Commit Graph

35 Commits

Author SHA1 Message Date
Randall Stewart
ded5ea6a25 This fixes a out-of-order problem with several
of the newer drivers. The basic problem was
that the driver was pulling the mbuf off the
drbr ring and then when sending with xmit(), encounting
a full transmit ring. Thus the lower layer
xmit() function would return an error, and the
drivers would then append the data back on to the ring.
For TCP this is a horrible scenario sure to bring
on a fast-retransmit.

The fix is to use drbr_peek() to pull the data pointer
but not remove it from the ring. If it fails then
we either call the new drbr_putback or drbr_advance
method. Advance moves it forward (we do this sometimes
when the xmit() function frees the mbuf). When
we succeed we always call advance. The
putback will always copy the mbuf back to the top
of the ring. Note that the putback *cannot* be used
with a drbr_dequeue() only with drbr_peek(). We most
of the time, in putback, would not need to copy it
back since most likey the mbuf is still the same, but
sometimes xmit() functions will change the mbuf via
a pullup or other call. So the optimial case for
the single consumer is to always copy it back. If
we ever do a multiple_consumer (for lagg?) we
will  need a test and atomic in the put back possibly
a seperate putback_mc() in the ring buf.

Reviewed by:	jhb@freebsd.org, jlv@freebsd.org
2013-02-07 15:20:54 +00:00
Gleb Smirnoff
c6499eccad Mechanically substitute flags from historic mbuf allocator with
malloc(9) flags in sys/dev.
2012-12-04 09:32:43 +00:00
Andre Oppermann
da2299c5c7 Remove unused and unnecessary CSUM_IP_FRAGS checksumming capability.
Checksumming the IP header of fragments is no different from doing
normal IP headers.

Discussed with:	yongari
MFC after:	1 week
2012-11-27 19:31:49 +00:00
Gleb Smirnoff
470797ab8e Utilize new macro to initialize if_baudrate. 2012-10-18 09:59:50 +00:00
Gleb Smirnoff
063efed28c The drbr(9) API appeared to be so unclear, that most drivers in
tree used it incorrectly, which lead to inaccurate overrated
if_obytes accounting. The drbr(9) used to update ifnet stats on
drbr_enqueue(), which is not accurate since enqueuing doesn't
imply successful processing by driver. Dequeuing neither mean
that. Most drivers also called drbr_stats_update() which did
accounting again, leading to doubled if_obytes statistics. And
in case of severe transmitting, when a packet could be several
times enqueued and dequeued it could have been accounted several
times.

o Thus, make drbr(9) API thinner. Now drbr(9) merely chooses between
  ALTQ queueing or buf_ring(9) queueing.
  - It doesn't touch the buf_ring stats any more.
  - It doesn't touch ifnet stats anymore.
  - drbr_stats_update() no longer exists.

o buf_ring(9) handles its stats itself:
  - It handles br_drops itself.
  - br_prod_bytes stats are dropped. Rationale: no one ever
    reads them but update of a common counter on every packet
    negatively affects performance due to excessive cache
    invalidation.
  - buf_ring_enqueue_bytes() reduced to buf_ring_enqueue(), since
    we no longer account bytes.

o Drivers handle their stats theirselves: if_obytes, if_omcasts.

o mlx4(4), igb(4), em(4), vxge(4), oce(4) and  ixv(4) no longer
  use drbr_stats_update(), and update ifnet stats theirselves.

o bxe(4) was the most correct driver, it didn't call
  drbr_stats_update(), thus it was the only driver accurate under
  moderate load. Now it also maintains stats itself.

o ixgbe(4) had already taken stats from hardware, so just
  - drop software stats updating.
  - take multicast packet count from hardware as well.

o mxge(4) just no longer needs NO_SLOW_STATS define.

o cxgb(4), cxgbe(4) need no change, since they obtain stats
  from hardware.

Reviewed by:	jfv, gnn
2012-09-28 18:28:27 +00:00
Mike Silbersack
66437e8374 Only stop the BXE controller if it was first started. Stopping
an uninitialized controller can cause IPMI bus errors on some
systems.

Reviewed by:	yongari
Obtained from:	McAfee, Inc.
MFC after:	2 weeks
2012-09-05 06:51:28 +00:00
Scott Long
b6f97155cc Convert a number of drivers to obtaining their parent DMA tag from their
PCI device attachment.
2012-03-12 08:03:51 +00:00
Ulrich Spörlein
004a39cd57 Remove spurious 8bit chars, turning files into plain ASCII. 2012-01-15 13:23:33 +00:00
Kevin Lo
5bbe0c5357 ether_ifattach() sets if_mtu to ETHERMTU, don't bother set it again
Reviewed by:	yongari
2012-01-07 09:41:57 +00:00
John Baldwin
b02a80a7c5 Update recently added drivers to use the if_*addr_r*lock() wrapper
functions instead of using the IF_ADDR_LOCK directly.  The wrapper
functions are the supported interface for device drivers.

Reviewed by:	bz, philip
MFC after:	1 week
2012-01-05 18:32:37 +00:00
Kevin Lo
1a7e07818d s/timout/timeout 2011-12-15 06:29:13 +00:00
Marius Strobl
4b7ec27007 - There's no need to overwrite the default device method with the default
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.
2011-11-22 21:28:20 +00:00
Ed Schouten
6472ac3d8a Mark all SYSCTL_NODEs static that have no corresponding SYSCTL_DECLs.
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.
2011-11-07 15:43:11 +00:00
Eitan Adler
36daf0495a - change "is is" to "is" or "it is"
- change "the the" to "the"

Approved by:	lstewart
Approved by:	sahil (mentor)
MFC after:	3 days
2011-10-16 14:30:28 +00:00
David Christensen
61b6cf3661 - Fix compiler warning in ADD_64() macro.
Approved by:	re
Obtained from:	dimitry@andic.com
MFC after:	One week
2011-09-13 15:49:28 +00:00
David Christensen
76dbe6498b - Major reorganization of mbuf handling throughout the driver to
increase robustness (no more calls to panic(9)) and simplify
  code.
- Allocate RX/TX data structures as a single buffer rather than
  an array of 4KB pages to simplify code.
- Fixed LRO (aka TPA) code.  Removed kernel module parameter and
  support enabling disabling LRO through ifconfig(8) command line.
  LRO is still disabled by default but should be enabled for best
  performance on an endpoint device.
- Fixed statistcs code and removed kernel module parameter (stats
  should just work).
- Added many software counters to help identify the cause of some
  performance issues.
- Streamlined adapter internal init/stop code paths.
- Fiddled with debug code (adding some here, removing some there).
- Continued style(9) adjustments.
2011-06-08 21:18:14 +00:00
David Christensen
8fedd5eb65 - Use bus_describe_intr() to describe interrupt usage.
- Use bus_bind_intr() to bind interrupt to a CPU when RSS/TSS is used.
- Use M_DONTWAIT for RSS/TSS buffer allocation.
- Add statistic to track max DRBR queue depth.
- Fix problem in bxe_change_mtu() which referenced the old MTU size
  in a debug print statement.

MFC after:	Two weeks
2011-05-12 23:26:53 +00:00
David Christensen
876379ca45 - Simplify multicast address programming.
- Fix an incorrect "uint32_t *" cast in bxe_set_rx_mode().

Submitted by:   yongari@
Approved by:    davidch@
MFC after:      Two weeks
2011-05-09 18:46:53 +00:00
David Christensen
f83b279588 - Fixed a typo in an if() statement when setting flow control for MTU
greater than 5000 bytes.

Submitted by:	yongari
2011-05-02 23:34:33 +00:00
David Christensen
e93ad740a0 - Re-committed r220603 which was accidentally backed out by an earlier
commit.
- Fixed a bug in an unused debug macro.

MFC after:	One week.
2011-05-02 20:43:24 +00:00
David Christensen
a7a0fa5f58 - Centralize driver tunables initialization/validation.
- Centralize PCI resource allocation/release.
- Enable flowid (TSS) support.
- Added "per-fastpath" locks and watchdog timeouts.
- Fixed problem where the CQ producer index was advanced beyond
  the size of the CQ ring during initialization.
- Replaced hard-coded debug levels in some debug print statements.
- More style(9) fixes.

MFC after:	Two weeks
2011-04-21 23:06:00 +00:00
Marius Strobl
4ef498c23e Remove a comment which is no longer true since r213878, apart from
this driver not using mii(4) in the first place.

Reviewed by:    yongari
2011-04-19 17:49:59 +00:00
Pyun YongHyeon
246419ba87 Remove mii(4) dependency and unneeded headers.
Reviewed by:	davidch
2011-04-13 16:35:43 +00:00
David Christensen
3a1594a579 - Fixed a problem when multiqueue mode was enabled which caused the CQ
chain to be corrupted.
- Removed many console print warnings and replaced with driver maintained
  counters.
- Several style(9) fixes.

MFC after:      One week.
2011-04-07 20:15:36 +00:00
David Christensen
07a558f997 - Removed multiple console error messages and replaced with statistic
counters to reduce spew.
- Fixed a TSO problem when an mbuf contains both header and payload in
  the same cluster.

MFC after:	One week.
2011-04-05 22:13:33 +00:00
Pyun YongHyeon
03daac8400 Make bxe(4) build with BXE_DEBUG. 2011-04-01 01:30:21 +00:00
David Christensen
a48d88c42d - Fixed build problem when not useing BXE_DEBUG.
MFC after:	One week.
2011-03-31 22:50:55 +00:00
David Christensen
b1403b1285 - Freshened debug support code.
- Renamed several RX variable for more consistent usage.
- Fixed a potential problem when masking RX CQ producer value.

MFC after:	One week.
2011-03-31 22:40:44 +00:00
David Christensen
68f867e9e1 - Fixed DMA engine errors by increasing timeouts to 200ms for reads/writes.
- Improved some error reporting calls to include file name/line number.
- Various style(9) fixes.

MFC after:	One week.
2011-03-31 22:04:00 +00:00
David Christensen
8b94099c81 - Added debug support to monitor mbuf defrag attempts/failures.
MFC after:	One week.
2011-03-31 21:30:00 +00:00
David Christensen
650e98d34b - Print number of queues when RSS is enabled.
- Improve reporting of media type (not always 10GBase-CX4).

MFC after:	One week
2011-03-31 21:01:10 +00:00
David Christensen
b105a5aed5 - Fixed a problem where the stack passed a TSO frame larger than the 64K
size allowed by the DMA descriptor for TSO frames.

MFC after:      One week
2011-03-31 20:46:50 +00:00
John Baldwin
3b0a4aef96 Do a sweep of the tree replacing calls to pci_find_extcap() with calls to
pci_find_cap() instead.
2011-03-23 13:10:15 +00:00
David Christensen
ee50cfe1f1 - Inadvertently committed files with +x attribute, fixed.
- Minor change to info output string.
2011-03-15 01:06:27 +00:00
David Christensen
dd46ab31de - Initial release of bxe(4) to support Broadcom NetXtreme II 10GbE.
(BCM57710, BCM57711, BCM57711E)

MFC after:	One month
2011-03-14 22:42:41 +00:00