The default e1000e parameters (interrupt throttling rate, MSI/MSI-X
mode) are optimized for desktop and server computers to optimize
user-space execution (i.e. what's typically referred to as "useful"
work). This assumption breaks on a router under load where most of
the "useful" work actually takes place either in hardware interrupt
handlers (IRQ) or at software IRQ (swirq) modes, so we try to reflect
that by overriding these parameters with more appropriate values.
Patch-by: Philip Prindeville <philipp@redfish-solutions.com>
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
There has been recent significant activity with the cake qdisc of late
but in the cobalt branch. Some of that effort is related to upstreaming
to kernel & iproute2 mainline but we're not quite there yet. Relevant
feature changes:
ingress mode: Instead of only counting packets that make it past the
shaper, include packets we've decided to drop as well, since they did
arrive with us on the link and took link capacity.
This mode is more suitable for shaping the ingress of a link
(e.g. from ISP) rather than the more normal egress.
ptm mode: Minor optimisation in packet overhead calculation.
dual-src/dsthost/triple-isolate: Optimise only calculating src or dst
host hashes only if required.
ack-filter/ack-filter-aggressive: Filter excessive TCP ACKS. Useful in
highly assymetric links (downstream v upstream capacity) where the
majority of upstream link capacity is occupied with ACKS for downstream
traffic.
A separate iproute2 patch to teach it about Cake's new features will
follow.
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
d02a05b mt7603: update firmware to version 20160107100755
4d4cd05 Partially revert "mt7603: use mcu command to set timing registers, fix OFDM timeout values"
170f334 mt76x2: remove MAC address limitation for multi-vif setups
3563b8f mt76x2: clean up MAC/BSSID address initialization
9de77e1 mt76x2: drop wiphy->addresses
a6a6e25 mt76x2: init: disable APCLI by default
c64633e mt76x2: configure rx filter based on monitor mode setting
ac815fa mt76x2: init: fix rx filter default value during init
e504656 mt7603: configure other-unicast drop based on monitor mode setting
Signed-off-by: Felix Fietkau <nbd@nbd.name>
11f42a8 mt76x2: add channel argument to eeprom tx power functions
3bd7e76 mt76x2: initialize channel power limits
19fff41 mt76x2: convert between per-chain tx power and combined output
737cf2b mt7603: rename mt7603_mac_reset to mt7603_pse_reset
8026638 mt7603: rename MT_PSE_RESET register
c4dd32a mt7603: remove watchdog reset on interface stop
d99092b mt7603: remove WARN_ON_ONCE for workaround checks
c8807b4 mt7603: simplify PSE reset
d8a5990 mt7603: warn if PSE reset fails
c079960 mt7603: clean up dma debug reads
96817d6 mt7603: make mt7603_mac_watchdog_reset() static
e953c78 mt7603: clear wtbl PS bit for powersave responses
57a2e33 mt7603: set tx-skip flag for powersave clients
c8e5ab1 mt7603: initialize wtbl ps flag on station add
b4034cf mt76x2: remove some harmless WARN_ONs in tx status and rx path
8e17d36 mt7603: remove some harmless WARN_ONs in rx path
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Latencies can be much higher on wifi devices, especially with
aggregation. Tune the network stack setting introduced in the previous
commit to account for that.
This commit reintroduces the previously reverted one with a fix for the
crash issues
Signed-off-by: Felix Fietkau <nbd@nbd.name>
This reverts commit 2dc485250d.
This patch needs some additional checks in order to avoid overwriting
unrelated fields for request sockets.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
dfb2f6c pkt_sched: make compile again
5ab7026 sch_cake: make compile again
6f28803 codel5: make more checkpatch compliant
bd426aa Fix build error on 4.12
e4a3628 Whitespace tidy up
Signed-off-by: Fushan Wen <qydwhotmail@gmail.com>
Fix the target dependency to make it possible to select this module also
on x86 target and its subtargets.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The firmware directory in the Linux kernel was removed in kernel 4.14,
take the e100 firmware files now from the linux-firmware repository
instead. To do so create the new package e100-firmware. This will also
work with older kernel versions.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Adds NFS4 client support:
1. Package kmod-fs-nfs is split into kmod-fs-nfs (nfs.ko) and
kmod-fs-nfs-v3 (nfsv3.ko).
2. A new package kmod-fs-nfs-v4 (nfsv4.ko) is created.
3. Package kmod-fs-nfs-common-v4 is renamed to kmod-fs-nfs-rpcsec
and includes additional module rpcsec_gss_krb5.ko.
CONFIG_NFS_V4 goes into kmod-fs-nfs-v4, CONFIG_NFSD_V4 (NFS4
server) is removed. Missing kernel module oid_registry.ko
needed by auth_rpcgss.ko is added to the package.
A new package kmod-crypto-cts needed by rpcsec_gss_krb5.ko is
also created.
Signed-off-by: Marcin Jurkowski <marcin1j@gmail.com>
[add dependency to kmod-crypto-ecb in fs-nfs-common-rpcsec]
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The module parameters "nogameport=1" and "nocir=1" are needed,
because this is not supported on recent chips and doesn't
really tell if the system is stable.
As this features will already be removed in linux-4.13 or newer,
this module parameters can be removed in the future.
Signed-off-by: Martin Schiller <ms@dev.tdt.de>
The DEFINE_PCI_DEVICE_TABLE macro was removed with upstream commit
7e9321599011 ("treewide: remove references to the now unnecessary
DEFINE_PCI_DEVICE_TABLE").
Use the pci_device_id struct to fix the acx-mac80211 build failure on
ramips.
Signed-off-by: Mathias Kresin <dev@kresin.me>
The kernel firmware/ is going away, so pull this firmware
from the linux-firmware git repo instead. No actual changes
to the installed files.
Signed-off-by: Bjørn Mork <bjorn@mork.no>
Significantly improves throughput on MT76x2, fixes some stability
issues, adds LED support.
Changes:
266ef38 mt76x2: mcu: remove unused parameter in mt76x2_mcu_msg_alloc signature
758376d mt7603: mcu: remove unused parameter in mt7603_mcu_msg_alloc() signature
e764787 Fix errors found by cppcheck
a6fce8a mt7603: add LED definition registers
f658dd2 mt76x2: add LED register definitions
f6a021d mt76x2: Support using PCI ID as chip ID
c9bdcd8 mt76: add led support using mac80211 led framework
58e9138 mt76x2: init: add ma80211 led callbacks
8ea8da3 mt7603: init: add ma80211 led callbacks
ded88cd mt76x2: Add PCI identifier for MT7602
51a6764 mt7603: remove unnecessary mcu register read function
fbdbf65 debugfs: add support for changing the LED pin
cc02e49 mac80211: move DT led configuration to the "led" child node
e4e7734 mt76x2: limit client WCID entries to 0-127
60172cc mt76x2: clear drop flag for all WCIDs on init
d8140b6 mt76x2: clear per-WCID tx rate lookup register
0ce7923 mt76x2: add helper function for setting drop mask
ccc4baf mt76x2: clear drop mask when sending a PS response
ff60d14 mt76: increase rx ring size for mt76x2
b57ada5 mt76x2: add rx statistics registers
af425de mt76x2: fix LNA gain register annotation
efd7724 mt76x2: sync channel gain value with latest reference driver
4af37bd mt76x2: implement dynamic AGC tuning based on false packet detection count
70f2002 mt76x2: add more gain tuning based on the latest reference driver
8f1c8ab mt76x2: sync tx power related values with reference driver
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Latencies can be much higher on wifi devices, especially with
aggregation. Tune the network stack setting introduced in the previous
commit to account for that
Signed-off-by: Felix Fietkau <nbd@nbd.name>
The wireless regdb is now loaded via firmware loading, CRDA support and
built-in regdb support have been removed.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
These adapters support SR-IOV. Thus the host can assign Virtual Functions
(VFs) to different VMs by the PCI-E Passthrough (e.g. VFIO for KVM), to
gain different advantages (performance, VF to VF communications, host
kernel offload, etc.).
Signed-off-by: Chris Blakely <cpblakely@gmail.com>
- Remove obsolete patch chunks regarding fixed_freq
- Instead of patching in custom HT40+/- parameters, use the standard
config syntax as much as possible.
- Use fixed_freq for mesh
- Fix issues with disabling obss scan when using fixed_freq on mesh
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Passing the ctrl iface to wpa_supplicant will automatically cause wpa_supplicant
to send "STOP_AP" messages to the hostapd. This breaks the AP interfaces.
Signed-off-by: Antonio Quartulli <ordex@autistici.org>
Due to limitations in the symvers treatment and the mei drivers
exporting the same funtions, modpost might use the wrong mei driver
to link against.
Work around this by renaming them all to the same name, making it
always the "right" module name even if the wrong file was used.
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Apearently we need to ensure mwlwifi loads before mwifiex on
the WRT3200ACM, else mwifiex will claim the wifi.
Fix this by reverting to AutoLoad, but keep the removal of
mac80211 line.
This partially reverts commit 471d5dc6e3.
Fixes: 471d5dc6e3 ("mwlwifi: switch to AutoProbe")
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Unset the default LED gpio pin if the same gpio pin is used by a button
defined via platform button. It prevents the change of the GPIO value
on wireless up/down or wireless traffic.
Fixes: FS#1129
Signed-off-by: Mathias Kresin <dev@kresin.me>
Commit 2e496876c6 fixed the generation of the depends line for external
kernel modules which makes it possible for the build system to
automatically detect this missing dependency.
This fixes the packaging of kmod-ltq-atm for the ar9, xway and
xway-legacy subtarget.
Fixes: FS#1124
Signed-off-by: Mathias Kresin <dev@kresin.me>
Newer kernels treat differing signatures an error, not just a warning,
so fix the signatures to match.
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Minimum supported kernel is 3.18, so we don't need to test for anything
older. In addition, the API hasn't changed since then, so we don't need
to check for any kernel version at all. This helps to keeps the amount
of changes more managable.
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Commit 2e496876c6 fixed the generation of the depends line for external
kernel modules which makes it possible for the build system to
automatically detect this missing dependency. This fixes the build bot
build of the lantiq/aes target.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The build system already defines KERNEL_CROSS which defaults to TARGET_CROSS.
Make use of this variable for kernel makefiles.
Signed-off-by: Karl Vogel <karl.vogel@gmail.com>
Now that we have working module dependency generation, we can switch to
AutoProbe and let modprobe handle loading required modules.
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Now that we have working module dependency generation, we can switch to
AutoProbe and let modprobe handle loading required modules.
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Now that we have working module dependency generation, we can switch to
AutoProbe and let modprobe handle loading required modules.
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Now that we have working module dependency generation, we can switch to
AutoProbe and let modprobe handle loading required modules.
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Reoder the build to build the glue module first and pass the glue module's
Module.symvers to the wl driver builds.
This allows modpost to properly store a wl_glue dependency in the driver.
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
