Return-path: Received: from Viola.Opus1.COM ([192.245.12.8]:3348 "EHLO Viola.Opus1.COM" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750823AbYIFEiH (ORCPT ); Sat, 6 Sep 2008 00:38:07 -0400 Received: from [172.17.0.188] ([61.8.118.66]) by Opus1.COM (PMDF V6.2-X27 #9830) with ESMTPSA id <01MZ7U16R87K8YG025@Opus1.COM> for linux-wireless@vger.kernel.org; Fri, 05 Sep 2008 21:38:03 -0700 (MST) Date: Fri, 05 Sep 2008 21:38:00 -0700 From: gavron@wetwork.net Subject: Re: Speed enhancement for BCM4306/2 In-reply-to: <48C1BA91.7050106@lwfinger.net> To: Larry Finger Cc: Broadcom Wireless , wireless Message-id: <48C20928.10202@wetwork.net> (sfid-20080906_063820_556844_9FAEB4A9) MIME-version: 1.0 Content-type: text/plain; charset=ISO-8859-1; format=flowed References: <48C1BA91.7050106@lwfinger.net> Sender: linux-wireless-owner@vger.kernel.org List-ID: Is this close? E --- /tmp/phy_g.c 2008-09-05 21:06:57.000000000 -0700 +++ phy_g.c 2008-09-05 21:36:18.000000000 -0700 @@ -2198,7 +2198,7 @@ else b43_phy_initb6(dev); - if (phy->rev >= 2 || phy->gmode) + if (phy->rev >= 2 && phy->gmode) b43_phy_inita(dev); if (phy->rev >= 2) { @@ -2216,9 +2216,9 @@ if (phy->gmode || phy->rev >= 2) { tmp = b43_phy_read(dev, B43_PHY_VERSION_OFDM); tmp &= B43_PHYVER_VERSION; - if (tmp == 3 || tmp == 5) { + if (tmp == 4 || tmp == 5) { b43_phy_write(dev, B43_PHY_OFDM(0xC2), 0x1816); - b43_phy_write(dev, B43_PHY_OFDM(0xC3), 0x8006); + b43_phy_write(dev, B43_PHY_OFDM(0xC3), 0x8606); } if (tmp == 5) { b43_phy_write(dev, B43_PHY_OFDM(0xCC), @@ -2226,7 +2226,7 @@ & 0x00FF) | 0x1F00); } } - if ((phy->rev <= 2 && phy->gmode) || phy->rev >= 2) + if ((phy->rev >=2) b43_phy_write(dev, B43_PHY_OFDM(0x7E), 0x78); if (phy->radio_rev == 8) { b43_phy_write(dev, B43_PHY_EXTG(0x01), Larry Finger wrote: > I recently found some places where the G-PHY initialization > specifications were wrong. I updated the relevant parts of the V3 > specifications at http://bcm-specs.sipsolutions.net - the pages for > GPHY, B6PHY and B5PHY were changed. > > When I prepared a patch containing these changes, the speed of the > BCM4306/2 for OFDM rates more than doubled. As an example, when > iwconfig is used to set the rate at 36M, the TX rate increased from > 6.8 to 15.8 Mb/s. > > Unfortunately, direct submission of my patch would violate the > clean-room rules. Michael Buesch has other things that are more > important. As a result, we need someone to look at the changes on the > web page and prepare a patch for submission. I will be happy to test > your submission. > > Larry > _______________________________________________ > Bcm43xx-dev mailing list > Bcm43xx-dev@lists.berlios.de > https://lists.berlios.de/mailman/listinfo/bcm43xx-dev >