This looks like a simple typo that has gone unnoticed for some time. The
impact is relatively low but it's clearly wrong.
Signed-off-by: John Heffner <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Effect:
Slave Interface: eth5
MII Status: up
Speed: 10000 Mbps
Duplex: full
Link Failure Count: 0
Permanent HW addr: XX:XX:XX:XX:XX:XX
Slave queue ID: 0
Signed-off-by: Krzysztof Piotr Oledzki <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
This is the second step for neighbour RCU conversion.
(first was commit d6bf7817 : RCU conversion of neigh hash table)
neigh_lookup() becomes lockless, but still take a reference on found
neighbour. (no more read_lock()/read_unlock() on tbl->lock)
struct neighbour gets an additional rcu_head field and is freed after an
RCU grace period.
Future work would need to eventually not take a reference on neighbour
for temporary dst (DST_NOCACHE), but this would need dst->_neighbour to
use a noref bit like we did for skb->_dst.
Signed-off-by: Eric Dumazet <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
When an interface was enslaved when it was down, bonding thinks
it has speed -1 even after it goes up. This leads into selecting
a wrong active interface in active/backup mode on mixed 10G/1G or
1G/100M environment.
before:
bonding: bond0: link status definitely up for interface eth5, 100 Mbps full duplex.
bonding: bond0: link status definitely up for interface eth0, 100 Mbps full duplex.
after:
bonding: bond0: link status definitely up for interface eth5, 10000 Mbps full duplex.
bonding: bond0: link status definitely up for interface eth0, 1000 Mbps full duplex.
Signed-off-by: Krzysztof Piotr Oledzki <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
before:
bonding: bond0: link status definitely up for interface eth5
bonding: bond0: link status definitely up for interface eth0
after:
bonding: bond0: link status definitely up for interface eth5, 100 Mbps full duplex.
bonding: bond0: link status definitely up for interface eth0, 100 Mbps full duplex.
Signed-off-by: Krzysztof Piotr Oledzki <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
This patch does not include any functional changes.
The changes are: empty lines, indentation and comments.
Signed-off-by: Dmitry Kravkov <[email protected]>
Signed-off-by: Eilon Greenstein <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Return proper L2 and L5 SPQ (slow path queue) credits. Previously, all
SPQ events were counted as L5 types.
Signed-off-by: Dmitry Kravkov <[email protected]>
Signed-off-by: Michael Chan <[email protected]>
Signed-off-by: Eilon Greenstein <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Decide which interrupt mode to use (MSI-X, MSI, INTa) only once in probe() and
initialize appropriate structures.
Signed-off-by: Dmitry Kravkov <[email protected]>
Signed-off-by: Eilon Greenstein <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
57712 HW supported with same set of features as for 57710/57711
Signed-off-by: Dmitry Kravkov <[email protected]>
Signed-off-by: Eilon Greenstein <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
MF (multi-function) mode supported not only by 57711E (E1H) devices,
but also by coming 57712E, then we use more generic names.
Signed-off-by: Dmitry Kravkov <[email protected]>
Signed-off-by: Eilon Greenstein <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
This is the new FW HSI blob and the relevant definitions without logic changes.
It also included code adaptation for new HSI. New features are not enabled.
New FW/HSI includes:
- Support for 57712 HW
- Future support for VF (not used)
- Improvements in FW interrupts scheme
- FW FCoE hooks (stubs for future usage)
Signed-off-by: Dmitry Kravkov <[email protected]>
Signed-off-by: Michael Chan <[email protected]>
Signed-off-by: Eilon Greenstein <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Initialize the rings only during cnic_uio_open() and shut them down
during cnic_uio_close(). Check for the new bit CNIC_LCL_FL_RINGS_INITED
before checking for ring interrupt.
Signed-off-by: Michael Chan <[email protected]>
Signed-off-by: Dmitry Kravkov <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
The pfid (physical function ID) is the same as PCI function on production
devices. The pfid for future devices will be different and will be used
for internal memory offsets.
Signed-off-by: Michael Chan <[email protected]>
Signed-off-by: Dmitry Kravkov <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Preparation work for upcoming firmware interface changes.
Signed-off-by: Michael Chan <[email protected]>
Signed-off-by: Dmitry Kravkov <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Commit 2ddaad397c entitled "tg3: Use
netif_set_real_num_{rx,tx}_queues()" added a new call to
netif_set_real_num_rx_queues in tg3_enable_msix(). This call also needs
to be added to the legacy path to correctly reflect the actual number of
rx queues.
Signed-off-by: Matt Carlson <[email protected]>
Acked-by: Eric Dumazet <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
net/mac80211/scan.c: In function ‘ieee80211_scan_cancel’:
net/mac80211/scan.c:794: warning: ‘finish’ may be used uninitialized in this function
Signed-off-by: John W. Linville <[email protected]>
When multiple interfaces are actively trying
to associate with the same BSS, they may both
find that the BSS isn't there and then try to
unlink it. This can cause errors since the
unlinking code can't currently deal with items
that have already been unlinked.
Normally this doesn't happen as most people
don't try to use multiple station interfaces
that associate at the same time too.
Fix this by using the list entry as a flag to
see if the item is still on a list.
Cc: [email protected]
Reported-by: Ben Greear <[email protected]>
Tested-by: Hun-Kyi Wynn <[email protected]>
Signed-off-by: Johannes Berg <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
This information is already available in mac80211, we just need to export it
via cfg80211 and nl80211.
Signed-off-by: Bruno Randolf <[email protected]>
Acked-by: Johannes Berg <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
This reverts commit 7acc7c683a. It was
applied to avoid possible warning in ieee80211_restart_hw, however
reason of the warning were races in mac80211, currently hopefully fixed.
Not reseting device when performing scan is bad for two reasons.
When forcing reset from iwl_check_stuck_queue(), in case of fail,
reset will be repeated until scan finish. But since firmware is in bad
shape, scan only finish after scan_check work (about 7s). So we will
delay the reset, what is not good behaviour.
When forcing reset from iwl_recover_from_statistics(), we will not
repeat the reset, so we will not perform reset at all when needed.
Signed-off-by: Stanislaw Gruszka <[email protected]>
Acked-by: Johannes Berg <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
We nulify local->scan_req on failure in __ieee80211_start_scan, so
__ieee80211_scan_completed will not call cfg80211_scan_done. Fix that.
Signed-off-by: Stanislaw Gruszka <[email protected]>
Acked-by: Johannes Berg <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
When performing hw scan and not abort it, __ieee80211_scan_completed()
is currently called from scan work, so does not need to reschedule work
to call drv_hw_scan().
Signed-off-by: Stanislaw Gruszka <[email protected]>
Acked-by: Johannes Berg <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
This is partial revert and fix for commit
85f72bc839 "mac80211: only cancel
software-based scans on suspend"
When cfg80211 request the scan and mac80211 perform some management work,
we defer the scan request. We do not canceling such requests when calling
ieee80211_scan_cancel(), because of SCAN_SW_SCANNING bit check just
before the call. So fix that problem.
Another problem, which commit 85f72bc839
tries to solve, is we can not cancel HW scan. Hence patch make
ieee80211_scan_cancel() ignore HW scan (see code comments). Keeping
local->mtx lock assures that the deferred scan will not become
"working" HW scan.
Signed-off-by: Stanislaw Gruszka <[email protected]>
Acked-by: Johannes Berg <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
We are taking local->mtx inside __ieee80211_scan_completed(), but just
before call to that function we drop the lock. Dropping/taking lock is not
good, because can lead to hard to understand race conditions.
Patch split scan_completed() code into two functions, first must be called
with local->mtx taken and second without it.
Signed-off-by: Stanislaw Gruszka <[email protected]>
Acked-by: Johannes Berg <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Use goto instruction to call __ieee80211_scan_completed only ones in
ieee80211_scan_work. This is prepare for the next patch.
Signed-off-by: Stanislaw Gruszka <[email protected]>
Acked-by: Johannes Berg <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Yikes! The error return keeps a netdev reference
and the rdev mutex locked, fix that!
Signed-off-by: Johannes Berg <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
This adds API to allow adding per-station GTKs,
updates mac80211 to support it, and also allows
drivers to remove a key from hwaccel again when
this may be necessary due to multiple GTKs.
Signed-off-by: Johannes Berg <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
When roaming while we have active BA session,
we can end up transmitting delBA frames to
the old AP while we're already on the new AP's
channel, which can cause warnings.
Simply avoid sending those frames, but still
tear down the internal session state, since
they are not really necessary anyway as we
will implicitly disassociate when sending the
association to the new AP.
Signed-off-by: Johannes Berg <[email protected]>
Acked-by: Luis R. Rodriguez <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
drv_config callback is called only after the ack for the nullframe
is received and so driver need not do anything special for this.
So remove NULLFUNC_COMPLETED, PS_ENABLED flags and bf_isnullfunc
flags from ath9k as mac80211 already handles them properly.
Signed-off-by: Senthil Balasubramanian <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
After the last rounds of cleanup, these functions are now functionally
equivalent and can thus be merged.
Also get rid of some excessive (and redundant) debug messages.
Signed-off-by: Felix Fietkau <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
The code gets more concise and readable when making the new ANI functions
fall back to the old ones if ANI v2 is disabled. This also makes further code
cleanup easier.
Signed-off-by: Felix Fietkau <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
Split out the PHY error counter update from ath9k_hw_ani_monitor_*, reuse
it in ath9k_hw_proc_mib_event (merged from ath9k_hw_proc_mib_event_old
and ath9k_hw_proc_mib_event_new).
Signed-off-by: Felix Fietkau <[email protected]>
Signed-off-by: John W. Linville <[email protected]>
John Heffner