#
303975 |
|
11-Aug-2016 |
gjb |
Copy stable/11@r303970 to releng/11.0 as part of the 11.0-RELEASE cycle.
Prune svn:mergeinfo from the new branch, and rename it to RC1.
Update __FreeBSD_version.
Use the quarterly branch for the default FreeBSD.conf pkg(8) repo and the dvd1.iso packages population.
Approved by: re (implicit) Sponsored by: The FreeBSD Foundation |
#
302408 |
|
08-Jul-2016 |
gjb |
Copy head@r302406 to stable/11 as part of the 11.0-RELEASE cycle. Prune svn:mergeinfo from the new branch, as nothing has been merged here.
Additional commits post-branch will follow.
Approved by: re (implicit) Sponsored by: The FreeBSD Foundation
|
#
302292 |
|
30-Jun-2016 |
wma |
ARM, ARM64: Workaround for buf_ring reordering
This patch offers a workaround to buf_ring reordering visible on armv7 and armv8. This is supposed to be removed once new buf_ring implementation is integrated into the tree.
Obtained from: Semihalf Reviewed by: alc,emaste Differential Revision: https://reviews.freebsd.org/D6986 Approved by: re (gjb)
|
#
298981 |
|
03-May-2016 |
pfg |
sys/sys: minor spelling fixes.
While the changes are minor, these headers are very visible.
MFC after: 2 weeks
|
#
296178 |
|
29-Feb-2016 |
sephe |
buf_ring/drbr: Add buf_ring_peek_clear_sc and use it in drbr_peek
Unlike buf_ring_peek, it only supports single consumer mode, and it clears the cons_head if DEBUG_BUFRING/INVARIANTS is defined.
The normal use case of drbr_peek for network drivers is:
m = drbr_peek(br); err = hw_spec_encap(&m); /* could m_defrag/m_collapse */ (*) if (err) { if (m == NULL) drbr_advance(br); else drbr_putback(br, m); /* break the loop */ } drbr_advance(br);
The race is: If hw_spec_encap() m_defrag or m_collapse the mbuf, i.e. the old mbuf was freed, or like the Hyper-V's network driver, that transmission- done does not even require the TX lock; then on the other CPU at the (*) time, the freed mbuf could be recycled and being drbr_enqueue even before the current CPU had the chance to call drbr_{advance,putback}. This triggers a panic in drbr_enqueue duplicated element check, if DEBUG_BUFRING/INVARIANTS is defined.
Use buf_ring_peek_clear_sc() in drbr_peek() to fix the above race.
This change is a NO-OP, if neither DEBUG_BUFRING nor INVARIANTS are defined.
MFC after: 1 week Sponsored by: Microsoft OSTC Differential Revision: https://reviews.freebsd.org/D5416
|
#
273866 |
|
30-Oct-2014 |
jpaetzel |
Plug memory ordering holes in buf_ring_enqueue. For at least some users this patch eliminates the races previously discussed on the mailing list.
Submitted by: oleg Reviewed by: kmacy MFC after: 2 weeks Tested by: kmacy,rpaulo
|
#
263954 |
|
30-Mar-2014 |
imp |
Remove instances of variables that were set, but never used. gcc 4.9 warns about these by default.
|
#
246482 |
|
07-Feb-2013 |
rrs |
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
|
#
244774 |
|
28-Dec-2012 |
attilio |
Improve bufring impl: - Remove unused br_prod_bufs member - Fixup r241037: buf_ring pads br_prod_* and br_cons_* members at 128 bytes, assuming a fixed cache line size for all the architectures. However, the above mentioned revision broke the padding. Use explicit padding to the CACHE_LINE_SIZE on the members that mark the initial new padded sections. Of course, the padding is not important for performance reasons in the DEBUG_BUFRING case, leaving br_cons members to share the cache line with br_lock. - Fixup r244732: by removing incorrectly added membar in buf_ring_dequeue_sc() where surrounding locking shoud be enough. - Drastically reduce the number of membar used (pratically reverting r244732) by switching rmb() in buf_ring_dequeue_mc() and wmb() in buf_ring_enqueue() to be complete barriers. This, along with br_prod_bufs departure, should fix ordering issues as explained in the provided comments.
This patch is not targeted for MFC.
Sponsored by: EMC / Isilon storage division Reviewed by: glebius
|
#
244733 |
|
27-Dec-2012 |
attilio |
Remove an unused var.
Sponsored by: EMC / Isilon storage division MFC after: 3 days
|
#
244732 |
|
27-Dec-2012 |
attilio |
br_prod_tail and br_cons_tail members are used as barrier to signal bug_ring ownership. However, instructions can be reordered around members write leading to stale values for ie. br_prod_bufs.
Use correct memory barriers to ensure proper ordering of the ownership tokens updates.
Sponsored by: EMC / Isilon storage division MFC after: 2 weeks
|
#
241037 |
|
28-Sep-2012 |
glebius |
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
|
#
207673 |
|
05-May-2010 |
joel |
Switch to our preferred 2-clause BSD license.
Approved by: kmacy
|
#
193848 |
|
09-Jun-2009 |
kmacy |
- add drbr routines for accessing #qentries and conditionally dequeueing - track bytes enqueued in buf_ring
|
#
191899 |
|
07-May-2009 |
kmacy |
No man page currently exists so comment the two uncommented non-trivial functions
|
#
186207 |
|
17-Dec-2008 |
kmacy |
merge in 2 buf_ring helper routines for enqueueing and freeing buf_rings
|
#
185543 |
|
02-Dec-2008 |
kmacy |
return ENOBUFS when ring is full
|
#
185193 |
|
23-Nov-2008 |
kmacy |
buf_ring_peek should return NULL if the ring is empty rather than whatever happened to be at cons_tail last time it was in use
|
#
185162 |
|
22-Nov-2008 |
kmacy |
- bump __FreeBSD version to reflect added buf_ring, memory barriers, and ifnet functions
- add memory barriers to <machine/atomic.h> - update drivers to only conditionally define their own
- add lockless producer / consumer ring buffer - remove ring buffer implementation from cxgb and update its callers
- add if_transmit(struct ifnet *ifp, struct mbuf *m) to ifnet to allow drivers to efficiently manage multiple hardware queues (i.e. not serialize all packets through one ifq) - expose if_qflush to allow drivers to flush any driver managed queues
This work was supported by Bitgravity Inc. and Chelsio Inc.
|