Date: Sun, 9 Aug 2015 12:28:22 +0200 From: Kristof Provost <kp@FreeBSD.org> To: freebsd-current@freebsd.org Cc: Gleb Smirnoff <glebius@freebsd.org> Subject: sysctl -a panic on VIMAGE kernels Message-ID: <B09833EB-ABE0-44BF-A53A-554207B92637@FreeBSD.org>
next in thread | raw e-mail | index | archive | help
Hi,
I=E2=80=99ve run into a reproducible panic on a VIMAGE kernel with =
=E2=80=98sysctl -a=E2=80=99.
Relevant backtrace bits:
#8 0xffffffff80e7dd28 in trap (frame=3D0xfffffe01f16b26a0)
at /usr/src/sys/amd64/amd64/trap.c:426
#9 0xffffffff80e5e6a2 in calltrap ()
at /usr/src/sys/amd64/amd64/exception.S:235
#10 0xffffffff80cea67d in uma_zone_get_cur (zone=3D0x0)
at /usr/src/sys/vm/uma_core.c:3006
#11 0xffffffff80cec029 in sysctl_handle_uma_zone_cur (
oidp=3D0xffffffff818a7c90, arg1=3D0xfffffe00010c0438, arg2=3D0,
req=3D0xfffffe01f16b2868) at /usr/src/sys/vm/uma_core.c:3580
#12 0xffffffff80a28614 in sysctl_root_handler_locked =
(oid=3D0xffffffff818a7c90,
arg1=3D0xfffffe00010c0438, arg2=3D0, req=3D0xfffffe01f16b2868)
at /usr/src/sys/kern/kern_sysctl.c:183
#13 0xffffffff80a27d70 in sysctl_root (arg1=3D<value optimized out>,
arg2=3D<value optimized out>) at =
/usr/src/sys/kern/kern_sysctl.c:1694
#14 0xffffffff80a28372 in userland_sysctl (td=3D0x0, =
name=3D0xfffffe01f16b2930,
namelen=3D<value optimized out>, old=3D<value optimized out>,
oldlenp=3D<value optimized out>, inkernel=3D<value optimized out>,
new=3D<value optimized out>, newlen=3D<value optimized out>,
retval=3D<value optimized out>, flags=3D0)
at /usr/src/sys/kern/kern_sysctl.c:1798
#15 0xffffffff80a28144 in sys___sysctl (td=3D0xfffff8000b1e49a0,
uap=3D0xfffffe01f16b2a40) at /usr/src/sys/kern/kern_sysctl.c:1724
In essence, what happens is that we end up in =
sysctl_handle_uma_zone_cur() and arg1 is a pointer to NULL,=20
so we call uma_zone_get_cur(zone); with zone =3D=3D NULL.
There=E2=80=99s been a bit of churn around tcp_reass_zone, and I think =
the latest version is wrong.
It marks the sysctl as CTLFLAG_VNET, but the exposed variable is not =
VNET_DEFINE().
The following fixes it for me:
diff --git a/sys/netinet/tcp_reass.c b/sys/netinet/tcp_reass.c
index 77d8940..3913ef3 100644
--- a/sys/netinet/tcp_reass.c
+++ b/sys/netinet/tcp_reass.c
@@ -84,7 +84,7 @@ SYSCTL_INT(_net_inet_tcp_reass, OID_AUTO, maxsegments, =
CTLFLAG_RDTUN,
"Global maximum number of TCP Segments in Reassembly Queue");
static uma_zone_t tcp_reass_zone;
-SYSCTL_UMA_CUR(_net_inet_tcp_reass, OID_AUTO, cursegments, =
CTLFLAG_VNET,
+SYSCTL_UMA_CUR(_net_inet_tcp_reass, OID_AUTO, cursegments, 0,
&tcp_reass_zone,
"Global number of TCP Segments currently in Reassembly Queue=E2=80=9D=
);
Regards,
Kristof=
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?B09833EB-ABE0-44BF-A53A-554207B92637>