Return-path: Received: from packetmixer.de ([79.140.42.25]:48343 "EHLO mail.mail.packetmixer.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932093AbaGRN0h (ORCPT ); Fri, 18 Jul 2014 09:26:37 -0400 From: Simon Wunderlich To: ath10k@lists.infradead.org Cc: linux-wireless@vger.kernel.org, kgiori@qca.qualcomm.com, sven@narfation.org, mathias.kretschmer@fokus.fraunhofer.de, Simon Wunderlich Subject: [PATCH 0/2] ath10k spectral scan support Date: Fri, 18 Jul 2014 15:26:30 +0200 Message-Id: <1405689992-8538-1-git-send-email-sw@simonwunderlich.de> (sfid-20140718_152645_766617_296D8FCF) Sender: linux-wireless-owner@vger.kernel.org List-ID: This is the first "PATCH" patchset iteration for the ath10k spectral scan feature. The feature is still experimental, but we made quite some progress. The fft_eval tool has also been updated based on the new patchset, and now supports both ath9k and ath10k sample formats in the same branch, which has been merged into master [1]. The bin issue has been solved thanks to the valuable input of Michal and Kathy! In my tests, most modes seem to support 64, 128 or 256 bins, and an option to configure that has been added. 512 bins should be possible as well under some circumstances according to Kathys mail, but I couldn't get that working. Also HT80/64 bin reports look bogus, so these are disabled for now. FFT_eval includes some data samples in case you are interested. The count issue is still open: Even when a count is specified, the hardware seems to send endless samples. It seems to work most of the time in VHT80 mode though, but in HT20 and HT40 the count value seems to be ignored. To reproduce this, start hostapd with the desired channel width and do: echo 8 > /sys/kernel/debug/ieee80211/phy0/ath10k/spectral_count echo manual > /sys/kernel/debug/ieee80211/phy0/ath10k/spectral_scan_ctl echo trigger > /sys/kernel/debug/ieee80211/phy0/ath10k/spectral_scan_ctl cat /sys/kernel/debug/ieee80211/phy0/ath10k/spectral_scan0 >> /tmp/fft.dump Repeating the last line and checking the filesize will easily show whether ath10k still sends samples or not. We would expect 8 samples in this configuration. See the patch for more information on the changes. Thanks again to Michal, Kathy, Kalle and Sven for their support! Cheers, Simon [1] https://github.com/simonwunderlich/FFT_eval Simon Wunderlich (1): ath10k: add spectral scan feature Sven Eckelmann (1): ath: Move spectral debugfs structs to shared header drivers/net/wireless/ath/ath10k/Kconfig | 1 + drivers/net/wireless/ath/ath10k/Makefile | 3 +- drivers/net/wireless/ath/ath10k/core.c | 4 + drivers/net/wireless/ath/ath10k/core.h | 7 + drivers/net/wireless/ath/ath10k/debug.c | 3 + drivers/net/wireless/ath/ath10k/mac.c | 8 + drivers/net/wireless/ath/ath10k/spectral.c | 558 +++++++++++++++++++++++++++++ drivers/net/wireless/ath/ath10k/spectral.h | 64 ++++ drivers/net/wireless/ath/ath10k/wmi.c | 104 +++++- drivers/net/wireless/ath/ath10k/wmi.h | 94 +++++ drivers/net/wireless/ath/ath9k/spectral.h | 71 +--- drivers/net/wireless/ath/spectral_common.h | 112 ++++++ 12 files changed, 958 insertions(+), 71 deletions(-) create mode 100644 drivers/net/wireless/ath/ath10k/spectral.c create mode 100644 drivers/net/wireless/ath/ath10k/spectral.h create mode 100644 drivers/net/wireless/ath/spectral_common.h -- 1.9.1