With the move to phylib (commit e6b043d) I was seeing sporadic "MDIO write
timeout" messages. Measure of the actual time spent showed latency times of
more than 1600us.
This patch uses the MII event indication of the FEC hardware to detect
completion of MDIO transactions.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: David S. Miller <davem@davemloft.net>
All these files use the big kernel lock in a trivial
way to serialize their private file operations,
typically resulting from an earlier semi-automatic
pushdown from VFS.
None of these drivers appears to want to lock against
other code, and they all use the BKL as the top-level
lock in their file operations, meaning that there
is no lock-order inversion problem.
Consequently, we can remove the BKL completely,
replacing it with a per-file mutex in every case.
Using a scripted approach means we can avoid
typos.
file=$1
name=$2
if grep -q lock_kernel ${file} ; then
if grep -q 'include.*linux.mutex.h' ${file} ; then
sed -i '/include.*<linux\/smp_lock.h>/d' ${file}
else
sed -i 's/include.*<linux\/smp_lock.h>.*$/include <linux\/mutex.h>/g' ${file}
fi
sed -i ${file} \
-e "/^#include.*linux.mutex.h/,$ {
1,/^\(static\|int\|long\)/ {
/^\(static\|int\|long\)/istatic DEFINE_MUTEX(${name}_mutex);
} }" \
-e "s/\(un\)*lock_kernel\>[ ]*()/mutex_\1lock(\&${name}_mutex)/g" \
-e '/[ ]*cycle_kernel_lock();/d'
else
sed -i -e '/include.*\<smp_lock.h\>/d' ${file} \
-e '/cycle_kernel_lock()/d'
fi
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: netdev@vger.kernel.org
Signed-off-by: David S. Miller <davem@davemloft.net>
All these files use the big kernel lock in a trivial
way to serialize their private file operations,
typically resulting from an earlier semi-automatic
pushdown from VFS.
None of these drivers appears to want to lock against
other code, and they all use the BKL as the top-level
lock in their file operations, meaning that there
is no lock-order inversion problem.
Consequently, we can remove the BKL completely,
replacing it with a per-file mutex in every case.
Using a scripted approach means we can avoid
typos.
file=$1
name=$2
if grep -q lock_kernel ${file} ; then
if grep -q 'include.*linux.mutex.h' ${file} ; then
sed -i '/include.*<linux\/smp_lock.h>/d' ${file}
else
sed -i 's/include.*<linux\/smp_lock.h>.*$/include <linux\/mutex.h>/g' ${file}
fi
sed -i ${file} \
-e "/^#include.*linux.mutex.h/,$ {
1,/^\(static\|int\|long\)/ {
/^\(static\|int\|long\)/istatic DEFINE_MUTEX(${name}_mutex);
} }" \
-e "s/\(un\)*lock_kernel\>[ ]*()/mutex_\1lock(\&${name}_mutex)/g" \
-e '/[ ]*cycle_kernel_lock();/d'
else
sed -i -e '/include.*\<smp_lock.h\>/d' ${file} \
-e '/cycle_kernel_lock()/d'
fi
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: Karsten Keil <isdn@linux-pingi.de>
Cc: netdev@vger.kernel.org
Signed-off-by: David S. Miller <davem@davemloft.net>
Avoid two extra instructions in sock_free(), to reload
skb->truesize and skb->sk
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
a new boolean flag no_autobind is added to structure proto to avoid the autobind
calls when the protocol is TCP. Then sock_rps_record_flow() is called int the
TCP's sendmsg() and sendpage() pathes.
Signed-off-by: Changli Gao <xiaosuo@gmail.com>
----
include/net/inet_common.h | 4 ++++
include/net/sock.h | 1 +
include/net/tcp.h | 8 ++++----
net/ipv4/af_inet.c | 15 +++++++++------
net/ipv4/tcp.c | 11 +++++------
net/ipv4/tcp_ipv4.c | 3 +++
net/ipv6/af_inet6.c | 8 ++++----
net/ipv6/tcp_ipv6.c | 3 +++
8 files changed, 33 insertions(+), 20 deletions(-)
Signed-off-by: David S. Miller <davem@davemloft.net>
xtsonic_open() doesn't check sonic_open() return code. If it is error
we must free requested IRQ.
Signed-off-by: Kulikov Vasiliy <segooon@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
macsonic_open() doesn't check sonic_open() return code. If it is error
we must free requested IRQs.
Signed-off-by: Kulikov Vasiliy <segooon@gmail.com>
Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
jazzsonic_open() doesn't check sonic_open() return code. If it is error
we must free requested IRQ.
Signed-off-by: Kulikov Vasiliy <segooon@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There was a missing "else" statement so the original code overflowed if
->master->name was too long. Also the ->slave and ->master buffers can
hold names with 9 characters and a NULL so I cleaned it up to allow
another character.
Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This is an off by one bug because strlen() doesn't count the NULL
terminator. We strcpy() addr into a fixed length array of size
UNIX_PATH_MAX later on.
The addr variable is the name of the device being mounted.
CC: stable@kernel.org
Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If tc35815_init_one() fails we must unmap mapped regions.
Signed-off-by: Kulikov Vasiliy <segooon@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix up TX Queue Host Flow Control to cause an Egress Queue Status Update to
be generated when we run out of TX Queue Descriptors. This will, in turn,
allow us to restart a stopped TX Queue.
Signed-off-by: Casey Leedom <leedom@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since some of xxx_register_driver() can return error we must unregister
already registered drivers.
Signed-off-by: Kulikov Vasiliy <segooon@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Use spin_lock_irqsave instead of spin_lock in ax_interrupt because
the interrupt handler can also be invoked from ei_watchdog.
Signed-off-by: Ken Kawasaki <ken_kawasaki@spring.nifty.jp>
Signed-off-by: David S. Miller <davem@davemloft.net>
Based upon a report by Randy Dunlap.
DSA needs PHYLIB, but PHYLIB needs NET_ETHERNET. So, in order
to select PHYLIB we have to make DSA depend upon NET_ETHERNET.
Signed-off-by: David S. Miller <davem@davemloft.net>
not all of the ICMP packets need an IP header payload, so we check the length
of the skbs only when the packets should have an IP header payload.
Based upon analysis and initial patch by Rodrigo Partearroyo González.
Signed-off-by: Changli Gao <xiaosuo@gmail.com>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
----
net/sched/act_nat.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)
Signed-off-by: David S. Miller <davem@davemloft.net>
Michael has been out of the scene for a while now,
but despite Michael's absence, p54 is still maintained.
Cc: Michael Wu <aluminum.tape@gmail.com>
Cc: Larry Finger <Larry.Finger@lwfinger.net>
Cc: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: Christian Lamparter <chunkeey@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
AR5008+ and AR9003 currently use two separate implementations of the
ath9k_hw_loadnf function. There are three main differences:
- PHY registers for AR9003 are different
- AR9003 always uses 3 chains, earlier versions are more selective
- The AR9003 variant contains a fix for NF load timeouts
This patch merges the two implementations into one, storing the
register array in the ath_hw struct. The fix for NF load timeouts is
not just relevant for AR9003, but also important for earlier hardware,
so it's better to just keep one common implementation.
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Occasionally the hardware can send out tx status information with the wrong
TID. In that case, the BA status cannot be trusted and the aggregate
must be retransmitted.
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
If the receive path gets stuck, a full hardware reset is necessary to
recover from it. If this happens during a scan, the whole scan might fail,
as each channel change bypasses the full reset sequence.
Fix this by resetting the fast channel change flag if stopping the
receive path fails.
This will reduce the number of error messages that look like this:
ath: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x40000020
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Cc: stable@kernel.org
Signed-off-by: John W. Linville <linville@tuxdriver.com>
PDADC values were only generated for values surrounding the target
index, however not for the target index itself, leading to a minor
error in the generated curve.
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Cc: stable@kernel.org
Signed-off-by: John W. Linville <linville@tuxdriver.com>
On AR9285, the antenna switch configuration register uses more than just
16 bits. Because of an arbitrary mask applied to the EEPROM value that
stores this configuration, diversity was broken in some cases, leading
to a significant degradation in signal strength.
Fix this by changing the callback to return a 32 bit value and remove
the arbitrary mask.
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Cc: stable@kernel.org
Signed-off-by: John W. Linville <linville@tuxdriver.com>
All callback functions are gathered in rt2x00dev->ops except
for the callback functions which are used in rt2800lib to
acces rt2800pci/usb.
Move the priv pointer from rt2x00dev to rt2x00dev->ops and
rename it to drv to make it obvious that it is the driver callback
structure.
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Large parts of the firmware initialization are shared
between rt2800pci and rt2800usb. Move this code into
rt2800lib.
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Currently rt2x00 uses the TSF_SYNC_BEACON mode for all beaconing
interface types. However, TSF_SYNC_BEACON is meant for IBSS networks and
thus implements TSF merging in the hardware. Rename TSF_SYNC_BEACON to
TSF_SYNC_ADHOC to better express its purpose and introduce the missing
TSF sync mode TSF_SYNC_AP_NONE which should be used for beaconing modes
that don't need TSF merging.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Activating the TBTT interrupt when a beacon interval of 0 is configured
results in an interrupt storm causing the machine to hang. Hence,
initialize the beacon interval to a reasonable default of 100TUs.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Using the set_tim callback without managing the DTIM count and the
broad- and multicast buffering in hw, fw or the driver results in wrong
DTIM count values being sent out in beacons. Since all PCI drivers
fetch new beacons periodically and hence get an updated TIM we can just
remove the set_tim callback from these.
The rt2x00 USB drivers don't update the beacon periodically and thus
rely on the set_tim callback to get a correct TIM for beacon
transmission. USB devices still suffer from the DTIM count being wrong
under some circumstances but removing the set_tim callback from these
would cause more harm then good.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Use separate mac80211_ops for rt2800pci and rt2800usb in preparation
for further fixes. This shouldn't introduce functional changes.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Updating the beacon on pre tbtt instead of beacondone allows much lower
latency in regard to TIM updates. Hence, use the pre tbtt interrupt for
updating the beacon in rt2800pci (older devices don't provide a pre tbtt
interrupt).
Also, add a new driver flag to indicate if a driver has pre tbtt support
or not and implement the according behavior in rt2x00lib.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Although mac80211 buffers broad- and mutlicast frames for us in AP mode
we still have to send them out after a DTIM beacon. Implement this
behavior by sending out the buffered frames when the beacondone
interrupt is processed.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Since all rt2x00 PCI drivers use threaded interrupts now we don't need
to schedule a work just to update the beacon. The only place where the
beacon still gets updated in atomic context is from the set_tim
callback. Hence, move the work scheduling there.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Use threaded interrupts for all rt2x00 PCI devices.
This has several generic advantages:
- Reduce the time we spend in hard irq context
- Use non-atmic mac80211 functions for rx/tx
Furthermore implementing broad- and multicast buffering will be
much easier in process context while maintaining low latency and
updating the beacon just before transmission (pre tbtt interrupt)
can also be done in process context.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
rt2800_write_beacon is the only function which uses
EXPORT_SYMBOL instead of EXPORT_SYMBOL_GPL. All symbols
in rt2x00 should however use the GPL restricted export.
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Acked-by: Gertjan van Wingerde <gwingerde@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Implement watchdog monitoring for USB devices (PCI support can
be added later). This will determine if URBs being uploaded to
the hardware are actually returning. Both rt2500usb and rt2800usb
have shown that URBs being uploaded can remain hanging without
being released by the hardware.
By using this watchdog, a queue can be reset when this occurs.
For rt2800usb it has been tested that the connection is preserved
even though this interruption.
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Currently vgc_level_reg and vgc_level are equal to eachother,
while the purpose of vgc_level_reg is the value last written
to the register and is remembered through link tuning resets.
The vgc_level is the currently active level, which is
reset during link tuning resets.
The usage of these variables depends on the drivers, some drivers
need both, while others need only one of the two.
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
While scanning the link tuner must be disabled. Otherwise
it will interfere with receiving all beacons for each channel
due to changing sensitivity levels.
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Rename CONFIG_DISABLE_LINK_TUNING to DRIVER_SUPPORT_LINK_TUNING
Link tuning support is not only based on EEPROM decisions, but
also if the device actually supports it.
Currently only rt2500usb doesn't support link tuning because
of hardware problems. But rt2800usb is also suspected of having
problems with link tuning.
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
The RSSI values in the RXWI descriptor aren't true RSSI
values. Instead they are more like the AGC values similar
to rt61pci. And as such, it needs the same conversion
before it can be passed to rt2x00lib/mac80211.
This requires the struct queue_entry to be passed to
rt2800_process_rxwi rather then the skb structure which
is contained in the queue_entry. This is required to
obtain the lna_gain information from the rt2x00_dev structure.
This fixes connection problems when using wpa_supplicant
which would try to connect to the worst AP's rather then the
best ones.
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Limit the txpower per rate by the approriate values in the eeprom.
This avoids too high txpower values resulting in bad tx performance.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
In hostap_add_interface() we do:
sprintf(dev->name, "%s%s", prefix, name);
dev->name has IFNAMSIZ (16) characters.
prefix is local->dev->name.
name is "wds%d"
strlen() returns the number of characters in the string not counting the
NULL so if we have a string with 11 characters we get "12345678901wds%d"
which is 16 characters and a NULL so we're past the end of the array.
Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
We must free priv->eeprom allocated in adm8211_read_eeprom().
Signed-off-by: Kulikov Vasiliy <segooon@gmail.com>
Acked-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This test is off by one because strlen() doesn't include the NULL
terminator.
Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This patch changes __attribute__ ((packed)) annotations to __packed in wl1271
code that was introduced in a recent patch in wireless-testing.
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Luciano Coelho <luciano.coelho@nokia.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Existing "ethtool -s ethX wol X" command configures hostsleep
parameters, but those are activated only during suspend/resume,
there is no way to configure host sleep without actual suspend.
This patch adds debugfs command to enable/disable host sleep based on
already configured host sleep parameters using wol command.
Signed-off-by: Amitkumar Karwar <akarwar@marvell.com>
Signed-off-by: Kiran Divekar <dkiran@marvell.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Ever since mac80211/drivers are no longer
fully in charge of keeping track of the
auth status, trying to make them do so will
fail. Instead of warning and reporting the
deauthentication to userspace, cfg80211 must
simply ignore it so that spurious
deauthentications, e.g. before starting
authentication, aren't seen by userspace as
actual deauthentications.
Cc: stable@kernel.org
Reported-by: Paul Stewart <pstew@google.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
