been the same since DTrace was imported. illumos > > > > > still has similar code. > > > > > > > > > > Note that drti.o is linked into any executable and shlib that defines > > > > > static probes, so the ABI change affects more than just dtrace(1). > > > > > Would it be possible to define a new value for RTLD_DI_LINKMAP, and > > > > > preserve the old behaviour for the old value? > > > > > > > > I think a bigger question is if Solaris/illumos treat l_addr as mapbase > > > > (absolute address) or relocbase (relative address). In the discussion > > > > in the phabricator I had assumed that all other OS's treated l_addr as > > > > the relative offset (relocbase). Does the code for illumos assume an > > > > absolute address or does it assume a relative address in l_addr? > > > > > > It is rather clear, since the dtrace code was pristine, that Solaris > > > provides the mapbase. I do not have Solaris/Illumos box anymore > > > (for quite some time), so I cannot check directly. > > > > > > My current PoV is that l_addr semantic must be restored, and relocbase > > > provided by newly added member. > > > > > > BTW, it is strange that perl triggers it, is it linked as PIE on HEAD ? > > > > Isn't the problem when perl is *not* linked as PIE? In this case > > relocbase is 0, so the ELF header access becomes a NULL pointer > > dereference. > drti checks for ET_DYN, only then it uses l_addr at all. The problem is before that, where it treats l_addr as a pointer to the ELF header so it can check the type. > PIE binaries are dso with non-zero base, non-PIE binaries are ET_EXEC, > which should make dtri.c ignore the l_addr value. ___ 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: r361303 - in head: lib/libc/gen libexec/rtld-elf sys/sys
On Thu, May 21, 2020 at 01:01:24PM -0400, Mark Johnston wrote: > On Thu, May 21, 2020 at 07:56:46PM +0300, Konstantin Belousov wrote: > > On Thu, May 21, 2020 at 09:03:44AM -0700, John Baldwin wrote: > > > On 5/21/20 8:12 AM, Mark Johnston wrote: > > > > On Thu, May 21, 2020 at 04:41:52PM +0300, Konstantin Belousov wrote: > > > >> On Thu, May 21, 2020 at 03:02:07PM +0200, Antoine Brodin wrote: > > > >>> On Thu, May 21, 2020 at 12:08 AM Konstantin Belousov > > > >>> wrote: > > > > > > Author: kib > > > Date: Wed May 20 22:08:26 2020 > > > New Revision: 361303 > > > URL: https://svnweb.freebsd.org/changeset/base/361303 > > > > > > Log: > > > Change the samantic of struct link_map l_addr member. > > > > > > It previously returned the object map base address, while all other > > > ELF operating systems return load offset, i.e. the difference > > > between > > > map base and the link base. > > > > > > Explain the meaning of the field in the man page. > > > > > > Stop filling the mips-only l_offs member, which is apparently > > > unused. > > > > > > PR: 246561 > > > Requested by: Damjan Jovanovic > > > Reviewed by: emaste, jhb, cem (previous version) > > > Sponsored by: The FreeBSD Foundation > > > MFC after:1 week > > > Differential revision:https://reviews.freebsd.org/D24918 > > > > > > Modified: > > > head/lib/libc/gen/dlinfo.3 > > > head/libexec/rtld-elf/rtld.c > > > head/sys/sys/link_elf.h > > > >>> > > > >>> Hi, > > > >>> > > > >>> After this commit, some ports fail to build with signal 11. > > > >>> For instance lang/perl5.30 fails to build with default options > > > >>> (DTRACE on) > > > >>> Disabling the DTRACE option makes it able to build again. > > > >>> > > > >> I see, thank you for reporting. > > > >> > > > >> So drti.c:dtrace_dof_init() does read l_addr, and the dtrace code > > > >> assumes > > > >> that l_addr is the base, not relocbase. > > > >> > > > >> Mark, was dofhp_addr initialization changed comparing to Solaris ? > > > > > > > > It appears it has been the same since DTrace was imported. illumos > > > > still has similar code. > > > > > > > > Note that drti.o is linked into any executable and shlib that defines > > > > static probes, so the ABI change affects more than just dtrace(1). > > > > Would it be possible to define a new value for RTLD_DI_LINKMAP, and > > > > preserve the old behaviour for the old value? > > > > > > I think a bigger question is if Solaris/illumos treat l_addr as mapbase > > > (absolute address) or relocbase (relative address). In the discussion > > > in the phabricator I had assumed that all other OS's treated l_addr as > > > the relative offset (relocbase). Does the code for illumos assume an > > > absolute address or does it assume a relative address in l_addr? > > > > It is rather clear, since the dtrace code was pristine, that Solaris > > provides the mapbase. I do not have Solaris/Illumos box anymore > > (for quite some time), so I cannot check directly. > > > > My current PoV is that l_addr semantic must be restored, and relocbase > > provided by newly added member. > > > > BTW, it is strange that perl triggers it, is it linked as PIE on HEAD ? > > Isn't the problem when perl is *not* linked as PIE? In this case > relocbase is 0, so the ELF header access becomes a NULL pointer > dereference. drti checks for ET_DYN, only then it uses l_addr at all. PIE binaries are dso with non-zero base, non-PIE binaries are ET_EXEC, which should make dtri.c ignore the l_addr value. ___ 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: r361336 - head/sys/powerpc/aim
On Thu, May 21, 2020 at 03:53:17PM +, Brandon Bergren wrote: > Author: bdragon > Date: Thu May 21 15:53:16 2020 > New Revision: 361336 > URL: https://svnweb.freebsd.org/changeset/base/361336 > > Log: > [PowerPC] Fix kernel boot on powerpc > > Recent changes have caused the vmspace objects to start coming from KVA > instead of direct-mapped memory on powerpc. As far as I can tell, this is > not actually a problem, so we should stop arbitrarily asserting that it is. > > I do not know why this was not being triggered before. UMA was recently changed in r357549 to use multi-page slabs if doing so would reduce internal fragmentation to a certain degree. In that case the slabs will be mapped into KVA. vmspace objects are quite large but smaller than a page, so they benefit from this. You can verify by checking the vm.uma.VMSPACE.keg.ppera sysctl, which gives the number of pages per slab. ___ 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: r361303 - in head: lib/libc/gen libexec/rtld-elf sys/sys
On Thu, May 21, 2020 at 07:56:46PM +0300, Konstantin Belousov wrote: > On Thu, May 21, 2020 at 09:03:44AM -0700, John Baldwin wrote: > > On 5/21/20 8:12 AM, Mark Johnston wrote: > > > On Thu, May 21, 2020 at 04:41:52PM +0300, Konstantin Belousov wrote: > > >> On Thu, May 21, 2020 at 03:02:07PM +0200, Antoine Brodin wrote: > > >>> On Thu, May 21, 2020 at 12:08 AM Konstantin Belousov > > >>> wrote: > > > > Author: kib > > Date: Wed May 20 22:08:26 2020 > > New Revision: 361303 > > URL: https://svnweb.freebsd.org/changeset/base/361303 > > > > Log: > > Change the samantic of struct link_map l_addr member. > > > > It previously returned the object map base address, while all other > > ELF operating systems return load offset, i.e. the difference between > > map base and the link base. > > > > Explain the meaning of the field in the man page. > > > > Stop filling the mips-only l_offs member, which is apparently unused. > > > > PR: 246561 > > Requested by: Damjan Jovanovic > > Reviewed by: emaste, jhb, cem (previous version) > > Sponsored by: The FreeBSD Foundation > > MFC after:1 week > > Differential revision:https://reviews.freebsd.org/D24918 > > > > Modified: > > head/lib/libc/gen/dlinfo.3 > > head/libexec/rtld-elf/rtld.c > > head/sys/sys/link_elf.h > > >>> > > >>> Hi, > > >>> > > >>> After this commit, some ports fail to build with signal 11. > > >>> For instance lang/perl5.30 fails to build with default options (DTRACE > > >>> on) > > >>> Disabling the DTRACE option makes it able to build again. > > >>> > > >> I see, thank you for reporting. > > >> > > >> So drti.c:dtrace_dof_init() does read l_addr, and the dtrace code assumes > > >> that l_addr is the base, not relocbase. > > >> > > >> Mark, was dofhp_addr initialization changed comparing to Solaris ? > > > > > > It appears it has been the same since DTrace was imported. illumos > > > still has similar code. > > > > > > Note that drti.o is linked into any executable and shlib that defines > > > static probes, so the ABI change affects more than just dtrace(1). > > > Would it be possible to define a new value for RTLD_DI_LINKMAP, and > > > preserve the old behaviour for the old value? > > > > I think a bigger question is if Solaris/illumos treat l_addr as mapbase > > (absolute address) or relocbase (relative address). In the discussion > > in the phabricator I had assumed that all other OS's treated l_addr as > > the relative offset (relocbase). Does the code for illumos assume an > > absolute address or does it assume a relative address in l_addr? > > It is rather clear, since the dtrace code was pristine, that Solaris > provides the mapbase. I do not have Solaris/Illumos box anymore > (for quite some time), so I cannot check directly. > > My current PoV is that l_addr semantic must be restored, and relocbase > provided by newly added member. > > BTW, it is strange that perl triggers it, is it linked as PIE on HEAD ? Isn't the problem when perl is *not* linked as PIE? In this case relocbase is 0, so the ELF header access becomes a NULL pointer dereference. ___ 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: r361303 - in head: lib/libc/gen libexec/rtld-elf sys/sys
On Thu, May 21, 2020 at 09:03:44AM -0700, John Baldwin wrote: > On 5/21/20 8:12 AM, Mark Johnston wrote: > > On Thu, May 21, 2020 at 04:41:52PM +0300, Konstantin Belousov wrote: > >> On Thu, May 21, 2020 at 03:02:07PM +0200, Antoine Brodin wrote: > >>> On Thu, May 21, 2020 at 12:08 AM Konstantin Belousov > >>> wrote: > > Author: kib > Date: Wed May 20 22:08:26 2020 > New Revision: 361303 > URL: https://svnweb.freebsd.org/changeset/base/361303 > > Log: > Change the samantic of struct link_map l_addr member. > > It previously returned the object map base address, while all other > ELF operating systems return load offset, i.e. the difference between > map base and the link base. > > Explain the meaning of the field in the man page. > > Stop filling the mips-only l_offs member, which is apparently unused. > > PR: 246561 > Requested by: Damjan Jovanovic > Reviewed by: emaste, jhb, cem (previous version) > Sponsored by: The FreeBSD Foundation > MFC after:1 week > Differential revision:https://reviews.freebsd.org/D24918 > > Modified: > head/lib/libc/gen/dlinfo.3 > head/libexec/rtld-elf/rtld.c > head/sys/sys/link_elf.h > >>> > >>> Hi, > >>> > >>> After this commit, some ports fail to build with signal 11. > >>> For instance lang/perl5.30 fails to build with default options (DTRACE on) > >>> Disabling the DTRACE option makes it able to build again. > >>> > >> I see, thank you for reporting. > >> > >> So drti.c:dtrace_dof_init() does read l_addr, and the dtrace code assumes > >> that l_addr is the base, not relocbase. > >> > >> Mark, was dofhp_addr initialization changed comparing to Solaris ? > > > > It appears it has been the same since DTrace was imported. illumos > > still has similar code. > > > > Note that drti.o is linked into any executable and shlib that defines > > static probes, so the ABI change affects more than just dtrace(1). > > Would it be possible to define a new value for RTLD_DI_LINKMAP, and > > preserve the old behaviour for the old value? > > I think a bigger question is if Solaris/illumos treat l_addr as mapbase > (absolute address) or relocbase (relative address). In the discussion > in the phabricator I had assumed that all other OS's treated l_addr as > the relative offset (relocbase). Does the code for illumos assume an > absolute address or does it assume a relative address in l_addr? It is rather clear, since the dtrace code was pristine, that Solaris provides the mapbase. I do not have Solaris/Illumos box anymore (for quite some time), so I cannot check directly. My current PoV is that l_addr semantic must be restored, and relocbase provided by newly added member. BTW, it is strange that perl triggers it, is it linked as PIE on HEAD ? ___ 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: r361303 - in head: lib/libc/gen libexec/rtld-elf sys/sys
On Thu, May 21, 2020 at 09:03:44AM -0700, John Baldwin wrote: > On 5/21/20 8:12 AM, Mark Johnston wrote: > > On Thu, May 21, 2020 at 04:41:52PM +0300, Konstantin Belousov wrote: > >> On Thu, May 21, 2020 at 03:02:07PM +0200, Antoine Brodin wrote: > >>> On Thu, May 21, 2020 at 12:08 AM Konstantin Belousov > >>> wrote: > > Author: kib > Date: Wed May 20 22:08:26 2020 > New Revision: 361303 > URL: https://svnweb.freebsd.org/changeset/base/361303 > > Log: > Change the samantic of struct link_map l_addr member. > > It previously returned the object map base address, while all other > ELF operating systems return load offset, i.e. the difference between > map base and the link base. > > Explain the meaning of the field in the man page. > > Stop filling the mips-only l_offs member, which is apparently unused. > > PR: 246561 > Requested by: Damjan Jovanovic > Reviewed by: emaste, jhb, cem (previous version) > Sponsored by: The FreeBSD Foundation > MFC after:1 week > Differential revision:https://reviews.freebsd.org/D24918 > > Modified: > head/lib/libc/gen/dlinfo.3 > head/libexec/rtld-elf/rtld.c > head/sys/sys/link_elf.h > >>> > >>> Hi, > >>> > >>> After this commit, some ports fail to build with signal 11. > >>> For instance lang/perl5.30 fails to build with default options (DTRACE on) > >>> Disabling the DTRACE option makes it able to build again. > >>> > >> I see, thank you for reporting. > >> > >> So drti.c:dtrace_dof_init() does read l_addr, and the dtrace code assumes > >> that l_addr is the base, not relocbase. > >> > >> Mark, was dofhp_addr initialization changed comparing to Solaris ? > > > > It appears it has been the same since DTrace was imported. illumos > > still has similar code. > > > > Note that drti.o is linked into any executable and shlib that defines > > static probes, so the ABI change affects more than just dtrace(1). > > Would it be possible to define a new value for RTLD_DI_LINKMAP, and > > preserve the old behaviour for the old value? > > I think a bigger question is if Solaris/illumos treat l_addr as mapbase > (absolute address) or relocbase (relative address). In the discussion > in the phabricator I had assumed that all other OS's treated l_addr as > the relative offset (relocbase). Does the code for illumos assume an > absolute address or does it assume a relative address in l_addr? It should be an absolute address, since dtrace appears to assume that an ELF header is mapped at l_addr. Here, "addr" is the value of l_addr for the RTLD_SELF linkmap: https://github.com/joyent/illumos-joyent/blob/master/usr/src/lib/libdtrace/common/dlink_common.c#L124 ___ 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: r361303 - in head: lib/libc/gen libexec/rtld-elf sys/sys
On 5/21/20 8:12 AM, Mark Johnston wrote: > On Thu, May 21, 2020 at 04:41:52PM +0300, Konstantin Belousov wrote: >> On Thu, May 21, 2020 at 03:02:07PM +0200, Antoine Brodin wrote: >>> On Thu, May 21, 2020 at 12:08 AM Konstantin Belousov >>> wrote: Author: kib Date: Wed May 20 22:08:26 2020 New Revision: 361303 URL: https://svnweb.freebsd.org/changeset/base/361303 Log: Change the samantic of struct link_map l_addr member. It previously returned the object map base address, while all other ELF operating systems return load offset, i.e. the difference between map base and the link base. Explain the meaning of the field in the man page. Stop filling the mips-only l_offs member, which is apparently unused. PR: 246561 Requested by: Damjan Jovanovic Reviewed by: emaste, jhb, cem (previous version) Sponsored by: The FreeBSD Foundation MFC after:1 week Differential revision:https://reviews.freebsd.org/D24918 Modified: head/lib/libc/gen/dlinfo.3 head/libexec/rtld-elf/rtld.c head/sys/sys/link_elf.h >>> >>> Hi, >>> >>> After this commit, some ports fail to build with signal 11. >>> For instance lang/perl5.30 fails to build with default options (DTRACE on) >>> Disabling the DTRACE option makes it able to build again. >>> >> I see, thank you for reporting. >> >> So drti.c:dtrace_dof_init() does read l_addr, and the dtrace code assumes >> that l_addr is the base, not relocbase. >> >> Mark, was dofhp_addr initialization changed comparing to Solaris ? > > It appears it has been the same since DTrace was imported. illumos > still has similar code. > > Note that drti.o is linked into any executable and shlib that defines > static probes, so the ABI change affects more than just dtrace(1). > Would it be possible to define a new value for RTLD_DI_LINKMAP, and > preserve the old behaviour for the old value? I think a bigger question is if Solaris/illumos treat l_addr as mapbase (absolute address) or relocbase (relative address). In the discussion in the phabricator I had assumed that all other OS's treated l_addr as the relative offset (relocbase). Does the code for illumos assume an absolute address or does it assume a relative address in l_addr? -- John Baldwin ___ 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: r361334 - in stable/12/sys: amd64/amd64 arm64/arm64 dev/acpica i386/i386 x86/acpica
Author: markj
Date: Thu May 21 15:28:35 2020
New Revision: 361334
URL: https://svnweb.freebsd.org/changeset/base/361334
Log:
MFC r361033:
Call acpi_pxm_set_proximity_info() slightly earlier on x86.
Modified:
stable/12/sys/amd64/amd64/mp_machdep.c
stable/12/sys/arm64/arm64/mp_machdep.c
stable/12/sys/dev/acpica/acpi_pxm.c
stable/12/sys/dev/acpica/acpivar.h
stable/12/sys/i386/i386/mp_machdep.c
stable/12/sys/x86/acpica/srat.c
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/sys/amd64/amd64/mp_machdep.c
==
--- stable/12/sys/amd64/amd64/mp_machdep.c Thu May 21 15:18:59 2020
(r361333)
+++ stable/12/sys/amd64/amd64/mp_machdep.c Thu May 21 15:28:35 2020
(r361334)
@@ -265,8 +265,9 @@ cpu_mp_start(void)
init_ops.start_all_aps();
set_interrupt_apic_ids();
-}
+ acpi_pxm_set_cpu_locality();
+}
/*
* AP CPU's call this to initialize themselves.
Modified: stable/12/sys/arm64/arm64/mp_machdep.c
==
--- stable/12/sys/arm64/arm64/mp_machdep.c Thu May 21 15:18:59 2020
(r361333)
+++ stable/12/sys/arm64/arm64/mp_machdep.c Thu May 21 15:28:35 2020
(r361334)
@@ -595,9 +595,7 @@ cpu_init_acpi(void)
acpi_unmap_table(madt);
#if MAXMEMDOM > 1
- /* set proximity info */
acpi_pxm_set_cpu_locality();
- acpi_pxm_free();
#endif
}
#endif
Modified: stable/12/sys/dev/acpica/acpi_pxm.c
==
--- stable/12/sys/dev/acpica/acpi_pxm.c Thu May 21 15:18:59 2020
(r361333)
+++ stable/12/sys/dev/acpica/acpi_pxm.c Thu May 21 15:28:35 2020
(r361334)
@@ -628,7 +628,8 @@ srat_walk_table(acpi_subtable_handler *handler, void *
}
/*
- * Setup per-CPU domain IDs from information saved in 'cpus'.
+ * Set up per-CPU domain IDs from information saved in 'cpus' and tear down
data
+ * structures allocated by acpi_pxm_init().
*/
void
acpi_pxm_set_cpu_locality(void)
@@ -651,6 +652,10 @@ acpi_pxm_set_cpu_locality(void)
printf("SRAT: CPU %u has memory domain %d\n", i,
pc->pc_domain);
}
+ /* XXXMJ the page is leaked. */
+ pmap_unmapbios((vm_offset_t)cpus, sizeof(*cpus) * max_cpus);
+ srat_physaddr = 0;
+ cpus = NULL;
}
int
@@ -662,20 +667,6 @@ acpi_pxm_get_cpu_locality(int apic_id)
if (cpu == NULL)
panic("SRAT: CPU with ID %u is not known", apic_id);
return (cpu->domain);
-}
-
-/*
- * Free data structures allocated during acpi_pxm_init.
- */
-void
-acpi_pxm_free(void)
-{
-
- if (srat_physaddr == 0)
- return;
- pmap_unmapbios((vm_offset_t)cpus, sizeof(*cpus) * max_cpus);
- srat_physaddr = 0;
- cpus = NULL;
}
/*
Modified: stable/12/sys/dev/acpica/acpivar.h
==
--- stable/12/sys/dev/acpica/acpivar.h Thu May 21 15:18:59 2020
(r361333)
+++ stable/12/sys/dev/acpica/acpivar.h Thu May 21 15:28:35 2020
(r361334)
@@ -536,7 +536,6 @@ voidacpi_pxm_parse_tables(void);
void acpi_pxm_set_mem_locality(void);
void acpi_pxm_set_cpu_locality(void);
intacpi_pxm_get_cpu_locality(int apic_id);
-void acpi_pxm_free(void);
/*
* Map a PXM to a VM domain.
Modified: stable/12/sys/i386/i386/mp_machdep.c
==
--- stable/12/sys/i386/i386/mp_machdep.cThu May 21 15:18:59 2020
(r361333)
+++ stable/12/sys/i386/i386/mp_machdep.cThu May 21 15:28:35 2020
(r361334)
@@ -199,6 +199,8 @@ cpu_mp_start(void)
start_all_aps();
set_interrupt_apic_ids();
+
+ acpi_pxm_set_cpu_locality();
}
/*
Modified: stable/12/sys/x86/acpica/srat.c
==
--- stable/12/sys/x86/acpica/srat.c Thu May 21 15:18:59 2020
(r361333)
+++ stable/12/sys/x86/acpica/srat.c Thu May 21 15:28:35 2020
(r361334)
@@ -56,13 +56,4 @@ parse_acpi_tables(void *dummy)
SYSINIT(parse_acpi_tables, SI_SUB_VM - 1, SI_ORDER_FIRST, parse_acpi_tables,
NULL);
-static void
-srat_set_cpus(void *dummy)
-{
-
- acpi_pxm_set_cpu_locality();
- acpi_pxm_free();
-}
-SYSINIT(srat_set_cpus, SI_SUB_CPU, SI_ORDER_ANY, srat_set_cpus, NULL);
-
#endif /* MAXMEMDOM > 1 */
___
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: r361332 - head/bin/ls
Author: kevans
Date: Thu May 21 15:15:50 2020
New Revision: 361332
URL: https://svnweb.freebsd.org/changeset/base/361332
Log:
ls: fix WITHOUT_LS_COLORS build
*sigh* references to colorflags should be gated by COLORLS.
Pointy hat to:kevans
Reported by: jenkins (rescue build)
X-MFC-With: r361318
Modified:
head/bin/ls/ls.c
Modified: head/bin/ls/ls.c
==
--- head/bin/ls/ls.cThu May 21 14:39:00 2020(r361331)
+++ head/bin/ls/ls.cThu May 21 15:15:50 2020(r361332)
@@ -270,8 +270,10 @@ main(int argc, char *argv[])
* For historical compatibility, we'll use our autodetection if CLICOLOR
* is set.
*/
+#ifdef COLORLS
if (getenv("CLICOLOR"))
colorflag = COLORFLAG_AUTO;
+#endif
while ((ch = getopt_long(argc, argv,
"+1ABCD:FGHILPRSTUWXZabcdfghiklmnopqrstuwxy,", long_opts,
NULL)) != -1) {
@@ -355,7 +357,9 @@ main(int argc, char *argv[])
* stdout isn't a tty.
*/
setenv("CLICOLOR", "", 1);
+#ifdef COLORLS
colorflag = COLORFLAG_AUTO;
+#endif
break;
case 'H':
fts_options |= FTS_COMFOLLOW;
___
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: r361331 - head/bin/ls
Author: kevans
Date: Thu May 21 14:39:00 2020
New Revision: 361331
URL: https://svnweb.freebsd.org/changeset/base/361331
Log:
ls(1): actually restore proper behavior
Highlights:
- CLICOLOR in the environment should imply --color=auto to maintain
compatibility with historical behavior
- -G should set CLICOLOR and imply --color=auto
The manpage has been updated to draw the connection between -G and --color;
the former is in-fact a sort of compromise between --color=always and
--color=auto, where we'll output color regardless of the environment lacking
CLICOLOR/COLORTERM assuming stdout is a tty.
X-MFC-With: r361318
Modified:
head/bin/ls/ls.1
head/bin/ls/ls.c
Modified: head/bin/ls/ls.1
==
--- head/bin/ls/ls.1Thu May 21 13:46:30 2020(r361330)
+++ head/bin/ls/ls.1Thu May 21 14:39:00 2020(r361331)
@@ -32,7 +32,7 @@
.\" @(#)ls.1 8.7 (Berkeley) 7/29/94
.\" $FreeBSD$
.\"
-.Dd May 20, 2020
+.Dd May 21, 2020
.Dt LS 1
.Os
.Sh NAME
@@ -135,7 +135,8 @@ This option is equivalent to defining
.Ev CLICOLOR
or
.Ev COLORTERM
-in the environment.
+in the environment and setting
+.Fl -color Ns = Ns Ar auto .
(See below.)
This functionality can be compiled out by removing the definition of
.Ev COLORLS .
Modified: head/bin/ls/ls.c
==
--- head/bin/ls/ls.cThu May 21 13:46:30 2020(r361330)
+++ head/bin/ls/ls.cThu May 21 14:39:00 2020(r361331)
@@ -265,6 +265,13 @@ main(int argc, char *argv[])
fts_options = FTS_PHYSICAL;
if (getenv("LS_SAMESORT"))
f_samesort = 1;
+
+ /*
+* For historical compatibility, we'll use our autodetection if CLICOLOR
+* is set.
+*/
+ if (getenv("CLICOLOR"))
+ colorflag = COLORFLAG_AUTO;
while ((ch = getopt_long(argc, argv,
"+1ABCD:FGHILPRSTUWXZabcdfghiklmnopqrstuwxy,", long_opts,
NULL)) != -1) {
@@ -342,7 +349,13 @@ main(int argc, char *argv[])
f_slash = 0;
break;
case 'G':
+ /*
+* We both set CLICOLOR here and set colorflag to
+* COLORFLAG_AUTO, because -G should not force color if
+* stdout isn't a tty.
+*/
setenv("CLICOLOR", "", 1);
+ colorflag = COLORFLAG_AUTO;
break;
case 'H':
fts_options |= FTS_COMFOLLOW;
___
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: r361330 - head/lib/libprocstat
Author: avg
Date: Thu May 21 13:46:30 2020
New Revision: 361330
URL: https://svnweb.freebsd.org/changeset/base/361330
Log:
libprocstat: fix reading of file descriptor table via kvm
This seems to have been broken since r247602 (from year 2013!).
Can be easily tested with
fstat -N /boot/kernel/kernel -M /var/crash/vmcore.last
MFC after:1 week
Sponsored by: Panzura
Modified:
head/lib/libprocstat/libprocstat.c
Modified: head/lib/libprocstat/libprocstat.c
==
--- head/lib/libprocstat/libprocstat.c Thu May 21 12:43:34 2020
(r361329)
+++ head/lib/libprocstat/libprocstat.c Thu May 21 13:46:30 2020
(r361330)
@@ -467,7 +467,7 @@ procstat_getfiles_kvm(struct procstat *procstat, struc
vm_map_entry_t entryp;
vm_object_t objp;
struct vnode *vp;
- struct file **ofiles;
+ struct filedescent *ofiles;
struct filestat *entry;
struct filestat_list *head;
kvm_t *kd;
@@ -554,25 +554,25 @@ procstat_getfiles_kvm(struct procstat *procstat, struc
}
nfiles = filed.fd_lastfile + 1;
- ofiles = malloc(nfiles * sizeof(struct file *));
+ ofiles = malloc(nfiles * sizeof(struct filedescent));
if (ofiles == NULL) {
- warn("malloc(%zu)", nfiles * sizeof(struct file *));
+ warn("malloc(%zu)", nfiles * sizeof(struct filedescent));
goto do_mmapped;
}
if (!kvm_read_all(kd, (unsigned long)filed.fd_ofiles, ofiles,
- nfiles * sizeof(struct file *))) {
+ nfiles * sizeof(struct filedescent))) {
warnx("cannot read file structures at %p",
(void *)filed.fd_ofiles);
free(ofiles);
goto do_mmapped;
}
for (i = 0; i <= filed.fd_lastfile; i++) {
- if (ofiles[i] == NULL)
+ if (ofiles[i].fde_file == NULL)
continue;
- if (!kvm_read_all(kd, (unsigned long)ofiles[i], ,
+ if (!kvm_read_all(kd, (unsigned long)ofiles[i].fde_file, ,
sizeof(struct file))) {
warnx("can't read file %d at %p", i,
- (void *)ofiles[i]);
+ (void *)ofiles[i].fde_file);
continue;
}
switch (file.f_type) {
___
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: r361303 - in head: lib/libc/gen libexec/rtld-elf sys/sys
On Thu, May 21, 2020 at 03:02:07PM +0200, Antoine Brodin wrote: > On Thu, May 21, 2020 at 12:08 AM Konstantin Belousov wrote: > > > > Author: kib > > Date: Wed May 20 22:08:26 2020 > > New Revision: 361303 > > URL: https://svnweb.freebsd.org/changeset/base/361303 > > > > Log: > > Change the samantic of struct link_map l_addr member. > > > > It previously returned the object map base address, while all other > > ELF operating systems return load offset, i.e. the difference between > > map base and the link base. > > > > Explain the meaning of the field in the man page. > > > > Stop filling the mips-only l_offs member, which is apparently unused. > > > > PR: 246561 > > Requested by: Damjan Jovanovic > > Reviewed by: emaste, jhb, cem (previous version) > > Sponsored by: The FreeBSD Foundation > > MFC after:1 week > > Differential revision:https://reviews.freebsd.org/D24918 > > > > Modified: > > head/lib/libc/gen/dlinfo.3 > > head/libexec/rtld-elf/rtld.c > > head/sys/sys/link_elf.h > > Hi, > > After this commit, some ports fail to build with signal 11. > For instance lang/perl5.30 fails to build with default options (DTRACE on) > Disabling the DTRACE option makes it able to build again. > I see, thank you for reporting. So drti.c:dtrace_dof_init() does read l_addr, and the dtrace code assumes that l_addr is the base, not relocbase. Mark, was dofhp_addr initialization changed comparing to Solaris ? ___ 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: r361275 - in head/sys: conf dev/hyperv/hvsock dev/hyperv/include dev/hyperv/vmbus modules/hyperv modules/hyperv/hvsock sys
On Thu, May 21, 2020 at 01:01:18PM +, Wei Hu wrote: > > -Original Message- > > From: Peter Holm > > Sent: Thursday, May 21, 2020 8:24 PM > > To: Wei Hu > > Cc: src-committ...@freebsd.org; svn-src-all@freebsd.org; svn-src- > > h...@freebsd.org > > Subject: Re: svn commit: r361275 - in head/sys: conf dev/hyperv/hvsock > > dev/hyperv/include dev/hyperv/vmbus modules/hyperv > > modules/hyperv/hvsock sys > > > > On Wed, May 20, 2020 at 11:03:59AM +, Wei Hu wrote: > > > Author: whu > > > Date: Wed May 20 11:03:59 2020 > > > New Revision: 361275 > > > URL: > > https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fsvnweb > > .freebsd.org%2Fchangeset%2Fbase%2F361275data=02%7C01%7Cweh% > > 40microsoft.com%7C61c524b5022b47b2c4e108d7fd81e75f%7C72f988bf86f14 > > 1af91ab2d7cd011db47%7C1%7C0%7C637256606689750658sdata=mw > > 4IXP3DnxICnK4U%2F8MzLbvMAzCuxih2f0waDyMSCTE%3Dreserved=0 > > > > > > Log: > > > HyperV socket implementation for FreeBSD > > > > > > This change adds Hyper-V socket feature in FreeBSD. New socket address > > > family AF_HYPERV and its kernel support are added. > > > > > > > Found this with a syscall fuzz test: > > > > panic: page fault > > cpuid = 2 > > time = 1590050529 > > KDB: stack backtrace: > > db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame > > 0xfe033d21d530 > > vpanic() at vpanic+0x182/frame 0xfe033d21d580 > > panic() at panic+0x43/frame 0xfe033d21d5e0 > > trap_fatal() at trap_fatal+0x387/frame 0xfe033d21d640 > > trap_pfault() at trap_pfault+0x99/frame 0xfe033d21d6a0 > > trap() at trap+0x2a5/frame 0xfe033d21d7b0 > > calltrap() at calltrap+0x8/frame 0xfe033d21d7b0 > > --- trap 0xc, rip = 0x80bcd3ba, rsp = 0xfe033d21d880, rbp = > > 0xfe033d21d910 --- > > _sx_xlock_hard() at _sx_xlock_hard+0x17a/frame 0xfe033d21d910 > > _sx_xlock() at _sx_xlock+0xba/frame 0xfe033d21d950 > > hvs_trans_close() at hvs_trans_close+0x42/frame 0xfe033d21d970 > > soclose() at soclose+0x161/frame 0xfe033d21d9e0 > > _fdrop() at _fdrop+0x1a/frame 0xfe033d21da00 > > closef() at closef+0x1db/frame 0xfe033d21da90 > > closefp() at closefp+0x96/frame 0xfe033d21dad0 > > amd64_syscall() at amd64_syscall+0x159/frame 0xfe033d21dbf0 > > fast_syscall_common() at fast_syscall_common+0x101/frame > > 0xfe033d21dbf0 > > --- syscall (6, FreeBSD ELF64, sys_close), rip = 0x8004283ca, rsp = > > 0x7fffe328, > > rbp = 0x7fffe460 --- > > > > https://nam06.safelinks.protection.outlook.com/?url=https:%2F%2Fpeople.free > > bsd.org%2F~pho%2Fstress%2Flog%2Fsetsockopt2- > > 2.txtdata=02%7C01%7Cweh%40microsoft.com%7C61c524b5022b47b2c > > 4e108d7fd81e75f%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C63 > > 7256606689750658sdata=RuBmWrBv7lGnhF2IHZ5NOP2rmV0c%2BJXuk > > RZl260KSIw%3Dreserved=0 > > > > Could this be yours? > > > Yes. Looks the lock was not initialized. The lock only gets initialized when > it is running > on HyperV. This type of socket only works on HyperV. > > How to reproduce it? Was it on HyperV? I am not sure how it can enter this > state. > > Wei The test is syscall() fuzzing, which typically flushes out missing parameter validations. This was *not* run on HyperV. You can find the test case here: https://svnweb.freebsd.org/base/user/pho/stress2/misc/setsockopt2.sh - Peter ___ 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: r361318 - head/bin/ls
On Thu, May 21, 2020 at 7:59 AM Rodney W. Grimes wrote: > > > Author: kevans > > Date: Thu May 21 03:50:56 2020 > > New Revision: 361318 > > URL: https://svnweb.freebsd.org/changeset/base/361318 > > > > Log: > > ls: fix a --color regression from r337956 > > > > The regression is in-fact that I flipped the default from never to auto. > > The > > incorrect impression was based on an alias that I failed to notice, > > installed by the Linux distribution that I used for testing compatibility > > here. Users that want the old default should be doing so with a shell > > alias > > as is done elsewhere, rather than making this decision in ls(1). > > > > Many thanks to rgrimes for pointing out the alias that I clearly > > overlooked > > that resulted in this; if you despised colors in your terminal from this, > > consider buying him a beer at the next venue that you see him at. > > Thanks Kyle, but this is likely to get me more rocks than beers :-) > FWIW- I received a not-insignificant number of valid complaints about the original default change because default color schemes are hard, and continued to receive an insignificant number of complaints as recent as a couple months ago (~2 years after the change?). This is considerably more flack than I expect to receive from other controversial proposals I've made and it was based on a fundamental misunderstanding of coreutils. Thanks, Kyle Evans ___ 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: r361303 - in head: lib/libc/gen libexec/rtld-elf sys/sys
On Thu, May 21, 2020 at 12:08 AM Konstantin Belousov wrote: > > Author: kib > Date: Wed May 20 22:08:26 2020 > New Revision: 361303 > URL: https://svnweb.freebsd.org/changeset/base/361303 > > Log: > Change the samantic of struct link_map l_addr member. > > It previously returned the object map base address, while all other > ELF operating systems return load offset, i.e. the difference between > map base and the link base. > > Explain the meaning of the field in the man page. > > Stop filling the mips-only l_offs member, which is apparently unused. > > PR: 246561 > Requested by: Damjan Jovanovic > Reviewed by: emaste, jhb, cem (previous version) > Sponsored by: The FreeBSD Foundation > MFC after:1 week > Differential revision:https://reviews.freebsd.org/D24918 > > Modified: > head/lib/libc/gen/dlinfo.3 > head/libexec/rtld-elf/rtld.c > head/sys/sys/link_elf.h Hi, After this commit, some ports fail to build with signal 11. For instance lang/perl5.30 fails to build with default options (DTRACE on) Disabling the DTRACE option makes it able to build again. Cheers, Antoine ___ 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: r361275 - in head/sys: conf dev/hyperv/hvsock dev/hyperv/include dev/hyperv/vmbus modules/hyperv modules/hyperv/hvsock sys
> -Original Message- > From: Peter Holm > Sent: Thursday, May 21, 2020 8:24 PM > To: Wei Hu > Cc: src-committ...@freebsd.org; svn-src-all@freebsd.org; svn-src- > h...@freebsd.org > Subject: Re: svn commit: r361275 - in head/sys: conf dev/hyperv/hvsock > dev/hyperv/include dev/hyperv/vmbus modules/hyperv > modules/hyperv/hvsock sys > > On Wed, May 20, 2020 at 11:03:59AM +, Wei Hu wrote: > > Author: whu > > Date: Wed May 20 11:03:59 2020 > > New Revision: 361275 > > URL: > https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fsvnweb > .freebsd.org%2Fchangeset%2Fbase%2F361275data=02%7C01%7Cweh% > 40microsoft.com%7C61c524b5022b47b2c4e108d7fd81e75f%7C72f988bf86f14 > 1af91ab2d7cd011db47%7C1%7C0%7C637256606689750658sdata=mw > 4IXP3DnxICnK4U%2F8MzLbvMAzCuxih2f0waDyMSCTE%3Dreserved=0 > > > > Log: > > HyperV socket implementation for FreeBSD > > > > This change adds Hyper-V socket feature in FreeBSD. New socket address > > family AF_HYPERV and its kernel support are added. > > > > Found this with a syscall fuzz test: > > panic: page fault > cpuid = 2 > time = 1590050529 > KDB: stack backtrace: > db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame > 0xfe033d21d530 > vpanic() at vpanic+0x182/frame 0xfe033d21d580 > panic() at panic+0x43/frame 0xfe033d21d5e0 > trap_fatal() at trap_fatal+0x387/frame 0xfe033d21d640 > trap_pfault() at trap_pfault+0x99/frame 0xfe033d21d6a0 > trap() at trap+0x2a5/frame 0xfe033d21d7b0 > calltrap() at calltrap+0x8/frame 0xfe033d21d7b0 > --- trap 0xc, rip = 0x80bcd3ba, rsp = 0xfe033d21d880, rbp = > 0xfe033d21d910 --- > _sx_xlock_hard() at _sx_xlock_hard+0x17a/frame 0xfe033d21d910 > _sx_xlock() at _sx_xlock+0xba/frame 0xfe033d21d950 > hvs_trans_close() at hvs_trans_close+0x42/frame 0xfe033d21d970 > soclose() at soclose+0x161/frame 0xfe033d21d9e0 > _fdrop() at _fdrop+0x1a/frame 0xfe033d21da00 > closef() at closef+0x1db/frame 0xfe033d21da90 > closefp() at closefp+0x96/frame 0xfe033d21dad0 > amd64_syscall() at amd64_syscall+0x159/frame 0xfe033d21dbf0 > fast_syscall_common() at fast_syscall_common+0x101/frame > 0xfe033d21dbf0 > --- syscall (6, FreeBSD ELF64, sys_close), rip = 0x8004283ca, rsp = > 0x7fffe328, > rbp = 0x7fffe460 --- > > https://nam06.safelinks.protection.outlook.com/?url=https:%2F%2Fpeople.free > bsd.org%2F~pho%2Fstress%2Flog%2Fsetsockopt2- > 2.txtdata=02%7C01%7Cweh%40microsoft.com%7C61c524b5022b47b2c > 4e108d7fd81e75f%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C63 > 7256606689750658sdata=RuBmWrBv7lGnhF2IHZ5NOP2rmV0c%2BJXuk > RZl260KSIw%3Dreserved=0 > > Could this be yours? Yes. Looks the lock was not initialized. The lock only gets initialized when it is running on HyperV. This type of socket only works on HyperV. How to reproduce it? Was it on HyperV? I am not sure how it can enter this state. Wei ___ 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: r361318 - head/bin/ls
> Author: kevans > Date: Thu May 21 03:50:56 2020 > New Revision: 361318 > URL: https://svnweb.freebsd.org/changeset/base/361318 > > Log: > ls: fix a --color regression from r337956 > > The regression is in-fact that I flipped the default from never to auto. The > incorrect impression was based on an alias that I failed to notice, > installed by the Linux distribution that I used for testing compatibility > here. Users that want the old default should be doing so with a shell alias > as is done elsewhere, rather than making this decision in ls(1). > > Many thanks to rgrimes for pointing out the alias that I clearly overlooked > that resulted in this; if you despised colors in your terminal from this, > consider buying him a beer at the next venue that you see him at. Thanks Kyle, but this is likely to get me more rocks than beers :-) > > MFC after: 1 week > Relnotes: yes > > Modified: > head/bin/ls/ls.1 > head/bin/ls/ls.c > > Modified: head/bin/ls/ls.1 > == > --- head/bin/ls/ls.1 Thu May 21 03:33:20 2020(r361317) > +++ head/bin/ls/ls.1 Thu May 21 03:50:56 2020(r361318) > @@ -32,7 +32,7 @@ > .\" @(#)ls.1 8.7 (Berkeley) 7/29/94 > .\" $FreeBSD$ > .\" > -.Dd August 18, 2018 > +.Dd May 20, 2020 > .Dt LS 1 > .Os > .Sh NAME > @@ -216,8 +216,8 @@ Output colored escape sequences based on > .Ar when , > which may be set to either > .Cm always , > -.Cm auto > -(default), or > +.Cm auto , > +or > .Cm never . > .Pp > .Cm always > @@ -252,6 +252,12 @@ environment variable is set and not empty. > .Pp > .Cm never > will disable color regardless of environment variables. > +.Cm never > +is the default when neither > +.Fl -color > +nor > +.Fl G > +is specified. > .Pp > For compatibility with GNU coreutils, > .Nm > > Modified: head/bin/ls/ls.c > == > --- head/bin/ls/ls.c Thu May 21 03:33:20 2020(r361317) > +++ head/bin/ls/ls.c Thu May 21 03:50:56 2020(r361318) > @@ -152,7 +152,7 @@ static int f_timesort;/* sort by time vice > name */ > int f_type; /* add type character for non-regular files */ > static int f_whiteout; /* show whiteout entries */ > #ifdef COLORLS > - int colorflag = COLORFLAG_AUTO; /* passed in colorflag > */ > + int colorflag = COLORFLAG_NEVER; /* passed in colorflag > */ > int f_color; /* add type in color for non-regular files */ > bool explicitansi;/* Explicit ANSI sequences, no termcap(5) */ > char *ansi_bgcol;/* ANSI sequence to set background colour */ > -- Rod Grimes rgri...@freebsd.org ___ 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: r361329 - releng/11.4/sys/kern
Author: kib
Date: Thu May 21 12:43:34 2020
New Revision: 361329
URL: https://svnweb.freebsd.org/changeset/base/361329
Log:
MFC r361037, r361056:
Fix spurious ENOTCONN from closed unix domain socket other' side.
Approved by: re (gjb)
Modified:
releng/11.4/sys/kern/uipc_socket.c
Directory Properties:
releng/11.4/ (props changed)
Modified: releng/11.4/sys/kern/uipc_socket.c
==
--- releng/11.4/sys/kern/uipc_socket.c Thu May 21 11:14:13 2020
(r361328)
+++ releng/11.4/sys/kern/uipc_socket.c Thu May 21 12:43:34 2020
(r361329)
@@ -1565,8 +1565,9 @@ restart:
m = so->so_rcv.sb_mb;
goto dontblock;
}
- if ((so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING)) == 0 &&
- (so->so_proto->pr_flags & PR_CONNREQUIRED)) {
+ if ((so->so_state & (SS_ISCONNECTING | SS_ISCONNECTED |
+ SS_ISDISCONNECTING | SS_ISDISCONNECTED)) == 0 &&
+ (so->so_proto->pr_flags & PR_CONNREQUIRED) != 0) {
SOCKBUF_UNLOCK(>so_rcv);
error = ENOTCONN;
goto release;
@@ -3516,8 +3517,17 @@ soisdisconnected(struct socket *so)
* SOCKBUF_LOCK(>so_rcv) are the same.
*/
SOCKBUF_LOCK(>so_rcv);
- so->so_state &= ~(SS_ISCONNECTING|SS_ISCONNECTED|SS_ISDISCONNECTING);
+
+ /*
+* There is at least one reader of so_state that does not
+* acquire socket lock, namely soreceive_generic(). Ensure
+* that it never sees all flags that track connection status
+* cleared, by ordering the update with a barrier semantic of
+* our release thread fence.
+*/
so->so_state |= SS_ISDISCONNECTED;
+ atomic_thread_fence_rel();
+ so->so_state &= ~(SS_ISCONNECTING|SS_ISCONNECTED|SS_ISDISCONNECTING);
socantrcvmore_locked(so);
SOCKBUF_LOCK(>so_snd);
sbdrop_locked(>so_snd, sbused(>so_snd));
___
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: r361275 - in head/sys: conf dev/hyperv/hvsock dev/hyperv/include dev/hyperv/vmbus modules/hyperv modules/hyperv/hvsock sys
On Wed, May 20, 2020 at 11:03:59AM +, Wei Hu wrote: > Author: whu > Date: Wed May 20 11:03:59 2020 > New Revision: 361275 > URL: https://svnweb.freebsd.org/changeset/base/361275 > > Log: > HyperV socket implementation for FreeBSD > > This change adds Hyper-V socket feature in FreeBSD. New socket address > family AF_HYPERV and its kernel support are added. > Found this with a syscall fuzz test: panic: page fault cpuid = 2 time = 1590050529 KDB: stack backtrace: db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame 0xfe033d21d530 vpanic() at vpanic+0x182/frame 0xfe033d21d580 panic() at panic+0x43/frame 0xfe033d21d5e0 trap_fatal() at trap_fatal+0x387/frame 0xfe033d21d640 trap_pfault() at trap_pfault+0x99/frame 0xfe033d21d6a0 trap() at trap+0x2a5/frame 0xfe033d21d7b0 calltrap() at calltrap+0x8/frame 0xfe033d21d7b0 --- trap 0xc, rip = 0x80bcd3ba, rsp = 0xfe033d21d880, rbp = 0xfe033d21d910 --- _sx_xlock_hard() at _sx_xlock_hard+0x17a/frame 0xfe033d21d910 _sx_xlock() at _sx_xlock+0xba/frame 0xfe033d21d950 hvs_trans_close() at hvs_trans_close+0x42/frame 0xfe033d21d970 soclose() at soclose+0x161/frame 0xfe033d21d9e0 _fdrop() at _fdrop+0x1a/frame 0xfe033d21da00 closef() at closef+0x1db/frame 0xfe033d21da90 closefp() at closefp+0x96/frame 0xfe033d21dad0 amd64_syscall() at amd64_syscall+0x159/frame 0xfe033d21dbf0 fast_syscall_common() at fast_syscall_common+0x101/frame 0xfe033d21dbf0 --- syscall (6, FreeBSD ELF64, sys_close), rip = 0x8004283ca, rsp = 0x7fffe328, rbp = 0x7fffe460 --- https://people.freebsd.org/~pho/stress/log/setsockopt2-2.txt Could this be yours? - Peter ___ 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"
Hot sale: 18W PD Charger ---From Andar Technology YolandaTung
___ 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: r361328 - stable/11/sys/kern
Author: kib
Date: Thu May 21 11:14:13 2020
New Revision: 361328
URL: https://svnweb.freebsd.org/changeset/base/361328
Log:
MFC r361037, r361056:
Fix spurious ENOTCONN from closed unix domain socket other' side.
Modified:
stable/11/sys/kern/uipc_socket.c
Directory Properties:
stable/11/ (props changed)
Modified: stable/11/sys/kern/uipc_socket.c
==
--- stable/11/sys/kern/uipc_socket.cThu May 21 11:12:27 2020
(r361327)
+++ stable/11/sys/kern/uipc_socket.cThu May 21 11:14:13 2020
(r361328)
@@ -1565,8 +1565,9 @@ restart:
m = so->so_rcv.sb_mb;
goto dontblock;
}
- if ((so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING)) == 0 &&
- (so->so_proto->pr_flags & PR_CONNREQUIRED)) {
+ if ((so->so_state & (SS_ISCONNECTING | SS_ISCONNECTED |
+ SS_ISDISCONNECTING | SS_ISDISCONNECTED)) == 0 &&
+ (so->so_proto->pr_flags & PR_CONNREQUIRED) != 0) {
SOCKBUF_UNLOCK(>so_rcv);
error = ENOTCONN;
goto release;
@@ -3516,8 +3517,17 @@ soisdisconnected(struct socket *so)
* SOCKBUF_LOCK(>so_rcv) are the same.
*/
SOCKBUF_LOCK(>so_rcv);
- so->so_state &= ~(SS_ISCONNECTING|SS_ISCONNECTED|SS_ISDISCONNECTING);
+
+ /*
+* There is at least one reader of so_state that does not
+* acquire socket lock, namely soreceive_generic(). Ensure
+* that it never sees all flags that track connection status
+* cleared, by ordering the update with a barrier semantic of
+* our release thread fence.
+*/
so->so_state |= SS_ISDISCONNECTED;
+ atomic_thread_fence_rel();
+ so->so_state &= ~(SS_ISCONNECTING|SS_ISCONNECTED|SS_ISDISCONNECTING);
socantrcvmore_locked(so);
SOCKBUF_LOCK(>so_snd);
sbdrop_locked(>so_snd, sbused(>so_snd));
___
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: r361327 - stable/12/sys/kern
Author: kib
Date: Thu May 21 11:12:27 2020
New Revision: 361327
URL: https://svnweb.freebsd.org/changeset/base/361327
Log:
MFC r361037, r361056:
Fix spurious ENOTCONN from closed unix domain socket other' side.
Modified:
stable/12/sys/kern/uipc_socket.c
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/sys/kern/uipc_socket.c
==
--- stable/12/sys/kern/uipc_socket.cThu May 21 06:40:51 2020
(r361326)
+++ stable/12/sys/kern/uipc_socket.cThu May 21 11:12:27 2020
(r361327)
@@ -1792,8 +1792,9 @@ restart:
m = so->so_rcv.sb_mb;
goto dontblock;
}
- if ((so->so_state & (SS_ISCONNECTED|SS_ISCONNECTING)) == 0 &&
- (so->so_proto->pr_flags & PR_CONNREQUIRED)) {
+ if ((so->so_state & (SS_ISCONNECTING | SS_ISCONNECTED |
+ SS_ISDISCONNECTING | SS_ISDISCONNECTED)) == 0 &&
+ (so->so_proto->pr_flags & PR_CONNREQUIRED) != 0) {
SOCKBUF_UNLOCK(>so_rcv);
error = ENOTCONN;
goto release;
@@ -3814,8 +3815,17 @@ soisdisconnected(struct socket *so)
{
SOCK_LOCK(so);
- so->so_state &= ~(SS_ISCONNECTING|SS_ISCONNECTED|SS_ISDISCONNECTING);
+
+ /*
+* There is at least one reader of so_state that does not
+* acquire socket lock, namely soreceive_generic(). Ensure
+* that it never sees all flags that track connection status
+* cleared, by ordering the update with a barrier semantic of
+* our release thread fence.
+*/
so->so_state |= SS_ISDISCONNECTED;
+ atomic_thread_fence_rel();
+ so->so_state &= ~(SS_ISCONNECTING|SS_ISCONNECTED|SS_ISDISCONNECTING);
if (!SOLISTENING(so)) {
SOCK_UNLOCK(so);
___
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: r361326 - head/sys/arm/xilinx
Author: jmg
Date: Thu May 21 06:40:51 2020
New Revision: 361326
URL: https://svnweb.freebsd.org/changeset/base/361326
Log:
Bring in support for single core Zynq devices. Turns out that real
hardware, the registers appear like there's two cores, but the second
core does not work, so base the number of cores upon the chip id.
Tested on a XC7Z007S.
also, previous commit was suppose to be D14429.
Submitted by: Thomas Skibo
Differential Revision: https://reviews.freebsd.org/D14429
Modified:
head/sys/arm/xilinx/zy7_mp.c
head/sys/arm/xilinx/zy7_slcr.h
Modified: head/sys/arm/xilinx/zy7_mp.c
==
--- head/sys/arm/xilinx/zy7_mp.cThu May 21 06:17:54 2020
(r361325)
+++ head/sys/arm/xilinx/zy7_mp.cThu May 21 06:40:51 2020
(r361326)
@@ -44,18 +44,59 @@ __FBSDID("$FreeBSD$");
#include
#include
+#include
-#defineZYNQ7_CPU1_ENTRY0xfff0
+#defineZYNQ7_CPU1_ENTRY0xfff0
-#defineSCU_CONTROL_REG 0xf8f0
-#define SCU_CONTROL_ENABLE (1 << 0)
+#defineSCU_CONTROL_REG 0xf8f0
+#define SCU_CONTROL_ENABLE 1
+#defineSCU_CONFIG_REG 0xf8f4
+#define SCU_CONFIG_N_CPUS_MASK 3
+#define SLCR_PSS_IDCODE0xf8000530
+
void
zynq7_mp_setmaxid(platform_t plat)
{
+ bus_space_handle_t slcr_handle;
+ int device_id;
+ bus_space_handle_t scu_handle;
- mp_maxid = 1;
- mp_ncpus = 2;
+ if (mp_ncpus != 0)
+ return;
+
+ /* Map in SLCR PSS_IDCODE register. */
+ if (bus_space_map(fdtbus_bs_tag, SLCR_PSS_IDCODE, 4, 0,
+ _handle) != 0)
+ panic("%s: Could not map SLCR IDCODE reg.\n", __func__);
+
+ device_id = bus_space_read_4(fdtbus_bs_tag, slcr_handle, 0) &
+ ZY7_SLCR_PSS_IDCODE_DEVICE_MASK;
+
+ bus_space_unmap(fdtbus_bs_tag, slcr_handle, 4);
+
+ /*
+* Zynq XC7z0xxS single core chips indicate incorrect number of CPUs in
+* SCU configuration register.
+*/
+ if (device_id == ZY7_SLCR_PSS_IDCODE_DEVICE_7Z007S ||
+ device_id == ZY7_SLCR_PSS_IDCODE_DEVICE_7Z012S ||
+ device_id == ZY7_SLCR_PSS_IDCODE_DEVICE_7Z014S) {
+ mp_maxid = 0;
+ mp_ncpus = 1;
+ return;
+ }
+
+ /* Map in SCU config register. */
+ if (bus_space_map(fdtbus_bs_tag, SCU_CONFIG_REG, 4, 0,
+ _handle) != 0)
+ panic("zynq7_mp_setmaxid: Could not map SCU config reg.\n");
+
+ mp_maxid = bus_space_read_4(fdtbus_bs_tag, scu_handle, 0) &
+ SCU_CONFIG_N_CPUS_MASK;
+ mp_ncpus = mp_maxid + 1;
+
+ bus_space_unmap(fdtbus_bs_tag, scu_handle, 4);
}
void
Modified: head/sys/arm/xilinx/zy7_slcr.h
==
--- head/sys/arm/xilinx/zy7_slcr.h Thu May 21 06:17:54 2020
(r361325)
+++ head/sys/arm/xilinx/zy7_slcr.h Thu May 21 06:40:51 2020
(r361326)
@@ -53,91 +53,91 @@
#define ZY7_SLCR_ARM_PLL_CTRL 0x0100
#define ZY7_SLCR_DDR_PLL_CTRL 0x0104
#define ZY7_SLCR_IO_PLL_CTRL 0x0108
-#define ZY7_SLCR_PLL_CTRL_RESET (1<<0)
-#define ZY7_SLCR_PLL_CTRL_PWRDWN (1<<1)
-#define ZY7_SLCR_PLL_CTRL_BYPASS_QUAL(1<<3)
-#define ZY7_SLCR_PLL_CTRL_BYPASS_FORCE (1<<4)
+#define ZY7_SLCR_PLL_CTRL_RESET (1 << 0)
+#define ZY7_SLCR_PLL_CTRL_PWRDWN (1 << 1)
+#define ZY7_SLCR_PLL_CTRL_BYPASS_QUAL(1 << 3)
+#define ZY7_SLCR_PLL_CTRL_BYPASS_FORCE (1 << 4)
#define ZY7_SLCR_PLL_CTRL_FDIV_SHIFT 12
-#define ZY7_SLCR_PLL_CTRL_FDIV_MASK (0x7f<<12)
+#define ZY7_SLCR_PLL_CTRL_FDIV_MASK (0x7f << 12)
#define ZY7_SLCR_PLL_STATUS0x010c
-#define ZY7_SLCR_PLL_STAT_ARM_PLL_LOCK (1<<0)
-#define ZY7_SLCR_PLL_STAT_DDR_PLL_LOCK (1<<1)
-#define ZY7_SLCR_PLL_STAT_IO_PLL_LOCK(1<<2)
-#define ZY7_SLCR_PLL_STAT_ARM_PLL_STABLE (1<<3)
-#define ZY7_SLCR_PLL_STAT_DDR_PLL_STABLE (1<<4)
-#define ZY7_SLCR_PLL_STAT_IO_PLL_STABLE (1<<5)
+#define ZY7_SLCR_PLL_STAT_ARM_PLL_LOCK (1 << 0)
+#define ZY7_SLCR_PLL_STAT_DDR_PLL_LOCK (1 << 1)
+#define ZY7_SLCR_PLL_STAT_IO_PLL_LOCK(1 << 2)
+#define ZY7_SLCR_PLL_STAT_ARM_PLL_STABLE (1 << 3)
+#define ZY7_SLCR_PLL_STAT_DDR_PLL_STABLE (1 << 4)
+#define ZY7_SLCR_PLL_STAT_IO_PLL_STABLE (1 << 5)
#define ZY7_SLCR_ARM_PLL_CFG 0x0110
#define ZY7_SLCR_DDR_PLL_CFG
svn commit: r361325 - head/sys/arm/xilinx
Author: jmg
Date: Thu May 21 06:17:54 2020
New Revision: 361325
URL: https://svnweb.freebsd.org/changeset/base/361325
Log:
minor cleanup of white space, and function name in panic...
This is a partial commit of the review.
Submitted by: Thomas Skibo
Differential Revision: https://reviews.freebsd.org/D23319
Reviewed by: andrew
Modified:
head/sys/arm/xilinx/zy7_mp.c
Modified: head/sys/arm/xilinx/zy7_mp.c
==
--- head/sys/arm/xilinx/zy7_mp.cThu May 21 05:34:02 2020
(r361324)
+++ head/sys/arm/xilinx/zy7_mp.cThu May 21 06:17:54 2020
(r361325)
@@ -58,7 +58,7 @@ zynq7_mp_setmaxid(platform_t plat)
mp_ncpus = 2;
}
-void
+void
zynq7_mp_start_ap(platform_t plat)
{
bus_space_handle_t scu_handle;
@@ -67,8 +67,8 @@ zynq7_mp_start_ap(platform_t plat)
/* Map in SCU control register. */
if (bus_space_map(fdtbus_bs_tag, SCU_CONTROL_REG, 4,
- 0, _handle) != 0)
- panic("platform_mp_start_ap: Couldn't map SCU config reg\n");
+ 0, _handle) != 0)
+ panic("%s: Could not map SCU control reg.\n", __func__);
/* Set SCU enable bit. */
scu_ctrl = bus_space_read_4(fdtbus_bs_tag, scu_handle, 0);
@@ -80,7 +80,7 @@ zynq7_mp_start_ap(platform_t plat)
/* Map in magic location to give entry address to CPU1. */
if (bus_space_map(fdtbus_bs_tag, ZYNQ7_CPU1_ENTRY, 4,
0, _handle) != 0)
- panic("platform_mp_start_ap: Couldn't map OCM\n");
+ panic("%s: Could not map OCM\n", __func__);
/* Write start address for CPU1. */
bus_space_write_4(fdtbus_bs_tag, ocm_handle, 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"
