Achiad Shochat 88a85f99e5 net/mlx5e: TX latency optimization to save DMA reads ... A regular TX WQE execution involves two or more DMA reads - one to fetch the WQE, and another one per WQE gather entry. These DMA reads obviously increase the TX latency. There are two mlx5 mechanisms to bypass these DMA reads: 1) Inline WQE 2) Blue Flame (BF) An inline WQE contains a whole packet, thus saves the DMA read/s of the regular WQE gather entry/s. Inline WQE support was already added in the previous commit. A BF WQE is written directly to the device I/O mapped memory, thus enables saving the DMA read that fetches the WQE. The BF WQE I/O write must be in cache line granularity, thus uses the CPU write combining mechanism. A BF WQE I/O write acts also as a TX doorbell for notifying the device of new TX WQEs. A BF WQE is written to the same I/O mapped address as the regular TX doorbell, thus this address is being mapped twice - once by ioremap() and once by io_mapping_map_wc(). While both mechanisms reduce the TX latency, they both consume more CPU cycles than a regular WQE: - A BF WQE must still be written to host memory, in addition to being written directly to the device I/O mapped memory. - An inline WQE involves copying the SKB data into it. To handle this tradeoff, we introduce here a heuristic algorithm that strives to avoid using these two mechanisms in case the TX queue is being back-pressured by the device, and limit their usage rate otherwise. An inline WQE will always be "Blue Flamed" (written directly to the device I/O mapped memory) while a BF WQE may not be inlined (may contain gather entries). Preliminary testing using netperf UDP_RR shows that the latency goes down from 17.5us to 16.9us, while the message rate (tested with pktgen) stays the same. Signed-off-by: Achiad Shochat <achiad@mellanox.com> Signed-off-by: Amir Vadai <amirv@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>		2015-07-27 00:29:17 -07:00
..
acpi
asm-generic	mm: clean up per architecture MM hook header files	2015-07-17 16:39:53 -07:00
clocksource
crypto
drm
dt-bindings
keys
kvm
linux	net/mlx5e: TX latency optimization to save DMA reads	2015-07-27 00:29:17 -07:00
math-emu
media
memory
misc
net	ip_tunnel: Call ip_tunnel_core_init() from inet_init()	2015-07-23 01:28:21 -07:00
pcmcia
ras
rdma	IB: Add rdma_cap_ib_switch helper and use where appropriate	2015-07-14 13:20:08 -04:00
rxrpc
scsi	IB/srp: Avoid using uninitialized variable	2015-07-14 13:20:09 -04:00
soc
sound
target
trace
uapi	lwtunnel: export linux/lwtunnel.h to userspace	2015-07-26 21:45:54 -07:00
video
xen
Kbuild