Re: [PATCH v3 1/2] mac80211: Implement Airtime-based Queue Limit (AQL)

2019-10-11 Thread Johannes Berg
On Thu, 2019-10-10 at 19:24 -0700, Kan Yan wrote: > > > + * ieee80211_txq_aql_check - check if a txq can send frame to device > > I wonder if this really should even be have "aql" in the name? It's also > > going to return NULL if there's nothing on the TXQ, for example, right? > > Renamed to

Re: [PATCH] ath10k: check data ack rssi enabled/disabled in htt rx event

2019-10-11 Thread Kalle Valo
Balaji Pothunoori wrote: > For all data packets transmitted, host gets htt tx completion event. Some > QCA9984 > firmware releases support WMI_SERVICE_TX_DATA_ACK_RSSI, which gives data > ack rssi values to host through htt event of data tx completion. Data ack rssi > values are valid if A0 bit

Re: [PATCH] ath10k: check data ack rssi enabled/disabled in htt rx event

2019-10-11 Thread Kalle Valo
Balaji Pothunoori wrote: > For all data packets transmitted, host gets htt tx completion event. Some > QCA9984 > firmware releases support WMI_SERVICE_TX_DATA_ACK_RSSI, which gives data > ack rssi values to host through htt event of data tx completion. Data ack rssi > values are valid if A0 bit

Re: [RFC PATCH 1/2] ath10k: report survey info as accumulated values

2019-10-11 Thread Kalle Valo
Sven Eckelmann writes: > From: Sven Eckelmann > > The survey report is expected to contain a counter which is increasing all > the time. But ath10k reports some kind of delta. This can either be the > difference to the last get_survey or the difference to some even older > get_survey because

Re: [PATCH][next] ath10k: fix null dereference on pointer crash_data

2019-10-11 Thread Kalle Valo
Colin King wrote: > Currently when pointer crash_data is null the present null check > will also check that crash_data->ramdump_buf is null and will cause > a null pointer dereference on crash_data. Fix this by using the || > operator instead of &&. > > Fixes: 3f14b73c3843 ("ath10k: Enable MSA

Re: [PATCH][next] ath10k: fix null dereference on pointer crash_data

2019-10-11 Thread Kalle Valo
Colin King wrote: > Currently when pointer crash_data is null the present null check > will also check that crash_data->ramdump_buf is null and will cause > a null pointer dereference on crash_data. Fix this by using the || > operator instead of &&. > > Fixes: 3f14b73c3843 ("ath10k: Enable MSA

Re: [RFC PATCH 2/2] ath10k: regularly fetch survey counters

2019-10-11 Thread Kalle Valo
Sven Eckelmann writes: > From: Sven Eckelmann > > The survey counters from firmwares like 10.2.4 are not actually using the > full 64 bit. Instead, they only use the lower 31 bit and overflow ever > 14-30s. The driver must frequently fetch the survey data and add it to the > survey data storage

Re: [RFC PATCH 0/2] ath10k: provide survey info as accumulated data

2019-10-11 Thread Kalle Valo
Sven Eckelmann writes: > it was observed that ath9k provides accumulated survey counters but ath10k > neither provides deltas nor accumulated counters. Instead it returns > some value which was returned at some point from the firmware. > > But as it turns out, this data is not reliable. To make

[PATCH v2] ath10k: Correct error handling of dma_map_single()

2019-10-11 Thread Bjorn Andersson
The return value of dma_map_single() should be checked for errors using dma_mapping_error() and the skb has been dequeued so it needs to be freed. Fixes: 1807da49733e ("ath10k: wmi: add management tx by reference support over wmi") Reported-by: Niklas Cassel Signed-off-by: Bjorn Andersson ---

Re: [PATCH] ath10k: Correct error check of dma_map_single()

2019-10-11 Thread Bjorn Andersson
On Fri 11 Oct 04:57 PDT 2019, Kalle Valo wrote: > Bjorn Andersson wrote: > > > The return value of dma_map_single() should be checked for errors using > > dma_mapping_error(), rather than testing for NULL. Correct this. > > > > Fixes: 1807da49733e ("ath10k: wmi: add management tx by reference

[ath6kl:pending-ath11k 316/345] drivers/net/wireless/ath/ath11k/debug.c:575 ath11k_write_simulate_fw_crash() error: we previously assumed 'ar' could be null (see line 549)

2019-10-11 Thread Dan Carpenter
tree: https://git.kernel.org/pub/scm/linux/kernel/git/kvalo/ath.git pending-ath11k head: 727f4796623292eb33ce560a3e25ba0040d17871 commit: 75a016fcc8f79a7d650462c69bb28aa886b4f09e [316/345] ath11k: cleanup and rename ath11k_send_crash_inject_cmd() If you fix the issue, kindly add following

Re: [PATCH] ath10k: Correct error check of dma_map_single()

2019-10-11 Thread Kalle Valo
Bjorn Andersson wrote: > The return value of dma_map_single() should be checked for errors using > dma_mapping_error(), rather than testing for NULL. Correct this. > > Fixes: 1807da49733e ("ath10k: wmi: add management tx by reference support > over wmi") > Cc: sta...@vger.kernel.org >