- More flexible cluster size selection, including the ability to fall
back to a safe cluster size (PAGE_SIZE from zone_jumbop by default) in
case an allocation of a larger size fails.
- A single get_fl_payload() function that assembles the payload into an
mbuf chain for any kind of freelist. This replaces two variants: one
for freelists with buffer packing enabled and another for those without.
- Buffer packing with any sized cluster. It was limited to 4K clusters
only before this change.
- Enable buffer packing for TOE rx queues as well.
- Statistics and tunables to go with all these changes. The driver's
man page will be updated separately.
MFC after: 5 weeks
hw.cxgbe.rsrv_noflow. When set, queue 0 of the port is reserved for
TX packets without a flowid. The hash value of packets with a flowid
is bumped up by 1. The intent is to provide a private queue for
link-level packets like LACP that is unlikely to overflow or suffer
deep queue latency.
Reviewed by: np
Obtained from: Netflix
MFC after: 3 days
in6addr_any and is not in the CLIP table either. This fixes a reported
TOE+IPv6 NULL-dereference panic in do_pass_open_rpl().
While here, stop creating hardware servers for any loopback address.
It's just a waste of server tids.
MFC after: 1 week
them up as part of firmware initialization (which the driver gets to do
only if it's the master driver).
Read the range of tids available for the ETHOFLD functionality if it's
enabled.
New is_ftid() and is_etid() functions to test whether a tid falls within
the range of filter tids or ETHOFLD tids respectively.
MFC after: 2 weeks
scheduling classes in the chip and to bind tx queue(s) to a scheduling
class respectively. These can be used for various kinds of tx traffic
throttling (to force selected tx queues to drain at a fixed Kbps rate,
or a % of the port's total bandwidth, or at a fixed pps rate, etc.).
Obtained from: Chelsio
to this event, adding if_var.h to files that do need it. Also, include
all includes that now are included due to implicit pollution via if_var.h
Sponsored by: Netflix
Sponsored by: Nginx, Inc.
and there are ifnets, that do that via counter(9). Provide a flag that
would skip cache line trashing '+=' operation in ether_input().
Sponsored by: Netflix
Sponsored by: Nginx, Inc.
Reviewed by: melifaro, adrian
Approved by: re (marius)
set to 15 to indicate that the peer did not send a window scale option
with its SYN. Do not send a window scale option in the SYN|ACK reply
in that case.
cards.
This is a T4 and T5 chip feature which lets the chip deliver multiple
Ethernet frames in a single buffer. This is more efficient within the
chip, in the driver, and reduces wastage of space in rx buffers.
- Always allocate rx buffers from the jumbop zone, no matter what the
MTU is. Do not use the normal cluster refcounting mechanism.
- Reserve space for an mbuf and a refcount in the cluster itself and let
the chip DMA multiple frames in the rest.
- Use the embedded mbuf for the first frame and allocate mbufs on the
fly for any additional frames delivered in the cluster. Each of these
mbufs has a reference on the underlying cluster.
- tom_uninit had to be reworked not to hold the adapter lock (a mutex)
around t4_deactivate_uld, which acquires the uld_list_lock.
- the ifc_match for the interface cloner that creates the tracer ifnet
had to be reworked as the kernel calls ifc_match with the global
if_cloners_mtx held.
all T4 and T5 based cards and is useful for analyzing TSO, LRO, TOE, and
for general purpose monitoring without tapping any cxgbe or cxl ifnet
directly.
Tracers on the T4/T5 chips provide access to Ethernet frames exactly as
they were received from or transmitted on the wire. On transmit, a
tracer will capture a frame after TSO segmentation, hw VLAN tag
insertion, hw L3 & L4 checksum insertion, etc. It will also capture
frames generated by the TCP offload engine (TOE traffic is normally
invisible to the kernel). On receive, a tracer will capture a frame
before hw VLAN extraction, runt filtering, other badness filtering,
before the steering/drop/L2-rewrite filters or the TOE have had a go at
it, and of course before sw LRO in the driver.
There are 4 tracers on a chip. A tracer can trace only in one direction
(tx or rx). For now cxgbetool will set up tracers to capture the first
128B of every transmitted or received frame on a given port. This is a
small subset of what the hardware can do. A pseudo ifnet with the same
name as the nexus driver (t4nex0 or t5nex0) will be created for tracing.
The data delivered to this ifnet is an additional copy made inside the
chip. Normal delivery to cxgbe<n> or cxl<n> will be made as usual.
/* watch cxl0, which is the first port hanging off t5nex0. */
# cxgbetool t5nex0 tracer 0 tx0 (watch what cxl0 is transmitting)
# cxgbetool t5nex0 tracer 1 rx0 (watch what cxl0 is receiving)
# cxgbetool t5nex0 tracer list
# tcpdump -i t5nex0 <== all that cxl0 sees and puts on the wire
If you were doing TSO, a tcpdump on cxl0 may have shown you ~64K
"frames" with no L3/L4 checksum but this will show you the frames that
were actually transmitted.
/* all done */
# cxgbetool t5nex0 tracer 0 disable
# cxgbetool t5nex0 tracer 1 disable
# cxgbetool t5nex0 tracer list
# ifconfig t5nex0 destroy
when the interface is up.
- Add a tunable to control the TOE's rx coalesce feature (enabled by
default as it always has been). Consider the interface MTU or the
coalesce size when deciding which cluster zone to use to fill the
offload rx queue's free list. The tunable is:
dev.{t4nex,t5nex}.<N>.toe.rx_coalesce
MFC after: 1 day
it is being installed). Improve other error messages while here.
- Select special FPGA specific configuration profile when appropriate.
MFC after: 3 days
includes support for the NIC and TOE features of the 40G, 10G, and
1G/100M cards based on the T5.
The ASIC is mostly backward compatible with the Terminator 4 so cxgbe(4)
has been updated instead of writing a brand new driver. T5 cards will
show up as cxl (short for cxlgb) ports attached to the t5nex bus driver.
Sponsored by: Chelsio
the firmware (instead of just the main firmware version) when evaluating
firmware compatibility. Document the new "hw.cxgbe.fw_install" knob
being introduced here.
This should fix kern/173584 too. Setting hw.cxgbe.fw_install=2 will
mostly do what was requested in the PR but it's a bit more intelligent
in that it won't reinstall the same firmware repeatedly if the knob is
left set.
PR: kern/173584
MFC after: 5 days
CSUM_TCP too. They are all set explicitly by the kernel usually.
While here, fix an unrelated bug where hardware L4 checksum calculation
was accidentally disabled for some IPv6 packets.
Reported by: alfred@
MFC after: 3 days
cannot be freed while do_pass_accept_req is running. This closes a race
where do_pass_establish on another CPU (the driver chose a different
queue for the new tid) expands the synq entry into a full PCB and then
releases the only hold on it, all while do_pass_accept_req is still
running.
MFC after: 3 days
This is the Compressed Local IPv6 table on the chip. To save space, the
chip uses an index into this table instead of a full IPv6 address in
some of its hardware data structures.
For now the driver fills this table with all the local IPv6 addresses
that it sees at the time the table is initialized. I'll improve this
later so that the table is updated whenever new IPv6 addresses are
configured or existing ones deleted.
MFC after: 1 week
- Teach find_best_mtu_idx() to deal with IPv6 endpoints.
- Install correct protosw in offloaded TCP/IPv6 sockets when DDP is
enabled.
- Move set_tcp_ddp_ulp_mode to t4_tom.c so that t4_tom.h can be included
without having to drag in t4_msg.h too. This was bothering the iWARP
driver for some reason.
MFC after: 1 week
- Add full support for IPv6 addresses.
- Read the size of the L2 table during attach. Do not assume that PCIe
physical function 4 of the card has all of the table to itself.
- Use FNV instead of Jenkins to hash L3 addresses and drop the private
copy of jhash.h from the driver.
MFC after: 1 week
on the fast data path) and use them instead of frobbing the adapter lock
and busy flag directly.
Other changes made while reworking all slow operations:
- Wait for the reply to a filter request (add/delete). This guarantees
that the operation is complete by the time the ioctl returns.
- Tidy up the tid_info structure.
- Do not allow the tx queue size to be set to something that's not a
power of 2.
MFC after: 1 week
resources are partitioned.
- Reduce the number of virtual interfaces reserved for PF4. This leaves
spare room in the source MAC table and allows the driver to setup
filters that rewrite the source MAC address.
- Reduce the number of filters and use the freed up space for the CLIP
(Compressed Local IPv6 addresses) table. This is a prerequisite for
IPv6 TOE support which will follow separately in a series of commits.
MFC after: 1 week
embryonic connection has been setup and never attempt to abort a tid
before this is done. This fixes a bad race where a listening socket is
closed when the driver is in the middle of step (b) here. The symptom
of this were "ARP miss" errors from the driver followed by tid leaks.
A hardware-offloaded passive open works this way:
a) A SYN "hits" the TCAM entry for a server tid and the chip delivers it
to the queue associated with the server tid (say, queue A). It waits
for a response from the driver telling it what to do.
b) The driver decides it is ok to proceed. It adds the new tid to the
list of embryonic connections associated with the server tid and then
hands off the SYN to the kernel's syncache to make sure that the kernel
okays it too. If it does then the driver provides an L2 table entry,
queue id (say, queue B), etc. and instructs the chip to send the SYN/ACK
response.
c) The chip delivers a status to queue B depending on how the third step
of the 3-way handshake goes. The driver removes the tid from its list
of embryonic connections and either expands the syncache entry or
destroys the tid. In any case all subsequent messages for the new tid
will be delivered to queue B, not queue A. Anything running in queue B
knows that the L2 entry has long been setup and the new flag is of no
interest from here on. If the listener is closed it will deal with
so_comp as normal.
MFC after: 1 week
counter) when the syncache doesn't want the driver to reply to an
incoming SYN. This fixes a harmless bug where tids_in_use would
go out of sync with the hardware counter.
MFC after: 3 days
This lets userspace read arbitrary information from the SFP+ modules
etc. on this bus.
Reading multiple bytes in the same transaction isn't possible right now.
I'll update the driver once the chip's firmware supports this.
MFC after: 3 days
#defines. This also has the advantage that it makes the names more
compact, iand also allows us to correct the non-uniform naming of
the PCIM_LINK_* defines, making them all consistent amongst themselves.
This is a mostly mechanical rename:
s/PCIR_EXPRESS_/PCIER_/g
s/PCIM_EXP_/PCIEM_/g
s/PCIM_LINK_/PCIEM_LINK_/g
When this is MFC'd, #defines will be added for the old names to assist
out-of-tree drivers.
Discussed with: jhb
MFC after: 1 week
evicted from the syncache but a later syncache_expand succeeds because
of syncookies. The TOE driver has to resort to more direct means to
install its hooks in the socket in this case.
the TOE driver reports that an active open failed. toe_connect_failed is
supposed to handle this but it should be provided the inpcb instead of the
tcpcb which may no longer be around.
Basically, this is automatic rx zero copy when feasible. TCP payload is
DMA'd directly into the userspace buffer described by the uio submitted
in soreceive by an application.
- Works with sockets that are being handled by the TCP offload engine
of a T4 chip (you need t4_tom.ko module loaded after cxgbe, and an
"ifconfig +toe" on the cxgbe interface).
- Does not require any modification to the application.
- Not enabled by default. Use hw.t4nex.<X>.toe.ddp="1" to enable it.
- Setup multiple DDP page sizes. When the driver attempts DDP it will
try to combine physically contiguous pages into regions of these sizes.
- Set the indicate size such that the payload carried in the indicate can
be copied in the header mbuf (and the 16K rx buffer can be recycled).
- Set DDP threshold to the max payload that the chip will coalesce and
deliver to the driver (this is ~16K by default, which is also why the
offload rx queue is backed by 16K buffers). If the chip is able to
coalesce up to the max it's allowed to, it's a good sign that the peer
is transmitting in bulk without any TCP PSH.
MFC after: 2 weeks
TCB. Filters are programmed by modifying the TCB too (via a different
routine) and the reply to any TCB update is delivered via a
CPL_SET_TCB_RPL. Figure out whether the reply is for a filter-write or
something else and route it appropriately.
MFC after: 2 weeks
interface's MTU. Initialize such freelists with correct values.
This wasn't a problem for common MTUs (1500 and 9000) as the buffers (2048
and 9216 in size) happened to have enough spare room. I ran into it when
playing around with unusual MTUs.
MFC after: 2 weeks
values).
- cong_drop specifies what to do on congestion: nothing, backpressure,
or drop.
- fl_pktshift specifies the padding before Ethernet payload.
- fl_pad specifies the boundary upto which to pad Ethernet payload.
- spg_len controls the length of the status page.
MFC after: 2 weeks
'fw_hdr_intfver' into an anonymous enum, which avoids a clang 3.2
warning about all the enum values being the same value.
Reviewed by: np
MFC after: 1 week
exported via PCI passthrough.
- Do not check for a specific physical function (PF) before claiming a device.
Different PFs have different device-ids so this check is redundant anyway.
- Obtain the PF# from the WHOAMI register instead of pci_get_function().
- Setup the memory windows using the real BAR0 address, not what the VM says it
is.
Obtained from: Chelsio Communications
- 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)
Significantly update tcp_lro for mostly two things:
1) introduce basic support for IPv6 without extension headers.
2) try hard to also get the incremental checksum updates right,
especially also in the IPv4 case for the IP and TCP header.
Move variables around for better locality, factor things out into
functions, allow checksum updates to be compiled out, ...
Leave a few comments on further things to look at in the future,
though that is not the full list.
Update drivers with appropriate #includes as needed for IPv6 data
type in LRO.
Sponsored by: The FreeBSD Foundation
Sponsored by: iXsystems
Reviewed by: gnn (as part of the whole)
MFC After: 3 days
- Device configuration via plain text config file. Also able to operate
when not attached to the chip as the master driver.
- Generic "work request" queue that serves as the base for both ctrl and
ofld tx queues.
- Generic interrupt handler routine that can process any event on any
kind of ingress queue (via a dispatch table).
- A couple of new driver ioctls. cxgbetool can now install a firmware
to the card ("loadfw" command) and can read the card's memory
("memdump" and "tcb" commands).
- Lots of assorted information within dev.t4nex.X.misc.* This is
primarily for debugging and won't show up in sysctl -a.
- Code to manage the L2 tables on the chip.
- Updates to cxgbe(4) man page to go with the tunables that have changed.
- Updates to the shared code in common/
- Updates to the driver-firmware interface (now at fw 1.4.16.0)
MFC after: 1 month
one. Interestingly, these are actually the default for quite some time
(bus_generic_driver_added(9) since r52045 and bus_generic_print_child(9)
since r52045) but even recently added device drivers do this unnecessarily.
Discussed with: jhb, marcel
- While at it, use DEVMETHOD_END.
Discussed with: jhb
- Also while at it, use __FBSDID.
The SYSCTL_NODE macro defines a list that stores all child-elements of
that node. If there's no SYSCTL_DECL macro anywhere else, there's no
reason why it shouldn't be static.
queues. Try to have a set of these per port when possible, fall back
to sharing a common pool between all ports otherwise.
- One control queue per port (used to be one per hardware channel).
- t4_eth_rx now handles Ethernet rx only.
- sysctls to display pidx/cidx for some queues.
MFC after: 1 week
filters working. (All other filters - switch without L2 info rewrite,
steer, and drop - were already fully-functional).
Some contrived examples of "switch" filters with L2 rewriting:
# cxgbetool t4nex0 iport 0 dport 80 action switch vlan +9 eport 3
Intercept all packets received on physical port 0 with TCP port 80 as
destination, insert a vlan tag with VID 9, and send them out of port 3.
# cxgbetool t4nex0 sip 192.168.1.1/32 ivlan 5 action switch \
vlan =9 smac aa:bb:cc:dd:ee:ff eport 0
Intercept all packets (received on any port) with source IP address
192.168.1.1 and VLAN id 5, rewrite the VLAN id to 9, rewrite source mac
to aa:bb:cc:dd:ee:ff, and send it out of port 0.
MFC after: 1 week
Reference code that shows how to get a packet's timestamp out of
cxgbe(4). Disabled by default because we don't have a standard way
today to pass this information up the stack.
The timestamp is 60 bits wide and each increment represents 1 tick of
the T4's core clock. As an example, the timestamp granularity is ~4.4ns
for this card:
# sysctl dev.t4nex.0.core_clock
dev.t4nex.0.core_clock: 228125
MFC after: 1 week
- Enable 5-tuple and every-packet lookup.
- Setup the default filter mode to allow filtering/steering based on IP
protocol, ingress port, inner VLAN ID, IP frag, FCoE, and MPS match
type; all combined together. You can also filter based on MAC index,
Ethernet type, IP TOS/IPv6 Traffic Class, and outer VLAN ID but you'll
have to modify the default filter mode and exclude some of the
match-fields in it.
IPv4 and IPv6 SIP/DIP/SPORT/DPORT are always available in all filter
rules.
- Add driver ioctls to get/set the global filter mode.
- Add driver ioctls to program and delete hardware filters. A couple of
the "switch" actions that rewrite Ethernet and VLAN information and
switch the packet out of another port may not work as the L2 code is not
yet in place. Everything else, including all "drop" and "pass" rules
with RSS or absolute qid, should work.
Obtained from: Chelsio Communications
that could have allowed the hardware pidx to reach the cidx even though
the freelist isn't empty. (Haven't actually seen this but it was there
waiting to happen..)
MFC after: 1 week
now a suitable base for all kinds of egress queues.
- Add control queues (sge_ctrlq) and allocate one of these per hardware
channel. They can be used to program filters and steer traffic (and
more).
MFC after: 1 week
down. The ingress queue lock was unused and has been removed as part of
these changes.
- An in-flight egress update from the SGE must be handled before the
queue that requested it is destroyed. Wait for the update to arrive.
- Interrupt handlers must stop processing rx events for a queue before
the queue is destroyed. Events that have not yet been processed
should be ignored once the queue disappears.
MFC after: 1 week
queue has its own interrupt. If the exact number that we need is not a
power of 2 and we're using MSI, then switch to interrupt multiplexing.
While here, replace the magic numbers with something more readable.
MFC after: 3 days
- everything related to LRO should be in #ifdef INET blocks
- reorder sge_iq's fields so that the most frequently used are all together
- pull all rx code into t4_intr_data directly
- let go of the ingress queue lock when passing up data
- refill the freelist only if it is short of at least 32 buffers
