Return-path: Received: from mail.atheros.com ([12.36.123.2]:57077 "EHLO mail.atheros.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751216Ab0CMGFS (ORCPT ); Sat, 13 Mar 2010 01:05:18 -0500 Received: from mail.atheros.com ([10.10.20.105]) by sidewinder.atheros.com for ; Fri, 12 Mar 2010 22:05:18 -0800 Date: Fri, 12 Mar 2010 22:05:16 -0800 From: "Luis R. Rodriguez" To: "linville@tuxdriver.com" CC: "linux-wireless@vger.kernel.org" , Sujith Manoharan , Vasanth Thiagarajan , Luis Rodriguez Subject: Re: [PATCH v3 04/12] ath9k_hw: fix RF analog setup for AR9271 Message-ID: <20100313060516.GA6579@tux> References: <1268459216-22372-1-git-send-email-lrodriguez@atheros.com> <1268459216-22372-5-git-send-email-lrodriguez@atheros.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" In-Reply-To: <1268459216-22372-5-git-send-email-lrodriguez@atheros.com> Sender: linux-wireless-owner@vger.kernel.org List-ID: On Fri, Mar 12, 2010 at 09:46:48PM -0800, Luis R. Rodriguez wrote: > From: Sujith > > AR9271 is a single chip and as such does not have external radios. > Devices with external radio require additional programming, skip > this for AR9271 as we do for other single chips. > > Signed-off-by: Sujith > Signed-off-by: Vasanthakumar Thiagarajan > Signed-off-by: Luis R. Rodriguez > --- > drivers/net/wireless/ath/ath9k/hw.c | 6 +++--- > drivers/net/wireless/ath/ath9k/phy.c | 6 +++--- > 2 files changed, 6 insertions(+), 6 deletions(-) > > diff --git a/drivers/net/wireless/ath/ath9k/hw.c b/drivers/net/wireless/ath/ath9k/hw.c > index b1a4a83..a0ef990 100644 > --- a/drivers/net/wireless/ath/ath9k/hw.c > +++ b/drivers/net/wireless/ath/ath9k/hw.c > @@ -517,7 +517,7 @@ static int ath9k_hw_post_init(struct ath_hw *ah) > ah->eep_ops->get_eeprom_ver(ah), > ah->eep_ops->get_eeprom_rev(ah)); > > - if (!AR_SREV_9280_10_OR_LATER(ah)) { > + if (!AR_SREV_9280_10_OR_LATER(ah) && !AR_SREV_9271(ah)) { > ecode = ath9k_hw_rf_alloc_ext_banks(ah); > if (ecode) { > ath_print(ath9k_hw_common(ah), ATH_DBG_FATAL, > @@ -917,7 +917,7 @@ int ath9k_hw_init(struct ath_hw *ah) > ath9k_hw_init_cal_settings(ah); > > ah->ani_function = ATH9K_ANI_ALL; > - if (AR_SREV_9280_10_OR_LATER(ah)) { > + if (AR_SREV_9280_10_OR_LATER(ah) || AR_SREV_9271(ah)) { > ah->ani_function &= ~ATH9K_ANI_NOISE_IMMUNITY_LEVEL; > ah->ath9k_hw_rf_set_freq = &ath9k_hw_ar9280_set_channel; > ah->ath9k_hw_spur_mitigate_freq = &ath9k_hw_9280_spur_mitigate; > @@ -1266,7 +1266,7 @@ void ath9k_hw_deinit(struct ath_hw *ah) > ath9k_hw_setpower(ah, ATH9K_PM_FULL_SLEEP); > > free_hw: > - if (!AR_SREV_9280_10_OR_LATER(ah)) > + if (!AR_SREV_9280_10_OR_LATER(ah) && !AR_SREV_9271(ah)) > ath9k_hw_rf_free_ext_banks(ah); > kfree(ah); > ah = NULL; > diff --git a/drivers/net/wireless/ath/ath9k/phy.c b/drivers/net/wireless/ath/ath9k/phy.c > index c3b5939..b761330 100644 > --- a/drivers/net/wireless/ath/ath9k/phy.c > +++ b/drivers/net/wireless/ath/ath9k/phy.c > @@ -839,7 +839,7 @@ int ath9k_hw_rf_alloc_ext_banks(struct ath_hw *ah) > > struct ath_common *common = ath9k_hw_common(ah); > > - BUG_ON(AR_SREV_9280_10_OR_LATER(ah)); > + BUG_ON(AR_SREV_9280_10_OR_LATER(ah) || AR_SREV_9271(ah)); > > ATH_ALLOC_BANK(ah->analogBank0Data, ah->iniBank0.ia_rows); > ATH_ALLOC_BANK(ah->analogBank1Data, ah->iniBank1.ia_rows); > @@ -870,7 +870,7 @@ ath9k_hw_rf_free_ext_banks(struct ath_hw *ah) > bank = NULL; \ > } while (0); > > - BUG_ON(AR_SREV_9280_10_OR_LATER(ah)); > + BUG_ON(AR_SREV_9280_10_OR_LATER(ah) || AR_SREV_9271(ah)); > > ATH_FREE_BANK(ah->analogBank0Data); > ATH_FREE_BANK(ah->analogBank1Data); > @@ -910,7 +910,7 @@ bool ath9k_hw_set_rf_regs(struct ath_hw *ah, struct ath9k_channel *chan, > * for single chip devices, that is AR9280 or anything > * after that. > */ > - if (AR_SREV_9280_10_OR_LATER(ah)) > + if (AR_SREV_9280_10_OR_LATER(ah) || AR_SREV_9271(ah)) > return true; > > /* Setup rf parameters */ > -- Hah actually now that I think about it, I bet the driver works just fine without this patch, its the same reason why AR_SREV_9280_10_OR_LATER(ah) adds AR9271... please test without this patch. This should be one less code change. When I reviewed the phy stuff last I meant to add an alias to AR_SREV_9280_10_OR_LATER() as "AR_SREV_SINGLE_CHIP(ah)" in hopes that it clarifies this code entry is usually meant for that, and then remove all the useless double checks spread all over the HAL. Luis