Hi everyone,
I'm parting ways with Tieto today so my email is going to become
defunct. I'll be reachable via: kazikcz at gmail dot com.
Michał
On 1 February 2017 at 09:33, Pali Rohár wrote:
> On Tuesday 31 January 2017 07:59:18 Tony Lindgren wrote:
>> * Kalle Valo [170130 22:36]:
[...]
>> > * before distro updates linux-firmware create yours own deb/rpm/whatever
>> > package "wl1251-firmware" which installs your flavor of nvs file (or
On 20 January 2017 at 13:51, Kalle Valo wrote:
> Michal Kazior wrote:
>> Firmware files are versioned to prevent older
>> driver instances to load unsupported firmware
>> blobs. This is reflected with a fallback logic
>> which attempts to load several firmware files.
[] (ath10k_pci_napi_poll [ath10k_pci]) from
[] (net_rx_action+0xac/0x160)
Reported-by: Mohammed Shafi Shajakhan
Signed-off-by: Michal Kazior
---
Notes:
v2:
* move sta->uploaded check to if (sta) {} block [johannes]
* explain how the change fixes the bug [johannes]
net/mac80211
On 13 January 2017 at 09:16, Johannes Berg wrote:
> On Thu, 2017-01-12 at 15:28 +0100, Michal Kazior wrote:
>> Station structure is considered as not uploaded
>> (to driver) until drv_sta_state() finishes. This
>> call is however done after the structure is
>> attached
On 13 January 2017 at 08:24, Johannes Berg wrote:
>
>> Unless you then continue to use that sta pointer after you release
>> data_lock.
>
> Ouch, ok. That's rather strangely hidden though.
>
>> Consider this:
>>
>> > CPU0 CPU1
>> > 1 synchronize_net()
>> > 2drv_sta_state
On 12 January 2017 at 16:46, Johannes Berg wrote:
> On Thu, 2017-01-12 at 16:14 +0100, Michal Kazior wrote:
>> Station pointers are RCU protected so driver must
>> be extra careful if it tries to store them
>> internally for later use outside of the RCU
>> section it o
r itself can't be set to
NULL/ERR_PTR because it is later used in
sta_state() for extra sanity checks.
Signed-off-by: Michal Kazior
---
drivers/net/wireless/ath/ath10k/core.h | 1 +
drivers/net/wireless/ath/ath10k/mac.c | 18 ++
2 files changed, 19 insertions(+)
diff --
_htt_txrx_compl_task [ath10k_core])
[] (ath10k_pci_napi_poll+0x54/0xe8 [ath10k_pci])
[] (ath10k_pci_napi_poll [ath10k_pci]) from
[] (net_rx_action+0xac/0x160)
Reported-by: Mohammed Shafi Shajakhan
Signed-off-by: Michal Kazior
---
net/mac80211/tx.c | 17 +++--
1 file changed, 7 insertions(+), 10 deleti
On 13 December 2016 at 19:37, Erik Stromdahl wrote:
>
>
> On 12/13/2016 06:26 PM, Valo, Kalle wrote:
>> Michal Kazior writes:
>>
>>> On 13 December 2016 at 14:44, Valo, Kalle wrote:
>>>> Erik Stromdahl writes:
>>>>
>>>
On 13 December 2016 at 14:44, Valo, Kalle wrote:
> Erik Stromdahl writes:
>
>> Code refactorization:
>>
>> Moved the code for ep 0 in ath10k_htc_rx_completion_handler
>> to ath10k_htc_control_rx_complete.
>>
>> This eases the implementation of SDIO/mbox significantly since
>> the ep_rx_complete c
On 5 December 2016 at 14:56, Johannes Berg wrote:
> On Tue, 2016-11-29 at 10:05 -0800, gree...@candelatech.com wrote:
>> From: Ben Greear
>>
>> This fixes OOM when using pktgen to drive a wifi station at more than
>> the station can transmit. pktgen uses ndo_start_xmit instead of
>> going
>> thr
On 2 December 2016 at 01:24, Ben Greear wrote:
> On 12/01/2016 02:52 PM, Ben Greear wrote:
>>
>> On 08/19/2016 06:34 AM, Ben Greear wrote:
>>>
>>>
>>>
>>> On 08/18/2016 11:59 PM, Michal Kazior wrote:
>>>>
>>>> On 19 Aug
On 2 December 2016 at 03:29, wrote:
> From: Ben Greear
>
> This appears to fix a problem where ath10k firmware would crash,
> mac80211 would start re-adding interfaces to the driver, but the
> iterate-active-interfaces logic would then try to use the half-built
> interfaces. With a bit of extra
lowing warning, fix it.
> drivers/net/wireless/ath/ath10k/wmi-tlv.c: In function
> ‘ath10k_wmi_tlv_op_gen_vdev_start’:
> drivers/net/wireless/ath/ath10k/wmi-tlv.c:1647:33: warning: variable ‘noa’
> set but not used [-Wunused-but-set-variable]
>
> Fixes: ca996ec56608 ("ath
On 22 November 2016 at 16:31, Pali Rohár wrote:
> On Tuesday 22 November 2016 16:22:57 Michal Kazior wrote:
>> On 21 November 2016 at 16:51, Pali Rohár wrote:
>> > On Friday 11 November 2016 18:20:50 Pali Rohár wrote:
>> >> Hi! I will open discussion about mac ad
On 21 November 2016 at 16:51, Pali Rohár wrote:
> On Friday 11 November 2016 18:20:50 Pali Rohár wrote:
>> Hi! I will open discussion about mac address and calibration data for
>> wl1251 wireless chip again...
>>
>> Problem: Mac address & calibration data for wl1251 chip on Nokia N900
>> are store
On 22 November 2016 at 10:43, Matteo Grandi wrote:
> Dear Bob, Michal, all
>
> I've finally managed to have a 80MHz channel bandwidth, thanks to your hint!
> The problem was related to the CRDA that even if it looks correctly
> installed, it actually doesn't work as supposed.
> I downloaded and re
On 21 November 2016 at 15:41, Zefir Kurtisi wrote:
> On 11/21/2016 03:04 PM, Benjamin Berg wrote:
>> In the case that a spectral scan is enabled the PHY errors sent by the
>> hardware as part of the scanning might trigger the radar detection and
>> channels might be marked as 'unusable' incorrectl
On 21 November 2016 at 15:04, Benjamin Berg wrote:
> In the case that a spectral scan is enabled the PHY errors sent by the
> hardware as part of the scanning might trigger the radar detection and
> channels might be marked as 'unusable' incorrectly. This patch fixes
> the issue by preventing the
On 21 November 2016 at 10:46, Matteo Grandi wrote:
> Dear Bob, Michal, all,
>
> I've just tried your advices (actually I already tried it following
> the wireless.wiki.kernel web pages) and I had a look at the syslog
> while I was typing the commands
> At the beginning I have this
[...]
> Other (h
On 20 November 2016 at 02:07, Bob Copeland wrote:
> On Fri, Nov 18, 2016 at 10:43:06PM +0100, Matteo Grandi wrote:
>> So the question is: who decide if use MIMO and higher MCS hopefully on
>> 80MHz channel or not? Is the firmware? Is there a way to force the
>> interface to use 80MHz and/or MIMO?
On 17 November 2016 at 17:32, Erik Stromdahl wrote:
>
>
> On 11/15/2016 11:19 AM, Michal Kazior wrote:
>> On 14 November 2016 at 17:33, Erik Stromdahl
>> wrote:
>>> Code refactorization:
>>>
>>> Moved the code for ep 0 in ath10k_htc_rx_completi
On 15 November 2016 at 18:07, Erik Stromdahl wrote:
> On 11/15/2016 11:13 AM, Michal Kazior wrote:
>> On 14 November 2016 at 17:33, Erik Stromdahl
>> wrote:
>>> This patch changes the order in which the driver waits for the
>>> target to become ready a
On 14 November 2016 at 17:33, Erik Stromdahl wrote:
> Signed-off-by: Erik Stromdahl
> ---
> drivers/net/wireless/ath/ath10k/hw.h |1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/drivers/net/wireless/ath/ath10k/hw.h
> b/drivers/net/wireless/ath/ath10k/hw.h
> index 46142e9..ef45ecf 100
On 14 November 2016 at 17:33, Erik Stromdahl wrote:
> Address definitions for SDIO/mbox based chipsets.
>
> Signed-off-by: Erik Stromdahl
> ---
> drivers/net/wireless/ath/ath10k/hw.h | 53
> ++
> 1 file changed, 53 insertions(+)
>
> diff --git a/drivers/net/wir
On 14 November 2016 at 17:33, Erik Stromdahl wrote:
> Added functions implementing the following BMI commands:
>
> BMI_READ_SOC_REGISTER
> BMI_WRITE_SOC_REGISTER
>
> Reading and writing BMI registers is sometimes needed for
> SDIO chipsets.
I didn't see ath10k_bmi_write_soc_reg nor ath10k_bmi_rea
On 14 November 2016 at 17:33, Erik Stromdahl wrote:
> Signed-off-by: Erik Stromdahl
> ---
> drivers/net/wireless/ath/ath10k/htc.h |2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/drivers/net/wireless/ath/ath10k/htc.h
> b/drivers/net/wireless/ath/ath10k/htc.h
> index 589800a..df16a0
On 14 November 2016 at 17:33, Erik Stromdahl wrote:
> Code refactorization:
>
> Moved the code for ep 0 in ath10k_htc_rx_completion_handler
> to ath10k_htc_control_rx_complete.
>
> This eases the implementation of SDIO/mbox significantly since
> the ep_rx_complete cb is invoked directly from the S
On 14 November 2016 at 17:33, Erik Stromdahl wrote:
> This patch changes the order in which the driver waits for the
> target to become ready and the service connect of the HTC
> control service.
>
> The HTC control service is connected before the driver starts
> waiting for the HTC ready control
On 14 November 2016 at 17:33, Erik Stromdahl wrote:
> The RX trailer parsing is now capable of parsing lookahead reports.
> This is needed by SDIO/mbox.
It'd be useful to know what exactly lookahead will be used for. What
payload does it carry.
> Signed-off-by: Erik Stromdahl
> ---
> drivers/
Command structure and event flow doesn't seem to
be any different compared to existing
implementation for other firmware branches.
This patch effectively adds in-driver support for
spectral scanning on QCA61x4 and QCA9377.
Tested QCA9377 w/ WLAN.TF.1.0-00267-1.
Signed-off-by: Michal K
percpu_down_read+0x12/0x60
[ 168.116680] [] vfs_write+0xb8/0x1a0
[ 168.118293] [] SyS_write+0x46/0xa0
[ 168.119912] [] entry_SYSCALL_64_fastpath+0x1a/0xa4
[ 168.121737] Code: Bad RIP value.
[ 168.123318] RIP [< (null)>] (null)
Signed-off-by: Michal Kazio
On 24 October 2016 at 15:42, Simon Wunderlich wrote:
> On Monday, October 24, 2016 3:33:24 PM CEST Johannes Berg wrote:
>> > > I think it would be reasonable only if the target channel is the
>> > > one we are using and we have done CSA. But when scanning non-
>> > > operative channels I don't thi
On 24 October 2016 at 12:18, Matthias Klein wrote:
> I also try to get a pcie wifi card (Compex WLE600VX) running in the clearfog
> pro board with kernel 4.4.
>
> As I read in this thread the "irg.mode=1" shoud help:
> https://www.spinics.net/lists/linux-wireless/msg155685.html
>
> But it is not w
On 21 October 2016 at 18:25, Oliver Zemann wrote:
> The problem is gone with kernel 4.7.3 on armbian - both compex cards work
> (wle600, wle900), unfortunately that kernel does not support sfp (seems like
> its kind of proprietary drivers from clearfog). So its a driver issue but i
> have no clue
On 13 September 2016 at 23:11, Marty Faltesek wrote:
[...]
> +int
> +ath10k_cal_data_alloc(struct ath10k *ar, void **buf)
> +{
> + u32 hi_addr;
> + __le32 addr;
> + int ret;
> +
> + vfree(*buf);
> + *buf = vmalloc(QCA988X_CAL_DATA_LEN);
Shouldn't you use ar->hw_param
On 21 September 2016 at 18:27, Matteo Grandi wrote:
> Hi Michal,
> thanks for the reply:
> I've already loaded the ath10k_core with the rowmode parameters, this
> is the modinfo result:
> root@Tam:~# modinfo ath10k_core
> filename:
> /lib/modules/3.14.48-g408ccb9/kernel/drivers/net/wireless/ath/at
On 16 September 2016 at 12:56, Matteo Grandi wrote:
> Hello all,
[...]
> [8.589474] ath10k_pci :07:00.0: qca988x hw2.0 (0x4100016c,
> 0x043222ff sub :) fw 10.2.4.70.54 fwapi 5 bdapi 1 htt-ver 2.1
> wmi-op 5 htt-op 2 cal otp max-sta 128 raw 0 hwcrypto 1 features
> no-p2p,raw-mode
T
On 19 September 2016 at 00:52, Steffen Arntz wrote:
> Hi at all,
>
> I have an QCA988X based card in my laptop and have some "issues" with
> it, that I was not experiencing with my previous ath9k based card.
>
> The main issue is, that after suspend to RAM the card seems to be
> "hung up" and only
On 17 September 2016 at 00:37, Hsu, Ryan wrote:
[...]
>> + /* WMI and HTT may use separate HIF pipes and are not guaranteed to be
>> + * serialized properly implicitly.
>> + *
>> + * Moreover (most) WMI commands have no explicit acknowledges. It is
>> + * possible to infer
On 1 September 2016 at 22:52, Ben Greear wrote:
> On 09/01/2016 11:53 AM, Johannes Berg wrote:
>> On Thu, 2016-09-01 at 11:23 -0700, Ben Greear wrote:
>>>
>>> Could easily be that others are corrupted too, but since probe resp
>>> is bad, the association will not proceed.
>>
>> makes sense.
>>
>>>
On 24 August 2016 at 19:20, Ben Greear wrote:
> On 07/19/2016 03:34 AM, Michal Kazior wrote:
>>
>> HW Rx filters and masks are not configured
>> properly by firmware during boot sequences. The
>> MAC_PCU_ADDR1 is set to 0s instead of 1s which
>> allows the HW to
On 19 August 2016 at 03:26, wrote:
> From: Ben Greear
>
> I was seeing kernel crashes due to accessing freed memory
> while debugging a 9984 firmware that was crashing often.
>
> This patch fixes the crashes. I am not certain if there
> is a better way or not.
>
> Signed-off-by: Ben Greear
> -
On 29 July 2016 at 17:09, Ben Greear wrote:
> On 07/29/2016 07:52 AM, Benjamin Berg wrote:
[...]
>> Yeah, I am aware of the fact that the firmware may do internal resets
>> from time to time. The interesting question (and one for which I do not
>> know the answer) is whether we get a wmi or other
On 27 July 2016 at 14:59, Thiagarajan, Vasanthakumar
wrote:
> On Wednesday 27 July 2016 06:13 PM, Michal Kazior wrote:
>> On 27 July 2016 at 14:36, Vasanthakumar Thiagarajan
[...]
>>> --- a/drivers/net/wireless/ath/ath10k/core.c
>>> +++ b/drivers/net/wireless/ath/ath10
On 27 July 2016 at 14:36, Vasanthakumar Thiagarajan
wrote:
> There are slight differences in Rx hw descriptor information
> among different chips. So far driver does not use those new
> information for any functionalities, but there is one important
> information which is available from QCA99X0 on
On 27 July 2016 at 10:33, Benjamin Berg wrote:
> Unfortunately ath10k does not generally allow modifying the coverage class
> with the stock firmware and Qualcomm has so far refused to implement this
> feature so that it can be properly supported in ath10k. If we however know
> the registers that
On 20 July 2016 at 17:24, Toke Høiland-Jørgensen wrote:
> Toke Høiland-Jørgensen writes:
>
>> Felix Fietkau writes:
>>
>>> - if I put a hack in the fq code to force the hash to a constant value
>>> (effectively disabling fq without disabling codel), the problem
>>> disappears and even multiple s
On 20 July 2016 at 13:43, Shajakhan, Mohammed Shafi (Mohammed Shafi)
wrote:
> Michal,
>
> Can you please let me know if this change is fine or not ?
> I am waiting infinitely for your reply long time
Sorry. I was absent for a while and this email slipped by.
Quoting your other email:
> is this
On 12 July 2016 at 12:09, Felix Fietkau wrote:
> Hi,
>
> With Toke's ath9k txq patch I've noticed a pretty nasty performance
> regression when running local iperf on an AP (running the txq stuff) to
> a wireless client.
>
> Here's some things that I found:
> - when I use only one TCP stream I get
On 12 July 2016 at 16:02, Dave Taht wrote:
[...]
>>> tcp_limit_output_bytes is?
>> 262144
>
> I keep hoping to be able to reduce this to something saner like 4096
> one day. It got bumped to 64k based on bad wifi performance once, and
> then to it's current size to make the Xen folk happier.
Not
firmware files making things even
more confusing.
Hence use request_firmware_direct() which does not
produce extra warnings. This shouldn't really
break anything because most modern systems don't
rely on udev/hotplug helpers to load firmware
files anymore.
Signed-off-by: Michal Kazior
---
d
Will be useful for implementing command barriers.
Signed-off-by: Michal Kazior
---
drivers/net/wireless/ath/ath10k/wmi-ops.h | 17 +
drivers/net/wireless/ath/ath10k/wmi-tlv.c | 29 +
drivers/net/wireless/ath/ath10k/wmi.c | 23
and only vdev create/delete
events can fix it. This is why the patch takes
dummy vdev approach.
This could be fixed in firmware as well but having
this fixed in driver is more robust, most notably
when thinking of users of older firmware such as
999.999.0.636.
Reported-by: Marek Puzyniak
Signed-o
This allows placing command barriers for explicit
serializing and synchronizing state.
Useful for future driver development.
Signed-off-by: Michal Kazior
---
drivers/net/wireless/ath/ath10k/core.h | 1 +
drivers/net/wireless/ath/ath10k/wmi.c | 31 +++
drivers/net
ore details.
I didn't do extensive testing but I can confirm
that I am no longer able to reroduce "no channel"
warnings and Marek tells me he no longer sees any
signal bumps on oscilloscope with his QCA9882.
Michal Kazior (4):
ath10k: implement wmi echo command
ath10k: implement
Will be useful for implementing command barriers.
Signed-off-by: Michal Kazior
---
drivers/net/wireless/ath/ath10k/wmi-ops.h | 12
drivers/net/wireless/ath/ath10k/wmi-tlv.c | 28
drivers/net/wireless/ath/ath10k/wmi.c | 29
237.336307] [] pci_device_remove+0x38/0xb0
...
Fixes: 5caa328e3811 ("mac80211: implement codel on fair queuing flows")
Fixes: fa962b92120b ("mac80211: implement fair queueing per txq")
Tested-by: Kalle Valo
Reported-by: Kalle Valo
Signed-off-by: Michal Kazior
---
Notes:
v1:
.
Therefore disable it unless it is really
beneficial with current codebase (which is when
firmware supports smart pull-push tx scheduling).
Signed-off-by: Michal Kazior
---
Notes:
v1:
- improve commit log
drivers/net/wireless/ath/ath10k/core.h | 1 +
drivers/net/wireless/ath/ath10k/mac.c | 24
On 27 June 2016 at 16:36, Mohammed Shafi Shajakhan
wrote:
> Hi Michal,
>
> thanks for the review ..
>
> On Mon, Jun 27, 2016 at 11:27:27AM +0200, Michal Kazior wrote:
>> On 23 June 2016 at 18:40, Mohammed Shafi Shajakhan
>> wrote:
>> > From: Mohammed Shafi Sh
aa328e3811 ("mac80211: implement codel on fair queuing flows")
Fixes: fa962b92120b ("mac80211: implement fair queueing per txq")
Reported-by: Valo, Kalle
Signed-off-by: Michal Kazior
---
Hi Kalle,
Can you verify this fixes call traces you're
seeing, please?
net/mac802
On 23 June 2016 at 18:40, Mohammed Shafi Shajakhan
wrote:
> From: Mohammed Shafi Shajakhan
>
> For chipsets like QCA99X0, IPQ4019 and later we are not getting proper
> NULL func status (always acked/successs !!) when hostapd does a
> PROBE_CLIENT via nullfunc frames when the station is powered of
On 10 June 2016 at 14:52, Bob Copeland wrote:
> Smatch warns about a number of cases in ath10k where a pointer is
> null-checked after it has already been dereferenced, in code involving
> ath10k private virtual interface pointers.
>
> Fix these by making the dereference happen later.
>
> Addresse
On 10 June 2016 at 11:08, Toke Høiland-Jørgensen wrote:
> Michal Kazior writes:
>
>> For A-MPDU all MPDU rx status (except last one) should share the same
>> timestamp. Last one has a different one so all you need is to
>> distinguish first and last MPDU. Non A-MPDU ob
On 10 June 2016 at 10:55, Felix Fietkau wrote:
> On 2016-06-10 10:50, Michal Kazior wrote:
>> On 9 June 2016 at 09:46, A. Benz wrote:
>>> Dear All,
>>>
>>> I am using LEDE on my IPQ806x (QCA9980) system (Archer C2600).
>>> With compat-wireless
On 10 June 2016 at 10:53, Toke Høiland-Jørgensen wrote:
>
>>> I initially thought that using the timestamp put into the frame by the
>>> hardware could be a way to get timing. But there's only a timestamp of
>>> the first symbol in rs_tstamp, and getting a time to compare it with is
>>> difficult;
On 9 June 2016 at 09:46, A. Benz wrote:
> Dear All,
>
> I am using LEDE on my IPQ806x (QCA9980) system (Archer C2600).
> With compat-wireless-2016-05-12, I observed traces attached below.
> The router is unstable and eventually reboots by itself (randomly).
>
> Upon reverting to compat-wireless-20
On 7 June 2016 at 13:12, Toke Høiland-Jørgensen wrote:
> Toke Høiland-Jørgensen writes:
>
>>> [snip]
>>>
>>> I also found one of my notes in my version of this - how can we
>>> estimate the duration of an A-MPDU, when we only get hardware
>>> de-encapsulated frames?
>>
>> Well in my case I'm side
On 10 June 2016 at 01:45, Ben Greear wrote:
> I'm trying to track down a tricky problem with a Mac book and power save
> with my ath10k firmware.
>
> At least part of the issue is that when the Mac sleeps, it goes into
> power-save
> state as far as AP is concerned, and then it simply turns itself
On 31 May 2016 at 14:12, Toke Høiland-Jørgensen wrote:
> Michal Kazior writes:
>
>> This patchset disables qdiscs for drivers
>> using software queuing and performs fq_codel-like
>> dequeuing on txqs.
>
> Hi Michal
>
> Is this version in a git repo somewh
On 26 May 2016 at 15:50, Toke Høiland-Jørgensen wrote:
> This is my attempt to add per-station airtime usage accounting to ath9k.
> For now I just export it to a new debugfs entry, but my plan is to use
> it to make (station) scheduling decisions. However, before attempting
> that I would like som
This adds a few debugfs entries to make it easier
to test, debug and experiment.
Signed-off-by: Michal Kazior
---
Notes:
v5:
* use the single "aqm" debugfs knob [Dave]
v4:
* stats adjustments (in-kernel codel has more of them)
net/mac80211/debu
This adds a debugfs entry to read and modify some
fq parameters and inroduces a module parameter to
control number of flows mac80211 shuold maintain.
This makes it easy to debug, test and experiment.
Signed-off-by: Michal Kazior
---
Notes:
v5:
* expose a single "aqm" debug
overflow.
Hence apply (a slighly modified, to fit the knobs)
CoDel5 on flow queues. This improves TCP
convergence and stability when combined with
wireless driver which keeps its own tx queue/fifo
at a minimum fill level for given link conditions.
Signed-off-by: Michal Kazior
---
Notes:
v4
Signed-off-by: Michal Kazior
---
Notes:
v4:
* removed internal fq.h and re-used in-kernel one
net/mac80211/agg-tx.c | 8 ++-
net/mac80211/ieee80211_i.h | 24 ++--
net/mac80211/iface.c | 12 ++--
net/mac80211/main.c| 7 +++
net/mac80211/rx.c
invalid ptr deref
* fix compilation for backports
Michal Kazior (5):
mac80211: skip netdev queue control with software queuing
mac80211: implement fair queueing per txq
mac80211: add debug knobs for fair queuing
mac80211: implement codel on fair queuing flows
mac80211: add debug knobs for
and
should allow software queuing based drivers
achieve better performance by utilizing airtime
more efficiently with big aggregates.
Signed-off-by: Michal Kazior
---
Notes:
v5:
* fix null-deref for non-txq drivers
* fix compilation after rebase
* fix powersave on ath9k w/ wak
Some setups suffer performance regressions with
current wake_tx_queue implementation.
Signed-off-by: Michal Kazior
---
Hi Roman,
Can you give this patch a try and see if it helps
with your performance problems, please?
drivers/net/wireless/ath/ath10k/core.h | 1 +
drivers/net/wireless/ath
On 10 May 2016 at 01:10, wrote:
> From: Ben Greear
>
> Might want to turn off verbose debug as soon as you
> see a firmware crash, for instance. Helps keep dmesg
> output from over-running the stuff you care about.
>
> Signed-off-by: Ben Greear
> ---
> drivers/net/wireless/ath/ath10k/debug.c
On 10 May 2016 at 01:11, wrote:
> From: Ben Greear
>
> They are not necessarily named in an intuitive manner,
> so at least add some comments to help the next person.
>
> Signed-off-by: Ben Greear
> ---
> drivers/net/wireless/ath/ath10k/core.h | 8
> 1 file changed, 4 insertions(+), 4
On 5 May 2016 at 13:00, Michal Kazior wrote:
[...]
> -static void ieee80211_drv_tx(struct ieee80211_local *local,
> -struct ieee80211_vif *vif,
> -struct ieee80211_sta *pubsta,
> -struct sk_buff *sk
Recent changes revolving around implementing
wake_tx_queue support introduced a significant
performance regressions on some (slower, uni-proc)
systems.
Signed-off-by: Michal Kazior
---
drivers/net/wireless/ath/ath10k/htt_rx.c | 3 ---
drivers/net/wireless/ath/ath10k/mac.c| 7
On 6 May 2016 at 07:51, Dave Taht wrote:
> On Thu, May 5, 2016 at 10:27 PM, Michal Kazior
> wrote:
>> On 5 May 2016 at 17:21, Dave Taht wrote:
>>> On Thu, May 5, 2016 at 4:00 AM, Michal Kazior
>>> wrote:
>>>> This adds a few debugfs entries
On 5 May 2016 at 17:21, Dave Taht wrote:
> On Thu, May 5, 2016 at 4:00 AM, Michal Kazior wrote:
>> This adds a few debugfs entries to make it easier
>> to test, debug and experiment.
>
> I might argue in favor of moving all these (inc the fq ones) into
> their own
and
should allow software queuing based drivers
achieve better performance by utilizing airtime
more efficiently with big aggregates.
Signed-off-by: Michal Kazior
---
Notes:
v4:
* make queue depth limit per interface instead of
per radio [Johannes]
include/net/mac80211.h |
This adds a few debugfs entries and a module
parameter to make it easier to test, debug and
experiment.
Signed-off-by: Michal Kazior
---
net/mac80211/debugfs.c| 77 +++
net/mac80211/debugfs_netdev.c | 28 +++-
net/mac80211
overflow.
Hence apply (a slighly modified, to fit the knobs)
CoDel5 on flow queues. This improves TCP
convergence and stability when combined with
wireless driver which keeps its own tx queue/fifo
at a minimum fill level for given link conditions.
Signed-off-by: Michal Kazior
---
Notes:
v4
This adds a few debugfs entries to make it easier
to test, debug and experiment.
Signed-off-by: Michal Kazior
---
Notes:
v4:
* stats adjustments (in-kernel codel has more of them)
net/mac80211/debugfs.c | 40
1 file changed, 40 insertions
Signed-off-by: Michal Kazior
---
Notes:
v4:
* removed internal fq.h and re-used in-kernel one
net/mac80211/agg-tx.c | 8 ++-
net/mac80211/ieee80211_i.h | 24 ++--
net/mac80211/iface.c | 12 ++--
net/mac80211/main.c| 7 +++
net/mac80211/rx.c
:
* split taildrop, fq and codel functionalities
into separate patches [Avery]
v2:
* fix invalid ptr deref
* fix compilation for backports
Michal Kazior (5):
mac80211: skip netdev queue control with software queuing
mac80211: implement fair queueing per txq
mac80211: add debug knob
On 19 April 2016 at 13:02, Johannes Berg wrote:
>> > How much of that could be done with nftables btw?
>
>> I'm not sure if I follow. Do you mean what I've been able to do with
>> relayd until now? Without link-local ipv6 routing DHCPv6 is broken
>> (could probably addressed with DHCPv6 Relay to a
On 19 April 2016 at 11:11, Johannes Berg wrote:
> On Mon, 2016-04-18 at 13:23 +0200, Michal Kazior wrote:
>
>> You can't really implement complete IPv6 support in relayd though.
>> Link-local routing is forbidden by the spec explicitly and a patch
>> I've cooked
On 19 April 2016 at 11:06, Johannes Berg wrote:
> On Mon, 2016-04-18 at 14:38 +0200, Michal Kazior wrote:
>> On 18 April 2016 at 07:31, Michal Kazior
>> wrote:
>> >
>> > On 17 April 2016 at 00:29, Johannes Berg > > > wrote:
>> > >
>>
On 18 April 2016 at 14:31, Eric Dumazet wrote:
> On Mon, 2016-04-18 at 07:16 +0200, Michal Kazior wrote:
>
>>
>> I guess .h file can give the compiler an opportunity for more
>> optimizations. With .c you would need LTO which I'm not sure if it's
>> availa
On 18 April 2016 at 07:31, Michal Kazior wrote:
> On 17 April 2016 at 00:29, Johannes Berg wrote:
>> On Thu, 2016-04-14 at 14:18 +0200, Michal Kazior wrote:
>>>
>>> + struct ieee80211_vif *vif;
>>> +
>>> +
On 17 February 2016 at 17:55, Felix Fietkau wrote:
> On 2016-02-17 12:55, Grzegorz Bajorski wrote:
>> Client interface briding was only possible when 4addr frames were used with
>> a 4addr/WDS aware AP. It was not possible to do it otherwise due to 3addr
>> frame limitation.
>>
>> The extap logic
On 17 April 2016 at 00:21, Johannes Berg wrote:
>> +static void ieee80211_txq_enqueue(struct ieee80211_local *local,
>> + struct txq_info *txqi,
>> + struct sk_buff *skb)
>> +{
>> + lockdep_assert_held(&txqi->queue.lock);
> [...]
>> +
On 17 April 2016 at 00:29, Johannes Berg wrote:
> On Thu, 2016-04-14 at 14:18 +0200, Michal Kazior wrote:
>>
>> + struct ieee80211_vif *vif;
>> +
>> + /* When packets are enqueued on txq
>> it's easy
>
On 17 April 2016 at 00:25, Johannes Berg wrote:
> On Sun, 2016-04-17 at 00:23 +0200, Johannes Berg wrote:
>> On Thu, 2016-04-14 at 14:18 +0200, Michal Kazior wrote:
>> >
>> >
>> > +++ b/net/mac80211/fq.h
>> >
>> Now that you've mostly rewri
1 - 100 of 841 matches
Mail list logo
