Commit Graph

328 Commits

Author SHA1 Message Date
Navdeep Parhar
c7dbd80213 cxgbe(4): Update T4 and T5 firmwares to 1.14.2.0.
Obtained from:	Chelsio Communications
MFC after:	3 days
2015-07-14 08:02:05 +00:00
Luigi Rizzo
847bf38369 Sync netmap sources with the version in our private tree.
This commit contains large contributions from Giuseppe Lettieri and
Stefano Garzarella, is partly supported by grants from Verisign and Cisco,
and brings in the following:

- fix zerocopy monitor ports and introduce copying monitor ports
  (the latter are lower performance but give access to all traffic
  in parallel with the application)

- exclusive open mode, useful to implement solutions that recover
  from crashes of the main netmap client (suggested by Patrick Kelsey)

- revised memory allocator in preparation for the 'passthrough mode'
  (ptnetmap) recently presented at bsdcan. ptnetmap is described in
        S. Garzarella, G. Lettieri, L. Rizzo;
        Virtual device passthrough for high speed VM networking,
        ACM/IEEE ANCS 2015, Oakland (CA) May 2015
        http://info.iet.unipi.it/~luigi/research.html

- fix rx CRC handing on ixl

- add module dependencies for netmap when building drivers as modules

- minor simplifications to device-specific routines (*txsync, *rxsync)

- general code cleanup (remove unused variables, introduce macros
  to access rings and remove duplicate code,

Applications do not need to be recompiled, unless of course
they want to use the new features (monitors and exclusive open).

Those willing to try this code on stable/10 can just update the
sys/dev/netmap/*, sys/net/netmap* with the version in HEAD
and apply the small patches to individual device drivers.

MFC after:	1 month
Sponsored by:	(partly) Verisign, Cisco
2015-07-10 05:51:36 +00:00
Navdeep Parhar
9af71ab3bc cxgbe(4): Add a new knob that controls the congestion response of netmap
rx queues.  The default is to drop rather than backpressure.

This decouples the congestion settings of NIC and netmap rx queues.

MFC after:	3 days
2015-07-06 20:56:59 +00:00
Navdeep Parhar
41f7622b64 cxgbe(4): Do not override the the global defaults for congestion drops.
The hw.cxgbe.cong_drop knob is not affected by this change because the
driver sets up congestion drop on a per-queue basis.

MFC after:	3 days
2015-07-06 20:28:42 +00:00
Navdeep Parhar
fd215e45eb cxgbe(4): request an automatic tx update when a netmap tx queue idles.
The NIC tx queues already do this.

MFC after:	1 week
Differential Revision:
2015-07-01 00:34:14 +00:00
Navdeep Parhar
dbbf46c40c cxgbe: get_fl_payload returns a header mbuf when successful.
MFC after:	3 days
2015-06-23 05:55:13 +00:00
Navdeep Parhar
0e4cd4a2e0 cxgbe(4): Add the ability to dump mailbox commands and replies. It is
enabled/disabled via bit 0 of adapter->debug_flags (which is available
at dev.t5nex.<n>.debug_flags).

MFC after:	1 week
2015-06-16 12:36:29 +00:00
Navdeep Parhar
a40200ca80 cxgbe: set the minimum burst size when fetching fl buffers to 128B for
netmap rx queues too.  This should have gone in as part of r283858.
2015-06-05 00:37:46 +00:00
Navdeep Parhar
a89409b817 cxgbe: no need to display the per-lane GT/s rating of the pcie link.
MFC after:	1 week
2015-06-01 03:24:39 +00:00
Navdeep Parhar
6af2071b47 cxgbe: set minimum burst size when fetching freelist buffers to 128B.
MFC after:	3 days
2015-06-01 00:55:15 +00:00
Jung-uk Kim
fd90e2ed54 CALLOUT_MPSAFE has lost its meaning since r141428, i.e., for more than ten
years for head.  However, it is continuously misused as the mpsafe argument
for callout_init(9).  Deprecate the flag and clean up callout_init() calls
to make them more consistent.

Differential Revision:	https://reviews.freebsd.org/D2613
Reviewed by:	jhb
MFC after:	2 weeks
2015-05-22 17:05:21 +00:00
Gleb Smirnoff
a5e0fa4063 Don't use ifm_data. It was used only for self checking debug.
Reviewed by:	np
2015-04-26 21:31:30 +00:00
Gleb Smirnoff
28ebe80cab Provide functions to determine presence of a given address
configured on a given interface.

Discussed with:	np
Sponsored by:	Nginx, Inc.
2015-04-17 11:57:06 +00:00
Navdeep Parhar
7ef00d7884 cxgbe/tom: return rx credits promptly if the socket buffer's low water
mark cannot be reached because the window advertised to the peer isn't
wide enough.  While here, tweak the normal credit return too.

MFC after:	1 month
2015-03-31 01:22:20 +00:00
Navdeep Parhar
70ca622987 cxgbe(4): provide the exact RSS hash type instead of a catch-all value
to the upper layers.
2015-03-26 18:45:51 +00:00
Navdeep Parhar
bb481a8edd cxgbe(4): Do not call sbuf_trim on an sbuf with a drain function.
MFC after:	1 week
2015-03-23 23:06:32 +00:00
John Baldwin
b12c0a9eeb Move special DDP handling for closing a connection into a new
handle_ddp_close() function in t4_ddp.c as the logic is similar
to handle_ddp_data().  This allows all knowledge of the special
DDP mbufs to be private to t4_ddp.c as well.
2015-03-16 15:56:06 +00:00
Ian Lepore
1eafc07856 Set the SBUF_INCLUDENUL flag in sbuf_new_for_sysctl() so that sysctl
strings returned to userland include the nulterm byte.

Some uses of sbuf_new_for_sysctl() write binary data rather than strings;
clear the SBUF_INCLUDENUL flag after calling sbuf_new_for_sysctl() in
those cases.  (Note that the sbuf code still automatically adds a nulterm
byte in sbuf_finish(), but since it's not included in the length it won't
get copied to userland along with the binary data.)

Remove explicit adding of a nulterm byte in a couple places now that it
gets done automatically by the sbuf drain code.

PR:		195668
2015-03-14 17:08:28 +00:00
Ian Lepore
b36424bd4b Revert r279934, r279938; this is going to be fixed in sbuf instead.
PR:		195668
2015-03-14 13:04:39 +00:00
Navdeep Parhar
070d262707 cxgbe(4): fix if_media handling for T520-BT cards. 1Gbps and 100Mbps
are valid for this card.

MFC after:	1 week
2015-03-14 00:02:53 +00:00
Ian Lepore
2f01da7886 Fix a paste-o, sb is already a pointer in this one. 2015-03-12 23:31:29 +00:00
Ian Lepore
9bc58e3d33 Nullterminate strings returned via sysctl.
PR:		195668
2015-03-12 18:22:20 +00:00
John Baldwin
69a088631e Resize receive socket buffers that support autosizing when receiving
TCP data via direct data placement.

Sponsored by:	Chelsio
MFC after:	1 week
2015-03-11 17:35:07 +00:00
Navdeep Parhar
08aeb15136 cxgbe(4): experimental rx packet sink for netmap queues. This is not
intended for general use.

MFC after:	1 month
2015-03-06 20:41:28 +00:00
Navdeep Parhar
1cdfce07df cxgbe(4): knobs to experiment with the interrupt coalescing timer for
netmap rx queues, and the "batchiness" of rx updates sent to the chip.

These knobs will probably become per-rxq in the near future and will be
documented only after their final form is decided.

MFC after:	1 month
2015-03-06 20:39:19 +00:00
Navdeep Parhar
7f93d696bf cxgbe(4): provide the correct size of freelists associated with netmap
rx queues to the chip.  This will fix many problems with native netmap
rx on ncxl/ncxgbe interfaces.

MFC after:	1 week
2015-03-06 16:05:20 +00:00
Navdeep Parhar
1c468ed975 cxgbe(4): allow tx hardware checksumming on the netmap interface.
It is disabled by default but users can set IFCAP_TXCSUM on the
netmap ifnet (ifconfig ncxl0 txcsum) to override netmap and force
the hardware to calculate and insert proper IP and L4 checksums in
outbound frames.

MFC after:	2 weeks
2015-02-24 21:31:13 +00:00
Navdeep Parhar
1605bac6fb cxgbe(4): set up congestion management for netmap rx queues.
The hw.cxgbe.cong_drop knob controls the response of the chip when
netmap queues are congested.
2015-02-24 18:40:10 +00:00
Navdeep Parhar
b91cbdda76 cxgbe(4): do not set the netmap rxq interrupts on a hair-trigger.
MFC after:	2 weeks
2015-02-24 18:32:17 +00:00
Navdeep Parhar
b46fad7b6a cxgbe(4): wait for the hardware to catch up before destroying a netmap txq.
MFC after:	2 weeks
2015-02-24 18:22:24 +00:00
Navdeep Parhar
f8e30f9cf1 cxgbe(4): request an automatic tx update when a netmap txq idles.
MFC after:	2 weeks
2015-02-24 18:19:25 +00:00
Navdeep Parhar
c5bb375553 cxgbe(4): there is no need to force an "unimplemented" panic needlessly.
The calls to free_nm_txq and free_nm_rxq are made just a few lines prior
to the panic.
2015-02-20 22:57:54 +00:00
Hans Petter Selasky
b5c1e0cb8d Update the infiniband stack to Mellanox's OFED version 2.1.
Highlights:
 - Multiple verbs API updates
 - Support for RoCE, RDMA over ethernet

All hardware drivers depending on the common infiniband stack has been
updated aswell.

Discussed with:	np @
Sponsored by:	Mellanox Technologies
MFC after:	1 month
2015-02-17 08:40:27 +00:00
Navdeep Parhar
d5d9fbbae2 cxgbe(4): allow the SET_FILTER_MODE ioctl to change the mode when it's
safe to do so.

MFC after:	1 month
2015-02-10 01:16:43 +00:00
Navdeep Parhar
b3d44a6800 cxgbe(4): tidy up some of the interaction between the Upper Layer
Drivers (ULDs) and the base if_cxgbe driver.

Track the per-adapter activation of ULDs in a new "active_ulds" field.
This was done pretty arbitrarily before this change -- via TOM_INIT_DONE
in adapter->flags for TOM, and the (1 << MAX_NPORTS) bit in
adapter->offload_map for iWARP.

iWARP and hw-accelerated iSCSI rely on the TOE (supported by the TOM
ULD).  The rules are:
a) If the iWARP and/or iSCSI ULDs are available when TOE is enabled then
   iWARP and/or iSCSI are enabled too.
b) When the iWARP and iSCSI modules are loaded they go looking for
   adapters with TOE enabled and enable themselves on that adapter.
c) You cannot deactivate or unload the TOM module from underneath iWARP
   or iSCSI.  Any such attempt will fail with EBUSY.

MFC after:	2 weeks
2015-02-08 09:28:55 +00:00
Navdeep Parhar
319f290030 cxgbe(4): adapter_full_init is always a synchronized operation.
MFC after:	1 week
2015-02-08 08:52:18 +00:00
Navdeep Parhar
d86a5ff917 cxgbe(4): a change to the synchronization rules within the the driver.
This is purely cosmetic because the new rules are already followed.

MFC after:	1 week
2015-02-08 08:42:45 +00:00
Navdeep Parhar
cb6c101a86 cxgbe(4): fix a test made while enabling TOE.
MFC after:	1 week
2015-02-07 01:50:32 +00:00
Navdeep Parhar
85dc477798 cxgbe(4): Add a minimal if_cxl module that pulls in the real driver as
a dependency.  This ensures "ifconfig cxl<n> ..." does the right thing
even when it's run with no driver loaded.

if_cxl.ko is the tiniest module in /boot/kernel.

MFC after:	2 weeks
2015-02-06 01:10:04 +00:00
Navdeep Parhar
a2355dd909 cxgbe(4): reserve id for iSCSI upper layer driver. 2015-02-05 08:52:20 +00:00
John Baldwin
86f05ea6cf Lock the socket buffer before jumping to the 'out' label if sblock()
fails in t4_soreceive_ddp().
2015-01-26 16:32:41 +00:00
John Baldwin
de5a10ecbc - Update a disabled KASSERT() to use sbused() instead of accessing
the no-longer existant sb_cc sockbuf member.
- Use sbavail() instead of sbused() in t4_soreceive_ddp() to match the
  usage in soreceive_stream() on which it is based.

Discussed with:	glebius (2)
2015-01-26 16:29:14 +00:00
John Baldwin
4f621933a5 Fix a couple of panics when detaching from a cxgbe/cxl interface that was
never brought up:
- Allow NULL to be passed to sglist_free().
- Don't try to stop an interface that was never fully initialized.

Reviewed by:	np
2015-01-26 16:26:28 +00:00
Hans Petter Selasky
d39d7c8636 Add missing linuxapi module dependencies and always use the FreeBSD
"MODULE_VERSION" macro definition. Remove the redefinition of the
"MODULE_VERSION" macro from the Linux kernel compatibility API.

MFC after:	1 month
Reported by:	np@
Sponsored by:	Mellanox Technologies
2015-01-19 21:53:00 +00:00
Navdeep Parhar
88d7f6bddf Allow cxgbe(4) to be built on i386. Driver attach will succeed only on a subset
of i386 systems.
2015-01-16 01:32:40 +00:00
Navdeep Parhar
e503548810 cxgbe/iw_cxgbe: fix whitespace nit in r277102.
Reported by:	stefanf@
2015-01-13 16:18:31 +00:00
Navdeep Parhar
b3e112f962 cxgbe/iw_cxgbe: allow any size during the initial MPA exchange.
MFC after:	1 month
2015-01-13 01:40:12 +00:00
Navdeep Parhar
db8bcd1b21 cxgbe/tom: allocate page pod addresses instead of ppod#.
MFC after:	2 weeks
2015-01-07 06:20:33 +00:00
Navdeep Parhar
f8c479085f cxgbe/tom: use vmem(9) as the DDP page pod allocator.
MFC after:	1 month
2015-01-06 01:30:32 +00:00
Navdeep Parhar
008015d2f4 cxgbe(4): fix the description of a strange bunch of counters.
MFC after:	1 week
2015-01-05 23:43:24 +00:00