Return-path: Received: from ebb05.tieto.com ([131.207.168.36]:63157 "EHLO ebb05.tieto.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753118Ab3G2Iji (ORCPT ); Mon, 29 Jul 2013 04:39:38 -0400 From: Michal Kazior To: CC: , , Michal Kazior Subject: [RFC 0/3] introduce chan_time parameter to scan request Date: Mon, 29 Jul 2013 10:39:15 +0200 Message-ID: <1375087158-22077-1-git-send-email-michal.kazior@tieto.com> (sfid-20130729_103941_885990_9756E761) MIME-Version: 1.0 Content-Type: text/plain Sender: linux-wireless-owner@vger.kernel.org List-ID: Hi, This patchset adds a new scan parameter. It allows specyfing for how long each channel should be visited during scan. I've tested this with ath9k (which doesn't have hw_scan) and ath10k (does have hw_scan). It's not super time-precise, but seems reasonable. The main motivation behind this is to allow easier ACS implementation in hostapd. This could also perhaps be useful for spectral scan and anyone interested in having ability to tune scan times (e.g. very fast initial channel scanning on mobile devices, long passive scans). I'm wondering if it would be okay to set swscan probe delay to 0 if chan_time is non-zero? Or does probe delay need to be non-zero? Or maybe it should also be configurable via a parameter? Pozdrawiam / Best regards, Michal Kazior. Michal Kazior (3): nl/cfg80211: add chan_time for scan request mac80211: add support for scan chan_time parameter ath10k: respect chan_time parameter in scan request drivers/net/wireless/ath/ath10k/mac.c | 6 ++++++ include/net/cfg80211.h | 3 +++ include/uapi/linux/nl80211.h | 7 +++++++ net/mac80211/scan.c | 22 ++++++++++++++++++++-- net/wireless/nl80211.c | 4 ++++ 5 files changed, 40 insertions(+), 2 deletions(-) -- 1.7.9.5