Helmut Schaa [Wed, 7 Sep 2011 18:11:26 +0000 (20:11 +0200)]
rt2x00: Avoid unnecessary uncached
Reading the TX desciptor words from coherent memory is always uncached
and potentially slow. Hence, don't read the TX descriptor prior to
writing it since we update all fields anyway.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com> Acked-by: Gertjan van Wingerde <gwingerde@gmail.com> Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Helmut Schaa [Wed, 7 Sep 2011 18:11:03 +0000 (20:11 +0200)]
rt2x00: Update some TX descriptor
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com> Acked-by: Gertjan van Wingerde <gwingerde@gmail.com> Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Helmut Schaa [Wed, 7 Sep 2011 18:10:45 +0000 (20:10 +0200)]
rt2x00: Remove incorrect led blink
rt2800 devices are not capable of configuring arbitrary LED on/off
periods. The LED_CFG register fields ON_PERIOD and OFF_PERIOD are only
used by the hw when the LED mode is set to "blink upon TX".
Hence, remove the blink callback. This will result in software emulation
for LED blinking.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com> Acked-by: Gertjan van Wingerde <gwingerde@gmail.com> Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Helmut Schaa [Wed, 7 Sep 2011 18:10:25 +0000 (20:10 +0200)]
rt2x00: Add LED_CFG register description
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com> Acked-by: Gertjan van Wingerde <gwingerde@gmail.com> Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Helmut Schaa [Wed, 7 Sep 2011 18:10:02 +0000 (20:10 +0200)]
rt2x00: Minor optimizazion in txdone path
We can save an indirect function call + some instructions for fetching
the actual function pointer by passing the driver specific txwi pointer
directly from rt2800pci/rt2800usb to rt2800lib instead of using the
rt2800_drv_get_txwi callback.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com> Acked-by: Gertjan van Wingerde <gwingerde@gmail.com> Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Trivial fixes to allow compilation without warnings when debug
compilation flag isn't set.
Also fix the compilation when debugfs flag isn't set.
Fix a warning: unused priv pointer on the way.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com> Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
iwlagn: iwl-trans.c can't dereference iwl_priv any more
This reaches encapsulation for this file. In order to reach this:
* move priv->valid_context to iwl_shared
* move the last_rejected initialization to the upper layer
* define a wrapper iwl_nic_config in the upper layer that calls to
cfg->lib->ops->nic_config
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com> Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
The traffic log debugfs handlers were mistakenly moved to the
transport layer because they print the pointers of the Tx / Rx
queues. The pointers of the queues can be fetched by another debugfs
handler.
So bring the traffic log back to the upper layer and remove the part
that reads the Tx / Rx queues' pointers.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com> Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Move a few declarations needed by the transport layer to iwl-shared.h
Move iwl_cmd_meta, iwl_tx_queue and friends to the internal transport
header file.
Move iwl_device_cmd iwl_host_cmd and friends to iwl-trans.h since
these structs are used in the API to the transport layer.
Move get_cmd_string to the upper layer with a declaration in
iwl-shared.h.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com> Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
iwlagn: remove priv dereferences from the transport layer
Another round of clean up on the transport layer.
Define a wrapper around wiphy_rfkill_set_hw_state to prevent the
transport layer from accessing priv->hw.
Also move wait_command_queue to the transport layer.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com> Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
The latter was the only place that called the first. The two
functions were split when the iwl_pci_probe called iwl_pci_down
upon failure in the probe. Since iwl_pci_probe undoes its work by
itself, there is no need to split between iwl_pci_down, and
iwl_pci_remove.
Thanks to Pavel Roskin for pointing that out.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com> Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Wey-Yi Guy [Tue, 6 Sep 2011 16:31:16 +0000 (09:31 -0700)]
iwlagn: only perform necessary calibration at init time
During init time, only the necessary calibration should be performed. This
not only save time, also avoid uCode crash because lack of necessary information.
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
ath9k_hw: Fix magnitude/phase average in TxIQ Calibration
The commit "ath9k_hw: Fix Tx IQ Calibration hang issue in
AR9003 chips" did not consider more than one potential sample
while calculating magnitude/phase average if more than one
sample has the same value which could affect post-processing
of outlier detection that causes an undesirable Tx IQ
correction value will be assigned to tx gain settings where
outlier happens.
Cc: Kai Shi <kaishi@qca.qualcomm.com> Reported-by: Paul Stewart <pstew@google.com> Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com> Signed-off-by: Paul Stewart <pstew@google.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Currently the samples debugfs which maintains the snapshorts of
mac/bb only on associated state. Hence to cover issues on idle
state, the samples are taken whenever the driver is ready.
Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Johannes Berg [Tue, 6 Sep 2011 12:13:06 +0000 (14:13 +0200)]
mac80211: further optimise buffer expiry timer
Juuso optimised the timer to not run all the
time in commit 3393a608c4979a94d1887efc05b7.
However, after that it will still run once
more even if all frames just expired. Fixing
that also makes the function return value a
little clearer in the process.
Also, while at it, change the return value
to bool (instead of int).
Cc: Juuso Oikarinen <juuso.oikarinen@nokia.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
b43: drop Copyright for not really copyrightable info
As discussed in "Licensing wlc_phy_radio.h and brcm80211" (Message-ID:
<AANLkTinQMRGkOvi4e4JMsHGV_BF_4Rf5oHCtZgQjwQ1g@mail.gmail.com>) content
of that file is not copyrightable, just names and numbers.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Minstrel HT tries very hard to establish a BA session with
each peer once there's some data on the way. However the
stack does not inform minstrel if an aggregation session
is already in place, so it keeps trying and wastes good
cycles in the tx status path.
[ 8149.946393] Open BA session requested for $AP tid 0
[ 8150.048765] Open BA session requested for $AP tid 0
[ 8150.174509] Open BA session requested for $AP tid 0
[ 8150.274376] Open BA session requested for $AP tid 0
...
Signed-off-by: Christian Lamparter <chunkeey@googlemail.com> Acked-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Felix Fietkau [Fri, 2 Sep 2011 23:40:27 +0000 (01:40 +0200)]
ath9k: implement .get_antenna and .set_antenna
On MIMO chips this can be used to enable/disable hardware chains, ensuring
that the MCS information is updated accordingly.
On non-MIMO chips with rx diversity (e.g. 9285), this configures the rx
input antenna.
Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Felix Fietkau [Fri, 2 Sep 2011 23:40:26 +0000 (01:40 +0200)]
ath9k: merge reset related functions
reduces unnecessary code duplication. Also takes the sc_pcu_lock within
ath_reset instead of callsites, which makes it possible to always cancel
all queued work items before the reset, possibly fixing a few race
conditions (work items vs reset) along with it.
Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Felix Fietkau [Fri, 2 Sep 2011 23:40:25 +0000 (01:40 +0200)]
ath9k: always call ath_reset from workqueue context
This makes it much easier to add further rework to avoid race conditions
between reset and other work items.
Move other functions to make ath_reset static.
Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
mac80211: stop tx before doing hw config and rate update
The assumption is that during the hw config, transmission was
already stopped by mac80211. Sometimes the AP can be switching
b/w the ht modes due to intolerant or etc where STA is in
the middle of transmission. In such scenario, buffer overflow
was observed at driver side. And also before updating the rate
control, the frames are continued to xmited with older rates.
This patch ensures that the frames are always xmitted with
updated rates and avoid buffer overflow.
Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com> Reviewed-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Javier Cardona [Tue, 6 Sep 2011 20:05:21 +0000 (13:05 -0700)]
mac80211: Stop forwarding mesh traffic when tx queues are full
Tx flow control for non-mesh modes of operation only needs to act on the
net device queues: when the hardware queues are full we stop accepting
traffic from the net device. In mesh, however, we also need to stop
forwarding traffic. This patch checks the hardware queues before
attempting to forward a mesh frame.
Signed-off-by: Javier Cardona <javier@cozybit.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Thomas Pedersen [Thu, 1 Sep 2011 19:32:14 +0000 (12:32 -0700)]
mac80211: check if mesh frame is in RMC after decrypt
To check whether a frame is in the RMC, we need access to the mesh
header. This header is encrypted in encrypted data frames, so make this
check after the frame has been decrypted.
Signed-off-by: Thomas Pedersen <thomas@cozybit.com> Signed-off-by: Javier Cardona <javier@cozybit.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Felix Fietkau [Sat, 20 Aug 2011 13:53:55 +0000 (15:53 +0200)]
mac80211: make ieee80211_send_bar available for drivers
To properly maintain the peer's block ack window, the driver needs to be
able to control the new starting sequence number that is sent along with
the BlockAckReq frame.
Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Sven Neumann [Tue, 30 Aug 2011 21:38:54 +0000 (23:38 +0200)]
cfg80211: reorder code to obsolete forward declaration
Reorder functions to remove the need for a forward declaration
introduced by the last commit.
Signed-off-by: Sven Neumann <s.neumann@raumfeld.com> Cc: John W. Linville <linville@tuxdriver.com> Cc: Luis R. Rodriguez <mcgrof@gmail.com> Cc: Daniel Mack <daniel@zonque.org> Cc: linux-wireless@vger.kernel.org Acked-by: Luis R. Rodriguez <mcgrof@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Sven Neumann [Tue, 30 Aug 2011 21:38:53 +0000 (23:38 +0200)]
cfg80211: hold reg_mutex when updating regulatory
The function wiphy_update_regulatory() uses the static variable
last_request and thus needs to be called with reg_mutex held.
This is the case for all users in reg.c, but the function was
exported for use by wiphy_register(), from where it is called
without the lock being held.
Fix this by making wiphy_update_regulatory() private and introducing
regulatory_update() as a wrapper that acquires and holds the lock.
Signed-off-by: Sven Neumann <s.neumann@raumfeld.com> Cc: John W. Linville <linville@tuxdriver.com> Cc: Luis R. Rodriguez <mcgrof@gmail.com> Cc: Daniel Mack <daniel@zonque.org> Cc: linux-wireless@vger.kernel.org Acked-by: Luis R. Rodriguez <mcgrof@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Pavel Roskin [Thu, 1 Sep 2011 15:54:12 +0000 (11:54 -0400)]
cfg80211: print bandwidth in chan_reg_rule_print_dbg()
Two spaces and the second "KHz" suggest that the code author meant to
print the bandwidth but forgot it. The code appears in commit e702d3cf
already with two spaces and "KHz" in place of the bandwidth.
Signed-off-by: Pavel Roskin <proski@gnu.org> Acked-by: Luis R. Rodriguez <mcgrof@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Luciano Coelho [Wed, 31 Aug 2011 13:01:48 +0000 (16:01 +0300)]
nl80211/cfg80211: add match filtering for sched_scan
Introduce filtering for scheduled scans to reduce the number of
unnecessary results (which cause useless wake-ups).
Add a new nested attribute where sets of parameters to be matched can
be passed when starting a scheduled scan. Only scan results that
match any of the sets will be returned.
At this point, the set consists of a single parameter, an SSID. This
can be easily extended in the future to support more complex matches.
Signed-off-by: Luciano Coelho <coelho@ti.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Dan Carpenter [Wed, 31 Aug 2011 06:38:30 +0000 (09:38 +0300)]
iwlwifi: signedness bug in iwl_trans_pcie_tx_agg_alloc()
unsigned shorts and unsigned chars are never == -1.
Signed-off-by: Dan Carpenter <error27@gmail.com> Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Dan Carpenter [Wed, 31 Aug 2011 06:37:42 +0000 (09:37 +0300)]
iwlwifi: fix double assign in iwl_start_tx_ba_trans_ready()
"vif" is assigned twice. We can remove the first one.
This silences a Smatch warning that "ctx" could be one step past the
end of the priv->contexts[] array.
Signed-off-by: Dan Carpenter <error27@gmail.com> Reviewed-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Dan Carpenter [Tue, 30 Aug 2011 19:16:15 +0000 (22:16 +0300)]
mac80211: handle allocation failures in mesh_pathtbl_init()
The calls to kzalloc() weren't checked here and it upsets the static
checkers. Obviously they're not super likely to fail, but we might
as well add some error handling.
Signed-off-by: Dan Carpenter <error27@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Felix Fietkau [Mon, 29 Aug 2011 16:57:54 +0000 (18:57 +0200)]
ath9k: fix checks for first subframe delimiter padding
The commit "ath9k_hw: Fix exceed transmission burst-time of 5GHz" added
a padding of 60 delimiters on the first subframe to work around an issue
on AR9380, but it lacked the checks to prevent it from being applied to
pre-AR9380, enterprise AR9380 or AR9580+
Signed-off-by: Felix Fietkau <nbd@openwrt.org> Cc: Rajkumar Manoharan <rmanohar@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
This will avoid mac80211 to trigger PS mode for connected station
based on the PM bit of incoming frames. AP firmware is capable of
handling such frames and buffering TX frames destined to the
stations that are in PS mode.
Signed-off-by: Yogesh Ashok Powar <yogeshp@marvell.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Vivek Natarajan [Mon, 29 Aug 2011 08:53:03 +0000 (14:23 +0530)]
cfg80211/nl80211: Indicate roaming feature capability to userspace.
When the rssi of the current AP drops, both wpa_supplicant and the
firmware may do a background scan to find a better AP and try to
associate. Since firmware based roaming is faster, inform
wpa_supplicant to avoid roaming and let the firmware decide to
roam if necessary.
For fullmac drivers like ath6kl, it is just enough to provide the
ESSID and the firmware will decide on the BSSID. Since it is not
possible to do pre-auth during roaming for fullmac drivers, the
wpa_supplicant needs to completely disconnect with the old AP and
reconnect with the new AP. This consumes lot of time and it is
better to leave the roaming decision to the firmware.
Signed-off-by: Vivek Natarajan <nataraja@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Felix Fietkau [Sun, 28 Aug 2011 19:11:01 +0000 (21:11 +0200)]
mac80211: retry sending failed BAR frames later instead of tearing down aggr
Unfortunately failed BAR tx attempts happen more frequently than I
expected, and the resulting aggregation teardowns cause performance
issues, as the aggregation session does not always get re-established
properly.
Instead of tearing down the entire aggr session, we can simply store the
SSN of the last failed BAR tx attempt, wait for the first successful
tx status event, and then send another BAR with the same SSN.
Signed-off-by: Felix Fietkau <nbd@openwrt.org> Cc: Helmut Schaa <helmut.schaa@googlemail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Rafał Miłecki [Sun, 28 Aug 2011 16:47:22 +0000 (18:47 +0200)]
ssb: use u16 for storing board rev
Specs say about size 2 (u16) and my 14e4:4727 has board rev 0x1211.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com> Acked-by: Larry Finger <Larry.Finger@lwfinger.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Felix Fietkau [Sat, 27 Aug 2011 23:52:10 +0000 (01:52 +0200)]
ath9k_hw: do not apply the 2.4 ghz ack timeout workaround to cts
It is only used to workaround interoperability issues related to longer
delays in receiving the block ack, so it is not necessary to apply it
to the CTS exchange.
Should improve throughput slightly, especially when there are lots
of retransmissions.
Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Javier Cardona [Tue, 6 Sep 2011 19:10:43 +0000 (12:10 -0700)]
mac80211: Defer tranmission of mesh path errors
Under failure conditions, the mesh stack sends PERR messages to the
previous sender of the failed frame. This happens in the tx feedback
path, in which the transmission queue lock may be taken. Avoid a
deadlock by sending the path error via the pending queue.
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Greg Dietsche [Sun, 28 Aug 2011 13:22:38 +0000 (08:22 -0500)]
iwlwifi: iwl-agn-rs.c: remove old comment
this comment refers to some code that was removed.
Signed-off-by: Greg Dietsche <Gregory.Dietsche@cuw.edu> Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Felix Fietkau [Sat, 27 Aug 2011 22:32:25 +0000 (00:32 +0200)]
ath9k: defer ath_tx_setup_buffer setup to the first tx attempt during aggr
With sequence number and buffer allocation deferred to when they're needed
for the first time, it becomes much easier to start dropping packets from
the tid queue if necessary, e.g. when latency suddenly increases. This can
lead to some future improvements in buffer management for better latency.
Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Felix Fietkau [Sat, 27 Aug 2011 22:32:23 +0000 (00:32 +0200)]
ath9k: move the sequence number from ath_frame_info to ath_buf
It is only necessary for BAW tracking and moving it to the ath_buf
makes it easier to add further improvements, such as deferring
seqno allocation in the aggregation path.
Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Felix Fietkau [Sat, 27 Aug 2011 22:32:22 +0000 (00:32 +0200)]
ath9k: clean up the aggregation tid queue
Use a sk_buff_head instead containing skbs instead of a list_head
containing ath_bufs. This makes it easier to decouple the aggregation
code from the ath_buf struct
Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
John W. Linville [Mon, 29 Aug 2011 19:47:58 +0000 (15:47 -0400)]
wl12xx/sdio_test.c: fix build breakage from WL127X_FW_NAME change
Commit c302b2c959164622558474871ae942da0e484a38 ("wl12xx: Use a single
fw for both STA and AP roles") changed the name of the firmware name
definition, breaking the build of wl12xx/sdio_test.c.
Signed-off-by: John W. Linville <linville@tuxdriver.com>
After analysis, this is not a bug, but a false warning. Nonetheless,
a cleanup is in order to prevent some future janitor proposing
the wrong fix, as I did in my original patch.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
ath9k: Add debugfs support for mac/baseband samples
This patch keep track of number of samples that includes
DMA debugs registers, PCU observe, CR, channel noise,
cycle conters, noisefloor history buffer and last N number
of tx and rx descriptor status. These samples are grouped
in table manner which dumping in debgufs.
Felix Fietkau [Sat, 27 Aug 2011 08:25:27 +0000 (10:25 +0200)]
ath9k: fix regression in sending aggregated packets
The recent commit "ath9k: Send legacy rated frames as unaggregated"
introduced a check to ensure that packets with non-MCS rates set in
the rate series will not be aggregated. However, it failed to check
if the rate series is valid before testing the flags, thus breaking
aggregation for normal MCS-only packets if the last series is unset.
Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
For AR9287 v1.3+ chips, MAC runs at 117MHz. But the initvals
IFS parameters are loaded based on 44/88MHz clockrate. So
eifs/usec from ini should not be used for AR9287 v1.3+.
The mentioned values are tested on 2 chain HT40 mode.
Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Larry Finger [Fri, 26 Aug 2011 18:41:38 +0000 (20:41 +0200)]
b43: Relax requirement for descriptors to be in the DMA zone
When 64-bit DMA was first used, there were problems with the
BCM4311 (14e4:4311). The problem was "fixed" by using the GFP_DMA
flag in the allocation of coherent ring descriptor memory.
The original problem is now believed to have been due to bugs in
the 64-bit DMA implementation in the rest of the kernel, and that
those bugs have been fixed. Accordingly, the requirement for the
descriptors to be in the DMA zone is relaxed.
Bounce buffers are left in the DMA zone.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net> Signed-off-by: Michael Buesch <m@bues.ch> Signed-off-by: John W. Linville <linville@tuxdriver.com>