Off by default, build behaves normally.
WITH_META_MODE we get auto objdir creation, the ability to
start build from anywhere in the tree.
Still need to add real targets under targets/ to build packages.
Differential Revision: D2796
Reviewed by: brooks imp
- Stateful TCP offload drivers for Terminator 3 and 4 (T3 and T4) ASICs.
These are available as t3_tom and t4_tom modules that augment cxgb(4)
and cxgbe(4) respectively. The cxgb/cxgbe drivers continue to work as
usual with or without these extra features.
- iWARP driver for Terminator 3 ASIC (kernel verbs). T4 iWARP in the
works and will follow soon.
Build-tested with make universe.
30s overview
============
What interfaces support TCP offload? Look for TOE4 and/or TOE6 in the
capabilities of an interface:
# ifconfig -m | grep TOE
Enable/disable TCP offload on an interface (just like any other ifnet
capability):
# ifconfig cxgbe0 toe
# ifconfig cxgbe0 -toe
Which connections are offloaded? Look for toe4 and/or toe6 in the
output of netstat and sockstat:
# netstat -np tcp | grep toe
# sockstat -46c | grep toe
Reviewed by: bz, gnn
Sponsored by: Chelsio communications.
MFC after: ~3 months (after 9.1, and after ensuring MFC is feasible)
checking the returned oldlen: when ENOMEM is due to the supplied
buffer being too short the return oldlen is equal to buffer size.
Without this additional check sockstat gets stuck in loop leaking the
memory if the returned ENOMEM was due the exceeded memorylocked
limit. This is easily can be observed running `limits -l 1k sockstat'.
Submitted by: Andrey Zonov <andrey zonov org>
MFC after: 1 week
Sockets not associated with a file descriptor include TCP TIME_WAIT states
and sockets created via the socket(9) API such as from rpc.lockd and the NFS
client.
PR: bin/164081
MFC after: 2 weeks
No objection: des
This makes a tiny percentage of entries in calendars ugly for latin1
users, but fixes them for UTF-8 users.
This badly needs a solution involving locale-dependent re-encoding.
addresses (most of them apart from ::1): put a whitespace
between local and remote address:port pairs.
PR: bin/145194
Submitted by: Fedor Dikarev
MFC after: 2 weeks
in6p_ip6_nxt
in6p_vflag
in6p_flags
in6p_socket
in6p_lport
in6p_fport
in6p_ppcb
Remove unused v6 macro aliases for inpcb flags:
IN6P_HIGHPORT
IN6P_LOWPORT
IN6P_ANONPORT
IN6P_RECVIF
IN6P_MTUDISC
IN6P_FAITH
IN6P_CONTROLOPTS
References to in6p_lport and in6_fport in sockstat are also replaced with
normal inp_lport and inp_fport references.
MFC after: 3 days
Reviewed by: bz
Using either one of the two would result in an empty protos[]
array, and no sockets were actually listed:
% sockstat -4
USER COMMAND PID FD PROTO LOCAL ADDRESS FOREIGN ADDRESS
% sockstat -6
USER COMMAND PID FD PROTO LOCAL ADDRESS FOREIGN ADDRESS
%
Fix this bug by tweaking appropriately the logic of handling opt_4,
opt_6, opt_u and protos_defined.
Submitted by: des
Pointy hat: keramida
behavior of sockstat(1) will still be to show "udp", "tcp" and
"divert" protocols, but we can now provide a (comma-separated)
list of protocols, as in:
% sockstat -P tcp
to list only TCP sockets, or we can filter more than one protocol
by separating the protocol names with a comma:
% sockstat -P tcp,udp
Protocol names are parsed with getprotobyname(3), so any protocol
whose name is listed in `/etc/protocols' should work fine.
Submitted by: Josh Carroll <josh.carroll@psualum.com>
Approved by: des
states that no longer have a corresponding file descriptor - until now,
sockstat would mostly randomly match null kern.file.*.xf_data fields
with the first mostly-closed socket.
This bugfix is a RELENG_5 candidate.
Approved by: andre
65536 / (sizeof(int) * CHAR_BITS) `int's instead of
65536 / (sizeof(int) * CHAR_BITS) bytes to avoid a possible
segmentation fault if ports above 16383 are specified via the
-p option on a platform with 4 byte wide ints.
Approved by: re (bmah)
Reported by: Marco Wertejuk <wertejuk@mwcis.com>
pointer types, and remove a huge number of casts from code using it.
Change struct xfile xf_data to xun_data (ABI is still compatible).
If we need to add a #define for f_data and xf_data we can, but I don't
think it will be necessary. There are no operational changes in this
commit.
wrote, and as such was not very pretty.
Changes that may cause problems for people who use sockstat in scripts:
- sockstat(8) now displays Unix domain sockets in addition to IPv4 and IPv6
- the last period in local and foreign addresses is changed to a colon to
make the port number easier to spot
- IPv4 and IPv6 sockets are listed separately (IPv4 first, then IPv6, then
Unix)
Sockstat use netstat tcp/udp socket print, and fstat tcp/udp
socket print, but it just specify all of IPPROTO_IP and
IPPROTO_IPV6 sockets for fstat.
Now IPv6 socket also use raw socket, but only netstat don't print it,
so now they are printed as empty entries in sockstat output.
Approved by: jkh
